here. - The Royal New Zealand College of General Practitioners

VOLUME 5 • NUMBER 2 • JUNE 2013OF PRIMARY HEALTH CARE‘There are a varietyof means by whichthe public can hold adoctor to account forpast wrongs.’See page 165Original Scientific PaperMeasles in previously vaccinatedchildrenSee page 93Original Scientific PaperPatient-centred approach to diabetescareSee page 114Original Scientific PaperDecision support for rapidmanagement of transient ischaemicattackSee page 138Original Scientific PaperDrugs stored in kitchens or bathroomsmay degradeSee page 146Back to BackBalancing sun exposure and Vitamin DneedsSee page 154ViewpointCentre for Adverse ReactionsMonitoring a source of practice-basedevidenceSee page 170

CONTENTSVOLUME 5 • NUMBER 2 • JUNE 2013OF PRIMARY HEALTH CAREISSN 1172-6164 (Print)ISSN 1172-6156 (Online)90 EditorialsFrom the Editor90 The path towards perfect practiceFelicity Goodyear-SmithGuest Editorial92 The public health implications of secondary measlesvaccine failureMary Ramsay, Kevin Brown93 Original Scientific PapersQuantitative Research93 Previous vaccination modifies both the clinical disease andimmunological features in children with measlesPeter Mitchell, Nikki Turner, Lance Jennings, Hongfang Dong99 The impact of patient and practice characteristics onretention in the diabetes annual review programmeRawiri Keenan, Janet Amey, Ross LawrensonQualitative Research105 Mental health promotion for gay, lesbian, bisexual,transgender and intersex New ZealandersJeffery Adams, Pauline Dickinson, Lanuola Asiasiga114 Understanding barriers to glycaemic control from thepatient’s perspectiveRon Janes, Janet Titchener, Joseph Pere, Rose Pere, Joy Senior123 Miscommunication between patients and generalpractitioners: implications for clinical practiceSonya MorganMixed Method Research129 Is it time to talk? Interpreter services use in generalpractice within CanterburyKara Seers, Lynley Cook, Gillian Abel, Philip Schluter,Paul BridgfordShort Reports138 Transient ischaemic attack and stroke risk: pilot of a primarycare electronic decision support toolAnnemarei Ranta141 Large increase in opportunistic testing for chlamydia during apilot project in a primary health organisationSunita Azariah, Stephen McKernon, Suzanne Werder146 Personal medicines storage in New ZealandCampbell Hewson, Chong Chi Shen, Clare Strachan, Pauline Norris151 Abortion services in a high-needs district: a communitybasedmodel of careSimon Snook, Martha Silva154 Back to Back154 All people should wear sunscreen or other protection fortheir skin whenever they are exposed to sunlightYes: John Kenealy; No: Ian Reid158 Continuing Professional Development158 String of PEARLS about preventive measures for cardiovasculardisease158 Cochrane Corner: Amitriptyline satisfactorily relieves pain inonly a minority of patients with fibromyalgiaMegan Arroll160 Vaikoloa: Keeping promises, measuring results: the PacificMaternal and Child Health Indicators ProjectFiona Langridge, Teuila Percival, Lani Stowers162 Nuggets of Knowledge: Sedating antihistamines in children—not a good choiceLinda Bryant164 Potion or Poison? Coenzyme Q10Shane Scahill165 Ethics165 Professional accountability of doctors in New ZealandKatharine Wallis170 Viewpoint170 The New Zealand Centre for Adverse Reactions Monitoring:a source of practice-based evidenceRuth Savage173 Book Review173 Doctor Colenso, I presume: An account of missionarymedical practice in New Zealand in the midnineteenthcentury—Ian St GeorgeReviewed by Derek Dow174 Letters to the Editor176 About the Journal of Primary Health CareVOLUME 5 • NUMBER 2 • JUNE 2013 J OURNAL OF PRIMARY HEALTH CARE 89

EDITORIALSFROM THE EDITORThe path towards perfect practiceFelicity Goodyear-Smith MBChB, MD,FRNZCGP, EditorCORRESPONDENCE TO:Felicity Goodyear-SmithProfessor and GoodfellowPostgraduate Chair,Department of GeneralPractice and PrimaryHealth Care, TheUniversity of Auckland,PB 92019 Auckland,New Zealandf.goodyear-smith@auckland.ac.nzIn our lead paper this issue, Mitchell and colleagueslook at the clinical severity of illness inchildren in a community outbreak of measlesand find that those who have been vaccinatedpreviously have significantly less severe illnessdespite inconclusive measles serology. 1 Thisindicates benefits from immunisation despite theapparent vaccine failure. In our guest editorial,Mary Ramsay, Head of Immunisation for PublicHealth England and her colleague Kevin Brownare of the opinion that such secondary infectionsfrom waning immunity are not very contagiousand unlikely to contribute to further measlestransmission, therefore probably will not impedestrategies towards global measles control. 2For many years New Zealand has funded theannual ‘Get Checked’ (now the ‘Diabetes CareImprovement Package’) general practice reviewof patients with diabetes. It has proved challengingto maintain patients within the programme,although research has shown that oncepatients have had two or three reviews, they aremore likely to continue participating. A studyby Keenan and colleagues found that youngerpatients (

EDITORIALSGUEST EDITORIALThe public health implications of secondarymeasles vaccine failureMary Ramsay MB BS, MRCP, FFPH; 1 Kevin Brown MD, MRCP, FRCPath 21Consultant Epidemiologistand Head, ImmunisationDepartment, Public HealthEngland, London, UK2Consultant MedicalVirologist, Virus ReferenceDepartment, Public HealthEnglandJ PRIM HEALTH CARE2013;5(2):92.CORRESPONDENCE TO:Mary RamsayImmunisation Department,Public Health England61 Colindale Avenue,London NW9 5EQ, UKMary.Ramsay@phe.gov.ukThe occurrence of primary measles vaccinefailure, where individuals fail to respond totheir first dose of vaccine, is known to occurin around 5% of those vaccinated. This observationhas led to the global policy for measles control recommendingtwo doses of measles vaccination. Secondaryvaccine failure, where individuals respondto vaccination but lose protective antibody overtime, is less well quantified. Studying secondaryfailure is difficult, because it requires documentationof the initial response to vaccination and becauseconfirmation of the diagnosis using standardserological assays is not straightforward. Despitethis, secondary infection, as defined by a boost inIgG with high avidity IgG, in a patient with a historyof previous infection or vaccination, has beenassociated with classical, mild atypical and evenasymptomatic measles infection. 1–4The routine availability of RT-PCR (reversetranscription–polymerase chain reaction) to detectlow levels of measles RNA, however, has madethe confirmation of secondary infections morestraightforward. The study by Mitchell andcolleagues in this issue 5 systematically comparesdisease severity in vaccinated and unvaccinatedcases. This study confirms previous observationsthat the course of measles infection is less severein vaccinated than unvaccinated cases. The highproportion of cases with documentation of twodoses of vaccine and the absence of IgM positivitysuggests that most of these cases were secondaryfailures. Although the authors make no attemptto estimate the frequency of secondary failure, itsuggests that most countries with high coverageof vaccination should expect to observe suchcases during measles outbreaks. Given the lesssevere presentation and the absence or low levelof IgM in many cases, however, detection of suchinfections through routine surveillance requiresthe use of a less specific case definition and theavailability of specialist microbiology, includingmeasles RT-PCR and IgG antibody avidity.The real public health question, however, iswhether secondary infection from waning immunitycould support measles transmission inhighly vaccinated communities. As it is likelythat unrecognised mild or unapparent secondaryinfections occur more frequently than observedin routine surveillance, secondary infectionhas potential to seriously impede global controlstrategies. No convincing evidence of secondaryfailures contributing to transmission has beenpublished, and it has been hypothesised that highattack rates in vaccinees only occur under conditionsof intense exposure. 6 This would suggestthat secondary infections may be less transmissible,as recently confirmed in a household study ofmumps cases in the Netherlands. 7 These conclusionswould be consistent with the observationof sustained measles elimination in populationswith high vaccination coverage, including thosewhere many individuals were vaccinated morethan 40 years earlier.References1. Pannuti CS, Morello RJ, Moraes JC, Curti SP, Afonso AM,Camargo MC, et al. Identification of primary and secondarymeasles vaccine failures by measurement of immunoglobulinG avidity in measles cases during the 1997 São Paulo epidemic.Clin Diagn Lab Immunol. 2004;11(1):119–22.2. Atrasheuskaya AV, Kulak MV, Neverov AA, Rubin S, IgnatyevGM. Measles cases in highly vaccinated population of Novosibirsk,Russia, 2000–2005. Vaccine. 2008;26(17):2111–8. Epub2008 Feb 27.3. Pedersen IR, Mordhorst CH, Glikmann G, von Magnus H. Subclinicalmeasles infection in vaccinated seropositive individualsin arctic Greenland. Vaccine. 1989;(4):345–8.4. Helfand RF, Kim DK, Gary HE Jr, Edwards GL, Bisson GP,Papania MJ, et al. Nonclassic measles infections in an immunepopulation exposed to measles during a college bus trip. J MedVirol. 1998;56(4):337–41.5. Mitchell P, Turner N, Jennings L, Dong H. Previous vaccinationmodifies both the clinical disease and immunological featuresin children with measles. J Prim Health Care. 2013;5(2):93–98.6. Paunio M, Peltola H, Valle M, Davidkin I, Virtanen M,Heinonen OP. Explosive school-based measles outbreak:intense exposure may have resulted in high risk, even amongrevaccinees. Am J Epidemiol. 1998;148(11):1103–10.7. Snijders BE, van Lier A, van de Kassteele J, Fanoy EB, RuijsWL, Hulshof F, et al. Mumps vaccine effectiveness in primaryschools and households, the Netherlands, 2008. Vaccine.2012;30(19):2999–3002.92 VOLUME 5 • NUMBER 2 • JUNE 2013 J OURNAL OF PRIMARY HEALTH CARE

ORIGINAL SCIENTIFIC PAPERQUANTITATIVE RESEARCHPrevious vaccination modifies both theclinical disease and immunological featuresin children with measlesPeter Mitchell MBBS, DPH; 1 Nikki Turner MBChB, FRNZCGP, MPH; 2 Lance Jennings PhD, FRCPath, FFSc(RCPA); 3 Hongfang Dong MSc (Statistics) 1ABSTRACTINTRODUCTION: Measles that develops in previously vaccinated cases has been reported to be associatedwith modified disease, although severity has usually been assessed by the presence or absence ofsymptoms. To date no studies have attempted to subjectively grade the severity of the clinical features.AIM: To investigate both the objective and subjective severity of measles in vaccinated and unvaccinatedcases in the context of a community outbreak.METHODS: A retrospective observational cohort study conducted in Christchurch in 2009 using notifieddata compared the presentation of measles in 14 confirmed cases that had received at least oneMMR (measles, mumps, rubella) vaccination and 14 age-matched unvaccinated confirmed cases. Additionaldetails on the subjective and objective severity of the illness were obtained from parents/guardiansusing a standardised telephone questionnaire.1Community and PublicHealth, Canterbury DistrictHealth Board, Christchurch,New Zealand2Immunisation AdvisoryCentre, Department ofGeneral Practice and PrimaryHealth Care, The Universityof Auckland, Auckland,New Zealand3Canterbury HealthLaboratories, CanterburyDistrict Health Board,ChristchurchRESULTS: The vaccinated group had significantly fewer clinical features on presentation (p=0.01,RR=1.3, 95% CI 1.1–1.6) and a less severe illness objectively, as measured by height and duration of fever,the number of days needing medication other than paracetamol and days required in bed. Unvaccinatedcases were 2.8 times more likely to have more severe clinical features than vaccinated cases (OR=2.8,95% CI 1.5–5.0). Unvaccinated cases were 3.0 times more likely to develop IgM antibody (RR=3.0, 95% CI0.9–9.3).DISCUSSION: Previously vaccinated children who develop measles are likely to have less severe diseaseand serology results that may be inconclusive, particularly for IgM antibody if tested in the first few daysafter the rash onset.KEYWORDS: Immunoglobulin M; measles; measles-mumps-rubella vaccine; polymerase chain reaction;vaccinationIntroductionMeasles vaccination is highly effective andprimary vaccine failure after two vaccinations israre, with less than 1% failing to seroconvert. 1Primary vaccine failure, following challenge withwild measles virus results in an illness of typicalseverity. 2 However, secondary vaccine failure,when measles develops after initial seroconversion,occurs in up to 6% of those vaccinated afterone dose 3,4 and has been reported to be associatedwith milder or modified disease 2,5–8 and a lowercase fatality rate. 9 In these studies, severity ofdisease has usually been assessed by the presenceor absence of symptoms and none have attemptedto subjectively grade the severity of individualclinical features.Case definitions for probable measles based onthe presence of clinical features alone 10 are notaccurate diagnostic guides in vaccinated communities2,11 because of the incidence of modifiedJ PRIM HEALTH CARE2013;5(2):93–98.CORRESPONDENCE TO:Peter MitchellMedical Officer,Community and PublicHealth, CanterburyDistrict Health Board,PO Box 1475, Christchurch8140, New ZealandPeter.Mitchell@cdhb.health.nzVOLUME 5 • NUMBER 2 • JUNE 2013 J OURNAL OF PRIMARY HEALTH CARE 93

ORIGINAL SCIENTIFIC PAPERQUANTITATIVE RESEARCHdisease; in vaccinated cases diagnosis may even bemore difficult if the symptoms that are presentare less florid. Our aim was to investigate boththe objective and subjective severity of measles invaccinated and unvaccinated cases, as significantdifferences between these groups would havepatient management and surveillance implicationsfor primary care and public health services.MethodsA measles outbreak occurring between June andSeptember 2009 in Christchurch, New Zealand,provided the opportunity to investigate whetherprevious vaccination modified the presentationand severity of measles. A retrospective observationalcohort study was conducted, with casesidentified from the notification database 12 thatcaptured the incidence of six presenting symptomsand laboratory serology (ELISA) results. Allsuspected probable and confirmed cases notifiedby general practitioners to the local public healthservice between June and September 2009 werereviewed. Cases were included if they met thestudy case definition for a confirmed case.Case definitionFor the purpose of this study, the definition of aconfirmed case of measles was based on the NewZealand Communicable Disease Control Manual.13 This definition is as follows:At least 12 months of age with either1. an illness characterised by a maculopapularrash and fever, plus at least one of thefollowing: cough, coryza, conjunctivitis orKoplik spots who was epidemiologicallylinked to a laboratory confirmed case, or2. an illness characterised by either amaculopapular rash or fever with at least oneof the following: cough, coryza, conjunctivitisor Koplik spots, that was confirmed bylaboratory testing as measles. Confirmatorylaboratory tests were demonstration of either:(i) measles virus RNA by PCR (polymerasechain reaction) except where this was within10 weeks of an MMR (measles, mumps andrubella) vaccination, or (ii) measles-specificIgM antibody, except where this was within12 weeks of an MMR vaccination.Cases were considered vaccinated if documentationof the date was provided for at least a singlemeasles vaccination given at over 12 months ofage. Cases were considered unvaccinated if theywere reported as having never been vaccinated.A total of 14 cases previously vaccinated withMMR met the case definition. They were all agedless than 17 years and were individually agematchedto control for age-related severity, 1 with14 unvaccinated cases.Evaluation of severityTo obtain information on the severity of theillness, a standard telephone questionnaire wasadministered by a doctor or health protectionofficer to parents/guardians of all cases. Thequestionnaire was developed, applying a standardrating scale for evaluating severity, usinginformation from the literature on measles signsand symptoms. The questionnaire covered thefollowing objective measures of severity:• height and duration of fever• requirement for analgesia/antipyretic(paracetamol) or other medication• days of confinement to bed or equivalent• hospitalisation• visits to a doctor• time to complete recovery.Parents/guardians were also asked to subjectivelygrade the severity 0 to 5 (with 5 being severe)of each of the following 14 clinical features:rash, cough, coryza, conjunctivitis, otitis media,bronch itis, pneumonia, laryngitis or croup, headache,irritability, confusion, vomiting, diarrhoeaand photophobia. Parents/guardians were unawareof the reason for the study.Ethics approval was not required for this studyunder Section 11 of the Ethical Guidelines forObservational Studies.Data analysisChi-square tests were used to compare the percentagesof cases with diagnostic clinical featuresin the vaccinated and unvaccinated groups. PairedStudent t tests were used to compare both the94 VOLUME 5 • NUMBER 2 • JUNE 2013 J OURNAL OF PRIMARY HEALTH CARE

ORIGINAL SCIENTIFIC PAPERQUANTITATIVE RESEARCHobjective and subjective severity of measles. Subjectivegrades of severity for the 14 clinical featureswere further grouped into three categories,with only the mild (0–1) and more severe (4–5)included in the analysis. The numbers of gradesin these two categories were compared by aChi-square test. The Chi-square test was used toexamine the difference of the immune responsebetween the two groups. The Student t test wasused to detect any association between the timingof the serology test and the presence of IgM antibody,14,15 and any difference in the timing of theserology test between the two groups (vaccinatedand unvaccinated cases) to determine if timingconfounded an apparent association between theimmune response and vaccination. All analyseswere conducted using SPSS 17.0 statistical package(SPSS Inc. Chicago, USA).ResultsOf 168 notified suspected, probable and confirmedmeasles cases, 64 met the study inclusioncriteria. Fourteen had documentation of previousvaccination and 50 reported being unvaccinated.The characteristics of the matched groups weresimilar (Table 1). On presentation, vaccinatedcases were less likely to have Koplik spots andhad significantly fewer clinical features (p=0.01,RR=1.3, 95% CI 1.1–1.6; see Table 2).WHAT GAP THIS FILLSWhat we already know: In previously vaccinated cases, measles signsand symptoms are likely to be less frequent compared with unvaccinated cases.There is some evidence that, in vaccinated cases, measles-specific IgM isless likely to be positive if tested in the first few days after the rash onset.What this study adds: In this study of a New Zealand communityoutbreak of measles, previously vaccinated children were likely to developsubjectively less severe disease compared with unvaccinated cases.Table 1. Characteristics of the study groupsVaccinatedn*=14Unvaccinatedn*=14Mean age 9.6 ± 5.9 years 9.3 ± 5.2 yearsRange 1–16 years 2–16 yearsMales: females 11: 3 10: 4Ethnicity—European: Maori 12:2 13:1Vaccination1 MMR 5 –2 MMR 9 –PCR positive 10 11IgM antibody positive † 2 ‡ 9Clinical criteria met pluscontact with confirmed case § 4 1* Number of cases† Nine cases in each group had serology completed‡ Another case was equivocal§ Does not include cases who were also laboratory confirmedTable 2. Clinical features of measles in the study groupsClinical featureDiagnostic clinical features asnotified by general practitionersn (%)Vaccinatedn=14Unvaccinatedn=14RR (95% CI)p-valueFever 10 (83)* 12 (86) 1.0 (0.7–1.4) 1.00Maculopapular rash 12 (86) 14 (100) 1.2 (0.9–1.4) 0.48Cough 10 (71) 13 (100) † 1.4 (1.0–2.0) 0.05Coryza 11 (85) † 9 (82) ‡ 1.0 (0.7–1.4) 1.00Conjunctivitis 6 (43) 6 (50)* 1.2 (0.5–2.7) 1.00Koplik spots 2 (17)* 9 (82) † 4.9 (1.3–17.9) 0.003Percentage of total clinicalfeatures present when notifiedRR relative risk* 2 cases unknown† 1 case unknown‡ 3 cases unknown65% 84% 1.3 (1.1–1.6) 0.01VOLUME 5 • NUMBER 2 • JUNE 2013 J OURNAL OF PRIMARY HEALTH CARE 95

ORIGINAL SCIENTIFIC PAPERQUANTITATIVE RESEARCHTable 3. Objective severity of measles as assessed by parents/guardiansClinical featureAverage objective severity levelVaccinatedn=14mean (standarddeviation)Unvaccinatedn=14mean (standarddeviation)Mean difference(95% CI)p-valueHeight of fever 38.8°C (0.9 )* 39.8°C (0.8) 1.0°C (0.2–1.7) 0.02Days of fever 2.3 (1.2) 4.3 (1.3) 2.0 (1.2–2.6)

ORIGINAL SCIENTIFIC PAPERQUANTITATIVE RESEARCHlost measurable circulating antibody or had noresponse at all.A potential bias of the study was the parents’/guardians’ reporting of the subjective severity ofthe illness, as that was not independently validated.However, the pattern of modified diseasein vaccinated cases seen in that reporting wasalso apparent in the objective measures (Table 3),as well as in general practitioners’ notificationdetails (Table 2). Another limitation of the studydesign was the small sample size, which limitedthe power of some of the statistical analysesto detect significance. For instance, only fourindividual symptoms were subjectively assessedas significantly less severe in vaccinated cases, althoughthe odds ratio for the unvaccinated grouphaving a greater number of subjectively moresevere symptoms was significant. On the otherhand, the strengths of the study were that itcompared two similar groups of confirmed casesin the same outbreak whose ages were within arelatively narrow age range, and the same laboratorywas used for serology and PCR tests.Of the 27 indicators (Methods and Case definition)reported, only three were either less frequentor less severe in unvaccinated cases. Theywere coryza as a presenting symptom and coryzaand bronchitis as subjectively assessed, but noneof these differences were statistically significant.The results support findings of previous studies2,5–8 that showed that measles associated withvaccine failure was likely to be less severe. Evenfollowing incidental exposure post-vaccination,measles symptoms were ameliorated, 16 a resultconsistent with the suggestion that partialimmunity may account for the findings. 2,16Although studies have found few or no differences,17,18 one of these studies 18 grouped childrenvaccinated prior to 12 months of age with unimmunisedcases.Case definitions for suspected measles in vaccinatedcommunities based only on clinical featureshave been shown to be unreliable, 2,11,19 in partbecause the symptoms in vaccinated cases maybe modified. This presents a quandary for bothsurveillance and public health management. Theintroduction of case definitions with less rigorousclinical criteria capturing presentations withfewer or milder symptoms would result in a shiftfrom the current situation of under-diagnosis toover-diagnosis, resulting in unnecessary publichealth intervention. We therefore suggest thatin previously vaccinated patients with suspectedmeasles who do not satisfy a clinical case definition,a PCR test (rather than serology) be done toestablish the diagnosis if presentation is withinthree to five days of the rash onset. In our experienceand that of others, 2,4,11 serology in thesecases and particularly within this timeframe 20may be associated with false negative IgM results.This study of measles cases in a communityoutbreak has shown that children previouslyvaccinated with MMR who develop measles arelikely to have less severe objective and subjectivedisease, and serology results that may notbe conclusive. The findings indicate gains fromimmunisation despite apparent vaccine failure.The study also identifies the importance of usingPCR to assist with the diagnosis, since casedefinitions for measles based on clinical criteriaalone can be unreliable in previously vaccinatedchildren. Diagnostic accuracy is important froma public health perspective, both for surveillancepurposes and to inform the response.References1. Strebel PM, Papania MJ, Dayan GH, Halsey NA. Measles vaccine.In: Plotkin SA, Orenstein WA, Offit PA, editors. Vaccines.5th ed. London, UK: Saunders-Elsevier; 2008. p.353–98.2. Edmonson MB, Addiss DG, McPherson JT, Berg JL, Circo RS,Davis JP. Mild measles and secondary vaccine failure during asustained outbreak in a highly vaccinated population. JAMA.1990;263:2467–71.3. Anders JF, Jacobson RM, Poland GA, Jacobsen SJ, Wollan PC.Secondary failure rates of measles vaccines: a meta-analysis ofpublished studies. Pediatr Infect Dis J. 1996;15:62–6.4. Mathias RG, Meekison WG, Arcand TA, Schechter MT. Therole of secondary vaccine failures in measles outbreaks. Am JPublic Health. 1989;79:475–8.5. No authors listed. Vaccination against measles: clinical trial oflive measles vaccine given alone and live vaccine preceded bykilled vaccine. Second Report to the Medical Research Councilby the Measles Vaccines Committee. BMJ. 1968;2:449–52.6. Cherry JD, Feigin RD, Shackleford PG, Hinthorn DR, SchmidtRR. A clinical and serologic study of 103 children with measlesvaccine failure. J Pediatr. 1973;82:802–8.7. Smith FR, Curran AS, Raciti A, Black FL. Reported measlesin persons immunologically primed by prior vaccination. JPediatr. 1982;101:391–3.8. Sheppeard V, Forssman B, Ferson MJ, Moreira C, Campbell-Lloyd S, Dwyer DE, et al. Vaccine failures and vaccineeffectiveness in children during measles outbreaks in NewSouth Wales, March–May 2006. Commun Dis Intell Q Rep.2009;33:21–6.VOLUME 5 • NUMBER 2 • JUNE 2013 J OURNAL OF PRIMARY HEALTH CARE 97

ORIGINAL SCIENTIFIC PAPERQUANTITATIVE RESEARCH9. Wolfson LJ, Grais RF, Luquero FJ, Birmingham ME, StrebelPM. Estimates of measles case fatality ratios: a comprehensivereview of community-based studies. Int J Epidemiol.2009;38:192–205.10. Kutty P, Rota J, Bellini W, Redd SB. Measles. In: Manual for thesurveillance of vaccine-preventable diseases. 5th Ed. [Internet].Atlanta, USA: Centers for Disease Control and Prevention;2012. [Cited 2012 May 10]. Available from: http://www.cdc.gov/vaccines/pubs /surv-manual/chpt07-measles.html11. Ferson MJ, Young LC, Robertson PW, Whybin LR. Difficultiesin clinical diagnosis of measles: proposal for modified clinicalcase definition. Med J Aust. 1995;163:364–6.12. EpiSurv, national notification database. Institute of EnvironmentalScience and Research Limited, Porirua, New Zealand.13. Ministry of Health. Communicable Disease Control Manual2012. Wellington, New Zealand: Ministry of Health. [cited2012 Nov 24]. Available from: http://www.health.govt.nz14. Centers for Disease Control and Prevention, Atlanta, USA.Current trends: serologic diagnosis of measles. MMWR MorbMortal Wkly Rep. 1982;31(396):401–2.15. Rossier E, Miller H, McCulloch B, Sullivan L, Ward K. Comparisonof immunofluorescence and enzyme immunoassay fordetection of measles-specific immunoglobulin M antibody. JClin Microbiol. 1991;29:1069–1071.16. Sakuta H, Sawada S, Kuroki Y. Severity of measles amongpatients with incidental postexposure vaccination. Jpn J InfectDis. 2008;61(4):304–6.17. Lee K-Y, Lee H-S, Hur J-K, Kang J-H, Lee B-C. Clinical featuresof measles according to age in a measles epidemic. Scand JInfect Dis. 2005;37:471–5.18. Nkowane BM, Bart SW, Orenstein WA, Baltier M. Measlesoutbreak in a vaccinated school population: epidemiology,chains of transmission and the role of vaccine failures. Am JPublic Health. 1987;77:434–8.19. Brown DW, Ramsay ME, Richards AF and Miller E. Salivarydiagnosis of measles: a study of notified cases in the UnitedKingdom, 1991–3. BMJ. 1994;308:1015–7.20. Leland DS. Measles and mumps. In: Dettrick B, Hamilton RG,Folds JD, editors. Manual of molecular and clinical laboratoryimmunology. 7th ed. ASM Press, American Society of Microbiology;2006. p.707–711.ACKNOWLEDGEMENTSWe would like to thankPaul Schooldermanand Sue McEwan fortheir assistance withadministering thequestionnaires. Alltesting was performedat Canterbury HealthLaboratories, CanterburyDistrict Health Board,Christchurch.COMPETING INTERESTSNone declared.98 VOLUME 5 • NUMBER 2 • JUNE 2013 J OURNAL OF PRIMARY HEALTH CARE

ORIGINAL SCIENTIFIC PAPERQUANTITATIVE RESEARCHThe impact of patient and practicecharacteristics on retention in the diabetesannual review programmeRawiri Keenan MBChB; 1 Janet Amey MSocSc (Hons); 1 Ross Lawrenson MBBS, MD, FRCGP, FFPH, FAFPHM 2ABSTRACTINTRODUCTION: Despite more than 10 years of the diabetes annual review (DAR) programme, ensuringthe annual return of diabetic patients for review remains a challenge for primary care. Regardlessof future arrangements for diabetes review programmes, regular review of patients remains clinicallyimportant.1Midlands Health Network,Hamilton, New Zealand2Department of GeneralPractice and Primary HealthCare, Waikato Clinical School,The University of Auckland,Hamilton, New ZealandAIM: To investigate the effect of patient and practice characteristics on the retention of patients continuouslyenrolled with the same practice in the DAR programme.METHODS: We undertook a retrospective, observational study of a cohort of enrolled diabetic patientswho had a DAR in the July 2006 – June 2007 reporting year and remained enrolled with the same practicefor the following three years. Controlling for death and migration, retention rates were calculated for age,gender, ethnicity, rurality, practice funding type and practice nurse (PN) to general practitioner (GP) ratio.RESULTS: The study included data from 78 practices and 6610 patients with Type 2 diabetes. Non-Maori and those aged 60 years and over were more likely to be retained in the programme. For practicefactors, those with a higher PN to GP ratio had a significant retention advantage. Rurality and fundingtype was not shown to have a significant role in retention.DISCUSSION: Results support the view that both patient and practice factors influence a patient’sretention within the DAR programme. The PN to GP ratio may be an important factor in the retention ofpatients in a DAR programme and warrants further research and consideration when planning futureprimary care models.KEYWORDS: Chronic disease; diabetes mellitus; general practice; nurses; primary health care; rural healthIntroductionThe burden of Type 2 diabetes in New Zealandis well documented. 1 The New Zealand ‘GetChecked’ diabetes annual review (DAR) programmewas implemented in 2000, aimed atimproving clinical outcomes for patients withdiabetes through a regular review. 2 Despite previousresearch looking at factors contributing toretention, 3,4 ensuring the annual return of peoplewith diabetes for review remains a challenge forprimary care. The Minister of Health announcedin September 2011 that the DAR programme willbe discontinued as it currently stands, with a newprogramme commencing in July 2012. Regardlessof such programme-level change, regular reviewremains important for people with diabetes and,therefore, retention issues remain a measure ofquality of care.One useful indicator of quality of care is the percentageof eligible patients having had a DAR inany year. This cross-sectional method, however,does not account for patient migration or deathJ PRIM HEALTH CARE2013;5(2):99–104.CORRESPONDENCE TO:Rawiri KeenanMidlands Health Network,PO Box 983, Hamilton3240, New Zealandrawiri.keenan@midlandshn.health.nzVOLUME 5 • NUMBER 2 • JUNE 2013 J OURNAL OF PRIMARY HEALTH CARE 99

ORIGINAL SCIENTIFIC PAPERQUANTITATIVE RESEARCHand identification of new patients with diabetes.A quality (or quality assurance) plan (QP)has been part of the Midlands Health Network(MHN, previously Pinnacle) for the last 14 years.The QP measurement of DARs is a cross-sectionalanalysis of practice performance over a financialyear. In terms of instigating clinical change, auditand feedback through programmes such as the QPhave been effective. However, the use of audit hasnot been universal and there is little evidence onhow to use it effectively. 5 Generally, a longitudinalapproach is superior to a cross-sectional onefor investigating factors affecting quality. 6The MHN is one of the largest primary health organisationsin the country, with general practicesin the Gisborne, Taupo, Waikato and Taranakiregions. This network of 100 general practiceteams is responsible for the primary care of almost450 000 enrolled people. This includes approximately17 500 patients with diabetes (2009/10), ofwhich 12 756 (71%) had their DAR in that singleyear. 7 A prior study in the MHN of a subset ofthree practices showed that the practice recordingof diabetes was complete and reliable. 8Previous work investigating retention has focusedon the overall health system and patient. It hasshown that once a patient with diabetes attendstheir second or third consecutive review, they aremore likely to continue to participate year afteryear. 4 That study followed the patient cohortacross practices and investigated the effect ofpatient factors such as age, gender and ethnicityon retention in the programme. This currentstudy looked at retention rates within individualpractices where the patient remained continuouslyenrolled. We aimed to look at the same patientfactors specifically for MHN patients, as well asfactors associated with the practice. The objectivewas to identify factors affecting retentionrates over a four-year time period for the samecohort of diabetic patients enrolled with the samegeneral practice.MethodsThis was a retrospective, observational study ofa cohort of MHN enrolled diabetic patients whohad a DAR in the July 2006 – June 2007 qualityreporting year and remained enrolled withthe same practice for the following three years,whether they had a subsequent DAR or not.MHN collects data from member general practicesfor the purposes of payment for service andquality reporting. The data are collected fromthe practice patient management systems usingclinical extracts and are held securely on MHNservers. All MHN staff, including any contractedstaff, sign a confidentiality agreement on commencementof employment, ensuring the privacyof health data. The review was conducted aspart of MHN quality assurance and continuousimprovement. Exemption from ethics committeereview was applied for and granted by theNorthern Y Regional Ethics Committee (EthicsRef. NTY/11/EXP/013).A new ‘retention dataset’ was created through thecombination of multiple datasets from the MHNdata warehouse (specifically ‘ever coded diabeticpatients’, ‘get checked’ and ‘enrolled patients’)and an updated ‘mortality dataset’ provided bythe Ministry of Health. The datasets were linkedby patient national health index (NHI) number.Aside from NHI number, any other identifyingdata were removed. Alongside the combinedretention dataset were data from the 2009 MHNWorkforce Census concerning general practitioner(GP) full-time equivalents (FTE) and practicenurse (PN) FTE per practice, practice rural rankingand practice funding type. 9The original dataset included 95 practices; 16were excluded for a combination of data-relatedissues based around the compatibility of differentpatient management software, as well as practicesthat left or joined MHN over the period ofinterest. We also excluded patients with Type 1diabetes, leaving 78 practices with 6875 individualswho were continuously enrolled according toour definition. A further group of 265 patientswere excluded because of missing ethnicity data,taking the study population to 6610.Although strictly speaking three DARs in threeyears is considered best practice, in order to reflectthe realities of general practice, retention was definedas ‘any individual from the 2006/07 cohortwho attended two further reviews in the followingthree years’. Patient deaths and transfers out100 VOLUME 5 • NUMBER 2 • JUNE 2013 J OURNAL OF PRIMARY HEALTH CARE

