Time to Intervene? - National Confidential Enquiry into Patient ...

acknowledgementsAcknowledgementsThis report, published by NCEPOD, could not havebeen achieved without the support of a wide range ofindividuals who have contributed to this study.Our particular thanks go to:The Expert Group who advised NCEPOD on what toassess during this study:Dr Andy Lockey Consultant in EmergencyMedicineDr David Pitcher Consultant CardiologistDr Jerry NolanConsultant Anaesthetist andResuscitation Council UKDr Linda Twohey Consultant in Anaestheticsand Intensive CareMr Ken Spearpoint Council for Professionals asResuscitation OfficersMrs Audrey Brightwell Lay representativeMs Dot TimmsClinical Operations ManagerResuscitation ServiceProfessor Tim Hendra Royal College of Physicians,NCEPOD Steering GroupMs Madeline Wang Patient Representative,NCEPOD Steering GroupThe Advisors who peer reviewed the cases:Professor Jennifer Adgey Consultant CardiologistDr Francis Andrews Consultant in CriticalCare/Emergency MedicineDr Meng Aw-Yong Associate Specialist inEmergency Medicine, MedicalDirector Metropolitan PoliceServiceDr Jeffrey BallConsultant Cardiologist/General PhysicianDr Ian BlakelyResuscitation OfficerDr James Bromilow Consultant in Anaesthesia andIntensive Care MedicineDr Kirsty Challen Specialist Registrar inEmergency MedicineDr Mukesh Chugh Consultant AnaesthetistProfessor Charles Deakin Consultant in CardiacAnaesthesia and IntensiveCare MedicineDr Michael Duffy Consultant Anaesthesia andIntensive Care MedicineDr David A Gabbott Consultant AnaesthetistMrs Sue Hampshire Senior Resuscitation OfficerDr Jeffrey Handel Consultant AnaesthetistMs Zoe Hayman Lead Resuscitation OfficerDr Katherine Henderson Consultant in EmergencyMedicineMs Kelly Hughes Resuscitation OfficerDr Katharine Hunt Consultant AnaesthetistMr Matthew Ibrahim Senior Resuscitation OfficerMr Dennis Jobling Resuscitation OfficerMr David Jones Resuscitation ManagerDr Natasha Joshi Consultant in Anaesthesia andIntensive Care MedicineDr Abhiram Mallick Consultant AnaesthetistDr Janet McComb Consultant Cardiologist3

forewordForewordThis may be the most important report that NCEPOD hasproduced in the last 10 years. I say this because it revealsthe need for a closer definition of the aims of treatment ofsick people in our hospitals.Our study reflects a belief that there is disparity betweenthe public and professional perception of the outcomeof cardiopulmonary resuscitation (CPR). Televisionmelodrama and the lay press convey a much moreoptimistic view of the whole process than professionalsbelieve to be realistic. The public believe that patientshave a 50:50 chance of surviving, where the professionalsaccept that survival to discharge is less than 15% † . Nordoes public appreciation factor in the chance that survivalwill often involve disability.Professionals may know otherwise, but they rarely find theopportunity to articulate their knowledge and when theydo, they frequently do not take sufficient account of themorbidity involved in their assessment of the prospectsof success. In short, too many of us have driftedinto an expectation that death will provoke a physicalintervention as part of a last ditch attempt to prolonglife. As a result the professions fail to give an appropriatepriority to their obligation to define the objects of theexercise, “the ceilings of treatment.” It is trite theory tosay that these should be decided by doctors and patientstogether where practical, and by doctors accepting theirresponsibility to take decisions in their patients’ bestinterests where it is not. In practice it seems that nodecision at all is taken in the overwhelming majority ofcases, and CPR is too commonly an instinctive responseto an unforeseen emergency.There are a number of reasons for hoping that this Reportwill prompt us all to step back and think again. Someof the findings are surprising, others reflect a failure tothink more carefully about things that are hardly counterintuitive.For example, this study found that the survivalrate of 10-15% is a very broad brush figure that cannotbe applied to individual cases without further inquiry.Cases that respond to defibrillation after being diagnosedimmediately are more likely to recover than people whoarrest as a result of some other condition, some daysafter being admitted to hospital or at a time when they arenot being closely monitored. No-one should be surprisedto find that the latter group’s chance of surviving todischarge is less than half of those who arrest in theEmergency Department (ED) or the Operating Theatre.A moment’s thought will also tell us that survival todischarge after such an experience does not necessarilyentail returning to live in one’s own home with the samedegree of autonomy as before. This study does not followup the quality of survival beyond noting the number whoreturned to their own home, which was greatly reduced.However it seems that if you have a cardiac arrest late atnight, a week or so after you were admitted to hospitalwith a non-cardiac disease, your chance of ever gettingback to your own home must be vanishingly small.If this study forces people to confront in a practical waythe limits of what is possible against that background, itwill have done us a service, but the true lessons go muchfurther.It is well established that surgeons who operate withoutthe informed consent of their patients are guilty of anassault and will be held to have acted unethically in theeyes of the General Medical Council. There is no basisfor asserting that different considerations apply to CPR:certainly there are emergency circumstances in which† Roberts D Hirschman D Scheltemak Adult & Pediatric CPR: Attitudes and Expectations of Health Professionals and Lay Persons,An J Emerge Med 2000; 18: 465/85