ORIGINAL SCIENTIFIC PAPERQUANTITATIVE RESEARCHTable 3. Retention rates by specific patient and practice factors (univariate analysis)EthnicityAgeDARin 2006/07*(n=6610)Continuouslyenrolled †(n=5861)Numberretained ‡ Total Odds ratio 95% CINon-Maori 5389 4786 4200 88% 1.70 || 1.43–2.02Maori 1221 1075 863 80%≥60 yrs 4369 3872 3501 90% 2.60 || 2.24–3.020–59 yrs 2241 1989 1562 79%Gender §Male 3289 2904 2516 87% 1.05 0.91–1.22Female 3319 2955 2545 86%PN to GP ratio≥1 3406 2999 2625 88% 1.33 || 1.15–1.54

ORIGINAL SCIENTIFIC PAPERQUANTITATIVE RESEARCHcare. 11,12 Using a study design that looks at patientand practice characteristics together in a multivariatemodel has allowed the factors of the mostsignificance overall, covering both patient andpractice, to be identified. Previously these werelooked at separately in any research on DAR retention.3,4,8,9,10,11 Combining them in this way alsoattempts to reflect the realities of general practice.As the study is designed, it has enabled a largeenough cohort to consider Maori versus non-Maori retention by age and a number of otherfactors. A weakness of this, though, was thegrouping of all non-Maori diabetic patients together.Although data on Maori are of interest ininvestigating ways to reduce health inequalities,it is known that Pacific people in New Zealandand those from the Indian subcontinent havesimilarly poor health outcomes as Maori. 13,14 Itis also documented that other ethnic minorities,such as Asian and Pacific people, have poorerretention in the DAR programme. 8 Overall, thenumbers of Asian and Pacific patients are low inour cohort, though by grouping with all non-Maori, the true difference between Maori andEuropean non-Maori may in fact be greater thanshown here. Since this study is looking at retentionwithin the programme and not outcomes,it is reasonable to group patients as has beendone. Another weakness of this study is the 17excluded practices. Our knowledge of the MHNdata, and of the practices themselves, gives us noreason to suggest that these practices are any differentfrom the ones included in our study.Practice characteristicsOur study shows some evidence for furtherinvestigating the effects of staffing levels andorganisation of general practice teams. This isof particular importance for future models ofcare currently being developed in New Zealandto cope with the rising health burden ofdiabetes. The ratio of 1:1 PN to GP was chosenfrom anecdotal evidence within the MHN ofhigher performance in those practices with morenurses. Our study had approximately a 50:50split between the practices with a PN:GP ratio of1:1 or more and those with less than this. Thosepractices with a high ratio of PN:GP were moreeffective at retaining their patients in this programme,supporting our anecdotal evidence. Thisresult is also in keeping with European studieslooking at multiple factors in quality of care,including practice nurse staffing. 14,15,16Potentially significant gains can be made in DARand other areas of patient care if we look at currentpractice models and find ways to improvethe organisation of general practice teams. Asservices are reconfigured to contain rising costsand reduce the current (and anticipated) burdenon secondary care services, it is important thatprimary care is appropriately supported to playits role as effectively and efficiently as possible.The piloting of new models of primary care andplanning for service configuration is underwayin the MHN, making these issues of particularrelevance in our network.The rural ranking scale used in this study is alocal, albeit arbitrary method of defining ruraland non-rural practices. 18 Other practice factorswere not found to be significant in the multivariateanalysis. It was interesting though to seethat rural practices in the univariate model weremore likely to retain patients than urban ones.A study from the UK showed no difference inprocess factors between rural, remote and urbanpractices. 19 As with that study, the definitionof rural was measured arbitrarily; here we usedthe rural ranking scale. The higher retention inrural practices when directly compared to urbanpractices in our study could be due to less patient‘churn’, or practice staff personally knowing theirenrolled patients. Anecdotal stories, such as of‘catching people’ in the supermarket, are commonfrom rural practice in New Zealand. However, wefound that being a rural practice was not associatedwith improved retention rates in a multivariateanalysis, after adjusting for age, ethnicity andPN to GP FTE ratio.We investigated practice funding for two reasons.Firstly, it was possible that the workload ofVLCA practices and their high needs populationcreated a difference in retention. Also, VLCAfunding can be an approximation of the socialdeprivation of a practice population. Practiceeligibility for VLCA funding is determined byethnicity and the deprivation level of the area itsenrolled patients live in (as measured by the NewVOLUME 5 • NUMBER 2 • JUNE 2013 J OURNAL OF PRIMARY HEALTH CARE 103

ORIGINAL SCIENTIFIC PAPERQUANTITATIVE RESEARCHACKNOWLEDGEMENTSWe would like to thank BenAmey for his assistancewith the SAS programme,Associate ProfessorDharmalingham (MonashUniversity) for his statisticaladvice and Dr GaryJackson for his supportand review of the study.FUNDINGThe Midlands HealthNetwork provided staffingsupport for this study.COMPETING INTERESTSRawiri Keenan and JanetAmey are employed by theMidlands Health Network.Zealand Deprivation Index 21 ). Separating the effectsof ethnicity is difficult and the deprivationscore of patients would perhaps have been a moreexact factor; but, nevertheless, VLCA provided auseful proxy for our study.Patient characteristicsPatient characteristics affecting diabetes care inNew Zealand have been described previously. 3Joshy et al. 4 showed older and non-Maori patientswere more likely to have regular diabetes review.In the MHN, and arguably across the country,the ethnicity of a patient has always, anecdotallyat least, been thought of as the main characteristicaffecting that patient’s retention. This showsthat while ethnicity is important, age is the mostsignificant variable. Gender made no impact onthe retention rates, although women in NewZealand in general have been shown to be morelikely to visit the doctor than men. 20Our results show that patient factors play astrong role in affecting a patient’s likelihood ofbeing retained in an annual review programme.The results, however, do support the idea thatpractice factors can play an important role. ThePN to GP ratio featured strongly and is somethingthat needs to be considered when planningpractice organisation in the future. Overall, thereis a continual need to focus on the review ofyounger and Maori patients, while continuing toacknowledge the role of the practice in determiningthe care patients receive. We expect this toremain true as the MHN transitions from thecurrent diabetes ‘Get Checked’ programme to theDiabetes Care Improvement Package which commencedin July 2012.References1. Joshy G, Simmons D. Epidemiology of diabetes in NewZealand: revisit to a changing landscape. N Z Med J.2006;119(1235):U1999.2. Health Funding Authority. Diabetes 2000. Wellington: HealthFunding Authority; March 2000.3. Tesa P, Le Lievre C, Lawrenson R. Why don’t patients with diagnoseddiabetes attend a free ‘Get Checked’ annual review?J Prim Health Care. 2009;1(3):222–225.4. Joshy G, Lawrenson RA, Simmons D. Retention of patients inthe Get Checked free annual diabetes review programme inNew Zealand. N Z Med J. 2008 Mar 4;121(1270):35–44.5. Foy R, Eccles MP, Jamtvedt G, Young J, Grimshaw JM, BakerR. What do we know about how to do audit and feedback?Pitfalls in applying evidence from a systematic review. BMCHealth Serv Res. 2005;5:50.6. Weiner M, Long J. Cross-sectional versus longitudinal performanceassessments in the management of diabetes. MedCare. 2004 Feb;42(2 Suppl):34–39.7. Midlands Health Network. QP13, 2009/10 Quality PlanReport. Hamilton: PGL Limited; 2010.8. Lawrenson R, Gibbons V, Joshy G, Choi P. Are there disparitiesin care in people with diabetes? A review of care provided ingeneral practice. J Prim Health Care. 2009;1(3):177–183.9. Pinnacle Group Limited. Workforce trends—summary ofresults from the 2009 census of general practitioners, practicenurses and practice management staff in the Pinnacle GeneralPractice Network. Hamilton: PGL Limited; 2010.10. Office of the Auditor-General. Effectiveness of the GetChecked Programme [Internet]. 2010; [cited 2011 June 24].Available from: http://www.oag.govt.nz/2010/diabetesprogramme11. Li R, Simon J, Bodenheimer T, Gillies RR, Casalino L,Schmittdiel J, et al. Organizational factors affecting the adoptionof diabetes care management processes in physicianorganizations. Diabetes Care. 2004;27(10):2312–2316.12. Janssen PG, Gorter KJ, Stolk RP, Rutten GE. Do characteristicsof practices and general practitioners influence the yield ofdiabetes screening in primary care? The ADDITION Netherlandsstudy. Scand J Prim Health Care. 2008;26(3):160–165.13. Hughes LO, Raval U, Raftery EB. First myocardialinfarctions in Asian and white men. BMJ. 1989 May20;298(6684):1345–1350.14. Paddison CA. Exploring physical and psychological wellbeingamong adults with Type 2 diabetes in New Zealand: identifyinga need to improve the experiences of Pacific peoples. N ZMed J. 2010;123(1310):30–42.15. Griffiths P, Maben J, Murrells T. Organisational quality, nursestaffing and the quality of chronic disease management inprimary care: observational study using routinely collecteddata. Int J Nurs Stud. 2011;48(10):1199–1210.16. Spigt M, Stefens C, Passage D, Van Amelsvoort L, Zwietering P.The relationship between primary health care organization andquality of diabetes care. Eur J Gen Pract. 2009;15(4):212–218.17. den Engelsen C, Soedamah-Muthu SS, Oosterheert NJ,Ballieux MJ, Rutten GE. Improved care of Type 2 diabetespatients as a result of the introduction of a practice nurse:2003–2007. Prim Care Diabetes. 2009;3(3):165–171.18. Variation of Advice Notice Pursuant to Section 51 of the Healthand Disability Service Act 1993 (Schedule 2, Appendix 11).Wellington: Health Funding Authority; 1999.19. McLean G, Guthrie B, Sutton M. Differences in the quality ofprimary medical care services by remoteness from urban settlements.Qual Saf Health Care. 2007;16(6):446–449.20. Santosh J, Crampton P. Gender differences in generalpractice utilisation in New Zealand. J Prim Health Care.2009;1(4):261–269.21. Salmond C, Crampton P, Atkinson J. NZDep2006 Index ofDeprivation. Wellington: Dept. of Public Health, University ofOtago; 20 07.104 VOLUME 5 • NUMBER 2 • JUNE 2013 J OURNAL OF PRIMARY HEALTH CARE

ORIGINAL SCIENTIFIC PAPERQUALITATIVE RESEARCHtions were not expected to have any interestin mental health and were unlikely to reply, aresponse rate has not been calculated.Key informant interviewsSeventeen key informants were interviewed byone of the research team, primarily by telephone,although some interviews were conducted face toface. Two informants provided written responsesto questions. Informants were typically peopleworking in, or having some other interest in, oneor more of the areas of GLBTI mental health. Keyinformants were identified through our knowledgeof the field, networking and in discussionwith Te Pou o Te Whakaaro Nui: The NationalCentre of Mental Health Research, Informationand Workplace Development (funder), and bysnowball sampling. A diverse range of informantswere interviewed, including people whoidentified as GLBTI, or worked with or had otherknowledge of these groups (see Appendix 2 in theweb version of this paper). Some informants identifiedas Maori or Pacific, and several informantsworked with young people. The interviews primarilysought informants’ knowledge of mentalhealth promotion and services targeted at GLBTIpeople, and areas for improvement (Appendix 3 inthe web version of this paper).Online qualitative surveyfor GLBTI individualsAn online qualitative survey was developed togather the views of GLBTI individuals (mentalhealth service users and non-users of services)and included questions about issues and gaps incurrent mental health promotion and services(Appendix 4 in the web version of this paper).The online submission process was promoted ongaynz.com (a news website aimed at GLBTI people),and an email with a hyperlink to the submissionform was sent directly to all GLBTI organisationsthat the authors were able to identify. Severalorganisations promoted the submission process totheir members via their mailing lists and websites.Data analysisThe email survey collected descriptive information.This information was reviewed by theWHAT GAP THIS FILLSWhat we already know: While gay, lesbian, bisexual, transgender, andintersex (GLBTI) individuals have the same basic mental health promotionneeds as members of the general population, they also experience additionalunique issues related to social discrimination, and to personal and communitysocial and behavioural risk factors. General practitioners play a central role inthese groups accessing mental health services.What this study adds: There is minimal policy in New Zealand in relationto GLBTI mental health promotion, and only a few mental health promotioninitiatives or services are directed at these populations. Appropriatelevels of mental health promotion and service provision should be availablefor GLBTI people.research team and collated to provide a descriptionof mental health promotion activities andservices. The survey also provided informationabout strategies and policies in relation to mentalhealth promotion. Copies or summaries of policiesand strategies were obtained and reviewed toassist in the description of mental health promotionactivities and services.The key informant interviews were audio recordedand transcribed; a further two were providedin written form. The online survey data were alsoprovided in written form from the online surveysoftware. Data were analysed using thematicanalysis. 25 The focus of the analysis was on thesemantic (explicit) level content of the data. Datafrom the key informants and GLBTI individualswere coded separately. The initial coding andprovisional themes were reviewed and discussedby the researchers. Due to the congruence betweencoding and provisional themes developedfrom the key informant and GLBTI individuals’datasets, the two sources of data were combinedat this point. The researchers undertook furtherrefinement of the coding and analysis.EthicsThis study was undertaken in accord with MasseyUniversity processes for ethical conduct ofresearch. The project was assessed by peer reviewto be low risk. Consequently, it was not reviewedby one of the University’s Human Ethics Committees.The researchers (authors of the article)VOLUME 5 • NUMBER 2 • JUNE 2013 J OURNAL OF PRIMARY HEALTH CARE 107

ORIGINAL SCIENTIFIC PAPERQUALITATIVE RESEARCHwere responsible for the ethical conduct of theresearch. Potential survey respondents wereadvised participation was entirely voluntary andanonymous, and withdrawal at any time beforecompleting and electronically submitting thesurvey without giving a reason was available. Keyinformants were provided with an informationsheet and a consent form that they signed if theyagreed to take part.FindingsA total of 124 GLBTI individuals completed theonline submission form (Table 1). The key findingsof the study are reported in two sections:1. current mental health promotion and serviceprovision; and2. key issues and gaps in mental healthpromotion and service provision.Current mental health promotionand service provisionSuccessful mental health promotion requires relevantpublic policy and supportive environments,but also requires services directed at populations atrisk for mental health problems. 24 In New Zealand,as there is limited policy developed to address thehealth needs of GLBTI populations, 26 it is necessaryto investigate mainstream policy, strategies,services and programmes as these are likely to havesome effect on the health and wellbeing of GLBTIpopulations. In addition to mainstream policies,some mental health services that are specific toGLBTI populations are identified.Policy and strategyThere are several overarching policy and strategydocuments to guide strengths-based mentalTable 1. Demographics of GLBTI respondents (n=124)Population group n Mental health service user nGay man 44 Yes—current or former 88Bisexual man 2 No 36Lesbian 28 Location (nearest city) nBisexual woman 13 Auckland 37Transgender 14 Wellington 35Other and more than one group 23 Dunedin 19Age (years) n Christchurch 620 and under 9 Hamilton 621–30 30 Tauranga 231–40 22 Whangarei 241–50 29 Paraparaumu 251–60 25 Nelson 261–70 5 New Plymouth 271 and over 4 Gisborne 1Ethnicity n Rotorua 1Pakeha/New Zealander 82 Levin 1Maori 11 Not stated 8Pasifika 3Asian/Indian 3Other European 15Other various 10 GLBTI gay, lesbian, bisexual, transgender and intersex people108 VOLUME 5 • NUMBER 2 • JUNE 2013 J OURNAL OF PRIMARY HEALTH CARE

ORIGINAL SCIENTIFIC PAPERQUALITATIVE RESEARCHhealth promotion and mental health serviceprovision in New Zealand, some of whichinclude mention of some or all of the GLBTIpopulations. These include: the New ZealandSuicide Prevention Strategy, The New ZealandSuicide Prevention Action Plan, Youth Health: AGuide to Action, Te Tahuhu, Health PromotingSchools, Health and Physical Education in theNew Zealand Curriculum, Youth DevelopmentStrategy Aotearoa, Professional Standards forTeachers and the National AdministrationGuidelines for schools (see Appendix 5 in theweb version on this paper).General mental health promotionTo translate policies and strategies into action,general population level services and programmesare provided. It is likely that some of thesemainstream programmes will benefit GLBTIpeople, but there is no evaluation evidence tosupport this. These services include: the NationalDepression Initiative; Like Minds, Like Mine;public health units funded by the Ministry ofHealth; the Mental Health Foundation; MentalHealth 101 and Travellers (see Appendix 6 in theweb version on this paper).GLBTI-focused mental health servicesA number of social and other GLBTI organisationsprovide general support, social, advocacyand information services that are likely tocontribute to the mental health and wellbeing ofGLBTI people. Examples of organisations providingthese services are: Body Positive; Women’sCentre (Christchurch); Pink Health Otautahi;Step Ahead Trust—Rainbow Group; WellingtonGay Switchboard; GenderBridge; and IntersexNZ. A range of specific community-based afterschooland school-based youth support serviceshave previously been identified. 27,28However, those organisations which have servicesand programmes with a specific health promotionfocus for some or all of the GLBTI populationsare much more limited (see Appendix 7 in theweb version on this paper). Identified servicesare provided by Auckland CADS (CommunityAlcohol and Drug Services), OUTLine NewZealand, New Zealand AIDS FoundationTable 2. GLBTI-focused mental health resourcesName of resourceCurious websiteYou, me, us. Our people, Our relationships(booklet and posters)Safety in our schoolsMaking schools safe for people ofevery sexualitySocial and ethical issues in sexuality(curriculum resource)Trans people: Facts & information(fact sheets)pridenz.com websiteSexuality, gender identity and depression(fact sheet)Column in express!(Alcohol and drug issues, and relatedhealth matters)Gay men’s mental healthRecommended reading on queermental healthBipolar bear blog(NZAF), Rainbow Youth, and City Associates.In relation to mental health, these services workmainly at a settings level (e.g. in schools) or atthe personal level (e.g. counselling). In addition,12 GLBTI-focused mental health promotionresources were identified, including mentalhealth promotion print and online resources (seeTable 2).A chief gap identified by the review was the lackof services, programmes or funding of GLBTIfocusedmental health initiatives by public healthunits at district health boards. The exceptionto this was funding allocated by the AucklandDistrict Health Board for a community projectworker based at OUTLine New Zealand. Theview of several district health boards contactedwas that mental health services are availableto all, and there is no need for GLBTI-specificservices or programmes. It should be noted thatsome mainstream health services appear to be pro-Producer of resourceNZ AIDS Foundation and RainbowYouthOUTLine NZ and Rainbow YouthOut There project (NZAF andRainbow Youth)PPTAGLBTI gay, lesbian, bisexual, transgender and intersex peoplePPTA Post Primary Teachers’ AssociationCADS Community Alcohol and Drug ServicePPTA and New Zealand SecondaryPrincipals’ CouncilHuman Rights Commissionpridenz.comNational Depression InitiativeDiana Rands, CADSChris Banks, Mental HealthFoundationMental Health FoundationChris BanksVOLUME 5 • NUMBER 2 • JUNE 2013 J OURNAL OF PRIMARY HEALTH CARE 109

ORIGINAL SCIENTIFIC PAPERQUALITATIVE RESEARCHviding services to GLBTI populations, but thesewere not identified through the survey sent todistrict health boards. For example, one informantnoted Kidz First Hospital run by CountiesManukau District Health Board provides supportfor transgender youth, and that Auckland SexualHealth Service (Greenlane Clinical Centre) hasprovided counselling when transgender people approachthem to gain access to hormone treatment.Key issues and gapsThe analyses of issues and gaps identified by keyinformants and GLBTI respondents who completedthe online submission form are presentedas three themes:• macro-social environment• social acceptance and connection, and• services and support.These different groupings acknowledge individualsinteract with different types of environmentalsystems and these interactions impact on theirhealth and wellbeing. 29 The themes are presentedas summaries; further elaboration and analysis isavailable. 23Macro-social environmentThe macro-social system relates to the cultureand broader social environment in which individualslive. A key issue for informants and respondentswas the negative impact on the mentalhealth of GLBTI people that arose from stigmaand homophobia or transphobia. While some ofthe actions that lead to GLBTI people experiencingstigma and homophobia or transphobia wereviewed as resulting from deliberate acts, theseactions were also often reported as being lessdeliberate. Education and general public awarenesscampaigns were suggested as one way toaddress these issues and to raise understanding inmainstream society of GLBTI issues.The informants and respondents also reported aneed to de-stigmatise mental health issues—bothwithin society as a whole and within the GLBTIcommunity. Awareness campaigns were suggestedas an appropriate way to address these issues.The need for all health promotion to include theneeds of GLBTI people was noted. Health promotionactivities need to recognise the diversitywithin the GLBTI population.Social acceptance and connectionThe negative effects of the immediate contextfor individuals were also identified by manyinformants and respondents. In particular, poorsocial acceptance and connection were identifiedas factors that contribute to hostile conditionsin which to achieve good mental health andwellbeing. Along with broader social acceptance,receiving support from friends and family,and ensuring support and safe environments foryoung people and older people were identified asimportant. The need for GLBTI people to addressnegative issues within their own communities,relating to supporting community members, wasalso discussed.Services and supportAccess to mental health services and the competencyof mental health services were the twooverarching issues for informants and respondents.For all respondents who are currentlyaccessing, or would like to access, mental healthcounselling and other services, the most widelyreported issue that hindered access to theseservices was cost. Several respondents reportedthat financial barriers meant they did not accessservices, even though they identified these asnecessary for their mental health and overallwellbeing. In some instances this was reported asprolonging the distress being experienced.The other main barrier was the lack of mentalhealth services provided by the public health system,particularly for those with mild to moderateneeds. Many respondents also reported a lack ofknowledge about available services.In relation to service competency, the chiefissue identified was that all services should beprovided in a culturally safe and appropriate way.For GLBTI people, culture may relate to issuesassociated with sexual or gender identity, or bodydiversity, as well as ethnic identity. Ensuringthat mental health staff displayed appropriateattitudes, had the necessary skills and abilities to110 VOLUME 5 • NUMBER 2 • JUNE 2013 J OURNAL OF PRIMARY HEALTH CARE

ORIGINAL SCIENTIFIC PAPERQUALITATIVE RESEARCHwork with GLBTI people, and did not make assumptionsaround sexual and gender identity wereidentified as important. In addition, it was seen asessential that services were GLBTI-friendly.DiscussionSeveral key points arise from the review ofmental health promotion and service delivery andfrom the key informant interviews and submissionsfrom GLBTI people.Building sector capacityIt is clear that there is very limited leadershipwith respect to mental health issues for GLBTIpeople—both from government agencies andfrom GLBTI communities. In addition, no GLBTIorganisation suitably positioned to take a nationalleadership role was identified by participants. Oneidea to improve coordination and leadership in thehealth area is a national health group which couldrepresent GLBTI concerns and issues.Reducing stigmaWhile some informants and respondents reportedthat social conditions for GLBTI people hadimproved in recent years, others noted that thisimprovement varied across various GLBTI groups.The literature clearly points out that the socialenvironment (including actions such as prejudice,stigma, discrimination, rejection and violence directedtowards GLBTI people) plays an importantrole in influencing the mental health of GLBTIpeople. 30 It is imperative that actions aimed at reducingGLBTI people’s exposure to such negativeexperiences and countering societal heterosexismare developed.Enhancing young people’s safetyFor many informants and respondents, a keyissue was the need to ensure the safety of youngpeople, particularly in schools. The wellbeing ofGLBTI students must be fostered by ensuringteachers are trained (pre-service and professionaldevelopment) in suicide prevention, mental healthpromotion, preventing bullying, and challenginghomophobia/transphobia. A review of the Healthand Physical Education Curriculum appearswarranted, with a particular focus on whether itis meeting the needs of young GLBTI people.Funder obligationsMost respondents reported that mainstreamhealth services should be able to provide competenthigh-quality services that are accessibleand acceptable to GLBTI people. This suggeststhat the Ministry of Health’s National HealthBoard as key funding agency, and district healthboards who are responsible for providing healthservices to their communities (as well as fundingothers to deliver services), need to prioritiseresources for GLBTI mental health and provideboth GLBTI-focused services and general servicesthat are inclusive of GLBTI people and recogniseany specific needs.GBLTI components within existing mainstreammental health promotion and service provisionshould be funded alongside GLBTI-focusedmental health promotion programmes andservices which promote community cohesiveness,provide support for young people coming out,and deliver information and support throughhelplines and websites etc. District health boardsneed to improve GLBTI access to mental healthservices by ensuring they are more inclusive ofGLBTI clients, including re-allocating resourcesfor this purpose as necessary.Research and information needsThere is very little research information availablein New Zealand about the epidemiology of mentalhealth issues for GLBTI people, or in-depthunderstanding of their experiences. Particulargaps include research that addresses the needs andexperiences of transgender and intersex people,and of older GLBTI people. Along with GLBTIfocusedresearch initiatives, there remains a needfor data to be routinely collected about sexualorientation in mainstream research.Supporting practitioners throughtraining and resourcesMany participants talked about ensuring thecompetency of practitioners (e.g. GPs, schoolcounsellors, counsellors, psychologists, psy-VOLUME 5 • NUMBER 2 • JUNE 2013 J OURNAL OF PRIMARY HEALTH CARE 111

ORIGINAL SCIENTIFIC PAPERQUALITATIVE RESEARCHchiatrists) who provide mental health services.In particular, they identified that practitionersneeded to have knowledge of GLBTI issues (bothmental health–related and wider issues), to deliverservices in respectful ways and avoid makingassumptions. Many participants considered thatpractitioners received inadequate training andeducation in GLBTI issues, both pre-service andin-service. A training programme is needed forprofessional bodies, and education and trainingproviders, to enhance GLBTI-related training andeducation opportunities for practitioners.International research has demonstratedthat GLBTI individuals experience higherlevels of mental health distress than theirheterosexual counterparts.Having GLBTI-inclusive services was identifiedby many participants as very important. As wellas having well-trained staff, it was noted thatorganisations require appropriate resources andother support to become truly inclusive. It wasalso suggested that an audit system be establishedto encourage services to review their policy, practicesand procedures, make changes as necessaryand maintain their inclusive practices.Conclusion and recommendationsWhile GLBTI individuals have the same basicmental health prevention and promotion needsas members of the general population, they alsoexperience additional unique issues related tosocial discrimination, personal and communitysocial and behavioural risk factors. 8 Internationalresearch has demonstrated that GLBTI individualsexperience higher levels of mental healthdistress than their heterosexual counterparts.This needs assessment research has confirmedthere is minimal policy in relation to GLBTImental health. Limited mental health promotionor prevention services directed at GLBTIpopulations in New Zealand were identified.While the review of existing services did notidentify robust evidence or any evaluations ofthe impact of existing programmes and serviceson the mental health and wellbeing of GLBTINew Zealanders, several GLBTI-focused services(e.g. telephone helplines, counselling) appearedwell utilised. There were many reports aboutgovernment-funded mainstream mental healthpromotion and prevention services that were notresponding appropriately to the needs of thesegroups. These findings mirror other New Zealandstudies, which have also reported limited servicedelivery and policy attention to the health needsof these groups. 26,31,32The New Zealand Government plays a lead rolein the policy development around GLBTI mentalheath, and it can easily be argued it should ensurethere are appropriate levels of mental healthpromotion and service provision for GLBTI people.Alongside this, the appropriate involvementof GLBTI representatives will encourage GLBTIcommunity engagement with mental healthpromotion and should result in more appropriateservice delivery responses being developed. 26,33 Acaveat for future action applies—while the needsfor GLBTI people as a group are often the same, atother times the needs of particular groups, such asgay men or lesbian women, may be different andmay require different solutions. In addition, theneeds of varying ethnic groups within these populations,and of all age groups including youngpeople and older people, need to be incorporated.References1. Wolitski RJ, Valdiserri RO, Stall R. Health disparities affectinggay and bisexual men in the United States: an introduction. In:Wolitski RJ, Stall R, Valdiserri RO, editors. Unequal opportunity:health disparities affecting gay and bisexual men in theUnited States. New York, NY: Oxford University Press; 2008.p. 3–32.2. Bakker F, Sandfort T, Vanwesenbeeck I, van Lindert H, WestertG. Do homosexual persons use health care services morefrequently than heterosexual persons: findings from a Dutchpopulation survey. Soc Sci Med. 2006;63:2022–30.3. Cochran SD, Mays VM. Relation between psychiatric symptomsand behaviorally defined sexual orientation in a sampleof the US population. Am J Epidemiol. 2000;151:516–23.4. Cochran SD, Sullivan JG, Mays VM. Prevalence of mental disorders,psychological distress, and mental health services useamong lesbian, gay, and bisexual adults in the United States. JConsult Clin Psychol. 2003;71:53–61.5. Tjepkema M. Health care use among gay, lesbian and bisexualCanadians. Health Rep. 2008;19:53–64.6. McNair R, Kavanagh A, Agius P, Tong B. The mental healthstatus of young adult and mid-life non-heterosexual Australianwomen. Aust NZ J Public Health. 2005;29:265–71.7. Kurtzman HS, Omoto A. Introduction: current and futureresearch on lesbian, gay, and bisexual people. In: Omoto A,112 VOLUME 5 • NUMBER 2 • JUNE 2013 J OURNAL OF PRIMARY HEALTH CARE

ORIGINAL SCIENTIFIC PAPERQUALITATIVE RESEARCHKurtzman HS, editors. Sexual orientation and mental health.Washington, DC: American Psychological Association; 2006.p. 3–9.8. Johnson C, Mimiaga M, Bradford J. Health care issues amonglesbian, gay, bisexual, transgender and intersex (LGBTI)populations in the United States: introduction. J Homosex.2008;54:213–24.9. Knox S, Kippax S, Crawford J, Prestage G, Van de Ven P. Nonprescriptiondrug use by gay men in Sydney, Melbourne andBrisbane. Drug Alcohol Rev. 1999;18:425–33.10. Trocki KF, Drabble LA, Midanik LT. Tobacco, marijuana,and sensation seeking: comparisons across gay, lesbian,bisexual, and heterosexual groups. Psychol Addict Behav.2009;23:620–31.11. Burgard SA, Cochran SD, Mays VM. Alcohol and tobaccouse patterns among heterosexually and homosexuallyexperienced Californian women. Drug Alcohol Depend.2005;77:61–70.12. Hatzenbuehler M, Keyes K, Hasin D. State-level policies andpsychiatric morbidity in lesbian, gay, and bisexual populations.Am J Public Health. 2009;99:2275–81.13. Jordan KM. Substance abuse among gay, lesbian, bisexual,transgender and questioning adolescents. School Psychol Rev.2000;29:201–6.14. Fergusson DM, Horwood LJ, Beautrais AL. Is sexual orientationrelated to mental health problems and suicidality in youngpeople? Arch Gen Psychiatry. 1999;56:876–80.15. Fergusson DM, Horwood LJ, Ridder EM, Beautrais AL. Sexualorientation and mental health in a birth cohort of young adults.Psychol Med. 2005;35:971–81.16. Skegg K, Nada-Raja S, Dickson N, Paul C, Williams S. Sexualorientation and self-harm in men and women. Am J Psychiatry.2003;160:541–6.17. Ministry of Health. What is primary mental health? 2012; [cited2012 May 10]. Available from: http://www.primarymentalhealth.org.nz/section/10330/what-is-primary-mental-health/.18. Mental Health Commission. How to access mental healthservices? 2009; [cited 2013 March 26]. Available from: http://www.hdc.org.nz/about-us/mental-health-and-addictions/mental-health-services.19. Neville S, Henrickson M. Perceptions of lesbian, gay andbisexual people of primary healthcare services. J Adv Nurs.2006;55:407–15.20. Australian Medical Association. AMA position statement:Sexual diversity and gender identity. Barton, ACT: AustralianMedical Association; 2002 [cited 2013 March 26]. Availablefrom: https://ama.com.au/position-statement/sexual-diversity-and-gender-identity-200221. Adams J, Braun V, McCreanor T. Gay men talking abouthealth. Am J Mens Health. 2012;6:182–93.22. Adams J, McCreanor T, Braun V. Doctoring New Zealand’s gaymen. N Z Med J. 2008;121(1287).23. Adams J, Dickinson P, Asiasiga L. Mental health promotionand prevention services to gay, lesbian, bisexual, transgenderand intersex populations in New Zealand: Needs assessmentreport. Auckland, New Zealand: Te Pou o Te Whakaaro Nui,The National Centre of Mental Health Research, Informationand Workforce Development; 2012.24. Barry MM, Jenkins R. Implementing mental health promotion.Edinburgh, UK: Elsevier; 2007.25. Braun V, Clarke V. Using thematic analysis in psychology. QualRes Psychol. 2006;3:77–101.26. Adams J, Braun V, McCreanor T. A critical analysis of gaymen’s health policy documents. GLIP Rev. 2010;6:42–59.27. Metzger N, Camburn F. Rainbow Youth national outreach scopingproject. Auckland, New Zealand: Point Research; 2010.28. Riches M. How do we make it better? Mapping the stepstowards a more supportive coming out environment for queeryouth in Aotearoa New Zealand (unpublished paper); 2011.29. Bronfenbrenner U. The ecology of human development:Experiments by nature and design. Cambridge, MA: HarvardUniversity Press; 1979.30. Clarke V, Ellis, SJ, Peel E, Riggs DW. Lesbian, gay, bisexual,trans and queer psychology: an introduction. Cambridge, UK:Cambridge University Press; 2010.31. Pega F. Health policies for lesbian, gay, bisexual, transgender,and intersex people [master’s thesis]. [Palmerston North(NZ)]: Massey University; 2007.32. Pega F, Gray A, Veale J. Sexual orientation data in probabilitysurveys: Improving data quality and estimating core populationmeasures from existing New Zealand survey data, 2010/2;[cited 2012 May 10]. Available from: www.statisphere.govt.nz2010.33. Adams J, Braun V, McCreanor T. Warning voices in a policyvacuum: professional accounts of gay men’s health in AotearoaNew Zealand. Soc Policy J NZ. 2007;30:199–215.FUNDINGThis research was fundedby Te Pou and the Ministryof Health. The viewsexpressed in this paper arethose of the authors andnot of these agencies.COMPETING INTERESTSNone declared.VOLUME 5 • NUMBER 2 • JUNE 2013 J OURNAL OF PRIMARY HEALTH CARE 113

ORIGINAL SCIENTIFIC PAPERQUALITATIVE RESEARCHUnderstanding barriers to glycaemic controlfrom the patient’s perspectiveRon Janes MD, CCFP, FRNZCGP; 1 Janet Titchener MD, AAFP, FRNZCGP; 2 Joseph Pere PhD; 3 Rose PerePhD; 3 Joy Senior RN 41Rural general practitioner,Wairoa, and Department ofGeneral Practice and PrimaryHealth Care, The Universityof Auckland, Auckland,New Zealand2Family physician anddiabetologist, Hastings,New Zealand, and Faculty,Lancaster General HospitalFamily Medicine ResidencyProgram, Lancaster, PA, USA3Lake Waikaremoana,New Zealand4Clinical Nurse Specialist—Diabetes, Wairoa Hospital,Hawke’s Bay DHB, Wairoa,New Zealand.ABSTRACTINTRODUCTION: To better understand barriers to glycaemic control from the patient’s perspective.METHODS: An interpretative phenomenological approach was used to study the experiences of 15adults with Type 2 diabetes. Participants each gave a semi-structured interview of their experiences ofliving with diabetes. Interviews were transcribed, and themes extracted and organised using a patientcentredframework.FINDINGS: Participants’ stories confirmed many of the barriers in the literature, particularly thoserelated to context, such as family, finances, work. Barriers also related to negative emotional reactionsto diabetes: fear of new events (diagnosis, starting pills/insulin); guilt about getting diabetes and notcontrolling it; and shame about having diabetes. Barriers also related to unscientific beliefs and personalbeliefs. There were additional barriers related to poor clinician–patient relationships. Overall, participantshad a poor understanding of diabetes, and complained that their clinician simply ‘told them what to do’.CONCLUSION: Using a patient-centred approach, this study identified many barriers to glycaemiccontrol. We suggest that a key barrier is clinician ignorance of their patients’ fears, beliefs, expectations,context; of what constitutes a positive therapeutic relationship; and of the limitations of a biomedical approachto patient non-adherence. Faced with both a worsening diabetes epidemic and increasing healthcare workforce shortages, clinicians urgently need to understand that it is they, not their patients, whomust change their approach if diabetes care is to be improved.KEYWORDS: Communication barriers; diabetes mellitus, type 2; medication adherence; patientcenteredcareJ PRIM HEALTH CARE2013;5(2):114–122.CORRESPONDENCE TO:Ron JanesWairoa Medical Centre,24 Kitchener St,PO Box 341, Wairoa,New Zealandronjanes@xtra.co.nzIntroductionDespite advances in medical management, manypeople with diabetes have poor glycaemic control,and many barriers to care have been identified.1 Non-adherence has been identified as abarrier: both the failure of clinicians to adhereto evidenced-based clinical guidelines, 2,3 andthe failure of patients to adhere to medicallyrecommended treatments. 4 However, biomedicalresearch has been unable to explain why it is thatclinicians and patients do not always adhere toexpert recommendations.While previous studies have directly askedpatients and clinicians about perceived barriersto diabetes care, 5,6 we chose an indirect phenomenologicalapproach to provide new insights intothe many factors that impact on how individualsmanage their diabetes. To explore and revealbarriers to glycaemic control from the patientperspective, themes were organised within theclinical framework of patient-centred medicine(PCM; see Figure 1). 7MethodsAn interpretative phenomenological methodof inquiry attempts to find meaning in, andlearn from, participants’ subjective experience.Thus, through their stories (already interpretedexperiences), and the interpretative lens of the114 VOLUME 5 • NUMBER 2 • JUNE 2013 J OURNAL OF PRIMARY HEALTH CARE

ORIGINAL SCIENTIFIC PAPERQUALITATIVE RESEARCHresearchers, an attempt is made to get as closeas possible to what it means to be a person withdiabetes. The research team contained individualsof different genders, countries of origin, ethnicitiesand educational backgrounds: RJ [male,Canada, Caucasian, medicine], JT [female, USA,Caucasian, medicine], RP [female, New Zealand(NZ), Maori, education], JP [male, NZ, Maori,sociology], and JS (female, NZ, Caucasian, nursing).Each participant was interviewed by two ofthe authors [excluding JS], and all team membersscrutinised the neutrality and defensibility ofthe analysis.A purposive sample of 15 adults living in Wairoa,NZ were recruited from patients attending thelocal diabetes clinic, and selected to vary by ethnicity(NZ Maori or NZ European). Participantsneeded to be able to put their experiences intowords, and give written, informed consent to participation.Participants were offered the opportunityto be interviewed in English or Maori.Participants’ experiences of living with diabeteswere audiotaped in semi-structured, face-to-faceinterviews. The interview guide contained broadlyfocused questions, permitted probes withinareas of inquiry, but also allowed for participants’views of ‘what mattered’. The interviewsbegan with an ‘icebreaker’ question. Succeedingquestions focused on revealing, within storytellingmode, participants’ actual experiences ofliving with diabetes for example, ‘Think back towhen you were first told you had diabetes. Tell usabout that experience?’ Probes included: ‘Who toldyou?’ ‘How did you feel/react to being told youhad diabetes?’ Participants were asked about anyexperiences of diabetes before they themselveswere diagnosed, their personal experience ofliving with diabetes (diagnosis, treatment, startinginsulin, complications) and the effects of allof these on themselves and their families. Theinterviews were transcribed, with identifyinginformation removed.Meanings of participants’ experiences werederived through all team members independentlyreading each interview several times, beforediscussion. The participants’ stories were thenindependently extracted by RJ and JT, using animmersion/crystallisation approach. 8 InterviewWHAT GAP THIS FILLSWhat we already know: Despite advances in diabetes management, manybarriers to glycaemic control, including non-adherence, have been identified.However, biomedical research has been unable to explain why both cliniciansand patients do not always comply with expert recommendations.What this study adds: This study identifies and explores barriers toglycaemic control from the patient perspective. The barriers have beenorganised within the clinical framework of patient-centred medicine, providinginsight into why both clinicians and patients may struggle to comply withexpert recommendations.Figure 1. Overview of the six components of patient-centred medicine 71. Disease and illness experienceUnderstanding the disease requires history, examination, and investigation.Understanding illness experience requires an exploration of four dimensions:a. Feelings/fears: the emotional/psychological responses to the illnessb. Ideas on causation: the intellectual response to the illnessc. Effects on functioning: the impact of the illness on body and lifestyled. Expectations: what the person expects of the provider.2. Understanding the whole personThe meaning of health and illness to a person varies according to their context.Just as the body is made up of a number of interlocking systems, so too, theindividual is a part of a family, a community, a culture, a country and an ecology.Clinical information only becomes useful knowledge when it is placed in thecontext of a particular patient’s world. Ignoring context will lead to errors in boththe interpretation and application of findings. Patient contexts include the person(life history, developmental stage), and their place in society (family, employment,leisure, finances, culture, as well as spiritual, social and health care supports).3. Finding common groundThe process through which the patient and doctor reach a mutual understandingand mutual agreement in three key areas:a. defining problems and prioritiesb. establishing goals of treatment and/or managementc. identifying the roles to be assumed by both the patient and the doctor.4. Incorporating disease prevention and health promotionThis involves health enhancement, risk avoidance, risk reduction, early identification,and complication reduction.5. Enhancing the doctor–patient relationshipEach consultation is considered an opportunity to improve the doctor–patient relationship:facilitating communication, growing compassion, and building trust, withthe ultimate goal of mutual respect and sharing of decision-making and power.6. Being realisticThis involves being realistic about time and timing, team-building and teamwork,and wise stewardship of resources.VOLUME 5 • NUMBER 2 • JUNE 2013 J OURNAL OF PRIMARY HEALTH CARE 115

ORIGINAL SCIENTIFIC PAPERQUALITATIVE RESEARCHTable 1. Participant characteristicsParticipantnumberGender* Ethnicity † Age in yearsYears since diagnosis(age at diagnosis in years)Years on insulintherapyYears from diagnosisuntil insulin started1 F E 61 11 (50) 6 52 F E 66 14 (52) 11 33 F E 90 30 (60) 30 04 F M-E 33 2 ‡ (31) 2 0 ‡5 F M-E 72 23 (49) 10 136 F M-E 71 14 (56) 1 137 F E 73 10 (63) 3 78 M M 46 16 (30) 3 139 F M 78 53 (25) 7 4610 M M 73 19 (54) 7 1211 M M 58 17 (41) 11 612 F Fijian-E 64 18 (46) 2 1613 F M 47 9 (38) 1 814 M M 60 20 (40) 1 1915 M M 57 28 (29) 8 20Mean – – 63.3 18.9 (44.3) 6.9 12.1* Gender: F female; M male† Ethnicity: E New Zealand European; M New Zealand Maori‡ Gestational diabetes mellitus in first pregnancy at age 20; remained on insulin after last pregnancy at age 31summaries, capturing the individual meanings ofwhat appeared to matter to participants in theirreported experiences, were created. Participantswere given the opportunity to read and commenton their interview summary, to ensure they hadnot been misheard. Barriers to glycaemic controlwere then identified and organised as themes andsubthemes within the PCM clinical framework(see Figure 1). 7 Agreement that the themes werecredible was achieved through discussion withinthe research team.This study had ethics approval from the CentralRegional Ethics Committee (Ref. CEN/07/22/EXP), which required signed informed consentfrom all participants.FindingsTable 1 outlines participants’ characteristics.Participant quotes are italicised and followed bytheir identification number from Table 1. Theheadings and subheadings of PCM, as outlinedin Figure 1, are used to organise the identifiedbarriers to glycaemic control. Only key barrierspertinent to the discussion are provided below.For content relating to headings and subheadingsnot included in this section see the Appendixpublished in the web version of this paper.Disease and illness experienceFeelings/fearsFear about the illnessParticipants expressed fear when they did notknow what was happening, what would happennext, and/or when they perceived a loss of controlover their lives—i.e. at specific transition events,such as at diagnosis, at the start of oral medication,and at the start of insulin therapy.At first, I thought, ‘oh here we go, I can’t work, Ican’t do anything’. (#2)116 VOLUME 5 • NUMBER 2 • JUNE 2013 J OURNAL OF PRIMARY HEALTH CARE

ORIGINAL SCIENTIFIC PAPERQUALITATIVE RESEARCH[Having diabetes] was no shock, until I went on tohave medications and then it was a bit of shock. (#14)I thought the end of the world was coming whenthey said I had to have injections. (#9)Treatment escalation was interpreted as gettingcloser to disability and death. Fear of treatmentescalation was greatest for the introduction ofinsulin.Adding gliclazide—I was not happy because I knowthat each time you need to add another drug youare moving along that continuum, and for me… Ifelt like once you hit insulin you are on a slide to….you know [death]. (#13)I thought, oh boy, once you are on that [insulin]you have not got far to go. I thought I must be onthe way out. (#14)Hypoglycaemia was a terrifying unknown. Somewould take active steps to avoid hypoglycaemia,such as missing insulin doses or overeating.When I had my first hypo, I hit the ground… Ithought, I am going to die here. (#13)To avoid hypos… I won’t have my insulin. (#4)…eat too much. Because I know that does not giveme a low. It might give me a high, but it does notgive me a low. The highs … are easier to deal withthan the lows. (#7)Guilt/self-blameParticipants blamed themselves for both gettingthe disease and not controlling it.I have type 2 diabetes, which is self-inflicted. (#1)A good diabetic is one who controls their diabetes…I am not a good diabetic. (#7)Accepting blame for both causing and not controllingthe disease, participants expressed feelings ofguilt and self-blame. Participants repeatedly talkedabout ‘right’ and ‘wrong’ ways to manage diabetes,‘good’ and ‘bad’ foods, and being ‘naughty’.You are always working on guilt. (#4)I have not managed to do anything that I shoulddo. (#7)I am very naughty when it comes to sausages. Youcan put all the meats in front of me and I willalways pick sausages. (#12)Lack of dietary self-control was policed by familymembers, by friends and workmates, by healthcare providers, and even by people who barelyknow them.I have a brother-in-law and sister who are doctors…and when they are around, that is when I get introuble if I am seen to be overindulging, they willremind me. (#13)Everyone brings the cake… and you know they go,oh, you are the diabetic, don’t give her any, she is adiabetic. They treat us like we are lepers. (#4)If you want someone to tell you you have beennaughty, you go to the doctor, or you go to [thenurse]. (#4)I went to the supermarket... This lady was there andI had some lollies in my trolley… I had lots of nicethings in my trolley. She proceeded to take themout, because I am not allowed to have them. (#9)ShameThere was significant reluctance to self-injectinsulin in public. Some expressed shame aroundexposing their body; others did not want peopleto know they had diabetes. For some, it was thefear of what others might think.I do not like lifting up my shirt and stabbing myselfin front of people. (#8)Well, there is so much of me that hangs out, thatdoes not get tucked in, I think I would get a wee bitembarrassed. (#5)Some people are quite ashamed of being diagnosedas a diabetic. (#1)There is a bit of a stigma. (#4)[I am] embarrassed to give needle in public… justlike a drug addict. (#12)VOLUME 5 • NUMBER 2 • JUNE 2013 J OURNAL OF PRIMARY HEALTH CARE 117

ORIGINAL SCIENTIFIC PAPERQUALITATIVE RESEARCHIdeas/beliefs about causationNon-scientific health beliefsA number of non-scientific health beliefs wereidentified.Diabetes is caused by eating sugarHe [husband] does not take sugar in his tea, andthen he says, I am not going to start that, havingsugar in my coffee and tea. I might get diabetes. (#9)Proper diet will control/cure diabetesDespite having had diabetes for many years andtaking medications, including insulin, participantswere still focused on diet as the key tocontrolling their disease. There appeared to beno understanding of underlying pathologies (e.g.pancreatic failure) that could not be addressedthrough diet. Even after years of having diabetessome participants still had the unrealistic hopethat the disease would ‘go away’.To me if I could get my diet under control, thediabetes would be controlled. (#7)I am eating the wrong food… that is why the damnthing is out of order. (#11)I have followed everything by the book. I wouldnot eat anything that I shouldn’t have. They saydon’t eat this, don’t eat that, so I have done it, but Istill have diabetes. (#2)I wanted to get cured… that is what I am lookingfor, to try to get over it. (#8)Only people with diabetes needto eat a healthy dietDespite participants believing that poor dietcaused their diabetes, there was the conflictingbelief that people without diabetes did not haveto eat healthily.He [husband] will take the tin and he will say tome, you are only supposed to have a couple [ofbiscuits]. He tells me he is the only one allowed toeat what he likes. (#2)Diabetes is like ‘a cold’Participants initially believed their diabetes,like previous acute illness experiences, wouldsimply go away, or could be managed simply bytaking a pill. This belief led some to just ignoretheir diabetes, especially if they weren’t takingmedication.It is just like having a cold… it will come right. (#5)It was just diabetes… take a pill and that will fixit. (#10)They [doctors] just says, you are diabetic andI go, nah, nah, because I was not taking no medication.(#8)Personal/cultural beliefsAbsence of symptoms equates to healthParticipants wanted to feel well but relied ontheir subjective assessment of how they feltin the present moment to decide if they werehealthy or not. If they felt healthy (had no symptoms),then they were healthy.Well, I just feel if I am feeling good, if I feel mybody is feeling good… I see myself as healthy,whether I have got diabetes or not. (#2)Participants noted that diabetes was differentfrom previous illnesses they had experiencedbecause it caused few, if any, symptoms.You know when you are sick, you know the sicknessis with you, coughing and spluttering and allthat kind of thing. (#8)Lack of symptoms meant some did not take theirdiabetes seriously.It is one of those illnesses that you are not aware ofit. You are not aware of the dangers of it, until all ofa sudden, bang, it [complications] has happened. (#7)In contrast, feeling unwell was a reason to takeaction.It was not until I got sick, I really started to dosomething about it [diabetes]. (#8)118 VOLUME 5 • NUMBER 2 • JUNE 2013 J OURNAL OF PRIMARY HEALTH CARE

ORIGINAL SCIENTIFIC PAPERQUALITATIVE RESEARCHFor many, it was not the disease but its treatmentthat caused symptoms, including hypoglycaemia.When I take the diabetes drugs, it makes me feelworse, even though it could be long term making melive longer, but short term, it makes me feel bad. (#13)I do not feel sick with my diabetes. The only thingthat affects me now is [when the] blood [glucose]goes down too quick. I do not get sick, but I feelweak. (#14)Drugs are chemicals to be avoidedWestern drugs were seen by some as ‘chemicals’to be avoided: some participants were averse totaking any medication.I can stand a headache without taking any tablets…I would… try some other alternative things. I wouldgo for a run… or have a feed. (#11)And with the drugs, I do not drink alcohol… I donot smoke, so having to put Western chemicals ormedicines, whatever, into the body is somethingthat I am thinking, why do I need to do this? Whydo I want to do this? (#13)Maori cultural beliefsMaori cultural beliefs were important to some.For participant #13, her cultural beliefs were indirect conflict with using drugs (see previousquote) and needles. She also relied on traditionalMaori beliefs and medicinal plants for healing.The body is tapu [restricted]… it makes me not likepoking holes in it [with needles]. (#13)I have often wondered whether this has been mystruggle with medication, that it is my wairua[spirit], knowing that you can do it without [drugs],because... medication has its place, but for me it hasbeen such a hard struggle to allow it in. (#13)Finding common groundParticipants’ comments suggested that clinicianshad assumed problems, priorities, goals oftreatment, and their respective roles (clinicians tomake recommendations; patients to comply withrecommendations) were already mutually agreed.Mutually defining problems and prioritiesPatients’ beliefs differ from clinicians’ beliefsParticipants held beliefs about health, disease,and acceptability of treatments that were fundamentallydifferent from those of clinicians, andwould make reaching mutual agreement difficult(see Non-scientific health beliefs section).Mutually defining the goalsGoals imposed by clinicianParticipants reported that clinicians just expectedthem to ‘do what they were told’.I do not remember that there were really any arguments,it was just that this is what you take whenyou have got diabetes and just, you know, get onand do it. (#13)Stick to your medication and take them at the righttimes and do a bit of exercise, your diet, and that isabout all. (#14)Participants believed that diabetes is causedby eating too much sugar and is primarilycontrolled through diet, which explains whythey felt guilty and ashamed for both causingtheir disease and not ‘properly’ controlling it.Cultural differences with regard to diet wereparticularly mentioned.I was sent to a dietitian, and I do not know thatbeing told to have something like a packet of raisinsand a yoghurt and a piece of fruit for morningtea… I do not know whether they [Maori] couldrelate to it, because we have a different style ofeating. (#13)There is some good, healthy [Maori] food… not likesome of the food dietitians and nurses say you haveVOLUME 5 • NUMBER 2 • JUNE 2013 J OURNAL OF PRIMARY HEALTH CARE 119

ORIGINAL SCIENTIFIC PAPERQUALITATIVE RESEARCHgot to eat, does not happen, because Maori do notlike those sort of food. (#14)As clinicians had not negotiated mutually agreedgoals, participants simply ignored what they hadbeen told to do, especially if it didn’t make senseto them. The patient’s main goal of wanting tofeel well was seemingly unexplored (see Diseaseand illness experience section).At the time, what they said, what you eat, ...I ignoredthat too. It was wrong. It cannot be, becausethis is what I have been eating all my life. (#15)Goals not individualisedNot only were goals imposed, but some cliniciansappeared to make identical recommendations toall patients, without enquiring what the personwas currently doing.I was walking from where Aunty [name] lives to[town], which is mountainous, every morning, andI was probably as fit as I had ever been, so to be toldto beef up your exercise, it is like—where do you gowith it? (#13)DiscussionThis is the first research investigating barriersto diabetes care that organises findings withinthe PCM clinical framework. 7 By listening tothe experiences of people with diabetes, multiplebarriers to glycaemic control were identified.Many of these have been previously reported, 1especially in relation to context (family, finances,work) and fears (insulin, hypoglycaemia). Byexamining barriers from the patient’s perspective,biomedicine’s labelling of patients as ‘nonadherent’can be challenged, and seen for whatit is—clinician ignorance: of their patients, ofwhat constitutes a positive therapeutic relationshipin chronic disease management, and of thecognitive bias within the biomedical approachto patient non-adherence. This ignorance is asignificant barrier to patients becoming chronicdisease ‘self-managers’.Our participants expressed negative emotions(fears, guilt, shame) and unscientific beliefsthat indicated a poor understanding of diabetes.Participants believed that diabetes is caused byeating too much sugar and is primarily controlledthrough diet, which explains why theyfelt guilty and ashamed for both causing theirdisease and not ‘properly’ controlling it. Thebelief ‘if I feel healthy, then I am healthy’, showsthat participants used symptoms to determineself-management behaviours. Symptom-motivatedself-adjustment of medication has been reportedin patients with hypertension, 9 rheumatoid arthritis,10 and even cancer. 11 The unscientific beliefthat diabetes was a self-limited illness that wouldjust ‘go away’ with time, was another reasonwhy participants chose to ignore medical recommendations.These emotions and beliefs may bemajor barriers to clinicians attempting to escalatemanagement for glycaemic control.This poor understanding of diabetes was despiteparticipants having lived with their disease formany years (Table 1), and despite having receivedongoing ‘diabetes education’. So while people arebeing given diabetes education, it would appearthat either the wrong information is being given,or the right information is being given but not ina format that allows the patient to understand itsufficiently to positively influence their emotions,beliefs, and self-management behaviours.As one participant said:I just could not understand what the hell was that.What is diabetes? She [nurse] was sitting theretelling me all about it and I was just going… yeah,yeah. (#11)The biomedical approach to disease management,where the ‘expert’ tells the patient whatto do, is still a common model in diabetes care,both for treatment and education. Rather thaneducating the patient by empowering them withknowledge that enables them to understand theirdisease sufficiently to make their own managementdecisions, patients are simply given a setof provider-chosen rules to follow (‘you should…eat these foods, take these pills’). As so aptly putby Hunt et al. 12The problem of promoting self-care behaviours isreduced to simply finding ways to educate and motivatepeople sufficiently so that they will pursuethe obviously [expert chosen] right course of action.120 VOLUME 5 • NUMBER 2 • JUNE 2013 J OURNAL OF PRIMARY HEALTH CARE

ORIGINAL SCIENTIFIC PAPERQUALITATIVE RESEARCHThis process of education explains why so manypatients with diabetes have such a poor understandingof their disease. Lack of knowledge andpoor understanding of specific diet plans, medication(action, side effects, schedules, adjustments),and glucose monitoring (i.e. HbA1c) have all beenidentified as significant barriers to self-management.6,7,13 A Cochrane review of 21 studies ofType 2 diabetes examining various interventionsto improve patient adherence concluded thatdiabetes education ‘showed a small effect on avariety of outcomes including HbA1c’. 14 Thus,while patients may undoubtedly be receivingdiabetes education, their resultant knowledge andunderstanding is insufficient to positively influencetheir emotional reactions to, beliefs about,and self-management of their diabetes.Clinician ignorance underpins all of the barriersdiscussed above—an ignorance of the wholepatient (i.e. the person’s illness experience andlife context); an ignorance of the critical role ofa therapeutic relationship based on mutual trustand respect; and an ignorance of how these twokey components positively interact. Without thiscritical understanding, clinicians are unable touse themselves as positive enablers for change. 15Within the biomedical paradigm, there is theunstated assumption that the disease expert’srole is to make clinical recommendations and thepatient’s role is to comply.I do not remember that there were really any arguments,it was just that this is what you take whenyou have got diabetes and just, you know, get onand do it. (#13)This thinking is ‘blind’ to how it devalues thepatient’s expert knowledge of self, interferes withshared decision-making, ‘problematises’ only thepatients’ perspective, and effectively preventspatients from making well-informed decisionsand actively self-managing their disease in thecontext of their life. This cognitive blind-spotof the biomedical paradigm has been repeatedlypointed out over a number of decades. 16,17We suggest that the source of this cognitiveblind-spot is to be found within positivism’sscientific method of inquiry that relies onquantitative methodologies to uncover the truth.The method assumes the objective observerdoes not influence the outcome of the experiment.Clinicians, educated within this paradigm,may see themselves as objective observers to anintervention (treating diabetes with drugs) andrecorders of the outcome (HbA1c). When analysingthe many potential factors responsible fortheir patients’ poor outcome, it simply does notthen occur to the clinician that the very mannerwith which they approach an investigation intobarriers to care will prevent them from seeingthat their relationship and interaction could be asignificant barrier to their patients self-managingtheir diabetes.Within the biomedical paradigm, there is theunstated assumption that the disease expert’srole is to make clinical recommendations andthe patient’s role is to comply.This study used interpretive phenomenology togain a better understanding of the lived experienceof people with diabetes. This enabled aninterpretation that moves beyond past attemptsto identify barriers to glycaemic control. Participantsstories speak to themes of potentialrelevance to people with diabetes in all societies,although the ability to generalise findingsexternally is not critical. 18 The small sample sizeis unproblematic for that reason and for threeother reasons: transferability depends on logical,rather than statistical inference; rich data do notrequire large numbers; and the concept of datasaturation is not part of interpretive phenomenology.18 This method aims merely to offer credibleinterpretations of phenomena in experience, andto generate further questions.This study identifies many barriers to glycaemiccontrol. However, by organising them withinthe PCM 7 framework, the greatest barrier wouldappear to be clinician ignorance of their patients’fears, beliefs, expectations, context; of the importanceof a trusting respectful therapeutic relationship;and of biomedicine’s blindspot regarding theVOLUME 5 • NUMBER 2 • JUNE 2013 J OURNAL OF PRIMARY HEALTH CARE 121

ORIGINAL SCIENTIFIC PAPERQUALITATIVE RESEARCHassumption that patients should simply complywith the recommendations of clinical experts.Faced with a worsening diabetes epidemic, cliniciansneed to understand that it is they, not theirpatients, who must change their behaviour.ACKNOWLEDGEMENTSWe would like to thank theparticipants for sharingtheir stories of livingwith diabetes. We alsogratefully acknowledgethe letters of support forour research from thefollowing organisations:Wairoa Primary HealthcareOrganisation, Te HauoraO Te Wheke a Nuku(Rongomaiwahine IwiTrust), KahungunuExecutive Ki Te WairoaCharitable Trust, Te IwiO Rakaipaaka Inc, andNgati Pahawera Inc. Wewould also like to thankViv Kerr and MarionHelmers at the Hawke’sBay District Health Boardlibrary, for assistanceobtaining articles.FUNDINGWe gratefully acknowledgea research grant fromThe Hawke’s Bay MedicalResearch Foundation Inc.to fund this research.COMPETING INTERESTSNone declared.References1. Nam S, Chesla C, Stotts NA, Kroon L, Janson SL. Barriers todiabetes management: patient and provider factors. DiabetesRes Clin Pract. 2011;93(1):1–9.2. Grant RW, Meigs JB. Overcoming barriers to evidence-baseddiabetes care. Curr Diabetes Rev. 2006;2(2):261–9.3. Schmittdiel JA, Uratsu CS, Karter AJ, Heisler M, SubramanianU, Mangione C, et al. Why don’t diabetes patientsachieve recommended risk factor targets? Poor adherenceversus lack of treatment intensification. J Gen Intern Med.2008;23(5):588–94.4. Haynes RB, Ackloo E, Sahota N, McDonald HP, Yao X.Interventions for enhancing medication adherence. CochraneDatabase Syst Rev. 2008, Issue 2. Art. No.: CD000011. DOI:10.1002/14651858.CD000011.pub3.5. Simmons D, Weblemoe J, Voyle J, Prichard A, Leakehe L,Gatland B. Personal barriers to diabetes care: lessons froma multi-ethnic community in New Zealand. Diabet Med.1998;15:958–64.6. Nagelkerk J, Reick K, Meengs L. Perceived barriers and effectivestrategies to diabetes self-management. J Adv Nurs.2006;54(2):151–8.7. Stewart M, Brown JB, Weston WW, McWhinney I, McWilliamC, Freeman T. Patient-centered medicine: transforming theclinical method. 2nd ed. Oxon, UK: Radcliffe Medical Press;2003. p. 360.8. Borkan J. Immersion/crystallization. In: Borkan JF, MillerWL, editors. Doing qualitative research. London, UK: SAGE;1999. p. 401.9. Benson J, Britten N. Patients’ decisions about whether ornot to take antihypertensive drugs: qualitative study. BMJ.2002;325:1–5.10. Pound P, Britten N, Morgan M, Yardley L, Pope C, Daker-White G, et al. Resisting medicines: a synthesis of qualitativestudies of medicine taking. Soc Sci Med. 2005;61(1):133–55.11. Carter S, Taylor D, Levinson R. A question of choice: compliancein medicine taking. 2nd ed. London: Medicine Partnership;2003.12. Hunt LM, Arar NH. An analytical framework for contrastingpatient and provider views of the process of chronic diseasemanagement. Med Anthropol Q. 2001;15(3):347–67.13. Dalewitz J, Khan N, Hershey CO. Barriers to control of bloodglucose in diabetes mellitus. Am J Med Qual. 2000;15(1):16–25.14. Vermeire EIJJ, Wens J, Van Royen P, Biot Y, Hearnshaw H,Lindenmeyer A. Interventions for improving adherence totreatment recommendations in people with type 2 diabetesmellitus. Cochrane Database Syst Rev. 2005; Issue 2. Art. No.:CD003638. DOI: 10.1002/14651858.CD003638.pub2.15. Balint M. The doctor, his patient and the illness. 2nd ed. NewYork: International Universities Press Inc.; 1972. p. 395.16. Kleinman A. The illness narratives: suffering, healing and thehuman condition. USA: Basic Books; 1988.17. Fisher L, Glasgow RE. A call for more effectively integratingbehavioral and social science principles into comprehensivediabetes care. Diabetes Care. 2007;30(10):2746–9.18. Winter G. A comparative discussion of the notion of ‘validity’in qualitative and quantitative research. The Qualitative Report[On-line serial] 2000;4(3/4) [cited 2012 Sept 2]. Availablefrom: http://www.nova.edu/ssss/QR/QR4-3/winter.html122 VOLUME 5 • NUMBER 2 • JUNE 2013 J OURNAL OF PRIMARY HEALTH CARE

ORIGINAL SCIENTIFIC PAPERQUALITATIVE RESEARCHMiscommunication between patients andgeneral practitioners: implications for clinicalpracticeSonya Morgan MHealScABSTRACTINTRODUCTION: Effective communication is integral to the general practice consultation, yet it isacknowledged that problems commonly occur. Previous research has shown that misunderstandings withpotentially significant consequences occur frequently, but does not provide a clear picture of how andwhy miscommunication occurs, or how such problems can be prevented or resolved. This study exploredthe occurrence and management of specific examples of miscommunication in two routine general practiceconsultations.Department of Primary HealthCare and General Practice,Wellington School ofMedicine and HealthSciences, University of Otago,Wellington, New ZealandMETHODS: A multi-method case study approach was used. The primary data collected for each caseincluded a video-recorded consultation and post-consultation interviews with each general practitioner(GP) and patient. Instances of communication mismatch were examined using in-depth interaction analysistechniques.FINDINGS: GPs and patients may not be aware when misunderstandings have occurred. In-depth analysisof the case studies revealed the complexity of miscommunication: it was not a straightforward matterto locate when or why instances of communication mismatch had occurred, and each of the mismatcheswas quite distinctive: (1) they were identified in different ways; (2) they occurred at different points in thecommunication process; (3) they arose because of problems occurring at different levels of the communication,and (4) they had different consequences.CONCLUSION: Given the frequency and complexity of miscommunication in general practice consultations,GPs need to consider adopting various strategies, at both the practice/systems level and the levelof the consultation interaction to minimise the risk of communication problems.KEYWORDS: Communication; general practice; physician-patient relationsIntroductionEffective communication is an integral part of thegeneral practice consultation. It is the primaryway information is exchanged, treatment decisionsare made and the therapeutic doctor–patientrelationship is established and maintained. 1 Yetresearch shows that communication problems area common feature of medical interactions, 2–4 andcan have significant adverse consequences forpatients’ quality of care, health outcomes, adherenceto treatment and satisfaction. 1,5,6 Further,miscommunication is the most common reasonfor patient medical complaints. 6–8Previous research has identified numerous andcomplex barriers to effective communication ingeneral practitioner (GP)–patient consultationsrelating to characteristics of GPs and patients,the nature of the GP–patient relationship, thestructure of the consultation and the nature ofthe different problems treated in primary healthcare. 6,9,10 However, much of the research reportedin the clinical literature relies on reported data,such as interviews or coding of consultations,and does not take account of the socioculturaland interactional contexts of GP–patient interaction.Research using social science methodologies,such as conversation analysis 11 does involveJ PRIM HEALTH CARE2013;5(2):123–128.CORRESPONDENCE TO:Sonya MorganDepartment of PrimaryHealth Care andGeneral Practice,Wellington Schoolof Medicine andHealth Sciences,University of Otago,PO Box 7343, WellingtonSouth, New Zealandsonya.morgan@otago.ac.nzVOLUME 5 • NUMBER 2 • JUNE 2013 J OURNAL OF PRIMARY HEALTH CARE 123

ORIGINAL SCIENTIFIC PAPERQUALITATIVE RESEARCHa close analysis of doctor–patient interaction.However, little research from either traditiondirectly examines miscommunication in naturallyoccurring consultations, and those that do tendto focus solely on either communication problemsidentified from post-hoc interviews with participants,12 or on misunderstandings that occur ininteraction during the consultation. 13,14 Existingresearch thus does not provide a complete pictureof the phenomenon.This paper describes the key findings of a study 15which used a triangulated case study approach toprovide a detailed and multi-layered analysis ofmiscommunication in two GP–patient consultations,and discusses the implications of the findingsfor clinical practice. Discourse analysis ofvideo-recorded naturally occurring consultations,along with in-depth interviews with both theGPs and patients, were used to identify instanceswhere communication mismatches had occurred,to explore how and why the mismatches occurred,and whether and how participants managedto resolve them.MethodsData collectionSeven consultations were selected from a dataset,video recorded between 2003 and 2005 as part ofa larger project: Clinical Decision Making whenRationing is Explicit (the Interaction Study). 16–19A total of 58 GPs in the wider Wellington regionof New Zealand were approached for the largerproject, using local networks and aiming for diversityof practice populations. Participating GPs’consultations were recorded for either a full dayor two half-day sessions. Written consent frompatients was sought by a research nurse whilepatients were seated in the waiting room.For the miscommunication sub-study reportedhere, data collection followed a sequential model(Figure 1). Consultation recordings were notviewed until after the interviews had taken place(i.e. at the time of the interview the researcherhad no knowledge of consultation content). Ofthe 14 patients approached for the current study,eight agreed to take part, with one patient subsequentlyexcluded due to equipment malfunction.Ethical Approval for the research was granted bythe Wellington Ethics Committee, New Zealand(Ref. 03/09/090).AnalysisThe theoretical framework for this research wasinteractional sociolinguistics. 20 The term ‘miscommunication’is an umbrella term used hereto refer to the overall process, while a specificinstance of miscommunication is termed a ‘communicationmismatch’. The analysis focuses onmismatches with potentially significant clinicalimplications. All seven GP–patient consultationsrecorded and the related interviews weresubjected to an initial content analysis; threeof the seven linked cases contained apparentcommunication mismatches determined by theresearcher or by the participants themselves. Twowere then purposively selected for in-depth casestudy analysis, as they were particularly richexemplars which offered detailed insight intothe sources and outcomes of different kinds ofmiscommunication. These consultations weretranscribed using adapted conversation analyticconventions (see the Appendix in the web versionof this paper), which capture both verbal andnon-verbal features, including overlaps in speech,pauses and interruptions. The consultation transcriptswere analysed using line-by-line discourseanalysis, 21 supported by contextual informationand post-consultation interviews. The interviewaudio-recordings were transcribed verbatim andwere analysed for thematic content.FindingsSummaryFour separate communication mismatches wereidentified, two in each of the two case studiesexamined. The findings suggest that apparentlyminor misunderstandings may have potentiallysignificant consequences, and that GPs andpatients may not even be aware that they haveoccurred. In-depth analysis of the case studiesrevealed the complexity of miscommunication. Itwas not a straightforward matter to locate whenor why instances of communication mismatchhad occurred in the dataset. Instead, each of themismatches was unique:124 VOLUME 5 • NUMBER 2 • JUNE 2013 J OURNAL OF PRIMARY HEALTH CARE