foreworda doctor is entitled to assume that the patient wouldwish an attempt at CPR to be made. But that cannotdefend the failure over a period of several days to findout what the patient’s wishes may be, or where this is notpossible, to determine the team’s view of the patients’best interests. The surgeon will rightly operate when wearrive in the ED unconconscious after a road traffic cash,but no-one supposes that as a result this entitles them tooperate without our consent on another occasion.It was in the hope of finding out how far that ethicalobligation sounds in modern medical practice that Iapproached this report. Alas, the results are profoundlydisappointing and as I read these pages I wonderedhow many of these interventions would be defensibleif charged as assaults before the criminal courts, or asprofessional misconduct before the GMC. The GMCrecognises that CPR should be administered in anemergency, but it is not good medical practice to fail toanticipate the needs of the patient before an emergencyarises. If the failure is deliberate or reckless then Isuggest that it is arguably criminal.Our advisors did find that in a substantial number ofcases resuscitation was attempted when they thoughta DNACPR (Do Not Attempt CPR) decision should havebeen made earlier. The first reason is that in too manycases the failure to formulate an appropriate plan reflectsthe fact that the assessment on admission was judgedto be deficient overall in 47% of these patients (Table3.23). This is the reality of modern medicine as revealedby other NCEPOD reports. Despite repeated calls for allacute admissions to be seen by a Consultant within 12hours, ‡ our Advisors could only identify an appropriatetimely Consultant review in 40% of cases. The dayswhen the assessment of acute admissions could beleft to juniors have gone, because the juniors are lessexperienced, the patients are often more ill and theexpectations of society are less tolerant.Even once that proposition is accepted, it is stillthoroughly dispiriting to find that in 38% of cases thecardiac arrest could have been prevented altogether.If the patient had been managed as they should havebeen, our Advisors thought that the arrest would not havehappened in the majority of cases. This is bad medicineinvolving the avoidable death of patients with consequentlessons for the professions.A decision about whether resuscitation should or shouldnot be attempted was documented on admission in only10% of cases. Our Advisors thought DNACPR shouldhave been documented in a further 20%. In short,resuscitation was wrongly attempted in many of thesecases because nobody had recognised that they were indanger of a cardiac arrest.When we looked at the decision tree the picture was evenworse. Of the 22% of patients whose resuscitation statushad been determined before their arrest, 42% had aDNACPR decision. This finding caused particular dismayto our advisors and authors, but it is hardly surprising.The fact that no decision has been taken may have fed anexpectation it need not be considered in the emergency.In the overwhelming majority of cases the question ofCPR was not raised with the patient before the arrest,which suggests that this may be cultural rather than theproduct of a deliberate decision in each case. Somecase reports said this happened because there wasno opportunity, a proposition our advisors struggled toreconcile with some of the intervals that elapsed betweenadmission and the arrest.This report suggests that today we stand at a crossroads.To the left lies a destiny familiar from America where60% of us will die in an ICU and we will spend 50% ofNHS expenditure in the last six months of life, much of itseeking to postpone the inevitable. This will happen, notbecause the patient has asked for it or because someonehas taken a calculated decision that it is in the patient’sinterest to make the attempt, but because the doctorsthink that they have a duty to do everything that they canto prolong the process of dying.‡ NCEPOD Studies: ‘A Journey in the Right Direction?’ 2007, ‘Adding Insult to Injury’ 2009 and the Royal College of Physicians‘Acute Care Toolkit 2’.6

forewordTo the right lies an acceptance of the limits of what ispractical and a recognition that the armamentarium ofmedicine should be deployed only where it is likely tobenefit the patient. In an age of unprecedented respectfor the patient’s autonomy, wherever possible a contractshould be formed. This means that we should be giventhe opportunity of deciding what is to be done in theevent of a cardiac arrest after admission to hospital. If weare unable to take part in such a discussion a decisionshould be taken by others as to whether it is in our bestinterests.As a lawyer, when I accept a client I have a professionalduty to agree with them in writing what the objectivesare, what the scope of my authority is to be and whatthe risks and benefits may be. Our increasing respect forthe autonomy of the patient ought to lead the medicalprofession to embrace a similar course. The right toconsent to treatment is not confined to surgery. The“ceilings of treatment” should be described and agreedwhenever possible. The patient’s views on resuscitationshould be canvassed wherever possible and appropriate.Where it is not possible and the doctors have to act inwhat they perceive to be the patient’s best interests, thatshould not be interpreted as an obligation to administerCPR to the dying. The doctor’s right to act in this contextis coterminous with his duty: a patient may only beresuscitated where they have consented to the processor the doctor is satisfied that it is in their best interests.This report should be a wake-up call to the NHS. It isover a century since Osler urged trainees to “listen to thepatient, he is telling you the diagnosis.” Today we mustadd that if you have the humility to ask, they will often tellyou what they want you to do about it as well.One final thought. The negative connotations of Do NotAttempt CPR orders may be associated with a concernthat other aspects of care will be compromised. It hasbeen suggested that what people are really trying say isthat when the inevitable occurs they should Allow NaturalDeath “AND,” or Allow Dignified Death “ADD”.Mr Bertie Leigh, Chair of NCEPOD7