ORIGINAL SCIENTIFIC PAPERQUALITATIVE RESEARCH1. they were identified in different ways2. they occurred at different points in thecommunication process3. they arose because of problems occurring atdifferent levels of the communication, and4. they had different consequences.The complexity of the mismatches is demonstratedin the following sections by drawing onillustrative data from two of the four mismatchesanalysed for this study, which provide usefulcontrasts in terms of the occurrence and managementof miscommunication.Identified in different waysCommunication mismatches in the study datawere identified in different ways and were notalways readily and/or immediately identifiable byeither one or both participants. For instance, inthe first case (Case 1, see Appendix in the webversion of this paper), the patient’s misunderstandingabout the urgency of specialist assessmentof her bleeding mole was not apparent tothe researcher from viewing the interaction dataalone, and identification of a mismatch requiredaccess to post-consultation interview data. TheGP explicitly expressed her concern about themole and her view that it needed to be removedand a misunderstanding would, therefore, nothave been expected. More importantly, themisunderstanding was not recognised by eitherthe GP or patient at any stage during the recordedinteraction. The patient was only alerted to themisunderstanding when she contacted the clinicabout a week later to enquire about delaying herprocedure, and was informed that the referralwas urgent.By contrast, in the second case, the patient’s misunderstandingabout the reasons for taking Cartiawas immediately apparent both to the participantsthemselves during the recorded interactionand to the researcher reviewing the recordingsubsequently (Case 2, see Appendix in the webversion of this paper). The misunderstandingbecame evident when the GP was reviewing thepatient’s medications and the patient reportedthat he had not been taking Cartia. As the GPin this case was alerted to the patient’s misunderstandingduring the interaction, she had theWHAT GAP THIS FILLSWhat we already know: Effective communication between doctor andpatient is an essential component of quality care, good health outcomes,adherence to treatment and patient satisfaction. However, communicationproblems are a common feature of medical interactions and can have significantadverse consequences.What this study adds: This study completed detailed analysis of tworoutine general practice consultations combined with participant interviewsto provide a clearer picture of how and why miscommunication occurs. GPsshould assume communication mismatches occur frequently, and work todevelop a mix of strategies to minimise the risk of more serious communicationproblems.Figure 1. Data collection sequenceInteraction StudyConsultations of seven general practitioners and their consenting patientswere audio/video recordedMiscommunication StudyThree GPs taking part in the Interaction Study agreed to have an additionalpost-consultation interview with the researcher (immediately after theirconsultations were recorded)Fourteen patients of the three GPs above were mailed study information andinvited to an additional post-consultation interview with the researcherIndividual interviews with eight consenting patients were undertaken(within 2.5 weeks of recorded consultation)Individual interviews with three GPs were undertaken(within 4 weeks of recorded consultation)Audio and videotapes were reviewed by the researcherFinal data synthesisVOLUME 5 • NUMBER 2 • JUNE 2013 J OURNAL OF PRIMARY HEALTH CARE 125

ORIGINAL SCIENTIFIC PAPERQUALITATIVE RESEARCHopportunity to respond at that time, and thepatient’s misunderstanding was immediately andsuccessfully addressed and resolved (as confirmedat the post-consultation interview).Occurred at different consultation pointsThe complexity of miscommunication wasfurther demonstrated in this dataset by the findingthat the mismatches occurred at differentpoints in the communication process, buildingup over time and across interactions. For example,although the patient’s misunderstandingabout taking Cartia was apparent to both GP andpatient during the consultation recorded for thisstudy, the origin of this misunderstanding was ina prior consultation. As mentioned above, it wasonly the GP’s review of medication adherence inthis subsequent consultation that alerted her tothe mismatch.Different reasons for occurrenceA detailed discourse analysis revealed themismatches came about because of multifacetedproblems occurring at different ‘levels’ of thecommunication. In the absence of a recording ofthe prior interactions, it is not possible to makeany claims about how the Cartia mismatch firstoccurred. However, in the case of the bleedingmole (Case 1), access to and analysis of the actualinteraction where the mole was first discussedrevealed some potential sources of the problem.These related to the high-level frames (assumptionsbased on background knowledge andexperiences) through which the GP and patientfiltered and interpreted information during theconsultation, as well as to localised aspects ofthe information delivery (such as the GP’s use ofauthoritative and persuasive language which potentiallyreduced the patient’s decision-making).Different consequencesFinally, the mismatches identified in this studydemonstrate that communication problems canhave different consequences. In both of the twocases illustrated in this paper, the observed misunderstandingscreated the potential for seriousadverse outcomes. For example, in Case 1, if thepatient concerned had not followed up with thespecialist clinic about having her mole removed(and no follow-up had occurred from the GP)serious negative consequences for the patient mayhave ensued. In Case 2, the patient’s misunderstanding(in the past) had already resulted in thenegative outcome of the patient misguidedly stoppinghis Cartia medication, which had potentiallyplaced him at an increased risk of stroke.DiscussionSome important lessons for clinical practicecan be taken from this detailed investigation ofmiscommunication in two actual general practiceconsultations. First of all, it is important torealise that communication mismatches occurfrequently and cannot be avoided altogether;what is more remarkable is that major communicationproblems are not documented more often.The GPs in this study were typical, experiencedclinicians. They did not make poor decisions intheir care of these patients, and yet, despite this,misunderstandings with potentially significantadverse consequences occurred. Furthermore, thefinding that a misunderstanding can go unrecognisedwhen both GP and patient felt the consultationwent well (mole case), highlights the factthat GPs should not assume their communicationhas been successful or understood as intended.General practice consultations are extremelycomplex interactional events, with numerouspotential barriers to effective communication, includingstrict time constraints. It is therefore essentialthat GPs work actively on strategies whichminimise the risk of more serious communicationproblems occurring. Yet minimising the risk ofmiscommunication in general practice is not astraightforward matter. Communication mismatchesare complex and multifaceted. They maybe identified in different ways (by GP, patient orboth), or may not be apparent at all during theconsultation. Mismatches may occur at any stageduring a single consultation, or they may developover time, surfacing in subsequent interactions.Mismatches may come about due to problems occurringat different levels of the communication,and they may have different consequences.Given the demonstrated complexity of miscommunication,different communication strategies126 VOLUME 5 • NUMBER 2 • JUNE 2013 J OURNAL OF PRIMARY HEALTH CARE

ORIGINAL SCIENTIFIC PAPERQUALITATIVE RESEARCHare needed to address different kinds of communicationproblems. In this study, for example,misunderstandings occurred in consultationsinvolving both chronic and acute conditions. Inchronic conditions, shared care plans are increasinglybeing used and can enable a common understandingabout self-management of medicationsand lifestyle activity. These may have potentialto reduce misunderstandings, and encouragepatients to take a central role in managing theirhealth and adhere to treatment plans. 22,23 In acutepresentations, where the patient may not be seenagain for some time, if ever, it is arguably morecrucial that patients leave the consultation witha clear, agreed action plan. Relying on patientsthemselves to take action subsequent to theconsultation (e.g. to make a specialist appointment)may not lead to expected outcomes andGPs should ensure system prompts are in place tomonitor agreed actions that have critical outcomesif not followed.In addition to these management/practice systemsapproaches, it is important that GPs identify andpractise strategies that will increase the chanceof patients both understanding and recallingkey information and decisions. Summarisingand repeating information with patients towardsthe end of the consultation is one important andcommonly used method, as is the ‘ask tell ask’strategy. 24Providing patients with a written summary orchecklist of the key consultation points maybe a useful communication aid. Although fewprimary care studies have addressed this issue, 25a recent Cochrane review found evidence thatpeople use written- or audio-recordings of consultationsto remind themselves of the informationcommunicated, to review information theymissed, or to share information with others 26 andthat such aids to recall are positively viewed bypatients. 26,27 Despite these positive findings, clinicianshave been reluctant to adopt such communication/recallaids, 28 perhaps challenged by theimplementation compliance.This study also suggests the importance of‘interactional checking’ as a safeguard whencommunicating with patients. For instance, inthe example of successful mismatch resolutionidentified here (Cartia case), the GP’s strategyof reviewing the patient’s medications allowedthe patient to disclose his non-adherence and hismisinterpretation of the Cartia prescription, thusallowing the misunderstanding to be identified.A well-established relationship between GP andpatient is likely to facilitate such ‘confessions ofnon-compliance’. 29 By contrast, no such ‘interactionalchecks’ occurred in the first case study andthe misunderstanding regarding the urgency ofthe patient’s mole referral did not become evidentuntil some time after the consultation.Line-by-line discourse analysis reveals theintricate details of the communication process. Itprovides an extension of analysis beyond participants’motivated recounting of events. 30 In addition,this study triangulated discourse analysis ofconsultations, with participants’ perceptions ofconsultations gained from participant interviews.Future research using this multi-method casestudy methodology could be considered to investigatethe effectiveness of different interventionsto avoid miscommunication (such as usingconsultation summaries or checklists).This study is based on just two selected casestudies. Although an obvious limitation, thishas also made it possible to undertake anin-depth analysis of linked data. It could alsobe argued that the GPs who were willing toparticipate had a particular interest in communication,and therefore may have behaved differentlyfrom other GPs who did not volunteer.However, misunderstandings still occurred inthese consultations, so it is reasonable to expectthat other GPs would experience similar communicationproblems. Whether the recordingequipment also affected the behaviour of theparticipants is unknown. However, modernaudio- and video-recording is a well-established,credible and unobtrusive data collection method,31,32 with participants reporting they quicklybecome accustomed to the equipment.ConclusionEffective communication is fundamental to thegeneral practice consultation, yet communicationproblems are frequent. Previous research does notprovide a clear picture of how and why miscom-VOLUME 5 • NUMBER 2 • JUNE 2013 J OURNAL OF PRIMARY HEALTH CARE 127

ORIGINAL SCIENTIFIC PAPERQUALITATIVE RESEARCHACKNOWLEDGEMENTSThe author would like tothank Maria Stubbe andEileen McKinlay for theirinput into the drafts ofthis article. The recordedconsultations examined inthis study were collectedby Lindsay Macdonaldas part of the InteractionStudy. The author isespecially grateful to theparticipating patients andgeneral practitioners,and would also like tothank Debbie McLeod,Tony Dowell, Kevin Dew,Libby Plumridge, and theresearch assistants whotranscribed the data, fortheir input and support.FUNDINGThis study was completedunder the umbrella ofa larger study by theApplied Research onCommunication in HealthGroup (ARCH), TheInteraction Study—ClinicalDecision Making whenRationing is Explicit (theInteraction Study), fundedby the Health ResearchCouncil of New Zealand.COMPETING INTERESTSNone declared.munication occurs, or how such problems canbe prevented or resolved. This study has shownthat miscommunication between GPs and theirpatients is an extremely complex phenomenon.Problems can occur at any point in the communicationprocess, often going unnoticed by theparticipants, and it is not always possible to determineprecisely when a communication breakdownhas occurred. In this study, the multi-methodcase study methodology enabled a detailedanalysis from both GP and patient perspectives.In the future, this methodology could be usedto examine the effectiveness and acceptability ofcommunication aids to avoid miscommunicationat both the systems level and within the consultationitself.References1. Ong LM, de Haes JC, Hoos AM, Lammes FB. Doctor-patientcommunication: a review of the literature. Soc Sci Med.1995;40:903–18.2. Lewin S, Skea Z, Entwistle V, Zwarenstein M, Dick J. Interventionsfor providers to promote a patient-centred approachin clinical consultations. Cochrane Database Syst Rev.2001;(4):CD003267.3. Simpson M, Buckman R, Stewart M, Maguire P, Lipkin M,Novack D, et al. Doctor-patient communication: the Torontoconsensus statement. BMJ. 1991;303:1385–7.4. Thistlethwaite J, Morris P. The patient-doctor consultation inprimary care: theory and practice. London: The Royal Collegeof General Practitioners; 2006.5. Kaplan SH, Greenfield S, Ware JE. Assessing the effects ofphysician-patient interactions on the outcomes of chronicdisease. Med Care. 1989;27:110–27.6. Stewart M, Brown J, Boon H, Galajda J, Meredith L, SangsterM. Evidence on patient-doctor communication. Cancer PrevControl. 1999;3:25–30.7. NZ Health and Disability Commissioner. General Practitioner,Dr B: an accident and medical clinic report by the Health andDisability Commissioner. New Zealand: Health and DisabilityCommission; 2004. Report No.: Case 03HDC06973.8. Wofford MM, Wofford JL, Bothra J, Kendrick SB, Smith A,Lichstein PR. Patient complaints about physician behaviors: aqualitative study. Acad Med. 2004;79:134–8.9. Heritage J, Maynard D. Introduction: analyzing interactionbetween doctors and patients in primary care encounters. In:Heritage J, Drew P, Maynard D, editors. Communication inmedical care: interactions between primary care physiciansand patients. Cambridge: Cambridge University Press; 2006.10. Williams S, Weinman J, Dale J. Doctor-patient communicationand patient satisfaction: a review. Fam Pract. 1998;15:480–92.11. Hutchby I, Wooffitt R. Conversation analysis: principles,practices and applications. Cambridge: Polity Press; 1998.12. Britten N, Stevenson F, Barry C, Barber N, Bradley C. Misunderstandingsin prescribing decisions in general practice:qualitative study. BMJ. 2000;320:484–8.13. Moss B, Roberts C. Explanations, explanations, explanations:how do patients with limited English construct narrative accountsin multi-lingual, multi-ethnic settings, and how can GPsinterpret them? Fam Pract. 2005;22:412–8.14. Roberts C, Moss B, Wass V, Sarangi S, Jones R. Misunderstandings:a qualitative study of primary care consultations inmultilingual settings, and educational implications. Med Educ.2005;39:465–75.15. Morgan S. Miscommunication in GP consultations: a microanalysisof communication mismatches in two case studies[master’s thesis]. [Wellington (NZ)]: University of Otago;2008.16. Dew K, Dowell A, Stubbe M, Plumridge E, Macdonald L.‘Treating’ patients differently: a qualitative study of how clinicaland social factors shape interactions between doctors andpatients. N Z Fam Physician. 2008;35.17. Dew K, Plumridge E, Stubbe M, Dowell A, Macdonald L,Major G. ‘You just got to eat healthy’: the topic of CAM in thegeneral practice consultation. Health Soc Rev. Special Issue:Integrative, Complementary and Alternative Medicine: Challengesfor Biomedicine? 2008;17.18. Dew K, Stubbe M, Macdonald L, Dowell A, Plumridge E. The(non) use of prioritisation protocols by surgeons. Sociol HealthIlln. 2010;32:545–62.19. Dowell T, Macdonald L, Stubbe M, Plumridge E, Dew K.Clinicians at work: what can we learn from interactions in theconsultation? N Z Fam Physician. 2007;34:345–50.20. Gumperz J. On interactional sociolinguistic method. In: SarangiS, Roberts C, editors. Talk, work, and institutional order:discourse in medical, mediation, and management settings.Berlin/New York: Mouton de Gruyter; 1999.21. Paltridge B. Making sense of discourse analysis. Queensland,Australia: Gerd Stabler; 2000.22. Russell G, Thille P, Hogg W, Lemelin J. Beyond fighting firesand chasing tails? Chronic illness care plans in Ontario,Canada. Ann Fam Med. 2008;6:146–53.23. Martin C, Peterson C. Improving chronic illness care: revisitingthe role of care planning. Aust Fam Physician. 2008;37:146–53.24. Kemp EC, Floyd MR, McCord-Duncan E, Lang F. Patientsprefer the method of ‘tell back-collaborative inquiry’ to assessunderstanding of medical information. J Am Board Fam Pract.2008;21:24–30.25. Liddell C, Rae G, Brown T, Johnston D, Coates V, MallettJ. Giving patients an audiotape of their GP consultation: arandomised controlled trial. Br J Gen Pract. 2004;54:667–72.26. Pitkethly M, MacGillivray S, Ryan R. Recordings or summariesof consultations for people with cancer. Cochrane DatabaseSyst Rev. 2008;(3): CD001539.27. Watson PWB, McKinstry B. A systematic review of interventionsto improve recall of medical advice in healthcare consultations.J R Soc Med. 2009;102:235–43.28. McConnell D, Butow PN, Tattersall MH. Audiotapes andletters to patients: the practice and views of oncologists, surgeonsand general practitioners. Br J Cancer. 1999;79:1782–8.29. Stubbe M, Dew K, Macdonald L, Plumridge E, Dowell A.Persuasion and resistance in general practice referral talk.Conference Proceedings: Presentation Communication,Medicine and Ethics Conference (COMET); 2007 28–30 June.Lugano, Switzerland: University of Lugano; 2007. p. 157.30. Bugge C, Jones A. Methods for studying patient participation.In: Collins S, Britten N, Ruusuvuori J, Thompson A, editors.Patient participation in health care consultations. England:Open University Press; 2007.31. Coleman T. Using video-recorded consultations for researchin primary care: advantages and limitations. Fam Pract.2000;17:422–7.32. Stewart M. Approaches to audiotape and videotape analysis:interpreting the interactions between patients and physicians.In: Crabtree B, Miller W, editors. Doing qualitative research.London: Sage Publications Inc.; 1992.128 VOLUME 5 • NUMBER 2 • JUNE 2013 J OURNAL OF PRIMARY HEALTH CARE

ORIGINAL SCIENTIFIC PAPERMIXED METHOD RESEARCHIs it time to talk? Interpreter services use ingeneral practice within CanterburyKara Seers BSc student; 1 Lynley Cook MBChB, MPH, FRNZCGP, FNZCPHM; 1 Gillian Abel PhD; 2Philip Schluter BSc (Hons), MSc, PhD; 3,4 Paul Bridgford BSc (Hons) 1ABSTRACTINTRODUCTION: Effective communication is fundamental to successful health care service delivery,and has a positive impact on access, quality of care, health outcomes, and patient satisfaction. Althoughthere are a growing number of New Zealanders who do not speak English proficiently, underutilisation oftrained interpreter services appears to be common in primary health care settings.AIMS: To describe the pattern of interpreter service need and utilisation by general practice services,and to identify key barriers and enabling factors to the use of trained interpreters.METHODS: A mixed methods study was employed. Census and Partnership Health Canterbury Te Keio Te Waka (PHC) databases were combined, and quantitative analysis used to derive interpreter serviceneed and utilisation patterns. Transcripts of focus groups and interviews from general practitioners, practicenurses and practice administration staff within the PHC were analysed, using qualitative methods toidentify barriers and enablers to interpreter service use.1Pegasus Health (Charitable)Ltd, Christchurch,New Zealand2Department of PublicHealth and General Practice,University of Otago,Christchurch, New Zealand3School of Health Sciences,University of Canterbury,Christchurch, New Zealand4School of Nursing andMidwifery, The Universityof Queensland, Brisbane,AustraliaRESULTS: For the years 2008–2010, approximately 10 742 consultations per year involved a non-Englishspeakingpatient, yet in only approximately 74.8 (0.7%) consultations per year were interpreter servicesutilised. Analysis of focus groups and interviews identified four global themes that represented barriers forinterpreter service utilisation; namely, practicalities, expectations, knowledge of service, and systems.DISCUSSION: The current use of interpreter services in PHC general practice appears to be significantlyless than the need. In order to maximise health outcomes and reduce risk, strategies must be initiated tocounter the barriers currently inhibiting interpreter service use, including adopting best practice policies.KEYWORDS: Communication; communication barriers; general practice; primary health careIntroductionEffective communication is essential to the accessand quality of health care services, and isrecognised as a health service user’s right in NewZealand. 1,2 It has been shown to have a positiveimpact on patient satisfaction, utilisation, qualityof care, and health outcomes. 3,4 However, increasingnumbers of New Zealanders are overseasborn,many originating from North-East Asiancountries, and with limited English proficiency(LEP). 5 For people with LEP, the use of trainedinterpreters is fundamental to ensure effectivecommunication and quality of care. 3 Despite this,underutilisation of trained interpreter servicesappears to be common in health care serviceswithin New Zealand 6 and overseas. 7–8Trained interpreter services are available toorganisations across New Zealand, includingPartnership Health Canterbury Te Kei o TeWaka (PHC)–affiliated general practice services(enrolled population of 369 674 and 94 generalpractice locations on 30 June 2010). Telephoneand face-to-face services are available throughLanguage Line (telephone-based service availablesince November 2007) and Interpreting Canterbury(telephone and face-to-face service availablesince February 2011) at no financial cost to cliniciansor patients, but uptake appears to be low.J PRIM HEALTH CARE2013;5(2):129–137.CORRESPONDENCE TO:Lynley CookPegasus Health(Charitable) Ltd,160 Bealey Avenue,Christchurch Central,Christchurch 8013,New Zealandlynley.cook@pegasus.org.nzVOLUME 5 • NUMBER 2 • JUNE 2013 J OURNAL OF PRIMARY HEALTH CARE 129

ORIGINAL SCIENTIFIC PAPERMIXED METHOD RESEARCHThe aims of this study were twofold: firstly, todescribe the pattern of interpreter service needand utilisation by PHC-affiliated general practices;and, secondly, to identify key barriers andenabling factors to the use of trained interpreters.Both quantitative and qualitative methods wereemployed.MethodsThis study used mixed methods, combining aquantitative cross-sectional analysis with a qualitativethematic content analysis.Quantitative methodsDesignA quantitative cross-sectional analysis wasundertaken combining the PHC 2008–2010 databaseof enrolled patients (held by PHC); NewZealand 2006 Census population database (heldby Statistics New Zealand); and Language Line(availability: 29/11/2007–31/10/2011) and InterpretingNew Zealand (availability: 1/02/2011–19/10/2011) billing information databases (bothheld by PHC).Target populationThe target population was adults (aged 15 yearsand older) enrolled in general practices affiliatedwith PHC (Canterbury, New Zealand) over the2008–2010 period.Database and variables definitionsInformation about non-English speakers wasgained in the 2006 Census, Question 13. 9 Thisquestion asked respondents to tick from a numberof options which language(s) they could have aconversation in about everyday things. Respondentswere explicitly reminded to tick English ifthey could have a conversation in English. For thepurpose of this analysis, those who were able tospeak a language but did not mark English weredefined as non-English speaking. Statistics NewZealand provided data on English and non-Englishspeaking variables by ethnicity (classified asEuropean/other, Maori, Pasifika, Asian, African,Middle Eastern), age (classified into 0–4, 5–14,15–24, 25–44, 45–64, and ≥65 year groups), andgender for the national population and for thegreater Christchurch region. This latter regionconsisted of Kaikoura, Hurunui, Waimakariri,Selwyn, Ashburton Districts and ChristchurchCity territorial authorities; designed to cover thegeographical region of enrolled PHC patients.People identifying with multiple ethnic groupsare represented multiple times in any ethnicspecificbreakdown in the Census 2006 figures,whereas PHC currently uses a single priorityclassification. 10Statistical analysisData were imported into the specialist statisticalpackage, SAS version 9.2 (SAS Institute Inc.,Cary, NC, USA), and then consistency and rangechecks were performed. Descriptive statisticswere calculated and reported for the demographicvariables and then for the enrolled population,consultancy numbers, and non-English-speakingproportion variables by ethnic, age and gendergroupings. These statistics were then used todetermine the expected number of non-Englishconsultations by taking each ethnic, age, and genderclassification combination, and multiplyingthe average patient numbers × average number ofconsultations/year × proportion of non-Englishspeakers (from the 2006 Census greater Christchurchregion database), and then summing overall classification combinations. For example, inthe 2006 Census there were 26 292 European/other women aged 15–24 years within the greaterChristchurch region. Of these, 48 (0.18%) werenon-English speakers. Over the study period,there was an average of 55 471 European/otherwomen aged between 15 and 24 years registeredwith the PHC who made an average of 2.54 consultations/year.Therefore, the expected numberof non-English consultancies for this group is55 471 × 2.54 × (48 / 26 292) = 257.26. Repeatingthis calculation over all age, ethnic and gendercategorisations, and then summing gives theexpected total. Stata version 12.0 (StataCorp, CollegeStation, Texas, USA) was used for all graphs.Qualitative methodsAn interpretive approach was taken to thequalitative arm of the study which focuses130 VOLUME 5 • NUMBER 2 • JUNE 2013 J OURNAL OF PRIMARY HEALTH CARE

ORIGINAL SCIENTIFIC PAPERMIXED METHOD RESEARCHon the meaning people give to phenomena orexperiences. 11 This approach is used frequentlyin health research, as it aims to gain an understandingrather than an explanation of theworld. 11 Purposive sampling was used to recruitgeneral practitioners (GPs), practice nurses andreceptionists into the qualitative arm of thestudy. Practices with high numbers of refugeeand migrants enrolled as patients were particularlytargeted. Participants were invited toparticipate in one of two focus groups. Ethicsapproval was not required for the study as it wasdeemed low risk.Focus groups are particularly useful whenthere is no depth of knowledge about the topicbut the researcher wishes to explore what andwhy people think the way they do. 12 A goodfocus group has few questions but relies onthe interaction within the group to elicit newknowledge or information. Two focus groupswere held in December 2011. One focus groupwas composed of four general practitioners, andthe other focus group of two practice nursesand three general practice administration staff.The questions within the semi-structured guideused in the focus groups were informed by thereading of the literature, but were broad enoughto allow for the exploration of new informationbased on the participants’ experiences. Thequestions focused on what the challenges are indealing with patients with LEP and how bestthese can be overcome, with a particular emphasison interpreter service use.As a follow-up to the focus groups, semistructuredin-depth interviews were conductedwith a practice nurse and a GP. These interviewswere conducted to explore the complex decisionmakinghealth providers engage in when seeinga patient with LEP, and to explore the meaningsand interpretations they give to this experience.The interviews allowed for some of the issuesbrought up in the focus groups to be explored inmore depth. However, it should be borne in mindthat saturation was not achieved through the useof only two in-depth interviews. Each data collectionsession was audio recorded and transcribedbefore being subjected to a thematic contentanalysis where common themes were searched forwithin the transcripts.WHAT GAP THIS FILLSWhat we already know: Good communication between provider andclient is fundamental to any health care provision service, and is recognisedas a health service user’s right in New Zealand. However, interpreter servicesare frequently underutilised in health care provision, with cost cited mostfrequently as the barrier to use.What this study adds: In a large urban/suburban region over years2008–2010, we estimate that approximately 10 742 consultations per yearinvolved a non-English-speaking patient, yet in only approximately 74.8(0.7%) consultations per year were interpreter services utilised. Analysis offocus groups and interviews identified four global themes that representedbarriers for interpreter service utilisation; namely, practicalities, expectations,knowledge of service, and systems.ResultsPopulation characteristicsIn 2010, the average registered population sizeof PHC totalled 366 075 individuals. Table 1 includesdemographics of this PHC population, togetherwith 2006 Census figures from the greaterCanterbury region and New Zealand. The ageand gender profiles are similar across all groups,and the ethnic and deprivation index profiles 13 aresimilar between PHC and greater Canterbury regiongroups—but different from national figures,a known demographic finding.Establishing interpreter service utilisationOver the calendar years 2008–2010, the PHC recorded2 669 586 consults for an average enrolledpopulation of 349 498 people. Figure 1 depictsthe average number of consultations per year overthis period for enrolled patients by ethnicity, age,and gender classifications. Clear ethnic, age, andgender differences emerged. In particular, malesgenerally had fewer consultations than females;consultation rates were markedly less for Asianpeople, and for Pasifika and Middle Easternwomen, compared to other ethnicities; and therewas a strong age-dependent skewed U-shapedrelationship in rates.The percentage of non-English-speaking peoplein the greater Christchurch region from the2006 Census by ethnicity, age, and genderVOLUME 5 • NUMBER 2 • JUNE 2013 J OURNAL OF PRIMARY HEALTH CARE 131

ORIGINAL SCIENTIFIC PAPERMIXED METHOD RESEARCHTable 1. Demographic data from the 2006 Census for New Zealand (N=4 028 247) and the greater Christchurch region(N=466 407), together with the 2010 PHC figures (N=366 075)Age (years)New Zealand(2006 Census)Greater Christchurch(2006 Census)PHC (2010)n (%) n (%) n (%)0–4 275 079 (6.8) 29 403 (6.3) 25 027 (6.8)5–14 592 500 (14.7) 62 289 (13.4) 47 258 (12.9)15–24 571 176 (14.2) 66 507 (14.3) 46 807 (12.8)25–44 1 134 255 (28.2) 131 667 (28.2) 97 888 (26.7)45–64 959 337 (23.8) 113 853 (24.4) 96 581 (26.4)≥ 65 495 606 (12.3) 62 688 (13.4) 52 514 (14.3)GenderFemale 2 062 626 (51.2) 238 956 (51.2) 192 022 (52.5)Male 1 965 621 (48.8) 227 451 (48.8) 174 053 (47.5)Ethnicity*European/other 3 080 361 (71.3) 414 414 (84.5) 308 916 (84.4)Maori 565 326 (13.1) 33 417 (6.8) 23 574 (6.4)Pasifika 287 658 (6.7) 11 037 (2.3) 8 964 (2.4)Asian 358 008 (8.3) 28 617 (5.8) 22 170 (6.1)African 10 647 (0.2) 1 209 (0.2) 1 267 (0.3)Middle Eastern 17 514 (0.4) 1 458 (0.3) 1 184 (0.3)Deprivation index †‡1–2 825 597 (20.5) 124 677 (26.7) 96 731 (28.7)3–4 810 849 (20.2) 104 499 (22.4) 74 587 (22.2)5–6 797 046 (19.8) 99 225 (21.3) 70 125 (20.8)7–8 791 388 (19.7) 79 989 (17.2) 52 679 (15.6)9–10 798 162 (19.8) 58 002 (12.4) 42 553 (12.6)PHC Partnership Health Canterbury Te Kei o Te Waka* Census figures give the total responses over all ethnic categories so individuals identifying with multiple ethnic groups will be countedmore than once† The Deprivation Index used here is the NZDep2006. The NZDep2006 is a scale from 1 to 10 that divides New Zealand meshblocksinto tenths with a value of 10 indicating that the meshblock is in the most deprived 10% of areas in New Zealand and, conversely, avalue of 1 indicates that it is in the least deprived 10% of New Zealand. 13‡ 4902 values missing from the New Zealand figures, 15 values missing from the greater Christchurch figures, and 29 400 values missingfrom the PHC figures.categorisations appears in Figure 2. Due to thesmall numbers, New Zealand Census figureswere used for the African percentages exceptthe male, 65-and-older year group, where noreliable estimate could be ascertained. Again,clear ethnic, age, and gender differences exist,with Asian people having the highest proportionof non-English speakers, followed by African,Pasifika and Middle Eastern peoples; a highernon-English-speaking proportion was seen infemales compared to males; and there was againa strong age-dependent skewed U-shaped relationshipfor Asian, African, Pasifika and MiddleEastern peoples.Given the similarity of the PHC and greaterCanterbury region profiles seen in Table 1, applicationof the 2006 Census non-English-speaking132 VOLUME 5 • NUMBER 2 • JUNE 2013 J OURNAL OF PRIMARY HEALTH CARE

ORIGINAL SCIENTIFIC PAPERMIXED METHOD RESEARCHFigure 1. Average number of consultations per year (over the period 2008-2010) for patients enrolled in Partnership Health Canterbury Te Kei o Te Wakaby ethnicity, age group, and genderproportions to the PHC population appearsreasonable. Combining the information containedin Figures 1 and 2 with the ethnicity, age, andgender characteristics for PHC summarised inTable 1, yielded an expected number of 10 742instances per year where non-English-speakingpatients aged 15 years and older would consulta PHC general practitioner. This equates to ap-proximately 1.5% of all consultations for patientsaged 15 years and older.Data were available from Language Line between29/11/2007 and 31/10/2011, and 131 serviceevents were recorded, at a rate of 33.4/year. ForInterpreting New Zealand, data were availablebetween 1/02/2011 and 19/10/2011, and 33 ser-Figure 2. Percentage of non-English speaking people in the greater Christchurch region from the 2006 Census by ethnicity, age group, and genderVOLUME 5 • NUMBER 2 • JUNE 2013 J OURNAL OF PRIMARY HEALTH CARE 133

ORIGINAL SCIENTIFIC PAPERMIXED METHOD RESEARCHvice events were recorded, at a rate of 41.4/year.Thus total utilisation of PHC-funded interpreterservices observed within the PHC enrolled populationequalled 33.4 + 41.4/year = 74.8/year, some0.7% of the expected number.Identifying key barriers and enablersto interpreter service useThere were several global themes identifiedwithin the qualitative arm of the study thatrepresented the barriers to the use of interpreterservices. These themes are discussed below withexcerpts from the transcripts presented to illustratethe analysis.Practicalities within a busy general practiceIssues such as additional time requirement, costand amenities were raised by the participants asbarriers to the use of interpreter services. Allparticipants emphasised that time was an issueand that if they went over time and this wasunscheduled, it caused delays for other patients.They indicated that the longer time taken in anappointment when an interpreter was used wasdisruptive to the practice.The other thing is timing of course you’ve then got,if you’ve got a third party involved in the consultationit’s much harder to be flexible with timing forother patients so you’ve got somebody else sittingthere waiting and so there’s pressure upon you to,to try to keep really hard to time which may meansome other patient’s consultation gets chopped. (GP)In addition, participants explained that setting upan appointment for a patient with LEP to have aninterpreter requires a greater deal of organisationthan for the average patient, and ideally shouldbe done in advance.The patient arrives… perhaps there isn’t the timeto set it up, again it would have to be planned.(Administrator)Perceptions of the financial costs both to thepractice and the taxpayer also present a practicalproblem and may inhibit the use of interpreterservices. One participant spoke of interpretersbeing an inefficient use of resources.There’s still a cost. It’s not free, um, in fact it’smore expensive overall if it goes through the DHB[district health board] than if it goes through theindividual practice, not to the practice directly,but to the taxpayer as a whole and I think as, I amacutely aware of the need in general practice to useresources as sparingly as possible um, and for themost appropriate cases at the most appropriate time.(GP)Other participants indicated that they did notrealise that there would be no direct cost to thepractice in using interpreters.There is no charge to the practice and there is no invoicingor paperwork or anything that the practiceneeds to do. (Interviewer)I mean, we are all probably quite high users and wehaven’t, I haven’t I didn’t pick that up till... (GP)I didn’t either. (GP)Many ways of ‘getting by’ without professionalinterpreters were identified. Some participantsfelt that they could cope sufficiently well usingtheir own communication techniques or familyor staff as interpreters. Some participants talkedabout managing communication with alternativetechniques such as mime, use of online translators,use of their own foreign language skills ornon-verbal techniques.Certainly when I’m speaking with someone,especially about something like diabetes with thefamily, I mean, I slow right down, and I’m alsowatching body language, and I can sense when I’mlosing them and so I change the way the words… Iuse my hands, I might … we can use pictures youknow it’s not just all about verbal communication…I just, always sort of managed it really. (Nurse)Cultural perceptionsPatients with LEP may have cultural valuesand beliefs that act as barriers to usinginterpreter services. Some of the participantsin this study perceived that some of theirpatients would take offence at being offered aninterpreter. As a result, they did not explicitlyoffer the service.134 VOLUME 5 • NUMBER 2 • JUNE 2013 J OURNAL OF PRIMARY HEALTH CARE

ORIGINAL SCIENTIFIC PAPERMIXED METHOD RESEARCHThey are actually quite offended because they thinkthat … you underestimate their ability of talking inEnglish. (Nurse)The participants indicated that many times,patients with LEP brought family members withthem to their appointments and these familymembers insisted on translating for the patient.Some of the participants felt that family membersoften tried to control how the consultationwas run and that this situation was therefore notideal. They indicated that in such situations aninterpreter would be preferable.…’cause the family often says: we will come andhelp. So sometimes, the family isn’t actually thehelp you need. (Administrator)People that share a common language do notnecessarily have similarities in terms of ethnicityor in their religion. The participants spokeof how one interpreter from a particular ethnicgroup may be acceptable to some patients but notothers, which made for additional difficulties.Privacy was also seen as a barrier to interpreterservice use as there were concerns about patientconfidentiality. Some ethnic groups are relativelysmall and ‘tightly-knit’ which makes the use ofan interpreter from that community problematic.There’s probably from the other side a degree ofreticence to use the interpreters as far as those thathave limited English as well. All to do with youknow, family information, information getting outinto the wider community, confidentiality or feelingsof confidentiality if you talk to interpreters,… and yeah a lot of people don’t want that person,or that person, or that person to know any of theirbusiness and you know, there’s often a ‘loss of face’especially if they’re talking about mental healthwhich is a difficult thing and they really want tokeep it confidential. (GP)Staff culture, including role responsibilities, mayinhibit interpreter service access. A participatingreceptionist explained how she didn’t feel thatshe had the authority to call in an interpreter fora patient when she thought it necessary.Yes I do think if we’re talking barriers to it [use ofinterpreter services] I think we really have to lookat the GPs and you know I take instruction. So ifthey’re prepared to instruct me to set it up—and Iknow how to do it now, it’s really simple—but itcomes down to if the GPs themselves are willing touse the service. (Nurse)KnowledgeMany of the participants were unaware of all theoptions that were available for patients with LEP.Participants indicated that health care providersmay also not access interpreter services simplybecause they do not think to use them.And I must admit that I’ve got a fairly ‘woolly’understanding of all the different agencies thatexist in Christchurch to provide health ’cause itsconstantly changing. (GP)Some participants felt that many patients withLEP were unaware of their rights as patients and,because of this lack of awareness, they hypothesisedthat patients also were unaware that theycould expect or ask for interpreter services withintheir consultations.A lot of these people have arrived in the countrythey know almost nothing about. They’re findingtheir way through the entire services, you know,how the education, social welfare system works.They’re completely at sea. (GP)SystemsAccording to participants, a lack of policy andinformation management poses a barrier toservice access. Individual practices are inhibitedby not having systematic recording of Englishproficiency, a lack of training policy regardingthe use of interpreter services, and by not havingtechnical set-ups or facilities conducive tointerpreter service use. The GPs, practice nursesand receptionists all indicated a need for furthertraining regarding interpreter service use.Maybe more education to the health professionals—let them be aware to use that [interpreter services]and maybe the importance of using it. (Nurse)Much of the information regarding patients withLEP is available only, or mostly, in English. ThisVOLUME 5 • NUMBER 2 • JUNE 2013 J OURNAL OF PRIMARY HEALTH CARE 135

ORIGINAL SCIENTIFIC PAPERMIXED METHOD RESEARCHis clearly a problem as the people who need toaccess it will have difficulties understandinginformation in a language they are not proficientin. The participants saw this as a problem andthat information should be printed in a variety oflanguages.I think it’s important to have this [informationabout interpreter services] more out and in differentlanguages … If it’s in English, how can you look atit? (Nurse)Health professionals commented on the difficultyof managing incoming information, includingthat on interpreter services, due to the sheerquantity of information. They identified a needfor a better system to manage this information.Every six months the DHB [district health board]send through another set of Language Line cards.They probably go straight to recycling… that’s thetrouble. (GP)DiscussionThis study is the first to attempt to quantify theextent of interpreter service underutilisationwithin a primary health care organisation in NewZealand. Based on reliable Census non-Englishspeakingdata, consistent demographic profilesbetween the PHC and Census figures, and the accurateconsultancy numbers contained within thePHC database, we would expect around 10 742instances per year where non-English-speakingadults aged 15 years and older would consult aPHC GP. However, only 74.8 (0.7%) consultationsper year involved interpreter services. Thesefigures represent a potentially massive underutilisationof services and, consequently, inferiorcare and outcomes for those affected. Whilethe expected consultation numbers were basedon some gross assumptions, ignoring importantnuances in health care service delivery in NewZealand (for instance, effective non-English consultationsand services), it does provide an insightinto the extent of this largely hidden problem.These findings are consistent with New Zealandand international studies reported elsewhere. 14–18Most recently, Gray and colleagues examined clinicians’pattern of use of interpreters in hospitalservices in the Wellington region. 14 They foundthat there was a high level of awareness of boththe clinical risk of not using interpreters for peoplewith LEP and the relevant policy. Yet, therewere low levels of trained interpreter utilisation.Instead, family members were often engaged tointerpret in consultations, a practice fraught withethical and moral difficulties.Previous studies have identified many of the barriersto interpreter service use identified by ourqualitative study. 7,16,19,20 Kale and colleagues identifiedinterpreter service and health care providercompetencies as a potential barrier to serviceutilisation, as well as general access issues. 7 BonacruzKazzi et al. 19 found that the main barrier toservice use was the poor identification of need foran interpreter, and Fatahi et al. 20 identified timing,practicalities and interpersonal issues as themain barriers to interpreter service use. Usingin-depth interviews from 20 internal medicineresident physicians from two urban teachinghospitals, Diamond and colleagues identified fourmain barriers to interpreter service use: providersjust ‘getting by’, time constraints, inconveniences,and normalisation of the problem. 16One barrier identified in our study, ‘knowledge’,has received scant attention in the literature. Thiswas one of the primary reasons given for interpreterservice underutilisation, and was due tomany providers being unaware of what systemsexist and how they function. This may reflect, inpart, the relatively recent establishment of interpreterservices within the greater Christchurchregion. Interestingly, while there is a significantliterature on this topic, issues surrounding use oftelephone interpreters as compared to face-to-faceinterpreters were not raised. 21–23While having salient strengths such as themixed method approach, this study also has someimportant limitations. The quantitative estimatesignored patients seeing GPs who both speak thesame non-English language, patients attendinggeneral practices who employ in-house interpreters,or those consultations that do not requireproficient language skills for an effective consultation.These factors are likely to partially off-setthe seemingly vast difference between the identifiedneed and supply of interpreter services. Forthe qualitative component, due to practical limita-136 VOLUME 5 • NUMBER 2 • JUNE 2013 J OURNAL OF PRIMARY HEALTH CARE

ORIGINAL SCIENTIFIC PAPERMIXED METHOD RESEARCHtions we were unable to run a greater number offocus groups and interviews. Ideally, we wouldhave liked to interview more health providers.Therefore, we cannot confirm that data saturationwas achieved in this study and recognise thatthere may be other barriers to interpreter serviceuse that we have not identified. Regardless, it islikely that a significant underutilisation of interpreterservices exists and that some of the keyidentified barriers are malleable to change.To redress this underutilisation, a regular comprehensivetraining and education programmefor health providers is recommended to overcomethe knowledge and perception barriers. Especiallyuseful would be a programme that assistedproviders through their first access of the service,thereby overcoming any initial set-up difficulties.Another recommendation is the need for systemsand policy development to guide the useof interpreter services. Gray and colleagues havedeveloped a toolkit to assist the use of interpretersin general practice. 24 Through a series of flowcharts, clinicians are guided on how to make decisionsabout when and what type of interpreter isneeded for a person with LEP. They also outlinepolicies and processes that will support patientswith LEP. For instance, they suggest that codingfor LEP is a basic requirement for patient records.The current underutilisation of interpreter servicesin Canterbury signifies that we are ignoringbest practice and failing many LEP patients. Inour quest to reduce inequities and improve healthoutcomes, concerted efforts are needed to diminishthe identified barriers for effective communicationand encourage appropriate and efficient useof interpreter services. These efforts will havecapacity and resource implications. In order to redressthe underutilisation of interpreter services,we recommend that priority is given to educationprogrammes on the use of interpreter serviceswithin general practice.References1. Health and Disability Commissioner. Code of Health andDisability Services Consumers’ Rights. Auckland: Health andDisability Commissioner; 2009.2. Street RL, Makoul G, Arora NK, Epstein RM. How doescommunication heal? Pathways linking clinician-patientcommunication to health outcomes. Patient Educ Couns.2009;74(3):295–301.3. Flores G. The impact of medical interpreter services on thequality of health care: a systematic review. Med Care Res Rev.2005;62(3):255–299.4. Karliner LS, Jacobs EA, Chen AH, Mutha S. Do professionalinterpreters improve clinical care for patients with limited Englishproficiency? A systematic review of the literature. HealthServ Res. 2007;42(2):727–754.5. Ministry of Social Development. The Social Report 2010. Wellington:Ministry of Social Development; 2010.6. Gray B, Hilder J, Donaldson H. Why do we not use trainedinterpreters for all patients with limited English proficiency?Is there a place for using family members? Aust J Prim Health.2011;17(3):240–249.7. Kale E, Syed HR. Language barriers and the use of interpretersin the public health services. A questionnaire-based survey.Patient Educ Couns. 2010;81(2):187–191.8. Kuo DZ, O’Connor KG, Flores G, Minkovitz CS. Pediatricians’use of language services for families with limited Englishproficiency. Pediatrics. 2007;119(4):e920–927.9. Statistics New Zealand. 2006 Census Questionnaires. [Cited2012 Aug 9]. Available from: http://www.stats.govt.nz/Census/about-2006-census/2006-census-questionnaires.aspx.10. Allan J-A. Review of the measurement of ethnicity: classificationand issues. Wellington: Statistics New Zealand; 2001.11. Green J, Thorogood N. Qualitative methods for health research.London: Sage; 2005.12. Liamputtong P. Qualitative research methods. South Melbourne:Oxford University Press; 2009.13. Salmond C, Crampton P, Atkinson J. NZDep2006 Index ofDeprivation. University of Otago, Wellington: Department ofPublic Health; 2007.14. Gray B, Stanley J, Stubbe M, Hilder J. Communication difficultieswith limited English proficiency patients: clinicianperceptions of clinical risk and patterns of use of interpreters.N Z Med J. 2011;124(1342):23–38.15. Rose DE, Tisnado DM, Malin JL, Tao ML, Maggard MA, AdamsJ, et al. Use of interpreters by physicians treating limitedEnglish proficient women with breast cancer: results from theprovider survey of the Los Angeles Women’s Health Study.Health Serv Res. 2010;45(1):172–194.16. Diamond LC, Schenker Y, Curry L, Bradley EH, Fernandez A.Getting by: underuse of interpreters by resident physicians. JGen Intern Med. 2009;24(2):256–262.17. Wearn A, Goodyear-Smith F, Everts H, Huggard P. Frequencyand effects of non-English consultations in New Zealandgeneral practice. N Z Med J. 2007;120(1264):U2771.18. Atkin N. Getting the message across-professional interpretersin general practice. Aust Fam Physician. 2008;37(3):174–176.19. Bonacruz Kazzi G, Cooper C. Barriers to the use of interpretersin emergency room paediatric consultations. J Paediatr ChildHealth. 2003;39(4):259–263.20. Fatahi N, Hellström M, Skott C, Mattsson B. General practitioners’views on consultations with interpreters: a triadsituation with complex issues. Scand J Prim Health Care.2008;26(1):40–45.21. Locatis C, Williamson D, Gould-Kabler C, Zone-Smith L,Detzler I, Roberson J, et al. Comparing in-person, video,and telephonic medical interpretation. J Gen Intern Med.2010;25(4):345–350.22. Crossman KL, Wiener E, Roosevelt G, Bajaj L, Hampers LC.Interpreters: telephonic, in-person interpretation and bilingualproviders. Pediatrics. 2010;125(3):e631–638.23. Gany F, Leng J, Shapiro E, Abramson D, Motola I, Shield DC, etal. Patient satisfaction with different interpreting methods: arandomized controlled trial. J Gen Intern Med. 2007;22(Suppl2):312–318.24. Gray B, Hilder J, Stubbe M. How to use interpreters in generalpractice: the development of a New Zealand toolkit. J PrimHealth Care. 2012;4(1):52–61, A1–8.ACKNOWLEDGEMENTSWe would like to thank DrBen Gray (University ofOtago, Wellington) andWayne Reid (PartnershipHealth Canterbury Te Keio Te Waka) for assistingwith this study and to thankthe study participants fortheir time and insights.FUNDINGThe work of KaraSeers was funded byPartnership HealthCanterbury Te Kei o TeWaka through a SummerStudent Scholarship.COMPETING INTERESTSNone declared.VOLUME 5 • NUMBER 2 • JUNE 2013 J OURNAL OF PRIMARY HEALTH CARE 137

ORIGINAL SCIENTIFIC PAPERSHORT REPORTTransient ischaemic attack and stroke risk: pilot ofa primary care electronic decision support toolAnnemarei Ranta MD, FRACPConsultant Neurologistand Lead Stroke Physician,MidCentral Health,Palmerston North, NewZealand, and AssociateDean of UndergraduateMedical Education, Universityof Otago, Wellington, atPalmerston NorthABSTRACTINTRODUCTION: Transient ischaemic attacks (TIAs) indicate high risk for stroke and rapid managementreduces stroke burden. Rapid specialist access to initiate timely management is often challenging toachieve.AIM: To assess the feasibility of implementing a TIA/Stroke electronic decision support (EDS) toolintended to aid general practitioners (GPs) in the timely management of TIAs.METHODS: An eight-week pilot provided access to the TIA/Stroke EDS to selected GPs in the MidCentraldistrict, with subsequent patient record review and a post-pilot user satisfaction survey.RESULTS: Eleven patients from eight practices were entered into the tool and when EDS-renderedadvice was followed, diagnosis was accurate and management was in accordance with New Zealand TIAguidelines. No adverse outcomes resulted and user feedback was positive.DISCUSSION: Results indicate that wider implementation of the TIA/Stroke EDS tool is feasible.KEYWORDS: Decision support systems; primary health care; software; stroke; transient ischaemic attackJ PRIM HEALTH CARE2013;5(2):138–140.CORRESPONDENCE TO:Annemarei RantaDepartment of Neurology,MidCentral Health,PB 11036, PalmerstonNorth 4442, New ZealandAnna.ranta@midcentraldhb.govt.nzIntroductionStroke is the second most common cause of deathworldwide and the most common cause of longtermadult disability in developed countries. 1,2Transient ischaemic attacks (TIA) identify peopleat high risk of stroke. This risk is greatest in thefirst 48 hours and then decreases over time. Thekey intervention that reduces subsequent strokeis same-day specialist review and initiation ofbest medical therapy at first point of contact, 3,4which has been associated with an 80% reductionin 90-day stroke risk from 10.3% to 2.1%. 4Providing 24-hour, seven-days-a-week rapidaccess to stroke specialists is a challenge throughoutNew Zealand and in particular in the smallersizeddistrict health boards (DHBs). To circumventthe problem of limited or delayed access tohospital specialist assessment, the MidCentralStroke Service, in collaboration with the Mid-Central DHB, and the Best Practice AdvocacyCentre Inc. (BPAC Inc.), developed a novel electronicdecision support (EDS) tool to aid generalpractitioners (GPs) in diagnosing, triaging, andtreating patients appropriately and expediently.The tool is based primarily on the New ZealandGuideline for the Assessment and Management ofPeople with Recent Transient Ischaemic Attack 5and its main objective is to prompt initiation ofbest medical therapy at first point of contact inthe community, rather than awaiting potentiallydelayed specialist review at the hospital. In orderto support rapid work-up in the community, GPsalso gain access to relevant diagnostics (e.g. headCT and carotid ultrasound) if deemed appropriateby the EDS tool. The tool is web-based, maintainedby BPAC Inc., and requires access to theMedTech32 practice management system.The purpose of this pilot was to assess the feasibilityof implementing the TIA/Stroke EDS inthe MidCentral DHB primary care sector prior toa district-wide launch.MethodsAt the time of the pilot there were 32 practices inthe MidCentral District using MedTech32. Thispilot involved eight (25%) of the 32 eligible GP138 VOLUME 5 • NUMBER 2 • JUNE 2013 J OURNAL OF PRIMARY HEALTH CARE

ORIGINAL SCIENTIFIC PAPERSHORT REPORTpractices and pilot practices were chosen based onthree factors: adequate numbers of GPs, currentcapability to access best practice EDS modules,and an overall representative patient mix of theMidCentral population. Practices were located inthe provincial centre of Palmerston North andthe smaller nearby town of Feilding. One practiceserves a predominantly Maori population. Practiceswere of median size ranging from two to fiveGPs per practice.The tool itself consists of a web-based data entryform requesting information about the presentingsymptoms and a brief examination. Some entryfields are self-populated through data extractionfrom the practice management system. Enteringpatient data takes approximately three to fiveminutes.The tool then runs the information through analgorithm with three main possible diagnosticoutcomes: (a) stroke, (b) TIA, or (c) ‘non-straightforwardneurological presentation.’The first two are further subdivided by riskcategory and anatomic localisation. Lastly, severaladditional stipulations to ‘diagnosis and triage’advice are provided if (a) the patient is young(

ORIGINAL SCIENTIFIC PAPERSHORT REPORTACKNOWLEDGEMENTSI would like to acknowledgethe support I havereceived from generalpractitioner colleagues inthe MidCentral district.In particular, I wouldlike to acknowledge DrsJonathon Morton andDavid Ayling for theircontribution to this project.COMPETING INTERESTSNone declared.have informed the GP that a diagnosis of TIAwas in fact unlikely, which may have led to arrivingat the correct diagnosis sooner. The secondpatient’s data was entered correctly into the EDStool and was correctly diagnosed by the tool ashaving suffered a stroke rather than a TIA. However,despite the EDS advising the GP to refer thepatient to the Accident & Emergency Department(A&E) for urgent specialist review, general practice–basedmanagement continued. This led toinappropriate delays in diagnostics and precludedtimely access to rehabilitation services. The authoris aware of one additional TIA patient whopresented to this cohort of GPs during the pilotperiod who was not entered into the EDS becauseof local IT difficulties.According to the post-pilot questionnaire, allparticipating GPs who had used the tool were satisfiedwith the TIA/Stroke EDS software and hadno major concerns regarding user-friendliness,time required to enter data, or the overall advicegiven by the tool. A few minor issues were raised,including a request to allow the GP more overrideoptions if the advice given by the tool appearedto be inappropriate. Other comments included amention that some medications were not recognisedby the EDS and a request to add a free-textbox to enter additional information to appear onthe referral form. In addition, some GPs voicedconcerns that A&E staff might turn down referralsfor patients with TIA as they would not bedeemed urgent enough by frontline hospital staff.However, those GPs who in fact used an EDSgeneratedA&E referral to send a patient to theA&E reported that having used the tool actuallyhelped the A&E referral process because it lentextra credence to the GP’s assessment.DiscussionTIAs are medical emergencies requiring urgentintervention in high-risk patients and this novelTIA/Stroke EDS tool is intended to improve appropriatenessand urgency of care. However, priorto launching this tool it was important to ensurethat there were no significant risks to patientsassociated with software use.Overall, this pilot did not identify any areas ofunacceptable risk associated with TIA/StrokeEDS use that would preclude wider implementation.In addition, participant feedback was positiveand suggested that the tool was user-friendlyand seen as potentially beneficial by treating GPs.The request to allow GPs more override optionsis a slightly difficult one. On the one hand, ifsufficient flexibility is not allowed, cliniciansmay see the tool as impinging on their autonomyand may simply not use it. On the other hand,the pilot data indicated that when GPs did notfollow the advice given by the EDS, managementwas less appropriate. To compromise, some additionaloverride options were added to the EDSfollowing the pilot; however, diagnostic accesscontinues to be available only for patients deemedto require them by the EDS tool. In addition, GPshave to enter a reason for overriding the adviceand are continuously reminded that they areveering away from the suggested and guidelinebasedtreatment plan.In conclusion, this pilot was judged sufficient toindicate acceptable usability and safety and theTIA/Stroke EDS has been launched in the Mid-Central District. Based on this pilot and preliminaryresults from district-wide post-implementationevaluations, the Health Research Councilhas funded a randomised controlled trial comparingEDS versus non-EDS assisted TIA managementin a number of New Zealand DHBs. Thistrial (FASTEST Trial: ACTRN12611000792921)is currently underway to assess feasibility of nationwidelaunch of this software tool and resultswill be available next year.References1. Johnston SC, Mendis S, Mathers CD. Global variation in strokeburden and mortality: estimates from monitoring, surveillance,and modelling. Lancet Neurol. 2009;8:345–354.2. Rothwell PM. The high cost of not funding stroke research:a comparison with heart disease and cancer. Lancet.2001;357:1612–1616.3. Lavallee PC, Meseguer E, Abboud H, Cabrejo L, Olivot JM,Simon O, et al. A transient ischaemic attack clinic with roundthe-clockaccess (SOS-TIA): feasibility and effects. LancetNeurol. 2007;6:953–960.4. Rothwell PM, Giles MF, Chandratheva A, Marquardt L,Geraghty O, Redgrave JN, et al. Effect of urgent treatment oftransient ischaemic attack and minor stroke on early recurrentstroke (EXPRESS study): a prospective population-basedsequential comparison. Lancet. 2007;370:1432–1442.5. Stroke Foundation of New Zealand. New Zealand guidelinefor the assessment and management of people with recenttransient ischaemic attack. Wellington: Stroke Foundation ofNew Zealand Inc; 2008.140 VOLUME 5 • NUMBER 2 • JUNE 2013 J OURNAL OF PRIMARY HEALTH CARE

ORIGINAL SCIENTIFIC PAPERSHORT REPORTLarge increase in opportunistic testing forchlamydia during a pilot project in a primaryhealth organisationSunita Azariah MBChB, FAChSHM, DPH; 1 Stephen McKernon MA, MSc, M DM; 2 Suzanne WerderRCpN, PGCertHealSc, PGCertProfSup 1ABSTRACTINTRODUCTION: The Auckland chlamydia pilot project was one of three funded by the Ministry ofHealth to trial implementation of the 2008 Chlamydia Management Guidelines. Chlamydia is the mostcommonly notified sexually transmitted infection in New Zealand.1Auckland Sexual HealthService, Auckland,New Zealand2Supplejack Ltd, Pt Chevalier,AucklandAIM: To increase opportunistic testing in under-25-year-olds and to improve documentation of partnernotification in primary care.METHODS: A four-month pilot was initiated in Total Healthcare Otara using a nurse-led approach. Laboratorytesting data was analysed to assess whether the pilot had any impact on chlamydia testing volumesin the target age-group. Data entered in the practice management system was used to assess follow-upand management of chlamydia cases.RESULTS: During the pilot there was a 300% increase in the number of chlamydia tests in the target agegroupfrom 812 to 2410 and the number of male tests increased by nearly 500%. Twenty-four percent ofpeople tested were positive for chlamydia, with no significant difference in prevalence by ethnicity. Thepilot resulted in better documentation of patient follow-up in the patient management system.DISCUSSION: There was a large increase in chlamydia testing during the pilot with a high prevalencefound in the population tested. Chlamydia remains an important health problem in New Zealand. Thecost benefit of increased chlamydia screening at a population level has yet to be established.KEYWORDS: Chlamydia; notification, partner; pilot project; prevalence; primary health careIntroductionThe Auckland chlamydia pilot was one of threepilots funded by the Ministry of Health to trialimplementation of the 2008 Chlamydia ManagementGuidelines. 1 Chlamydia is the most commonlynotified sexually transmitted infection(STI) in New Zealand 2 and can result in significantadverse sequelae including peri-hepatitis andpelvic inflammatory disease (PID). 3–5 EnvironmentalScience and Research services (ESR) laboratorysurveillance data indicate that over 70%of chlamydia cases are diagnosed in those agedunder 25 years and rates of diagnosis are higherin those of Maori and Pacific ethnicity. 2 TheESR data also indicate disproportionate numbersof chlamydia cases are diagnosed in females. Datafrom the Waikato pilot confirmed a lower testuptake in males. 6The Auckland pilot had multiple aims and objectives.This paper describes outcomes of two aims:firstly, to increase opportunistic testing for chlamydiain those aged under 25 years (particularlymales) and, secondly, to improve documentationof follow-up and partner notification of diagnosedcases. As previous research has indicatedthat nurse-led opportunistic testing is very successfulat increasing chlamydia testing rates, 7 weopted to utilise this approach.J PRIM HEALTH CARE2013;5(2):141–145.CORRESPONDENCE TO:Sunita AzariahAuckland Sexual HealthService, Building 7,Greenlane Clinical Centre,Auckland District HealthBoard, Greenlane West,PB 92024, Auckland,New ZealandSunitaA@adhb.govt.nzVOLUME 5 • NUMBER 2 • JUNE 2013 J OURNAL OF PRIMARY HEALTH CARE 141

ORIGINAL SCIENTIFIC PAPERSHORT REPORTTable 1. Chlamydia testing: laboratory dataAllN (%)Malen (%)Pre-pilotFemalen (%)Age14–19n (%)Age20–24n (%)AllN (%)Malen (%)PilotFemalen (%)Age14–19n (%)Age20–24n (%)Total no. tests 812 (100) 144 (17.7) 668 (82.3) 281 (34.6) 509 (62.7) 2410 (100) 695 (28.8) 1715 (71.2) 916 (38.0) 1466 (60.8)No. individualstestedNo. positivetests% individualspositive** After exclusion of duplicates.760 (93.6) 130 (90.3) 630 (94.3) 254 (90.4) 484 (95.1) 2157 (89.5) 658 (94.7) 1499 (87.4) 815 (89.0) 1315 (89.7)194 (23.9) 59 (41.0) 135 (20.2) 83 (32.7) 109 (22.5) 618 (76.1) 155 (21.3) 370 (22.0) 251 (30.8) 262 (19.9)25.5 45.4 21.4 32.7 22.5 24.3 23.6 24.7 30.8 19.9MethodsA South Auckland primary care setting was chosenfor the pilot because of the youthful populationdemographics and the relatively high proportion ofyoung Maori and Pacific people in the region. TotalHealthcare Otara (THO) agreed to implementa four-month pilot project in their 10 primary carepractices. The limited time period was imposeddue to Ministry of Health time constraints.Chlamydia testing in primary care is alreadyknown to be acceptable and to be regardedpositively by young people if given in theright environment.THO operates by a nurse-triage process and duringthe pilot period a chlamydia test was to be offeredby the nurses to all sexually active under-25-yearolds,using criteria recommended in the chlamydiamanagement guidelines. If consent was given,males provided a first-pass urine specimen andfemales obtained a self-collected vaginal swab fortesting. Laminated instruction cards were providedto explain how to collect their vaginal sample.All patients testing positive for chlamydia wereto be recalled by a registered nurse for treatmentand to discuss partner notification. It was recommendedthat all cases should receive a follow-uptelephone call one week after treatment to checkwhether there had been any risk of re-infection,to check whether sexual contacts had been notifiedand to offer a repeat chlamydia test in threemonths’ time.Laboratory testing data was obtained from thecommunity laboratory. A reference sample of datafrom a four-month period 12 months prior to theintroduction of the pilot (pre-pilot period) and asecond set of chlamydia testing data during thepilot implementation period (pilot period) wasrequested for all tests processed in the target agegroup. The laboratory data included age and gender,but not ethnicity. During the pilot period,ethnicity data was extracted from the practicemanagement system (PMS); THO had introduceda new MedTech32 template to the PMS specificallyfor the pilot.A positive chlamydia test was only counted onceif there was more than one positive test for an individualwithin a one-month period (as PCR testscan remain positive for several weeks after treatment)and individuals were counted only once ifthey were re-tested within a one-month period.There was some contamination of the laboratorydata, as the reports also contained data from threepractices in the same region not included in thepilot because they used the same laboratory referencecode.Statistical analysisThe two-proportion z-score test was used todetermine whether observed differences between142 VOLUME 5 • NUMBER 2 • JUNE 2013 J OURNAL OF PRIMARY HEALTH CARE

ORIGINAL SCIENTIFIC PAPERSHORT REPORTthe pre-pilot and pilot data were significant. Probabilityvalues were calculated from the z-scoresusing a normal distribution calculator and asignificance level of 0.05 was used.Ethical approval was obtained for the projectfrom the Northern Regional Ethics Committee(Ref. NTX/10/EXP/169). The pilot project implementationcommenced 6 December 2010 andfinished 31 March 2011.ResultsDuring the pre-pilot period, THO provided consultationsto 2746 patients in the target age groupand 812 chlamydia tests were processed by thelaboratory, which corresponded to 760 individualsafter exclusion of duplicates. Table 1 showsthe laboratory testing data. As noted in themethods section, some of these individuals wouldhave been tested outside of the pilot setting. Themajority of chlamydia tests were requested infemales (82%) and, overall, 26% of individualstested were positive. The percentage of positivetests in males was nearly double that of females.Males and females in the 14- to 19-year age grouptested positive more frequently than those in the20- to 24-year age group, but this difference wasonly significant for females (p=0.006).During the pilot period, data from the PMS indicatedthat 3687 patients in the target age groupwere triaged by THO and 1715 of these patients(46%) were tested for chlamydia. The numbersof chlamydia tests processed by the laboratoryincreased dramatically from 812 tests to 2410(just over 300%) and the number of male testsincreased by nearly 500%, from 144 to 695 tests.The proportion of male tests increased from 18%to 29% of total tests. The PMS data revealed thatWHAT GAP THIS FILLSWhat we already know: Chlamydia is the commonest sexually transmittedinfection notified in New Zealand. Opportunistic testing is recommendedin the under-25-year-old age group. However, test uptake in males is muchlower than females in primary health care settings.What this study adds: Nurse-led opportunistic testing for chlamydiain primary health care is successful at increasing testing in both males andfemales. It is possible to improve documentation of partner notification inprimary health care, but further research is needed in this area.a similar proportion of males (47%) and females(45%) presenting to THO during the pilot periodwere tested for chlamydia. Overall, 24% of peopletested were positive which was similar to thepre-pilot period. The test positivity rate in males,however, declined considerably from pre-pilotlevels (p

ORIGINAL SCIENTIFIC PAPERSHORT REPORTthan urine specimens; prior to the pilot, manyTHO clinical staff had thought this approach totesting would not be acceptable to young women.The numbers of male urine specimens increasedfrom 12% to 27% of total specimens (p

ORIGINAL SCIENTIFIC PAPERSHORT REPORTWhilst the major shift from diagnostic to opportunistictesting during the pilot project wasassociated with a decreased prevalence of positivetests, particularly in males, overall in the under-25-year age group, the positive test rate was morethan twofold higher than the 9% reported in the2011 ESR laboratory surveillance data. 2 Femalesaccounted for the majority of chlamydia tests inthe pilot and this was a similar finding to testingpatterns in other primary care settings. Datafrom the much bigger Waikato pilot found therewas much lower chlamydia test uptake in malesthan in females. 6 In contrast to the Waikato pilotand other data, 2 we did not find the prevalenceof chlamydia to be higher in those of Maori andPacific ethnicity compared with Europeans.There was better documentation of partner notificationand follow-up during the pilot and thiswas encouraging. This is an area of case managementof STIs that urgently needs addressing. 11It should not be too difficult to improve withappropriate training and systems, as it has beenshown that trained practice nurses can achieve asgood outcomes for partner notification for chlamydiaas specialist sexual health clinic advisors. 12Good follow-up of cases is important, as there isa high rate of re-infection. 13In conclusion, there was a large increase inchlamydia testing during the pilot project, but itis doubtful that this is sustainable and the costbenefit of the testing has yet to be established. 14However, the very high prevalence of chlamydiaand recent evidence of much higher rates ofhospital admissions for chlamydia-related PID inNew Zealand compared with other countries, 15indicates chlamydia continues to be a very significanthealth problem for young people presentingto primary care.References1. Ministry of Health. Chlamydia Management Guidelines. Wellington:Ministry of Health; 2008.2. Institute of Environmental Science and Research Ltd. Sexuallytransmitted infections in New Zealand: Annual surveillancereport 2011. Porirua, New Zealand: Institute of EnvironmentalScience and Research Ltd; 2012.3. Mardh PA, Ripa T, Svensson L, Westrom L. Chlamydia trachomatisinfection in patients with acute salpingitis. N Engl JMed. 1977;296:1377–1379.4. Wolner-Hanssen P, Westrom L, Mardh PA. Peri-hepatitis andchlamydial salpingitis. Lancet. 1980;1(8174):901–3.5. Morgan J, Colonne C, Bell A. Trends of reported chlamydiainfections and related complications in New Zealand,1998–2008. Sex Health. 2011;8(3):412–8.6. Morgan J, Bell A. The highs and lows of opportunistic chlamydiatesting: uptake and detection in Waikato, New Zealand.Sex Transm Infect. 2009;85(6):452–454.7. Lawton BA, Rose SB, Elley CR, Bromhead C, MacDonald EJ,Baker MC. Increasing the uptake of opportunistic chlamydiascreening: a pilot study in general practice. J Prim Health Care.2010;2(3):199–207.8. Hogan AH, Howell-Jones RS, Pottinger E, Wallace LM, Mc-Nulty CAM. ‘…they should be offering it’: a qualitative study ofyoung peoples’ attitudes towards chlamydia screening in GPsurgeries. BMC Public Health. 2010;10:616.9. Rose SB, Lawton BA, Bromhead C, MacDonald EJ, Lund KA.Self-obtained vaginal swabs for PCR chlamydia testing: a practicalalternative. Aust NZ J Obstet Gynaecol. 2007;47:415–418.10. Morgan J, Haar J. General practice funding to improve provisionof adolescent primary sexual health care in New Zealand:results from an observational intervention. Sex Health.2009;6(3):203–207.11. Morgan J, Donnell A, Bell A. Is everyone treated equally?Management of genital Chlamydia trachomatis infection inNew Zealand. Int J STD AIDS. 2010;21:595–600.12. Low N, McCarthy A, Roberts TE, Huengsberg M, Sanford E,Sterne JA, et al. Partner notification of chlamydia infectionin primary care: randomised controlled trial and analysis ofresource use. BMJ. 2006;332(7532):14–19.13. Niccolai LM, Livingston KA, Laufer AS, Pettigrew MM.Behavioural sources of repeat Chlamydia trachomatis infections:importance of different sex partners. Sex Transm Infect.2011;87:248–253.14. Low N. Screening programmes for chlamydial infection: whenwill we ever learn? BMJ. 2007;334:725–728.15. Bender N, Herrmann B, Andersen B, Hocking JS, van Bergen J,Morgan J, et al. Chlamydia infection, pelvic inflammatory disease,ectopic pregnancy and infertility: cross-national study.Sex Transm Infect. 2011;87:601–608.ACKNOWLEDGMENTSThe authors would liketo thank the staff of TotalHealthcare Otara for theirparticipation in this pilot,in particular, RichardHulme (Clinical Director)and Gillian Davies (NurseLeader). We wouldalso like to thank ourproject manager KarenHolland for her steerageof this pilot project.FUNDINGThis pilot project wasfunded by the NewZealand Ministry of Health.COMPETING INTERESTSNone declared.VOLUME 5 • NUMBER 2 • JUNE 2013 J OURNAL OF PRIMARY HEALTH CARE 145

ORIGINAL SCIENTIFIC PAPERSHORT REPORTPersonal medicines storage in New ZealandCampbell Hewson; 1 Chong Chi Shen; 1 Clare Strachan PhD, 1,2 Pauline Norris PhD 11School of Pharmacy,University of Otago, Dunedin,New Zealand2Division of PharmaceuticalTechnology, Faculty ofPharmacy, University ofHelsinki, FinlandABSTRACTINTRODUCTION: Poor storage of medicines can reduce their efficacy, yet little is known about howpeople store medicines in their homes and elsewhere, why these locations are chosen, and whether theconditions are suitable for medicines storage.AIM: To investigate where medicines are commonly stored in New Zealand households, why, and thetypical conditions—temperature and relative humidity—in those places of storage.METHODS: Data from a large qualitative study on the meanings of medicines were analysed to explorewhere people store medicines in their households, and why. A data logger was used to log temperatureand relative humidity in common medicine storage places, such as homes and cars.RESULTS: Kitchens and bathrooms were the most commonly reported storage places, with peopleinfluenced by convenience, desire to remember to take medicines, and child safety when deciding whereto store medicines. High temperatures and humidity were found in kitchens and bathrooms, extremetemperatures in a car and a backpack, and extremely low temperatures in checked-in luggage on a plane.DISCUSSION: Temperature- and humidity-sensitive medicines should not be stored long-term in commonstorage locations, such as kitchens and bathrooms. Conditions in these places may not comply withthe recommended storage conditions given by the manufacturer. Furthermore, medicines should not beleft in backpacks or cars, especially if the vehicle is in the sun. Medicines that may degrade upon freezingand thawing—such as protein-containing medicines, emulsions, suspensions and some solutions—should not be stored in the cargo hold of a plane.KEYWORDS: Drug storage; humidity; New Zealand; temperatureJ PRIM HEALTH CARE2013;5(2):146–150.CORRESPONDENCE TO:Pauline NorrisProfessor, School ofPharmacy, Universityof Otago, PO Box 56,Dunedin 9054,New Zealandpauline.norris@otago.ac.nzIntroductionAppropriate storage is essential to ensure thesafety, quality and efficacy of medicines. However,apart from a few studies in developingcountries, 1–4 there is little research on where peoplestore medicines and why, and the conditionsin these common locations.The shelf life of a medicine established by amanufacturer only applies when products arestored under conditions outlined on the label. 5According to guidelines, medicines should alsobe stored out of reach and sight of children, inorder to prevent unintentional poisonings. 6,7Medicines should be stored in their originalpackaging. Medicines to be stored at room temperatureshould be kept in a well-ventilated place,between 15 and 25°C, or up to 30°C, dependingon climatic zone. 8Temperature greatly influences medicine degradationchemically, physically and microbiologically.Chemical degradation, such as oxidationor hydrolysis, increases with temperature, oftenexponentially. 9 Physical degradation occurs atboth high and low temperatures, and may beirreversible. Proteins may denature and aggregateat high temperatures, or because of freezing.Emulsions can crack and, in solutions, medicinesmay precipitate at low temperatures and particlesin suspensions may grow, making them hard tore-suspend upon shaking. 9,10146 VOLUME 5 • NUMBER 2 • JUNE 2013 J OURNAL OF PRIMARY HEALTH CARE

ORIGINAL SCIENTIFIC PAPERSHORT REPORTMoisture can also be a problem for medicinestability. Exposure to high humidity increaseschemical degradation of water-labile medicinesand excipients, for example cellulose. 11 Allplastics used in medicine packaging are somewhatpermeable to water vapour. Moisture alsopromotes the growth of microbes, 11 particularlyin hot conditions (e.g. greater than 30°C and 75%relative humidity).Studies in other settings have found problemswith medicines storage. Temperatures in emergencymedical vehicles, such as helicopters andambulances, were found to exceed recommendedstorage conditions on hot days, 12,13 as did thosein a doctor’s bag in two cars. 10 This study wasundertaken to investigate where people in NewZealand store their medicines and why, and todescribe the conditions (temperature and relativehumidity) in such places.MethodsPersonal medicine storage in householdsData from the ongoing project Medications inEveryday Life: Understandings and Social Practiceswas used to investigate where people store theirmedicines and why. Transcripts of 36 householdinterviews that included questions about medicinesused and where they were stored, and mapsand photographs of storage locations were used.Households were located in four cities in NewZealand. The study methods are further describedelsewhere. 14 Information relevant to medicinestorage locations and reasons for these locationswas extracted from transcripts. A count of howmany rooms in each household had medicines inthem was produced from the household maps.Personal medicine storage conditionsTwo data loggers (Hobo U10 TemperatureRelative Humidity Data Logger, U10-003 andH8-001-02. Onset Computer Corporation, CapeCod, USA) were placed in six different locationsthat were identified as common storage places(but not in the households in the Medicationsin Everyday Life study). Three kitchens, twobathrooms and one bedroom were used; three inDunedin and three in Palmerston North. TheWHAT GAP THIS FILLSWhat we already know: Excessive heat and humidity can damage medicinesand reduce their efficacy. Studies in developing countries and in healthcare settings have shown that storage conditions are sometimes suboptimal.What this study adds: People in New Zealand often store medicines inkitchens, bathrooms, and also sometimes store them in cars and carry themin bags. For convenience, and in order to remember to take their medicines,New Zealanders often store medicines in places where excessive heat andhumidity may cause problems with medicines.duration of placement varied from one day to sixdays. Measurements were taken every 10 to 15minutes. Since many of the participants in thefirst part of the study mentioned forms of mobilestorage of medicines, such as handbags and cars,data loggers were also placed in a backpack inthe sun in Dunedin, two cars in Dunedin, andsuitcases in the cargo hold of planes flying fromDunedin to Christchurch, Palmerston North, andCopenhagen.ResultsMedicine storage in householdsParticipants were 104 people (40% men and 60%women), with an average age of 34 years. Arange of ages, ethnic groups and households ofdifferent sizes and composition were included.The kitchen was the most common location inthe house for storage of medicines (33 householdshad medicine in the kitchen), followed by thebathroom (21 households) and the bedroom (19households). Forms of mobile storage, such ashandbags and backpacks were also common (18households). Other storage locations included acar glove box, under a mattress, and in a garage.The average number of rooms where medicineswas stored per household was three.Reasons given for storage decisions included:convenience, as a cue to remember to take them,and safety of children. Many participants saidthey stored their medicines in particular locationsbecause of convenience. This was oftenbased on their daily routine and where theywould be when the medicines needed to be taken.VOLUME 5 • NUMBER 2 • JUNE 2013 J OURNAL OF PRIMARY HEALTH CARE 147

ORIGINAL SCIENTIFIC PAPERSHORT REPORTMedicines to be consumed with food were oftenstored in particular locations in the kitchen tomake them convenient to take at a meal time.Medicines needed to be taken first thing in themorning were stored in the bedroom.Placement of medicines was often used as a cueto remember to take them. Many people putmedicines in a place that was part of their dailyroutine in order to be reminded to take them.Studies in other settings have found problemswith medicines storage. Temperatures inemergency medical vehicles, such as helicoptersand ambulances, were found to exceedrecommended storage conditions on hot days,as did those in a doctor’s bag in two cars.For most households with children, storingmedicines out of children’s reach was also an importantconsideration. Few households mentionedtemperature or humidity as considerations inmedicine storage.The use of some form of intermediary storagewas common, such as putting the next dose of amedicine into another container in preparationfor consumption. This both reminded them totake the medicines, and allowed them to checkwhether they had taken them. Some peoplemixed several medicines together to create ahomemade daily or weekly medicine dispenser.Medicine storage conditionsTable 1 presents the temperature and relative humidity(RH) results for the household locationsstudied. Although the mean conditions of all thestorage places studied were similar, the maximumand minimum conditions differed markedly.Kitchens had the highest maximum temperature,and bathrooms had the highest maximumhumidity. The bedrooms had relatively constanttemperature and humidity conditions.Table 2 shows the conditions in mobile storage locations.Temperatures in the backpack rose above60°C after only around 20 minutes in the sunaround midday. The storage temperature reached54.5°C in one of the cars when in the sun. Thestudy was carried out on mild Dunedin summerdays. 15,16Although the temperature of the data loggerdid not drop very much during the flights toChristchurch and Palmerston North (possiblybecause of clothing acting as insulation), on thelong sequence of flights from Dunedin to Copenhagen,the temperature dropped to -3.32°C.Table 1. Temperature and relative humidity (RH) in different household locations and outside temperatures for the sametime periodTemperature and humidity °C (RH) Outside temperature °CCity Location Maximum Minimum Mean Maximum MinimumPalmerston NorthDunedinDunedinPalmerston NorthPalmerston NorthDunedinBathroomBathroomBedroomKitchen(cabinet)Kitchen(above the oven)Kitchen31.5(100)26.0(95.0)23.5(69.2)27.1(85.2)32.8(57.5)36.3(81.6)20.6(33.0)13.8(51.4)18.8(50.1)18.7(36.2)18.7(27.2)16.0(34.5)23.6(47.1)18.4(70.2)21.7(59.2)21.9(47.2)23.5(45.0)21.0(68.2)22.4 10.920.0 14.722.2 8.620.2 11.523.5 13.524.9 14.8148 VOLUME 5 • NUMBER 2 • JUNE 2013 J OURNAL OF PRIMARY HEALTH CARE

ORIGINAL SCIENTIFIC PAPERSHORT REPORTTable 2. Storage conditions in mobile storage and outside temperatures for the same time periodTemperature and humidity °C (RH) Outside temperature °CCity Location Maximum Minimum Mean Maximum MinimumDunedinBackpack(in sun)67.3(51.5)18.0(15.0)36.4(22.1)21.8 9.5DunedinCar(silver colour)42.5(75.7)14.2(15.0)24.3(40.7)21.5 15.9DunedinCar (dark greencolour)54.5(66.1)7.1(15.0)19.7(46.4)17.8 11.8Dunedin toChristchurchand return flightsSuitcase (cargohold of plane)26.1(70.4)22.1(63.1)24.2(66.6)– –Dunedin toPalmerston Northand return flightsSuitcase (cargohold of plane)25.6(49.3)14.5(34.2)20.4(42.1)– –Dunedin toCopenhagenand return flightsSuitcase (cargohold of plane)23.4(63.8)-3.3(42.0)15.0(52.0)– –DiscussionFew people in this study mentioned temperatureor humidity as considerations in where tostore their medicines. Our study suggests thatmedicines may be stored in unsuitable places.Kitchens and bathrooms are likely to be unsuitableplaces for long-term storage of medicines,and cars and bags left in the sun can reachextremely high temperatures. These can rapidlylead to significant chemical and physical degradationof medicines. 10 Protein-containing productsare especially prone to degradation at temperatureextremes, with insulin—especially in the rapidactingsolution form—potentially almost completelydenaturing within hours at the observedmaximum temperature. 17 Very high temperatures,such as those of the backpack (above 60°C), mayalso affect the integrity of the packaging of themedicines. 12The extremely low temperatures of checkedluggage on aeroplanes could affect the physicalstability of some liquid preparations, such assolutions, suspensions and emulsions. If temperaturesbecame sufficiently cold for protein-basedformulations to freeze, this would very likelylead to protein denaturation. 10In the first part of the study, data from a small,non-random sample of participants was used.This was sufficient to generate ideas about wherepeople store medicines and why, but resultsshould be interpreted with caution and a largerquantitative study would be needed if it wasimportant to establish population rates for eachstorage location or reason. The number of storagelocations that were studied was limited and theduration of the study was short, therefore thestorage conditions measured in the studied locationsmay not be typical.We recommend that temperature- and humiditysensitiveitems should not be stored long-term inkitchens and bathrooms. Medicines, especiallyprotein-based medicines, should not be stored inbags or cars for any longer than strictly necessary,and care should be taken to keep these out of thesun. Medicines susceptible to freezing—such asemulsions, solutions and suspensions—shouldnot be stored in the cargo hold of a plane on longflights. Health care professionals, particularlypharmacists, should advise people where to storemedicines, in order to ensure their safety, qualityand efficacy, and poor storage should be consideredas a possible cause when medicines are notproviding the desired effect. More research isVOLUME 5 • NUMBER 2 • JUNE 2013 J OURNAL OF PRIMARY HEALTH CARE 149

ORIGINAL SCIENTIFIC PAPERSHORT REPORTneeded, in a wider variety of locations, to establishwhether conditions in usual storage locationsactually reduce the effectiveness of medicines.This should focus on high-risk medicines, suchas insulin, and be used to guide a broader publiceducation campaign about medicines storage.ACKNOWLEDGEMENTSWe would like tothank Professors KerryChamberlain, DarrinHodgetts, Kevin Dew,Dr Linda Nikora, and HelenMadden for allowingaccess to this data and MrRichard German, FacultyLibrarian, Health SciencesLibrary, University ofOtago, for his help withliterature searching.FUNDINGThis project was fundedby the New ZealandPharmacy Education andResearch Foundation(NZPERF), and DouglasPharmaceuticals. TheMedications in EverydayLife: Understandings andSocial Practices projectwas funded by the HealthResearch Council andthe Marsden Fund.References1. Obitte N, Chukwu A, Odimegwu D, Nwoke V. Surveyof drug storage practice in homes, hospitals and patentmedicine stores in Nsukka, Nigeria. Scientific Res Essays.2009;4:1354–9.2. Kiyingi KS, Lauwo JA. Drugs in the home: danger and waste.World Health Forum. 1993;14:381–4.3. Yousif M. In-home drug storage and utilization habits: a Sudanesestudy. East Mediterr Health J. 2002;8:422.4. Jassim AM. In-home drug storage and self-medication withantimicrobial drugs in Basrah, Iraq. Oman Med J. 2010;25:79.5. World Health Organization (WHO). Quality Assurance ofPharmaceuticals. A compendium of guidelines and relatedmaterials. WHO: Geneva, Switzerland; 1997.6. Medsafe: New Zealand Medicines and Medical Devices SafetyAuthority. Safe use of medicines. [Cited 2011 Jan 27]. Availablefrom: http://www.medsafe.govt.nz/consumers/safe.asp7. Therapeutic Goods Administration (TGA). You and yourhealthcare products. [Cited 2011 Jan 27]. Available from:http://www.tga.gov.au/meds/healthcare.htm8. Kopp S. Stability testing of pharmaceutical products in a globalenvironment. Regul Aff J. 2006:291–4.9. Sinko PJ, Martin AN. Martin’s physical pharmacy and pharmaceuticalsciences: physical chemical and biopharmaceuticalprinciples in the pharmaceutical sciences. 5th ed. Philadelphia:Lippincott Williams & Wilkins; 2006, p.xiv, 794.10. Crichton B. Keep in a cool place: exposure of medicines tohigh temperatures in general practice during a British heatwave.J R Soc Med. 2004;97:328–9.11. Rhodes CT, Carstensen JT. Drug stability: principles andpractices. 3rd ed. New York: Marcel Dekker, 2000, p.vii, 773.12. Helm M, Castner T, Lampl L. Environmental temperaturestress on drugs in prehospital emergency medical service.Acta Anaesthesiol Scand. 2003;47:425–9.13. Brown LH, Krumperman K, Fullagar CJ. Out-of-hospital medicationstorage temperatures: a review of the literature anddirections for the future. Prehosp Emerg Care. 2004;8:200–6.14. Chamberlain K, Madden H, Gabe J, Dew K, Norris P. Formsof resistance to medications within New Zealand households.Medische Antropologie. 2011;23:299.15. Meteorological Service of New Zealand Ltd. Metservice:Ten Day Forecast. 2011; [cited 2011 Dec 25]. Available from:http://www.metservice.com/towns-cities/dunedin16. National Institute of Water and Atmospheric Research(NIWA). The National Climate Database. [Cited 2011 Jan31]. Available from: http://cliflo.niwa.co.nz/pls/niwp/wgenf.genform117. Oliva A, Fariña JB, Llabrés M. Influence of temperature andshaking on stability of insulin preparations: degradation kinetics.Int J Pharm. 1996;143:163–70.COMPETING INTERESTSNone declared.150 VOLUME 5 • NUMBER 2 • JUNE 2013 J OURNAL OF PRIMARY HEALTH CARE

ORIGINAL SCIENTIFIC PAPERSHORT REPORTAbortion services in a high-needs district:a community-based model of careSimon Snook MBChB, MRCGP (dist), FRNZCGP, DFFP, DRCOG; 1 Martha Silva MPH PhD 2ABSTRACTINTRODUCTION: In 2009, a high-deprivation district health board in New Zealand set up a communitybasedabortion clinic in order to provide a local service and to avoid out-of-region referrals. The serviceoffers medical abortions for women with pregnancies of up to 63 days’ gestation, and surgical abortionwith local anaesthetic for women with pregnancies of up to 14 weeks’ gestation.1Clinical Lead, CommunityClinic for Sexual Health andContraception, Gisborne,New Zealand2Health Researcher, TheUniversity of Auckland,Auckland, New ZealandAIM: To describe the services developed and assess safety and timeliness for the first year of community-basedservices.METHODS: An audit of clinical records for patients seen in 2010 was performed in order to obtain dataon location of services, timeliness, safety and complications.RESULTS: Eighty-two percent of locally provided abortions in 2010 were medical abortions, completedon average less than two days after referral to the service. One percent of patients experienced haemorrhagingpost abortion, and 4% had retained products. These rates are within accepted standards for anabortion service.DISCUSSION: This report illustrates that a community-based model of care can be both clinically andculturally safe, while providing a much-needed service to a high-needs population.KEYWORDS: Abortion, induced; community health services; delivery of health care; New ZealandIntroductionAbortion services in New Zealand are designatedas a core service that is publicly funded andmust be made accessible by all district healthboards (DHBs) around the country. 1 Despite this,abortion services have tended to concentrate inlarger urban areas, forcing many women to travelconsiderable distances to access these servicesoutside their communities. 2 In 2009, a local abortionservice was established in a high-deprivationDHB. This report aims to describe the community-basedabortion service offered by this DHB,and show through the data of its first full year ofoperation that a community-based population responsiveservice can offer safe and timely servicesto a high-needs community.The DHB’s population is mostly urban, with alarge Maori population (44%), and with the highestlevel of deprivation of any other district, withtwo thirds of the population (65%) living in decile1–3 (with decile 1 indicating the highest level ofdeprivation on a scale of 1 to 10). 3 This trend isfurther exacerbated when split by ethnicity, with77% of Maori living within deciles 1–3.Until the opening of the local abortion servicewomen were transferred out of region for theirabortion.The community clinicThe community clinic is a small clinic set in acommercial area near the centre of town offeringsexual health and contraception services. Prior tothe establishment of the local abortion service,patients considering termination of pregnancywould visit the community clinic either directlyor via referral from a general practitioner (GP).J PRIM HEALTH CARE2013;5(2):151–153.CORRESPONDENCE TO:Simon SnookPO Box 22, Greytown.Wairarapa 5712,New Zealandsimon@snip.co.nzVOLUME 5 • NUMBER 2 • JUNE 2013 J OURNAL OF PRIMARY HEALTH CARE 151

ORIGINAL SCIENTIFIC PAPERSHORT REPORTThe community clinic would organise counsellingand referral to the abortion service, ensureall clinical investigations required, includingblood tests, microbiological assessment, andreferral for gestational dating ultrasound. Themidwife/counsellor would then liaise with thepatient/whanau (family/support) with regard totravel assistance. Post-abortion counselling andcare was also offered locally.After the abortion service was established withinthe community clinic in 2009, patient flowremained the same through the early stages, butabortion was offered locally. For women preferringmedical abortion, second certification iseither undertaken by telephone or in person bythe visiting doctor completing the abortion.Surgical abortion with local anaesthetic is offeredby a visiting doctor one morning a week, threeweeks out of four, for women with pregnanciesup to 14 weeks’ gestation. The procedure is undertakenin a consultation room on a gynaecologicalexamination couch. During their wait beforeand after the procedure, each woman has privateuse of one of the other consultation rooms inwhich they can rest in a reclining chair.Medical abortion is offered to women with pregnanciesup to 63 days’ gestation. Mifepristone isgiven initially, followed by 800 mcg misoprostolgiven via the buccal route 24 hours later. Womengenerally return home after the misoprostolfor abortion to occur. Follow-up of patients formedical abortion is primarily undertaken by amidwife/counsellor, who telephones all patientson the day of misoprostol administration and theday following abortion at a minimum. There isalso active follow-up by the midwife/counsellorto ensure that post-abortion serum beta-HCG(human chorionic gonadotropin) is undertakento confirm abortion. Secondary medical adviceis offered by telephone by the visiting doctorwho completed the abortion via a contact numbergiven to all patients. The patients also have theoption of visiting the clinic after abortion formedical or counselling follow-up. The clinic offersfull contraception assistance to all patientsattending for abortion, including the fitting ofintrauterine devices (IUDs) or contraceptiveimplants.MethodsA manual record search was performed on thenotes of all women entering the abortion servicein 2010. Data included date of presentation; dateof termination; method of termination; laboratoryresults, including beta-HCG and ferritin;and complications.A search was done for each patient for hospitaladmissions since termination, ultrasound scanof uterus and laboratory investigations. This wasperformed to ensure that complications that didnot present at the clinic and that had not beenreported to the clinic would not be omitted.There is a small possibility that complicationsthat occurred out of region may have been missedfrom this audit.ResultsA total of 180 women from the DHB had anabortion in 2010. Of these, 81% (n=145) had anabortion locally, while the remaining 19% wentoutside the region for the service. Of the patientswho had an abortion at the local clinic, the major-Table 1. Safety of medical termination of pregnancyTairawhiti DHB Accepted Standard (RANZCOG)Blood loss requiring transfusion 0%

ORIGINAL SCIENTIFIC PAPERSHORT REPORTity (82%, n=120) had a medical abortion. Womenchoosing medical abortions have the shortest waitfor the service, with an average of less than twodays from referral to termination. Surgical abortionsperformed out of region have an averagewait of two days longer (7.9 days) than surgicalabortions performed locally (5.6 days).For all abortions performed locally, there wereno serious complications. Table 1 shows therate of complications specifically for medicalabortions carried out locally, compared tothe accepted standard outlined by the RoyalAustralian and New Zealand College of Obstetriciansand Gynaecologists (RANZCOG). 4Complication rates were very low and withinaccepted standards.DiscussionThe retrospective case review revealed that mostwomen from the DHB accessing abortion serviceshad medical abortions locally, with minimal waitingtime from referral to completion of service.Women accessing surgical abortions locally alsoaccessed the service more rapidly than those travellingto a different district for the service. In alow decile area, the time and expense saved fromavoiding travel to other regions can be significantfor patients. Safety outcomes also achievedrecognised standards, and loss to follow-up forcompletion of post-abortion serum beta-HCG wasextremely low at 1%.Setting up of an abortion service in the DHBprovided a number of challenges. Conscientiousobjection amongst hospital staff hindered developmentof the service within the secondary carefacility. This led to the setting of a communityclinic being raised as an alternative. Thiswas the first time in New Zealand that such aservice has been offered from a sexual healthclinic. The experience has shown that this communityapproach is an excellent fit for a servicethat is naturally an integral part of communitysexual health.Despite concerns to the contrary, abortion provisionin a community setting was shown to be safeand effective, with low and acceptable complicationrates. There is evidence that once a womanWHAT GAP THIS FILLSWhat we already know: Abortion services in New Zealand are mostlyavailable through larger hospital-based clinics, and community-based modelsof abortion care are rare.What this study adds: This short study provides some evidence that awell-structured, community-based service in a low resource setting can offersafe and timely abortion services for a high-needs population.is sure of her decision, termination should be undertakenas soon as possible for the benefit of thewoman’s emotional wellbeing. 5 Local provisionof service and availability of medical terminationcan assist greatly in improving timeliness. 6References1. Ministry of Health. Service Coverage Schedule 2008/09. NewZealand Ministry of Health. [Cited 2012 June 3]. Availablefrom: http://www.nsfl.health.govt.nz/apps/nsfl.nsf/pagesmh/189.2. Silva M, McNeill R. Geographic access to termination ofpregnancy services in New Zealand. Aust N Z J Public Health.2008;32(6):519–521.3. Statistics New Zealand. 2006 Census. [Cited 2012June 3]. Available from: http://www.stats.govt.nz/Census/2006CensusHomePage.aspx.4. Royal Australian and New Zealand College of Obstetriciansand Gynaecologists (RANZCOG). Termination of pregnancy:a resource for health professionals. RANZCOG; 2005.5. Silva M, McNeill R, Ashton T. Ladies in waiting: the timelinessof first trimester abortion services in New Zealand. ReprodHealth. 2010;7:19.6. Silva M, Ashton T, McNeill R. Improving termination of pregnancyservices in New Zealand. NZ Med J. 2011;124(1339).ACKNOWLEDGEMENTSChristine Hannah,Midwife, GisborneCommunity Clinic,undertook the datacollection for this audit.COMPETING INTERESTSDr Snook is a Director ofISTAR Ltd, a listed charityinvolved in the importationand distribution ofmifepristone within NewZealand and Australia.VOLUME 5 • NUMBER 2 • JUNE 2013 J OURNAL OF PRIMARY HEALTH CARE 153

BACK TO BACKAll people should wear sunscreen or otherprotection for their skin whenever they areexposed to sunlightJohn Kenealy MBChB,FRACS (Pl Rec Surg)Auckland Regional PlasticReconstructive and HandSurgery, MiddlemoreHospital, PB 93311,Auckland 1640,New ZealandJohn.Kenealy@middlemore.co.nzKenealy J. All people shouldwear sunscreen or otherprotection for their skinwhenever they are exposedto sunlight—the ‘yes’ case.J Prim Health Care.2013;5(2):154–156.YESAny discussion regarding sun protection relatesto those with vulnerable skin (Fitzpatrick typesI–IV) and does not relate to darker types V andVI, whose skin has ‘built-in’ sunscreen in theform of significant amounts of melanin in thestratum corneum.Ultraviolet (UV) radiation has been recognisedas a mutagen since 1936. 1 Indeed as early as thenineteenth century, sunlight was recognised asthe cause of the higher incidence of skin cancerin rural outdoor workers and sailors. 2 Morerecent work shows that the different wavelengthsof UV light interact with the skin in multipleadverse ways. 3Throughout history deliberate sun exposure hasgone in and out of fashion. In Western countries,until the beginning of the twentieth century,tanned skin was associated with the lower (rural)classes, and women went out of their way topreserve their pale skin. In some Asian culturesthis is still prevalent. Outdoor clothing wasdesigned to avoid sun exposure with long sleeves,large brimmed hats or sun bonnets, and parasols.It is only in the mid- to late-twentieth centurythat deliberate sun exposure has become widespread,and increased leisure time associated withincreasing affluence has resulted in ever increasinglevels of ‘accidental’ sun exposure. This, andincreasing longevity, may, in part, be responsiblefor the increasing prevalence of all types of skincancer. There has also been a coincident increasein incident UV radiation, due to a decrease inatmospheric ozone, resulting in less absorptionwithin the lower stratosphere.The National Toxicology Program Report onCarcinogens from the (US) Department of Healthand Human Services considers broad-spectrumUV radiation to be a carcinogen contributing tomost of the estimated 1.5 million skin cancersand 8000 deaths due to melanoma that occureach year in the United States. 4,5 New Zealandhas much worse incidence than this. It is notknown whether there is a safe level of regularsun exposure that imposes no (or minimal) skincancer risk over time. 6Cumulative lifetime sun exposure is also responsiblefor much of the adverse cosmetic changes tothe skin that we associate with ageing, includingwrinkling, thinning, loss of elasticity, scaliness,dryness, telangiectasia and dyspigmentation.It therefore makes sense from the point of viewof causation, and applying the precautionary prin-BACK TO BACK this issue:John KenealyIan ReidWhile evidence can help inform best practice, it needs to be placed in context.There may be no evidence available or applicable for a specific patient withhis or her own set of conditions, capabilities, beliefs, expectations and socialcircumstances. There are areas of uncertainty, ethics and aspects of care for whichthere is no one right answer. General practice is an art as well as a science. Qualityof care also lies with the nature of the clinical relationship, with communication andwith truly informed decision-making. The BACK TO BACK section stimulatesdebate, with two professionals presenting their opposing views regarding a clinical,ethical or political issue.154 VOLUME 5 • NUMBER 2 • JUNE 2013 J OURNAL OF PRIMARY HEALTH CARE

BACK TO BACKciple, that sun exposure should be reduced fromits current levels.Unfortunately, due to factors including the longlead time in the development of both mutagenicchanges and chronic sun damage, the difficultyof accurately quantifying sun exposure and sunprotection measures retrospectively, and changesin social attitudes and habits over time, it hasbeen difficult to produce hard scientific proof ofthe benefit of sun avoidance/sun protection. Additionally,different skin cancers have a differentrelationship to timing of sun exposure. Epidemiologicalevidence suggests that melanoma incidenceis most influenced by childhood sunburn,whereas non-melanoma skin cancer incidence appearsto correlate with total cumulated exposure.Despite that, one prospective randomised studyhas shown that even relatively short periods ofthe use of a moderately protective sunscreen(SPF15) does reduce the subsequent incidence ofsquamous cell carcinoma (SCC) and actinic keratois(an SCC precursor). 7 This prospective studyalso suggested a reduction in melanoma. 8 Basalcell carcinoma was unaffected.Reduction of sun exposure can be achieved bya number of methods, which are best combinedand matched to individual circumstances.• Sun avoidance (e.g. seeking shade, deferringexposure to a time of daywhen UV levels are lower)• Sun protective clothing (e.g. hat, long sleeves,one-piece swimsuit rather than bikini)• Sunscreen (physical or chemical).RisksSun avoidance and protective clothing carry norisks in themselves other than any risk associatedwith a reduction in exposure below the levelsnecessary to synthesise vitamin D. Therefore, itis a sensible first part of any strategy to reduceUV dose.A number of potential risks have been raised regardingthe routine use of sunscreens. Sunscreenstake the form of chemicals which absorb UVlight, physical particles that reflect UV light, or acombination of both.Numerous chemicals are used in modern sunscreens,many can be absorbed, but have highsafety indices, and although some have beenshown to have potential systemic effects, these areat levels many times higher than can be observed,even when applied to the entire body surface ofadults. Additionally, all active substances in sunscreensused in the USA are subject to FDA approval.9 A review of the evidence by Burnett andWang from Memorial Sloan Kettering, New Yorkin 2011 concluded ‘..none of the data published todate conclusively demonstrates adverse effects onthe health of humans from the use of sunscreen’. 10Children, particularly infants, have a largersurface to volume ratio, and have a skin structurethat is both more vulnerable to UV-inducedchanges, and also higher absorptive potential.This is why sun avoidance and protective clothingshould be used as the predominant means ofreducing sun exposure in the young, and physicalsunblocks preferred to chemical blocks.Physical sunblocks generally comprise zinc oxide,titanium dioxide, or a combination of both. In recentyears, to improve the cosmetic acceptabilityand physical properties of these compounds, theparticle size has been significantly reduced downto ‘nano’ level. This has raised concerns aboutthe potential for these compounds to be absorbedand have a systemic effect. This is unfounded asin fact they are only absorbed into the accelular/avascular stratum corneum. 11Vitamin DEarly in the twentieth century vitamin D deficiencywas recognised as the cause of rickets, which wascured by regular sun exposure. UV exposure ofvitamin D precusors in the skin is an initiating stepin the metabolic pathway. Vitamin D can, however,be taken and absorbed orally. Several studies havelooked at sunscreen use and its impact on vitaminD levels and have failed to show that there is aclinically significant reduction in levels. 12,13Apart from its long-known effects on bonemetabolism, vitamin D has more recently beenmooted as having an effect on mortality, cardiovasculardisease, multiple sclerosis, malignancy,and immunity. None has been proven to date.VOLUME 5 • NUMBER 2 • JUNE 2013 J OURNAL OF PRIMARY HEALTH CARE 155

BACK TO BACKIn conclusion, there is clear evidence that bothacute and chronic UV exposure causes damage tothe skin. There is clear evidence that sun protectionmeasures, including avoidance, sun protectiveclothing, and sunscreen, will all reduce thedose of radiation that the skin receives. Thereis some evidence that the reduction in exposurethat is currently achievable does reduce some ofthe risks associated with such exposure. There islittle evidence to support a real harm associatedwith these sun avoidance measures, and therefore,on balance, it is still advisable that individualswith skin types I–IV take measures to protectthemselves from the sun whenever they areexposed or likely to be exposed.References1. Stadler LJ, Sprague GF. Genetic effects of ultra-violet radiationin maize: I. Unfiltered radiation. Pro Natl Acad Sci U S A.1936;22(10):572–578.2. Hockberger PE. A history of ultraviolet photobiology forhumans, animals and microorganisms. Photochem Photobiol.2002;76(6):561–579.3. Kullavanjaya P, Lim HW. Photoprotection. J Acad Dermatol.2005;52:937–58.4. Wolpowitz D, Gilchrest BA. The vitamin D questions: howmuch do you need and how should you get it? J Am AcadDermatol. 2006;54(2):301–17.5. Ultraviolet (UV) radiation, broad spectrum and UVA, UVB,and UVC—National Toxicology Program. [Cited 2013Apr 11]. Available from: http://ntp.niehs.nih.gov/index.cfm?objectid=BD4CD88D-F1F6-975E-792094AC1CE4B062.6. American Academy of Dermatology. Position statement onvitamin D. November 1, 2008; [cited 2013 Apr 11]. Availablefrom: http://www.aad.org/forms/policies/uploads/ps/psvitamin%20d.pdf7. Van der Pols JC, Williams GM, Pandeya N, Logan V, Green AC.Prolonged prevention of squamous cell carcinoma of the skinby regular sunscreen use. Cancer Epidemiol Biomarkers Prev.2006;15:2546–2548.8. Green AC, Williams GM, Logan V, Strutton G. Reduced melanomaafter regular sunscreen use: randomized trial follow-up.J Clin Oncol. 2011;29(3):257–263.9. Gonzalez S, Fernandez-Lorente M, Gilaberte-Calzada Y. The lateston skin photoprotection. Clin Dermatol. 2008;26:614–626.10. Burnett M, Wang S. Current sunscreen controversies: acritical review Photodermatol Photoimmunol Photomed.2011;27:58– 67.11. Pflucker F, Wendel V, Hohenberg H, Gartner E, Will T, PfeifferS, et al. The stratum corneum layer: an effective barrier againstdermal uptake of different forms of topically applied micronisedtitanium dioxide. Skin Pharmacol Appl Skin Physiol.2001;14(Suppl 1)92–97.12. Marks R, Foley PA, Jolley D, Knight KR, Harrison J, ThompsonSC. The effect of regular sunscreen use on vitamin D levels inan Australian population. Results of a randomized controlledtrial. Arch Dermatol. 1995;131:415–421.13. Farrerons J, Barnadas M, Rodriguez J, Renau A, Yoldi B,Lopez-Navidad A, et al. Clinically prescribed sunscreen(sun protection factor 15) does not decrease serum vitaminD concentration sufficiently either to induce changes inparathyroid function or in metabolic markers. Br J Dermatol.1998;139:422–427.All people should wear sunscreen or otherprotection for their skin whenever they areexposed to sunlightIan R Reid MDProfessor, Departmentof Medicine, Facultyof Medical and HealthSciences, The Universityof Auckland, PB 92019,Auckland, New Zealandi.reid@auckland.ac.nzReid I. All people should wearsunscreen or other protectionfor their skin whenever theyare exposed to sunlight—the‘no’ case. J Prim Health Care.2013;5(2):156–157.NOBone doctors seem prone to contradicting colleaguesfrom other disciplines when it comes topublic health messages. The first obvious exampleis advice regarding body weight—most doctorsbadger their patients to remain thin but, in bonehealth, excessive thinness is a significant risk factorfor osteoporotic fractures. Sunlight exposurerepresents a similar set of contradictions. NewZealand has many fair-skinned residents and asunny climate, resulting in one of the world’shighest rates of skin cancer, so sunlight avoidanceseems logical. However, mineral metabolism iscritically dependent on adequate levels of vitaminD which, despite its name, is absent from mostdiets and is in fact a pro-hormone made in theskin as a result of ultraviolet (UV) light exposure.Thus, vitamin D deficiency is usually a result ofpoor sunlight exposure and the cheapest strategyfor its prevention is encouragement of regulartime in the sun. Is this compatible with the sunsafemessages promoted by dermatologists?The answer is probably yes. In temperate countries,the individuals most at risk of vitamin D156 VOLUME 5 • NUMBER 2 • JUNE 2013 J OURNAL OF PRIMARY HEALTH CARE

BACK TO BACKdeficiency are those with dark skins, since theUV light is absorbed by melanin and thereforenot available for the creation of vitamin D. Theseindividuals (particularly those of African and Indianorigin) have much lower risks of skin cancerthan Europeans, and provide the majority of thecases presenting with rickets and osteomalacia inNew Zealand. 1 They and their children need tobe made aware of their need for regular sunlightexposure or, if this is not possible, be providedwith oral vitamin D supplements to enablegrowth and development to progress normally.Frail elderly Europeans present a more difficultsituation. Their frailty may confine them indoors,yet their pale skin and long history of sun exposureplaces them at high risk of skin cancers. Inthe absence of any intervention, clinical osteomalacia(including myopathy) does occur in thisgroup, and oral vitamin D supplements (1–2 multivitamintablets daily, each typically containing400–800 IU of vitamin D, or calciferol 1.25 mg/month) are probably the simplest interventions.Putting aside these two high-risk groups, whatrecommendations should we provide to the restof the population? The first issue here is whatlevels of vitamin D are necessary for health.In recent years, there has been an explosion ofreports of multiple disease associations withvitamin D. This includes various forms of cancer,cardiovascular disease, many infections, fractures,autoimmune diseases, neurological conditions andsimply being admitted to hospital. Many authorshave inferred causation from these descriptionsof association. Such inferences cannot be drawnfrom observational studies, which properly providehypothesis generation. Available trial datado not suggest that achieving very high levels ofvitamin D (>70 nmol/L) is helpful, and the recentreport from the Institute of Medicine in theUnited States was that minimum levels of serum25-hydroxyvitamin D should be 40–50 nmol/L. 2Such levels are achievable with modest amountsof sunlight exposure that fall well short of thosethat are likely to cause skin damage. For instance,we have demonstrated that elderly individualswho spend 15–30 minutes outside daily inAuckland in October can achieve these levels. 3Based on UV light intensity, it has been calculatedthat fair-skinned individuals in Aucklandand Christchurch need to expose their arms andhands (or equivalent skin area) to mid-morningor mid-afternoon sunshine for only 6–8 minutesin the summer to achieve satisfactory vitamin Dstatus. 4 These levels of exposure produces lessthan one third of a minimal erythemal dose, suggestingthat the risk of significant skin damageis very low. In the winter, exposure at noon isrequired for 24 minutes in Auckland and for >40minutes in Christchurch, to maintain these levels.Individuals with highly pigmented skin haveexposure times 3–6 times greater than this. 5Thus, we should be providing a balanced message,encouraging regular sunlight exposure, butemphasising that in summer this can be brief,and it should not be in the middle of the daywhen the risk of skin damage is high. In winterin the South Island, some level of vitamin D deficiencyis almost inevitable without supplementation,unless substantial reserves of vitamin D(which is stored in adipose tissue) have been builtup during the warmer months. This is reflectedin the 2012 Consensus Statement from theMinistry of Health, encouraging daily outdooractivity in the early morning and late afternoonduring summer, and in the middle of the dayduring winter. 6 Sunshine is neither saviour nordemon, but exposure in moderation is an importantpart of healthy living.References1. Blok BH, Grant CC, McNeil AR, Reid IR. Characteristics ofchildren with florid vitamin D deficient rickets in the Aucklandregion in 1998. NZ Med J. 2000;113(1117):374–6.2. Ross AC, Manson JE, Abrams SA, Aloia JF, Brannon PM,Clinton SK, et al. The 2011 Report on Dietary ReferenceIntakes for Calcium and Vitamin D from the Institute of Medicine:what clinicians need to know. J Clin Endocrinol Metab.2011;96(1):53–8.3. Reid IR, Gallagher DJA, Bosworth J. Prophylaxis against vitaminD deficiency in the elderly by regular sunlight exposure.Age Ageing. 1986;15:35–40.4. Nowson CA, McGrath JJ, Ebeling PR, Haikerwal A, DalyRM, Sanders KM, et al. Vitamin D and health in adults inAustralia and New Zealand: a position statement. Med J Aust.2012;196(11):686 –7.5. Springbett P, Buglass S, Young AR. Photoprotection andvitamin D status. J Photochem Photobiol. 2010;101:160–8.6. Ministry of Health and Cancer Society of New Zealand.Consensus Statement on Vitamin D and Sun Exposure in NewZealand. Wellington: Ministry of Health; 2012.VOLUME 5 • NUMBER 2 • JUNE 2013 J OURNAL OF PRIMARY HEALTH CARE 157

CONTINUING PROFESSIONAL DEVELOPMENTPEARLSCOCHRANE CORNERString of PEARLSPractical Evidence About Real Life SituationsPreventive measures for cardiovascular diseasePEARLS are succinct summaries of Cochrane Systematic Reviews forprimary care practitioners—developed by Prof. Brian McAvoy for theCochrane Primary Care Field (www.cochraneprimarycare.org), NewZealand Branch of the Australasian Cochrane Centre at the Departmentof General Practice and Primary Health Care, University of Auckland(www.auckland.ac.nz/uoa), funded by the Ministry of Health (www.health.govt.nz), and published in NZ Doctor (www.nzdoctor.co.nz.).Limited evidence for statins in primary prevention ofCVD in people at low riskLimited benefit from statins in acute coronarysyndromeReduced and/or modified dietary fat may preventcardiovascular diseaseNo clear benefit of salt reduction on mortality andcardiovascular morbiditySome evidence that organisation of secondaryprevention of ischaemic heart disease in primary careis effectiveNo evidence for benefits of homocysteine-loweringinterventions for preventing cardiovascular eventsUltrasound screening for abdominal aortic aneurysmmay reduce mortality in men aged 65 to 79 yearsAmitriptyline satisfactorilyrelieves pain in only a minority ofpatients with fibromyalgiaMegan Arroll PhD, FHEA, CPsychol, CSci, AFBPsS; Visiting ResearchFellow, Chronic Illness Research Team, University of East London,Stratford Campus, Water Lane, London, E15 4LZ, United Kingdom;Email m.a.arroll@sa.uel.ac.ukTHE PROBLEM: Fibromyalgia is a chronic condition thatis characterised by widespread pain, often accompanied bysevere fatigue, depression and sleep disturbance. Fibromyalgiacan be challenging to treat in primary care and is associatedwith high levels of disability and low quality of life. 1 Conventionalanalgesics are limited in effectiveness; hence, moreunconventional treatments such as antidepressants are oftenprescribed. Health care utilisation by those with a diagnosisof fibromyalgia is high. 2CLINICAL BOTTOM LINE: This review demonstrates thatami triptyline can provide satisfactory pain relief to somepatients with fibromyalgia, but only a minority of them; amitriptylinewill not work for most people with this condition.Treatment for pain in fibromyalgia: Amitriptyline vs placeboAmitriptylinevs placebo infibromyalgiaSuccess Evidence HarmsEffective:NNT=4.8(range 1.7to 15)Cochrane review 31 in 4 (25%) morepeople thanplacebo reporthaving at leastone adverseevent 3NNT = numbers needed to treat. An NNT of 4.8 means that for every 4–5 peoplegiven the treatment, 1 person will find the treatment effective.DISCLAIMER: PEARLS are for educational use only and are not meantto guide clinical activity, nor are they a clinical guideline.References1. Verbunt JA, Pernot D, Smeets R. Disability and quality of life in patients withfibromyalgia. Health Qual Life Outcomes. 2008;6:8.2. Hughes G, Martinez C, Myon E, Taïeb C, Wessely S. The impact of a diagnosisof fibromyalgia on health care resource use by primary care patients in theUK: an observational study based on clinical practice. Arthritis Rheum.2006;54(1):177–83.3. Moore RA, Derry S, Aldington D, Cole P, Wiffen PJ. Amitriptyline for neuropathicpain and fibromyalgia in adults. Cochrane Database of Systematic Reviews2012, Issue 12. Art. No.: CD008242. DOI: 10.1002/14651858.CD008242.pub2.All people residing in New Zealand have access to the Cochrane Libraryvia the Ministry website www.health.govt.nz/cochrane-library158 VOLUME 5 • NUMBER 2 • JUNE 2013 J OURNAL OF PRIMARY HEALTH CARE

Latest CVrisk toolsCV risk assessment and management is a key wayof preventing people having a heart attack or stroke.Two new resources to support CV risk assessmentand management have just been released.Take advantageof these freeresourcesnow!Taking Control Patient ResourceA self-management workbook for people atelevated risk of CVD. Evidence based andco-created with primary care over two years.E-LearningImprove Heart Health – CV Risk Assessmentand Management. An e-learning programme forhealth professionals to help successfully designand deliver CV risk assessment andmanagement services.Want more information?Be sure to take advantage of these free tools toimprove the care and management of your patients.Visit www.heartfoundation.org.nz for more information.

CONTINUING PROFESSIONAL DEVELOPMENTVAIKOLOAKeeping promises, measuring results: the PacificMaternal and Child Health Indicators ProjectFiona Langridge BHSc, MSc; Teuila Percival MBChB, FRACP; Lani Stowers BBus Health Admin (HIM)Pacific Maternal and Child Health Indicators Project Team , Pacific Health Section, School of Population Health,The University of Auckland, Auckland, New ZealandVAIKOLOAPacific PrimaryHealth CareTreasuresVai (water)is a symbol of‘life-source’ andkoloa (treasures)to shareThere can be no keener revelation of a society’s soul than the way in which it treats its children.—Nelson MandelaThe key indicators used to measure maternaland child health globally are theMillennium Development Goals (MDGs),especially MDG 4 and MDG 5 (see Box 1). 1These latter two goals are the furthest frombeing achieved by 2015. 2 All countries that aresignatory to the 1989 United Nations Conventionon the Rights of the Child should be awareit is the right of every child to good health andprotection from harm. The future of any youngand developing nation depends on the wellbeingof its most important resource—the children,who will be the next leaders. Horton 3 outlines10 reasons why the needs of women and childrencontinue to remain ‘invisible’. One of the reasonshe cites concerns the importance of evaluation.Not enough time has been spent measuring theeffects of policies and there is a deficit of data toassist in decision-making.Box 1. Relevant global maternal and child healthmillennium development goals. 1MDG 1: Eradicate extreme poverty and hungerMDG 4: Reduce child mortalityMDG 5: Improve maternal healthMDG 6: Combat HIV/AIDS, malaria and other diseaseslacking but very much needed. 5 Not a great dealhas changed since Finau discussed the challengesfor health information systems in the Pacific in1994 (Box 2). 6 Growing disparities and emerginghealth problems in the Pacific are not highlightedby generalised global health indicators,such as the MDGs and mortality. More sensitivematernal and child health (MCH) indicators willunearth the hidden and actual MCH issues. Thiscan potentially create a policy environment whereinvestment in MCH will occur in the Pacific.Growing disparities and emerging health problems in the Pacificare not highlighted by generalised global health indicators, such asthe MDGs and mortality. More sensitive maternal and child healthJ PRIM HEALTH CARE2013;5(2):160–161.(MCH) indicators will unearth the hidden and actual MCH issues.CORRESPONDENCE TO:Fiona LangridgeSchool of PopulationHealth, The Universityof Auckland, PB 92019,Auckland, New Zealandf.langridge@auckland.ac.nzThe Pacific region has been neglected with regardto measuring health development and much ofthe research is anecdotal. 4 There is a paucity ofdata, highlighted by the need to monitor MDGs.Alongside this, there are challenges such as therapid transition between communicable andnon-communicable diseases in some populations.Civil registration and vital statistics systems areIn terms of child mortality improvements, thePacific does not appear to be faring well incomparison to the rest of the world. A reportfrom UNICEF and other organisations indicatesthat of all the MDG regions, Sub-Saharan Africaand Oceania have the furthest to go to achieveMDG 4. 7 They have only achieved around a 30%reduction in under-five mortality, and are there-160 VOLUME 5 • NUMBER 2 • JUNE 2013 J OURNAL OF PRIMARY HEALTH CARE

CONTINUING PROFESSIONAL DEVELOPMENTVAIKOLOAfore not on track to reach this goal. It is worthmentioning that Sub-Saharan Africa has the challengeof the HIV/AIDS pandemic. This is not amajor factor for Oceania. Improving maternal andchild health is a priority for Pacific governments,policy makers, and global health and donororganisations.The United Nations Commission on Informationand Accountability for Women’s and Children’sHealth has developed a framework for monitoringand reporting on MCH progress, outcomesand investment. 2 This has included developing acore set of indicators to measure MCH, alongsideimproving health information and vital statistics. 1The Pacific Maternal and Child Health IndicatorsProject (CHIP) Keeping Promises, MeasuringResults is reviewing the appropriateness andfunctionality of the 11 core MCH indicators forthe Pacific region. The effectiveness of indicatorsand government and donors’ ability to measureprogress and investment in a meaningful way isreliant on the quality of that data and information.Equally, indicators need to be relevantto the Pacific Island nation context reflectinglocal health profiles. Some indicators may lackrelevance in the Pacific context given persistingdata gaps and health profiles at variance withother regions in the world. The Pacific CHIPteam is looking at available data sources for the11 core indicators and the relevance and functionalityof those indicators for the Pacific. Afterreview and consultation with expert informants,a framework of MCH indicators for the Pacificwill be developed. This will build on globalindicators, with recommended modificationsor additions. The steps in health informationsystems or policy needed to achieve this will beproposed. This will culminate in an increasedfocus on and advocacy for the health of Pacificmothers and children.Box 2. Finau’s comments on the challenges of Pacifichealth information systems. 6Health statistics and medical records of many hospitalsare tucked away in dingy corners with insufficientroom, manual card systems and poorly trained staff.Careers in these offices are either temporary untilsomething better comes along or terminal in that nobetter career options can be found. (p. 165)References1. World Health Organization. Monitoring maternal newborn andchild health: understanding key progress indicators. Countdownto 2015 and Health Metrics Network. 2011; [cited 2012Aug 28]. Available from: http://www.who.int/healthmetrics/news/monitoring_maternal_newborn_child_health.pdf.2. World Health Organization. Commission on information andaccountability for Women’s and Children’s Health. Keepingpromises, measuring results. 2011; [cited 2012 Aug 30]. Availablefrom: http://www.who.int/topics/millennium_development_goals/accountability_commission/Commission_Report_advance_copy.pdf.3. Horton R. The continuing invisibility of women and children.The Lancet. 2010. 375(9730):1941–1943.4. Taylor R, Bampton D, Lopez A. Contemporary patterns ofPacific Island mortality. Int J Epidemiol. 2005;34:207–214.5. Health Information Systems Knowledge Hub. 2012. Improvingvital statistics in the Pacific 2011–2014. Pacific Health Dialog.2012;18(1):63–64.6. Finau SA. National health information systems in thePacific Islands: in search of a future. Health Policy Planning.1994;9(2):161–170.7. UNICEF, WHO, UNDP, The World Bank. 2012 Levels andtrends in child mortality, report 2012; [cited 2012 Nov 17]Available from: http://www.who.int/maternal_child_adolescent/documents/levels_trends_child_mortality_2012.pdf.ACKNOWLEDGEMENTSWe would like to thankthe Health InformationSystems Knowledge Hub,School of PopulationHealth, The University ofQueensland, AustraliaVOLUME 5 • NUMBER 2 • JUNE 2013 J OURNAL OF PRIMARY HEALTH CARE 161

CONTINUING PROFESSIONAL DEVELOPMENTNUGGETS OF KNOWLEDGESedating antihistamines in children—not a good choiceLinda Bryant MClinPharm, PGDipHospPharmAdmin, PhD, FNZHPA, FNZCP, FPSNZ, MCAPAYou wouldn’t give your children coffee due to the adverse CNS effects—so why treat children withsedating antihistamines?KEY POINTS• First-generation(sedating)antihistaminesare relativelynon-specific andhave marked CNSeffects.• Second-generation(non-sedating)antihistamines arewell tolerated overa long period andhave no adverseeffects on learning.• Only cetirizine islicensed in NZ foruse in very youngchildren (one yearand older).• Routine use ofantihistamines,with or withoutdecongestants, isnot recommendedfor children withotitis media.‘Social medication’ is a term used for medicationthat gives parents control over children’s behaviourthat they perceive as fractious and irritatingor that reduces the inconvenience of a sick child. 1Sedating antihistamines must not be used as acoping strategy for the family.First generation (sedating) antihistamines are notlicensed for use in children under two years old.They are ‘dirty’ drugs in that they are relativelynon-specific and act on histaminic, serotonergicand cholinergic receptors with marked effectson the central nervous system (CNS). They canact as anti-emetics, hypnotics and tranquilisers.Promethazine and trimeprazine are antipsychoticderivatives and are contraindicated for use inchildren less than two years of age due to the riskof marked sedation and respiratory depression.Conversely, these drugs may cause paradoxicalexcitation in children.Use in coughs and coldsFor coughs and colds in children less than sixyears old, first generation antihistamines showedno significant benefit in small studies. 2–5 Thereis a paucity of studies in 6–12-year-old chil-Table 1. Antihistamines: licensed use in childrendren, but a similar lack of benefit is noted. 6 Asystematic review of antihistamines for chroniccough in children found that they could not berecommended as empirical therapy due to the potentialadverse effects, especially in very youngchildren. 7Use in otitis mediaRepeated systematic reviews of antihistamines,usually with decongestants, for children withotitis media have concluded that for acute otitismedia the routine use of antihistamines in childrencannot be supported as the harms outweighany benefit. 8 Similarly, for children with otitismedia with effusion, there was no benefit demonstratedbut there was potential harm and soantihistamines, with or without decongestants,are not recommended. 9Second-generation (non-sedating)antihistaminesSecond generation antihistamines act primarilyon peripheral histaminic receptors and are nonsedatingbecause they do not readily cross theblood–brain barrier. They have few adverse drugJ PRIM HEALTH CARE2013;5(2):162–163.CORRESPONDENCE TO:Linda BryantClinical Manager, ClinicalAdvisory Pharmacist,East Health Trust PHOPO Box 38248, HowickAuckland, New Zealandl.bryant@auckland.ac.nzFirst generation Trade name Licensed in NZ for childrenChlorpheniramine 2 mg/5 mL Histafen ≥ 6 years oldDexchlorpheniramine 2 mg/5 mL Polaramine ≥ 2 years oldPromethazine 5 ml/5 mL Phenergan; Allersoothe ≥ 2 years oldTrimeprazine 30 mg/5 mL Vallergan Forte ≥ 2 years oldSecond generation Trade name Licensed in NZ for childrenLoratadine 1 mg/1 mL Lorapaed ≥ 2 years oldCetirizine 1 mg/1 mL Cetirizine AFT ≥ 1 year oldNUGGETS of KNOWLEDGE provides succinct summaries of pharmaceutical evidence abouttreatment of common conditions presenting in primary care and possible adverse drug reactions.162 VOLUME 5 • NUMBER 2 • JUNE 2013 J OURNAL OF PRIMARY HEALTH CARE

CONTINUING PROFESSIONAL DEVELOPMENTNUGGETS OF KNOWLEDGETable 2. Antihistamines: symptoms in overdoseFirst generation (e.g. promethazine)Unconsciousness, commonly delayed. In addition,convulsions, hallucinations, delirium, acute anxiety,psychotic reactions, extreme hyperaesthesia andhyperalgesia with extensor plantar responses may occur.Anticholinergic action may cause tachycardia, flushed skin,dry mouth and sometimes mydriasis and urinary retention.In adults, CNS depression is more common, withdrowsiness, coma, convulsions, progressing to respiratoryfailure or cardiovascular collapse.In infants and children, CNS stimulation predominatesover CNS depression causing ataxia, excitement, tremors,psychoses, hallucinations, convulsions and possiblyhyperpyrexia, which may be followed by deepening comaand cardiorespiratory collapse (and death).Second generation (e.g. loratadine)Somnolence, tachycardia, headachereactions, are well tolerated over a long periodand have no adverse effects on learning, but theyare still not licensed for children under two yearsold, except for cetirizine, which is licensed forchildren one year or older.References1 Allotey P, Reidpath DD, Elisha D. ‘Social medication’ and thecontrol of children: a qualitative study of over-the-countermedication among Australian children. Paediatrics. 2004;114(3):e378–83.2 Hutton N, Wilson MH, Mellits ED, Baumgardner R, WissowLS, Bonuccelli C, et al. Effectiveness of an antihistaminedecongestantcombination for young children with thecommon cold: a randomized, controlled clinical trial. J Pediatr.1991;118(1):125–30.3 Clemens CJ, Taylor JA, Almquist JR, Quinn HC, Mehta A,Naylor GS. Is an antihistamine-decongestant combination effectivein temporarily relieving symptoms of the common coldin preschool children? J Pediatr. 1997;130(3):463–6.4 Sakchainanont B, Ruangkanchanasetr S, Chantarojanasiri T,Tapasart C, Suwanjutha S. Effectiveness of antihistamines incommon cold. J Med Assoc Thai. 1990;73(2):96–101.5 Paul IM, Yoder KE, Crowell KR, Shaffer ML, McMillan HS,Carlson LC et al. Effect of dextromethorphan, diphenhydramine,and placebo on nocturnal cough and sleep qualityfor coughing children and their parents. Pediatrics. 2004Jul;114(1):e85–90.6 Yoder KE, Shaffer ML, La Tournous SJ, Paul IM. Child assessmentof dextromethorphan, diphenhydramine, and placebofor nocturnal cough due to upper respiratory infection. ClinPediatr (Phila). 2006 Sep;45(7):633–40.7 Chang AB, Peake J, McElrea MS. Anti-histamines forprolonged non-specific cough in children. CochraneDatabase Syst Rev. 2008 Apr 16;(2):CD005604. doi:10.1002/14651858.CD005604.pub3.8 Coleman C, Moore M. Decongestants and antihistaminesfor acute otitis media in children. Cochrane Database SystRev. 2008 Jul 16;(3):CD001727. doi: 10.1002/14651858.CD0 01727.pub4.9 Griffin G, Flynn CA. Antihistamines and/or decongestantsfor otitis media with effusion (OME) in children. CochraneDatabase Syst Rev. 2011 Sep 7;(9):CD003423. doi:10.1002/14651858.CD003423.pub3.ACKNOWLEDGEMENTSThanks to Kwee Goh,Clinical AdvisoryPharmacist, for assistingwith this Nugget.VOLUME 5 • NUMBER 2 • JUNE 2013 J OURNAL OF PRIMARY HEALTH CARE 163

CONTINUING PROFESSIONAL DEVELOPMENTPOTION OR POISON?Coenzyme Q10J PRIM HEALTH CARE2013;5(2):164.Shane Scahill BPharm, MMgt, PhD, RegPharmNZ; Honorary Senior Lecturer, School of Pharmacy, The University of Auckland, PB 92019,Auckland, New Zealand. Email: s.scahill@auckland.ac.nzGPs have questioned the benefit of co-administrationof coenzyme Q10 with statinsto reduce the associated myalgia/myopathy.Where does the evidence lie?SOME AVAILABLE BRANDS: Blackmores,BioBalance, BioOrganics, Clinicians,Good Health, MICROgenics,Nutra-Life, Radiance, SANDERSON,Solgar, Swanson Ultra, Thompson’s.ACTIVE CONSTITUENTS: CoenzymeQ10 is also known as: CoQ10, Q10, vitaminQ10, Ubiquinone, Ubidecarenone,Mitoquinone, Andelir, Heartcin, Neuquinone,Taidecanone and 2, 3 dimethoxy-5methyl-6-decaprenyl benzoquinone.MANUFACTURER CLAIMS: Promotingheart health, increasing energy levels,enhancing the immune system, supportinghealthy gums, providing antioxidantactivity and decreasing side effectsassociated with certain prescriptiondrugs. Periodontal gum disease, allergies,bronchial asthma, chronic microbialinfections and low sperm motility are allindications. Supplementation for womenwith breast cancer is claimed to shrinktumours, reduce pain and cause partialremission in some individuals. CoenzymeQ10 is also a potential treatmentfor neurodegenerative disorders such asParkinson’s, ALS and Alzheimer’s.EVIDENCE FOR EFFICACY: Along withother agents, coenzyme Q10 appearsin 26 Cochrane Database SystematicReviews (CDSR) covering a range ofindications including: neurological,cardiovascular, muscular soreness and atrophy,fertility and cancer-related fatigueand disease associated with mitochondrialfunction. Full systematic reviewsspecific to coenzyme Q10 include primaryhypertension (2009), heart failure(2010) and Parkinson’s disease (2012).A CDSR review of the concomitant useof coenzyme Q10 and statins to reduceassociated myalgia/myopathy has notbeen undertaken. A systematic reviewpublished in the Journal of the AmericanCollege of Cardiology (JACC) suggeststhere is little evidence to support theroutine use of coenzyme Q10 in conjunctionwith statins for statin-associatedmyopathy. The authors suggest it is possibleto hypothesise about the benefits ofcoenzyme Q10 at a molecular level; however,large clinical studies are required toallow robust systematic review.ADVERSE EFFECTS: No serious adverseevents reported in high doses (up to900 mg/day) in a short-term study inhealthy volunteers. Adverse effectsinclude nausea, upper abdominal pain,rashes, dizziness, sensitivity to light, irritability,fatigue, headache, heartburn. Indoses over 100 mg taken in the evening,coenzyme Q10 may cause mild insomnia.Changes in haematological, biochemicaland urinalysis parameters have occurredbut are not deemed clinically significant.CONTRAINDICATIONS: Limited informationavailable. Pregnant or lactatingwomen.PRECAUTIONS: Limited informationavailable. Allergy to coenzyme Q10 orexcipients.Summary MessageThere are no Cochrane DatabaseSystematic Reviews (CDSR) reportingthe clinical efficacy of coenzyme Q10 instatin-associated myopathy. CoenzymeQ10 is generally well tolerated. A systematicreview suggests that rather thanhypothesising about potential clinicalbenefits, large clinical studies are requiredto allow robust systematic review.Further, there is insufficient evidence torecommend the routine administrationof concomitant coenzyme Q10 for preventionof statin-associated myopathy.DRUG INTERACTIONS: Theoretically,chemotherapy and radiotherapy. Warfarin,due to the effect of coenzyme Q10on clotting and bleeding. Natural levelsof coenzyme Q10 may possibly be depletedwhen taking statins, anti-diabeticagents and beta blockers.Key referencesCochrane Library. [Cited 2013 Jan 22]. Available from:www.health.govt.nz/cochrane-library.Cooke H. CAM-Cancer Consortium. Co-enzyme Q10[Internet]. [Cited 2012 Dec 14]. Available from:http://www.cam-cancer.org/CAM-Summaries/Dietary-approaches/Co-enzyme-Q10.Ikematsu H, Nakamura K, Harashima S, Fujii K,Fukutomi N. Safety assessment of coenzyme Q10(Kaneka Q10) in healthy subjects: a double-blind,randomized, placebo-controlled trial. Regul ToxicolPharmacol. 2006; 44(3):212–8.Healthpost. Coenzyme Q10 [Internet]. [Cited 2013Feb 25]. Available from: http://www.healthpost.co.nz/supplements-and-natural-health/a-z-of-natural-health/coenzyme-q10.htm?gclid=CLqL1v_hz7UCFSRxQgodggwAKQ.Marcoff L, Thompson PD. The role of coenzyme Q10in statin-associated myopathy. J Am Coll Cardiol.2007;49(23):2231–7.Herbal medicines are a popular health care choice, but few have been tested to contemporary standards.POTION OR POISON? summarises the evidence for the potential benefits and possible harms of well-known herbal medicines.164 VOLUME 5 • NUMBER 2 • JUNE 2013 J OURNAL OF PRIMARY HEALTH CARE

ETHICSProfessional accountability of doctors inNew ZealandKatharine Wallis MBChB, MBHL, FRNZCGPIn the good old days: ‘A doctor (unlike apolitician or an actor) [was] judged only by hispatients and immediate colleagues, that is, behindclosed doors, man to man.’ 1 How times havechanged. Medicine today is more effective, moredangerous and more expensive than ever before,and the public demand for professional accountabilityis (rightly) greater than ever. It may nothave mattered much in the past if a doctor misseddiagnosing a leaking aneurysm, because the outlookwas bleak regardless, but today, because suchpatients can be treated and likely saved, makingthe diagnosis matters. With such power comesresponsibility.Responsibility may be considered synonymouswith accountability, although responsibilityimplies being morally accountable for one’s actionswhile accountability implies being merelyaccountable. 2 Accountability may be individual orcollective, retrospective or prospective. Retrospectiveaccountability is backward looking and isabout holding someone to account for past actionsand present consequences. The process of identifyingrisk is central to the process of accountabilityand allocating blame when things go wrong. 3Prospective accountability, rather than lookingback to assign blame, attempts to ensure that theright thing happens going forward.There are a number of organisations in New Zealandwith a role to play in satisfying the publicdemand for professional accountability. While inmost countries a tort-based malpractice system isused to both provide compensation for medicalinjury and to hold doctors to account, in NewZealand suing is barred by the no-fault accidentcompensation scheme and doctors are held to accountthrough separate accountability processes.As Douglas and Wildavsky 4 pointed out, the typeof society generates the type of accountability.Most medical professional accountabilityprocesses judge according to the process of carerather than the outcome. This might be appropriate,given the high degree of uncertainty inhealth care and the highly variable outcomes thesame treatment can have in different individuals,2 but it is important that there is also someoutcomes-based accountability to ensure thatmedicine is delivering more good than harmoverall, to ensure that doctors are not seeking towork perfectly in a system that is delivering moreharm than benefit.Medical professional accountability is importantto maintain standards and to foster trust in theprofession. To be accountable is to be responsible—forpast actions and for future actions. Tobe accountable is to inform patients about (pastor future) actions and decisions, to justify thesedecisions, and to suffer punishment in the case ofeventual misconduct. 2Retrospective accountabilityRetrospective accountability is about holdingsomeone to account for something in the past.The ETHICS column explores issues around practising ethically in primary health care and aims toencourage thoughtfulness about ethical dilemmas that we may face.THIS ISSUE: Our guest ethicist and GP Katharine Wallis explores the various agencies that deal withretrospectively and prospectively holding doctors accountable for their actions.J PRIM HEALTH CARE2013;5(2):165–169.CORRESPONDENCE TO:Katharine WallisKatharine.wallis@otago.ac.nzVOLUME 5 • NUMBER 2 • JUNE 2013 J OURNAL OF PRIMARY HEALTH CARE 165

ETHICSThe Health and Disability Commissionerand the Code of Consumers’ RightsThe Health and Disability Commissioner and theCode of Health and Disability Services Consumers’Rights 1996 5,6 (the Code) came into beingfollowing recommendations made by Judge DameSilvia Cartwright in The Report of the Committeeof Inquiry into Allegations Concerning theTreatment of Cervical Cancer at National Women’sHospital and into Other Related Matters. 7 DameSilvia found the system of medical self-regulationin New Zealand wanting. She identified, amongother factors, ‘a failure of peer review andconsequential dominance of clinical freedom’,a collective abdication by medical staff of theircollective ethical and professional responsibilities,and a ‘pervading atmosphere of defensiveness andeven arrogance’. 7 She concluded that she couldnot ‘leave the encouragement of new habits andpractices to the medical profession alone’ and recommendedlegislative changes to increase awarenessof patients’ rights and public scrutiny of themedical profession. Hence the introduction of theHealth and Disability Commissioner Act. 5 TheCommissioner’s role is to promote and protect therights of consumers as set out in the Code:1. Right to be treated with respect2. Right to freedom from discrimination,coercion, harassment, and exploitation3. Right to dignity and independence4. Right to services of an appropriate standard5. Right to effective communication6. Right to be fully informed7. Right to make an informed choice and giveinformed consent8. Right to support9. Rights in respect of teaching or research10. Right to complain. 6The Commissioner has a number of availableoptions for dealing with complaints, althoughhis powers are limited to reporting, recommending,and referring. 8 The Commissioner may refera complaint to the Medical Council for possiblerehabilitation, and may investigate a complaint todetermine whether there has been a breach of theCode. The Commissioner may find a breach evenwhen a patient has suffered no harm. If the Commissionerfinds a breach, the Commissioner mayrefer the complaint to the Director of Proceedingsfor possible discipline.The Commissioner receives about 1300 complaintsper year against all types of providersand investigates less than 10% of these. He refersfew providers each year for either rehabilitationor discipline. 9 In recent years, the Commissionerhas tended to reserve the disciplinary route forcomplaints raising ethical issues (such as boundarytransgressions and inappropriate relationships)and the rehabilitative route for complaints raisingcompetence issues. The Commissioner’s decisionsare final and neither consumers nor providers canappeal his decisions.The Medical Council of New ZealandIn New Zealand, in contrast to many other countries,the competence and disciplinary processeshave been separated so that the Medical Councilis no longer prosecutor, judge, and beneficiary offines. The Medical Council oversees competenceissues while the Health Practitioners’ DisciplinaryTribunal deals with disciplinary matters.Colleagues have a discretion to report a doctor tothe Medical Council if they believe the doctorposes a risk of harm to the public, but employers,the Health and Disability Commissioner, theAccident Compensation Corporation (ACC), andthe Courts all have a duty to do so (s.34). 10 Allpatient complaints to the Medical Council mustbe referred to the Health and Disability Commissionerin the first instance. The processes fordealing with complaints and adverse events werestreamlined following recommendations from the2001 Cull Report, which inquired into repeatedcomplaints and disciplinary proceedings against aNorthland gynaecologist. 11In response to competence or fitness to practisereferrals, the Medical Council may order interimsuspension, and/or refer the doctor to its HealthCommittee, a Performance Assessment Committee(for possible performance assessmentand rehabilitation), or to a Professional ConductCommittee (for possible discipline). The MedicalCouncil receives 40 to 50 referrals per year, andconducts performance reviews on about half ofthese. 12166 VOLUME 5 • NUMBER 2 • JUNE 2013 J OURNAL OF PRIMARY HEALTH CARE

ETHICSThe Health PractitionersDisciplinary TribunalUnder the Health Practitioners CompetenceAssurance Act, the purpose of the disciplinaryprocess is not to punish the doctor but to ‘protectthe health and safety of members of the public’(section 3). 10 Both the Director of Proceedingsand a Medical Council Professional ConductCommittee have the power to bring a disciplinarycharge against a doctor before the HealthPractitioners Disciplinary Tribunal.The Tribunal must consider the evidence placedbefore it and decide whether, on balance ofprobabilities, a charge of professional misconducthas been proven. If the Tribunal finds a chargeproven, the Tribunal may order that the doctor beremoved from the register, suspended for up tothree years, censured, have conditions on practiceimposed, be fined up to $30,000 and/or pay costs.All Tribunal hearings are held in public unlessthere are grounds for the Tribunal to order otherwise(usually charges of a sexual nature). Thereare only about 10 medical disciplinary hearings inNew Zealand each year. 13The Privacy CommissionerIssues to deal with health information privacycome under the Privacy Act 1993, which givesthe patient control over access to his or herpersonal health information and imposes a dutyof non-disclosure on health practitioners. 14 Thereare situations when disclosure is permitted tothe extent necessary for the particular purposeto protect public interest considerations or thepatient’s own safety and these are set out in theHealth Information Privacy Code. 15The Privacy Commissioner investigates complaintsalleging a breach of information privacy.If the Commissioner finds a breach, she maysettle the complaint through conciliation (mostcomplaints end here) or refer the complaint tothe Director of Human Rights Proceedings, whomay bring a charge before the Human RightsReview Tribunal. This Tribunal has the power toaward damages to the complainant for pecuniaryloss suffered, loss of benefit, humiliation, loss ofdignity, and injury to feelings.The Coroner and the CourtsThe Coroner and Courts play only a minor rolein professional accountability in New Zealand.The Coroner has the power to investigate thecircumstances and causes of a patient’s death andto make recommendations, and the Courts mayhear cases of medical manslaughter. 16,17 Followingreform of the Crimes Act in 1997, when thethreshold for medical manslaughter was liftedfrom ordinary negligence to gross negligence,defined as a major departure from the standard ofcare expected of a reasonable person, there havebeen very few cases of medical manslaughter inNew Zealand. 18 There has been only one case ofalleged medical manslaughter since 2000, whena midwife was found not guilty in 2006 for hermanagement of a breech delivery which ended inthe death of the baby.Prospective accountabilityProspective accountability is about ensuring thatthe right thing happens going forward. Prospectiveaccountability is linked to moral deliberationand extends beyond legal duty. It is concernedwith the roles we occupy in society and theobligations these roles entail: as doctors we havea responsibility to safeguard the best interests ofour patients and to work for the public good. 19While the aforementioned medical professionalaccountability processes might help a patient tochoose wisely when to place and when to withdrawtrust, these processes do not do away withthe patient’s need to trust. As Paul 20 has noted,external controls are ‘blunt instruments in particularcases and require a functioning internalmorality to interpret them’.Ultimately, a patient must still rely on a doctorhaving a functioning internal morality, or a commitmentto professionalism, to integrity, compassion,altruism, and continuous improvement. 21According to Baier:Rights do define a sort of individualist tip of theiceberg of morality, one that takes no extra organisationto stay afloat, but that is because it is supportedby the submerged floating mass of cooperatively dischargedresponsibilities and socially divided labour. 22VOLUME 5 • NUMBER 2 • JUNE 2013 J OURNAL OF PRIMARY HEALTH CARE 167

ETHICSNew Zealand Medical AssociationThe New Zealand Medical Association (NZMA)attempts to capture the ‘submerged floating mass’of cooperatively discharged responsibilities in itsCode of Ethics. 23 The NZMA Code of Ethics setsout 67 recommendations to guide professional behaviour.The recommendations, unlike the dutiesimposed by the Code of Consumers’ Rights, arenot legally enforceable. The recommendations arebased on the following 12 Principles of EthicalBehaviour:1. Consider the health and wellbeing of thepatient to be your first priority.2. Respect the rights, autonomy and freedom ofchoice of the patient.3. Avoid exploiting the patient in any manner.4. Practise the science and art of medicine tothe best of your ability with moral integrity,compassion and respect for human dignity.5. Protect the patient’s private informationthroughout his/her lifetime and followingdeath, unless there are overridingconsiderations in terms of public interest orpatient safety.6. Strive to improve your knowledge and skillsso that the best possible advice and treatmentcan be offered to the patient.7. Adhere to the scientific basis for medicalpractice while acknowledging the limits ofcurrent knowledge.8. Honour the profession, including its traditions,values, and its principles, in the ways that bestserve the interests of the patient.9. Recognise your own limitations and thespecial skills of others in the diagnosis,prevention and treatment of disease.10. Accept a responsibility to assist in theprotection and improvement of the health ofthe community.11. Accept a responsibility to advocate foradequate resourcing of medical services andassist in maximising equitable access to themacross the community.12. Accept a responsibility for maintaining thestandards of the profession.Medical Council of New ZealandThe Medical Council has developed a number ofprospective accountability processes designed toprotect the public and to ensure that the rightthing happens going forward. The Council specifiesscopes of practice, prescribes the qualificationsand experience required for registration,and issues annual practising certificates to doctorswhom the Council considers are competent andfit to practise. The Medical Council acceptssatisfactory participation in approved continuingprofessional development (CPD) programmes assufficient proof of a practitioner’s competence andfitness to practise.The Royal New Zealand Collegeof General PractitionersThe Royal New Zealand College of GeneralPractitioners (RNZCGP) has developed a CPDprogramme, the Maintenance of ProfessionalStandards programme, for general practitioners.CPD programmes usually comprise continuingmedical education, continuous quality improvement,and peer review activities.Successful peer review entails being ‘assessed bythose who are both sufficiently informed to judgewhat they assess and sufficiently independent tojudge it objectively.’ 24 Peer review, if it is to be effectiveand to identify strengths and weaknessesand determine competence, must be judgmentaland demanding while also being supportive, andmust overcome self-protecting etiquette. AlthoughNew Zealand’s professional accountabilityprocesses include elements of peer review, currentpeer review processes have more educational valueand/or provide collegial support. They will needto be strengthened if they are to provide satisfactoryaccountability. It is to be hoped that theproposed practice visits are up to the task.The RNZCGP does not confine itself to theongoing competence of vocationally registeredgeneral practitioners, but is also interested inthe context in which practitioners work. TheRNZCGP has set out the standards expected inpractices in its Aiming for Excellence publicationand has developed the CORNERSTONE practiceaccreditation programme. 25There are a number of other organisations with arole to play in ensuring that the right thing happensin health care, including the now defunct168 VOLUME 5 • NUMBER 2 • JUNE 2013 J OURNAL OF PRIMARY HEALTH CARE

VIEWPOINTACKNOWLEDGEMENTSThe Centre for AdverseReactions Monitoringin the New ZealandPharmacovigilance Centreis funded by Medsafe, NewZealand Ministry of Health.COMPETING INTERESTSNone declared.continued reporting thereafter. 13 The reportsoften described elderly patients with multiplecomorbidities, and higher doses of simvastatin,the most commonly prescribed statin at this time;also, co-prescription with diltiazem featured inseveral reports. This medicine is only a weakinhibitor of CYP3A4 and was not considered tointeract sufficiently with statins to be a problem.However, there was increasing use of statins athigher doses in keeping with new guidelinesfor intensive lowering of LDL cholesterol forprimary and secondary prevention of ischaemiccardiovascular events. A published case report 14and a combination of NZ and Australian adversereaction reports 10,11 suggested that diltiazem wascontributing to inhibition of simvastatin metabolismas daily doses of simvastatin increased.In the context of increasing reports of rhabdomyolysis,an analysis of clinical trials of high-doseversus moderate-dose statin therapy revealed thatthe rate of myopathy and rhabdomyolysis with80 mg simvastatin daily was approximately fourtimes greater than with 80 mg atorvastatin orlower doses of simvastatin. 15These developments indicate the need to continuemonitoring throughout what is calledthe ‘life cycle’ of a medicine and for reportingserious adverse reactions, even if they are alreadyknown. In 2011, the US FDA advice limitingthe use of simvastatin 80 mg daily was based onthe accumulated evidence concerning the risk ofrhabdomyolysis together with evidence of littleextra benefit with this dose compared with lowerdoses. 16 Thus, practice-based evidence from adversereactions reports and the research they havestimulated has led to advice that can minimisethe risk of a very serious adverse reaction so thatthose most likely to benefit are prescribed thesemedicines and medicine interactions are avoided.This is not the end of the story. Trisha Greenhalgh,in her article on evidence-based medicine,discussed Aristotle’s concept of phronesis orpractical wisdom. 17 In the context of this conceptand statin use, we have observed in New Zealandreports that very elderly or very ill patients hadbeen taking a statin for many years before theydeveloped rhabdomyolysis and that it appearedto have been triggered by concomitant diseaseor medicine interactions as the patients becameolder or more unwell. Some of these patientshad malignancies and had urgently required amacrolide antibiotic or imidazole antifungal agentwhile taking an interacting statin. This led usto suggest that careful consideration be givento the relevance of the five-year risk estimateof cardiovascular events to these patients andthe need to consider on a case-by-case basis theneed for a statin and at what point the risks of avery painful distressing event might outweighpotential benefit.Assessing causality and report qualityIt can be correctly argued that for each individualpatient it is impossible to know all the variablesthat may have led to their adverse experience andtheir relative contributions. For example, wouldan individual patient have experienced haemorrhagewith warfarin even if they were not takingaspirin? Did they also take an over-the-counternon-steroidal anti-inflammatory drug? ProfessorArroll and colleagues 18 in the June 2012 issue ofthis journal discussed probabilistic reasoning indiagnosis and management of individual patientsand this is what is also applied to assessment ofindividual and grouped adverse reaction reports.However, within the patient consultation, wetailor our questions, clinical examination andinvestigations to increase or decrease the likelihoodof a particular diagnosis. This opportunityis afforded only to the clinician who sees thepatient and not to those assessing adverse reactionreports. Adverse reaction reports that include theclinician’s reasoning, as well as details about variablessuch as other medicines and comorbidities,make these reports extremely valuable as practicebasedevidence.Flexibility not hierarchyIn summary, the practice-based evidence derivedfrom adverse drug reaction reports can, occasionally,be used alone to identify serious adversereactions. More often they generate or strengthenhypotheses that need testing in formal studies.Nevertheless, formal studies do not necessarilydiscount hypotheses arising from adverse reactionreports if they are insufficiently powered or notdesigned to detect the adverse effect. It is now172 VOLUME 5 • NUMBER 2 • JUNE 2013 J OURNAL OF PRIMARY HEALTH CARE

BOOK REVIEWapparent that adverse reaction reports also providegood insights into the environment in whichmedicines are used, the risk factors that mightlead to serious adverse effects and the effect ofadvice and guidelines in practice.References1. Parsonson B. The case for practice-based evidence tosupport evidence-based practice. J Primary Health Care.2012;4(2):98–99.2. Kunac DL, Harrison-Woolrych M, Tatley MV. Pharmacovigilancein New Zealand: the role of the New Zealand PharmacovigilanceCentre in facilitating safer medicines use. N Z MedJ. 2008;121(1283):76–89.3. Adverse Drug Reactions Advisory Committee (ADRAC).Withdrawal of lumiracoxib in Australia. Australian AdverseDrug Reactions Bulletin. 2008;27(2):6–7.4. Bombardier C, Laine L, Reicin A, Shapiro D, Burgos-Vargas R,Davis B et al. Comparison of upper gastrointestinal toxicity ofrofecoxib and naproxen in patients with rheumatoid arthritis.N Engl J Med. 2000;343:1520–8.5. Kearney PM, Baigent C, Godwin J. Do selective cyclooxygenase-2inhibitors and traditional nonsteroidal anti-inflammatorydrugs increase the risk of atherothrombosis? Meta-analysisof randomised trials. BMJ. 2006;332:1302–1308.6. McGettigan P, Henry D. Cardiovascular risk and inhibition ofcyclo-oxygenase: a systematic review of the observationalstudies of selective and non-selective inhibitors of cyclooxygenase2. JAMA. 2006;296:1633–1644.7. New Zealand Guidelines Group. Management of dyspepsia andheartburn. Wellington: New Zealand Guidelines Group; 2004.8. Camm AJ, Kirchhof P, Lip GYH, Schotten U, Savelieva I, Ernst S,et al. The task force for the management of atrial fibrillation forthe European Society of Cardiology. Guidelines for the managementof atrial fibrillation. Eur Heart J. 2010; 31:2369–2429.9. Graham DJ, Staffa JA, Shatin D, Andrade SE, Schech SD,La Grenade L, et al. Incidence of hospitalized rhabdomyolysisin patients treated with lipid-lowering drugs. JAMA.2004;292(21):2585–2590.10. Savage R, Tatley M. Myopathy with statins: check CK levelsand interactions. Prescriber Update. 2004;25(1):4–5.11. Adverse Drug Reactions Advisory Committee (ADRAC). Riskfactors for myopathy and rhabdomyolysis with statins. AustralianAdverse Drug Reactions Bulletin. 2004;23(1):2.12. Maggo SDS, Kennedy MA, Clark DWJ. Clinical implications ofpharmacogenetic variation on the effects of statins. Drug Saf.2011;34(1):1–19.13. Beggs PW, Clark DJW, Williams SM, Coulter DM. A comparisonof the use, effectiveness and safety of bezafibrate,gemfibrozil and simvastatin in normal clinical practice usingthe New Zealand Intensive Medicines Monitoring Programme(IMMP). Br J Clin Pharmacol. 1999;47(1):99–104.14. Gladding P. Potentially lethal interaction between diltiazemand statins [Letter]. Ann Int Med. 2004;140(8):676.15. Davidson MH, Robinson JG. Safety of aggressive lipid management.J Am Coll Cardiol. 2007;49(17):1753–62.16. FDA Drug Safety Communication. New restrictions, contraindications,and dose limitations for Zocor (simvastatin)to reduce the risk of muscle injury. [cited 2012 November2]. Available from: http://www.fda.gov/Drugs/DrugSafety/ucm256581.htm.17. Greenhalgh T. Why do we always end up here? Evidencebased medicine’s conceptual cul-de-sacs and some off roadalternative routes. J Prim Health Care. 2012;4(2):92–97.18. Arroll B, Allan GM, Elley CR, Kenealy T, McCormack J, HudsonB, et al. Diagnosis in primary care: probabilistic reasoning. JPrim Health Care. 2012;4(2):166–73.Doctor Colenso, I presume:An account of missionary medicalpractice in New Zealand in themidnineteenth centuryIllustrated by the work of the Rev. William Colenso FLS FRS inthe Wairarapa and Hawke’s BayIan St GeorgeReview by Derek Dow, The University of AucklandEmail: d.dow@auckland.ac.nzIan St George’s tribute to William Colenso’s medical work was publishedto mark the bicentenary in 2011 of its subject’s birth. Colensois presented as a nineteenth century polymath, described by the authoras a ‘printer, missionary, explorer, politician, botanist, educationalist,liberation theologist—and importantly herein, healer and dispenserof medicines’. The extent of his medical involvement is highlightedby Colenso’s claim in 1897 that at one time he had the ‘most completesurgery in NZ’.The text consists of lengthy extracts from Colenso’s writing, bothpublished and unpublished, and from secondary sources, with linkingpassages by the author. This is underpinned by extensive footnotes, contextualisingthe story. St George outlines Colenso’s medical education,such as it was, and his practice while employed by the Church MissionarySociety from 1834 to 1852. It also includes details of his own illhealth, and his opposition to the consumption of tobacco and alcohol.Colenso’s entry in the Dictionary of New Zealand Biography makes nomention of Colenso’s role as a healer. Ian St George’s work is a welcomecorrective to this omission.Place of publication: WellingtonDate of publication: 2012No. of pages: 59ISBN 978-0-9876604-1-1Copies are available for $10 from the Colenso Society, 32 Hawkestone St, Thorndon,Wellington 6011 (Email: istge@yahoo.co.nz)VOLUME 5 • NUMBER 2 • JUNE 2013 J OURNAL OF PRIMARY HEALTH CARE 173

LETTERS TO THE EDITORAspirin for primary prevention: Noread with interest the article in the Journal of Primary HealthI Care about aspirin and its use in primary prevention. 1 I notethat there is still a recommendation to consider aspirin inthose with a cardiovascular risk over 15%. I am not sure wherethe evidence for this statement comes from.In most other countries, the use of aspirin in primaryprevention has fallen rapidly. In the UK it is no longer used,following the BMJ article ‘Don’t use Aspirin for PrimaryPrevention’. 2 This position has been reaffirmed by KausikRay’s meta-analysis 3 —here are his thoughts on the matter inan interview with Medscape. 4Medscape: After the results of the ATT meta-analysis were published,and now with these new data just published, the prevailingmessage already seems to be that aspirin should be abandonedas primary prevention altogether. But other investigators havesuggested that the benefits of long-term use of daily aspirin forprevention of chronic disease may outweigh the consequencesassociated with the increased risk for bleeding, particularly gastrointestinalbleeding.Prof. Ray: They are wrong. If you had a bleed in your eye I wouldsay that is pretty important. If you had a bleed into your brain thatdidn’t kill you, I would say that is also important, and obviouslyfatal bleeds are included as well. And if you come into hospitalneeding a blood transfusion, are you likely to take aspirin again?No. What is your risk? It is preventing a heart attack that wouldn’thave killed me; that is how people need to think about this. If yougive someone a statin, you are reducing cardiovascular deaths andyou are also reducing nonfatal MIs. There is really no flip side,apart from dysglycemia and nonfatal side effects like myalgia. Thatis not the same as bleeding. This is the information patients shouldbe given. Hopefully, it is very clear: aspirin is not the same as theantihypertensive, and it is not the same as the statin.Recent European Society of Cardiology guidelines alsodon’t recommend using aspirin for primary prevention forcoronary heart disease. 5 This is the line also adopted by theAustralian Heart Foundation. 6I wonder if it could be explained how the figure of a CVrisk of 15% as the threshold for prescribing aspirin was arrivedat and where the evidence of benefit for it is derived from?Bill Cartledge, general practitioner, Avon Medical CentreStratford, Taranaki; bill.cartledge@phcl.health.nzReferences1. Bryant L. Aspirin—yes?—no?—maybe? J Prim Health Care. 2012;4(4):344.2. Barnett H, Burrill P, Iheanacho I. Don’t use aspirin for primary prevention. BMJ2010;340:c1805.3. Seshasai SR,Wijesuriya S,Sivakumaran R, Nethercott S, Erqou S, Sattar N, elal. Effect of aspirin on vascular and non vascular outcomes; meta-analysis ofrandomized controlled trials. Arch Intern Med. 2012 Feb 13;172(3):209–16.4. Brookes L, Ray K. Aspirin in primary prevention. [Cited 2013 Feb]. Availablefrom: www.medscape.com/viewarticle/7598835. Perk J, De Backe G, Gohlke H, Graham I, Reiner Z, Verschuren M, et al.European Guidelines on cardiovascular disease prevention in clinical practice(version 2012). The Fifth Joint Task Force of the European Society of Cardiologyand Other Societies on Cardiovascular Disease Prevention in Clinical Practice(constituted by representatives of nine societies and by invited experts). EurHeart J. 2012 33:1635–1701.6. Australian Heart Foundation. Coronary heart disease. [Cited 2013 Feb].Available from: www.heartfoundation.org.au/information-for -professionals/Clinical-Information/pages/coronary-heart-disease.aspx.Author’s ResponseThank you, Dr Cartledge, for raising this controversial aspect of aspirinfor primary prevention. There is good evidence for aspirin after a cardiovascularevent, and evidence that risk outweighs benefit for peoplewho have not had an event. The dilemma is around those people whoare at high risk of a cardiovascular event, but have not had an event yet.For many years the New Zealand guidelines have taken a holistic approachto cardiovascular risk and when to treat along the continuumof risk, recommending the addition of pharmacological treatment ata calculated cardiovascular risk greater than 15%. This has added tothe complexity because the research usually focuses on one medicalcondition—high blood pressure, dyslipidaemia or use of an antithrombotic.This means that in primary prevention trials, and particularly theolder trials, there is no analysis according to level of holistic cardiovascularrisk—an unfortunate gap in the evidence.Bringing this into the New Zealand concept of treating high cardiovascularrisk, a pragmatic approach is to treat cardiovascular risk greaterthan 15% with statins and/or blood pressure–lowering medicines, andif the risk continues above 15%, then add aspirin to reduce the risk afurther 1%. 1The work by Selak et al. 1 reviewed the risks and benefits of aspirin accordingto age. A rider that was not included in the article, but shouldhave been, is that harm may outweigh benefit for primary preventionand a cardiovascular risk over 15% for men over 80 years old. Usualclinical judgement applies for those with a history of peptic ulcerdisease or bleeding disorders.Linda BryantReference1. Selak V, Elley C, Wells S, Rodgers A, Sharpe N. Aspirin for primary prevention:yes or no? J Prim Health Care. 2010;2(2):92–99.Letters may respond to published papers, briefly report original research or case reports, or raise matters of interest relevant toprimary health care. The best letters are succinct and stimulating. Letters of no more than 400 words may be emailed to:editor@rnzcgp.org.nz. All letters are subject to editing and may be shortened.174 VOLUME 5 • NUMBER 2 • JUNE 2013 J OURNAL OF PRIMARY HEALTH CARE

LETTERS TO THE EDITORCommunication with non-English-speaking patientsThe article from Ete-Rasch and Nelson 1 was a great additionto the literature examining the reasons behind theadmission of Pacific children to hospital because of severeskin infection, and noting the ethnic disparities in admissions.The researchers noted that four of the 11 mothers hadEnglish as a second language, and that the research interviewwas conducted in Samoan for three of those women. Unfortunatelyno information was provided about how thesewomen managed to communicate with their primary careservice. Did they believe that their limited English interferedwith the consultation? Did they have a language-concordantclinician? Was an interpreter used; if so, was it a familymember or a professional? One of the conclusions related tothe provision of health information. This is extraordinarilydifficult if the clinician does not share a language withthe patient. Written information is not necessarily of use,because without a shared language an assessment of literacycannot be done. Interpreter services are significantly underutilisedin New Zealand. 2 With DHB funding of interpretersin primary care in Auckland and PHO funding in manyother areas it is important that, in research projects such asthis, the important issue of adequacy of communication isproperly assessed, and that we all expect nothing less thanthe same level of communication with our Pacific patientsas we expect with English-speaking patients, by the use of aprofessional interpreter if necessary.Dr Ben Gray, FRNZCGP MBChBSenior Lecturer, Primary Health Care and General PracticeUniversity of Otago WellingtonReferences1 Ete-Rasch E, Nelson K. Management of skin infections of Pacific childrenprior to hospitalisations. J Prim Health Care. 2013;5(1):43–51.2. Gray B, Hilder J, Donaldson H. Why do we not use trained interpreters forall patients with limited English proficiency? Is there a place for using familymembers? Aust J Prim Health. 2011;17(3):240–249.Authors’ ResponseDr Gray has raised a relevant point that was not covered in our article.However, while the study didn’t pursue the use of interpreters forthe four mothers with English as their second language, mothersthemselves didn’t mention this during the interviews. Therefore, wecan only assume interpreters were not made available to them. In ourview, the use of interpreters would probably not have made a differencein the outcome of the study as mothers’ experiences were similarwhether English was their first or second language.We acknowledge that health literacy is a barrier to health care and theuse of interpreters is one way of addressing this. However, as writtenhealth information available in other languages is usually translatedfrom the English versions, we consider clinicians should be familiarwith the information provided. Such written information should notonly be available at consultation but in the community generally. Weapplaud Dr Gray’s support in ensuring that people with English as asecond language are provided with the same level of care as everyoneelse by encouraging the use of interpreters. While spoken languageis important in communication and understanding, culture also hasa role. Language and culture go hand in hand. Understanding andinterpreting something spoken in one language (English, for example)can be understood and interpreted differently by people from differentcultures and background. Shared language together with sharedunderstanding is better for better health outcomes.Elaine Ete-Rasch and Katherine NelsonAn alarming symptomwas asked to visit an 80-year-old lady complaining of unilateraltinnitus. She suffered with cardiac failure and had beenIreceiving home visits for a while on account of her reducedmobility. She also had a past history of contralateral mastoidsurgery and subsequent hearing loss. New auditory symptomsin her one good ear were understandably of particular concernfor her.Upon letting me in to her three-bedroom semi-detachedhouse she described her symptoms. She had been experiencingan intermittent high-pitched beeping sound over the past 24hours. Intrigued I unpacked my auroscope to examine further.As I peered towards a healthy looking tympanic membraneI was surprised to hear a beep for myself. Looking directlyabove the patient’s sofa there was a smoke alarm flashing; asecond beep a minute later clinched the diagnosis.The role of home visits has been a subject of ongoingdebate. 1 However, this case served as a reminder of the possiblerole of home visits not only in serving our less mobile patients,but also in securing an unusual diagnosis that otherwise mighthave led to unnecessary further investigation. It also enabledthe important public health preventative measure of ensuringa functioning battery in the smoke alarm, even though itrequired the patient’s son to come round and fit it.Geryl Rees and John Whitaker, Heaton MerseyMedical Practice, 460 Didsbury Road, Stockport,Cheshire, SK4 3BT. johnwhitaker@doctors.net.ukReferences1. Theili G, Kruschinski C, Buck M, Müller CA, Hummers-Pradier E. Home visits—central to primary care, tradition or an obligation? A qualitative study. BMC FamPract. 2011;12:24.VOLUME 5 • NUMBER 2 • JUNE 2013 J OURNAL OF PRIMARY HEALTH CARE 175

ABOUT THE JOURNAL OF PRIMARY HEALTH CAREThe Journal of Primary Health Care(JPHC) began publishing in 2009,superseding the previous RNZCGPjournal the New Zealand Family Physician.It is a interdisciplinary publication aimed atmoving research into primary health carepractice and practice into research. Thisincludes the fields of family practice, primaryhealth care nursing and community pharmacyas well as areas such as health care delivery,health promotion, epidemiology, public healthand medical sociology of interest to a primaryhealth care provider audience. It is positionedas relevant to countries within the Pacific rim.JPHC publishes peer-reviewed quantitativeand qualitative original research, systematicreviews, papers on improving performanceand short reports that are relevant to itsprimary health care practitioners. For the aim,scope, instructions to authors and templatesfor publications see www.rnzcgp.org.nz/journal-of-primary-health-care/.JPHC includes pithy digests of the latestevidence including a String of PEARLS(Practical Evidence About Real LifeSituations), Potion or Poison? (evidence forthe potential benefits and possible harms ofwell-known herbal medicines), CochraneCorner (a summary of a Cochrane review),Nuggets of Knowledge (succinct synopsesof pharmaceutical evidence for primary care)and Pounamu and Vaikoloa, (Maori and Pacificprimary health care treasures respectively).JPHC publishes viewpoints, commentariesand reflections that explore areas ofuncertainty on aspects of care for whichthere is no one right answer. Debate isstimulated in Back to Back, where twoprofessionals present their opposing viewson a topic. There is a regular Ethics column.Letters to the Editor are welcomed.INDEXINGThe Journal is indexed in MEDLINE, ExcerptaMedica (EMBASE), Cumulative Indexto Nursing and Allied Health Literature(CINAHL), Scopus and Index New Zealand(INNZ). It is also included in the Directoryof Open Access Journals (DOAJ), http://www.doaj.org). Complete text of the Journalis available online at www.rnzcgp.org.nz/journal-of-primary-health-care/ and throughvarious aggregators including PubMedCentral and EBSCO.EDITORProf. Felicity Goodyear-Smith: Professor andGoodfellow Postgraduate Chair, Departmentof General Practice and Primary Health Care,University of Auckland, Auckland, NewZealand; editor@rnzcgp.org.nzEDITORIAL BOARDThe Editorial Board comprises renownedand active primary care clinicians, clinicaland scientific academics and health policyexperts with both New Zealand andinternational representation.Prof. Bruce Arroll: Professor, Departmentof General Practice & Primary Health Care,University of Auckland, NZProf. Jenny Carryer: Professor of Nursing,School of Health and Social Services, MasseyUniversity, Palmerston North, NZProf. Peter Crampton: Pro-Vice-Chancellor,Division of Health Sciences, School ofMedicine and Health Sciences, University ofOtago, Dunedin, NZDr Ofa Dewes: Research fellow, Departmentof Pacfic Health, School of Population Health,University of Auckland.Prof. Tony Dowell: Professor, Departmentof Primary Health Care and General Practice,Wellington School of Medicine, University ofOtago, NZMs Eileen McKinlay: Senior Lecturer inPrimary Health Care, Department of PrimaryHealth Care and General Practice, Universityof Otago Wellington, NZProf. Pauline Norris: Profesor and Chairin Social Pharmacy, University of Otago,Dunedin, NZDr Barry Parsonson: Psychologist for NZMinistry of Education, Napier, NZDr Shane Reti (QSM): International ProgramDirector Clinical Informatics and CEO ofClinical Informatics Industrial Research,Harvard Medical School, USAProf. Kurt Stange: Professor of FamilyMedicine, Case Western Reserve University,Cleveland, OH, USA and Editor, Annals ofFamily MedicineSUBMISSIONSFull instructions for authors can be found at http://www.rnzcgp.org.nz/information-for-authorsPlease send all submissions to the Editor: editor@rnzcgp.org.nzFor queries about submitted articles please contact the Managing Editor: managingeditor@rnzcgp.org.nzJPHC ADMINISTRATIONFor subscription and advertising queries, or to sign up for email alerts, please contact the Publications Coordinator:RNZCGP, PO Box 10440, The Terrace, Wellington 6143, New Zealand; jphcnz@rnzcgp.org.nzJPHC is printed on uncoated, acid-free paper which meets the archival requirements of ANSI/NISO Z39.48-1992 (Permanence of Paper) and isForest Stewardship Council (FSC)–certified which meets the highest environmentally responsible standards.The Journal of Primary Health Care is the official journal of the RNZCGP. However, views expressed are not necessarily those of theCollege, the Editor, or the Editorial Board. ©The Royal New Zealand College of General Practitioners 2013. All Rights Reserved.176 VOLUME 5 • NUMBER 2 • JUNE 2013 J OURNAL OF PRIMARY HEALTH CARE