Back to contentsPrincipalRecommendationsPrincipal RecommendationsPatient population, initial assessment and firstconsultant reviewCPR status must be considered and recorded for allacute admissions, ideally during the initial admissionprocess and definitely at the initial consultant reviewwhen an explicit decision should be made in this group ofpatients, and clearly documented (for CPR or DNACPR).When, during the initial admission, CPR is considered asinappropriate, consultant involvement must occur at thattime. (All Doctors)Care before the cardiac arrestWhere patients continue to deteriorate after nonconsultantreview there should be escalation of patientcare to a more senior doctor. If this is not done, thereasons for non-escalation must be documented clearlyin the case notes. (All Doctors)Resuscitation attemptEach hospital should ensure there is an agreed plan forairway management during cardiac arrest. This mayinvolve bag and mask ventilation for cardiac arrests ofshort duration, tracheal intubation if this is within thecompetence of members of the team responding tothe cardiac arrest or greater use of supraglottic airwaydevices as an alternative. (Medical Directors)Period after the cardiac arrestEach hospital should audit all CPR attempts and assesswhat proportion of patients should have had a DNACPRdecision in place prior to the arrest and should not haveundergone CPR, rather than have the decision made afterthe first arrest. This will improve patient care by avoidingundignified and potentially harmful CPR attempts duringthe dying process. (Medical Directors)Resuscitation statusHealth care professionals as a whole must understandthat patients can remain for active treatment but that inthe event of a cardiac arrest CPR attempts may be futile.Providing active treatment is not a reason not to considerand document what should happen in the event of acardiac arrest. (All Health Care Professionals)9

10PrincipalRecommendations

Back to contentsIntroductionIntroductionCardiopulmonary resuscitation (CPR) of patients canbe an important, life-sustaining intervention. It shouldbe remembered that CPR was originally developed tosave the lives of younger people dying unexpectedly,mostly from primary cardiac disease (the phrase ‘heartstoo young to die’ is often used). However CPR hascome to be seen as a procedure that should be usedfor patients as a therapy to restore cardiopulmonaryfunction and prolong life, irrespective of the underlyingcause of cardiac arrest. A high proportion of in-hospitaldeaths now involve CPR attempts, even when theunderlying condition and general health of the patientmakes success unlikely. In addition, even when thereis clear evidence that cardiac arrest or death are likely,decisions about the patient’s CPR status are notalways documented clearly. The result is that patientsmay undergo futile attempts at CPR during their dyingprocess. Improved knowledge, training, and do notattempt cardiopulmonary resuscitation (DNACPR)decision-making should improve patient care and preventthese futile and undignified procedures at the end of life.Patients for whom CPR cannot prolong life, but merelyprolong the dying process should be identified early.Rates of survival and complete physiological recoveryfollowing in-hospital cardiac arrest are poor. This appliesto patients in all age groups. It has been shown that fewerthan 20% of adult patients having an in-hospital cardiacarrest will survive to discharge. 1In the large study by Meaney et al 1 , the highest survivalrates were found in patients who have a ventricularfibrillation (VF) cardiac arrest (37% survival to hospitaldischarge). The commonest underlying cause of cardiacarrest in patients having a VF cardiac arrest is primarymyocardial ischaemia, however this group is the minorityof patients who have an in-hospital cardiac arrest(8718/51919 (17%) of cardiac arrests studied). Mostcardiac arrests occur in patients in general ward areasand are often unmonitored. The underlying cardiac arrestrhythm is usually asystole or pulseless electrical activity(PEA), and the chance of survival to hospital discharge isextremely poor (11%).Many in-hospital cardiac arrests are predictable eventsnot caused by primary cardiac disease. 2 In this group,cardiac arrest often follows a period of slow andprogressive physiological derangement that is oftenpoorly recognised and treated. 3 Identification of obviousmarkers of deterioration in patients who have a cardiacarrest was shown as far back as 1990 by Schein andcolleagues 4 and has been subsequently demonstratedin other publications. 5-8 Following work by NCEPOD 2 theNational Institute for Clinical Excellence produced clinicalguidance to promote recognition and management ofthe acutely unwell patient (NICE CG 50). 9 This guidelinewas based on the evidence of delayed recognition ofillness and that intervention could improve outcomes andreduce cardiac arrest rates. 10-13It was hoped that the changes put in place as a resultof NICE CG 50 would have improved processes of carefor acutely ill patients and that this would be evident inpatients who had a cardiac arrest.This study was originally planned to start in 2007,however, due to the publication of NICE CG 50 anda report by the National Patient Safety Agency onrecognising the acutely ill patient, 14 the study wasdeferred to allow these documents to be embedded intopractice.NB: Throughout this report the term ‘cardiac arrest’or ‘arrest’ will be used interchangeably to representcardiorespiratory arrest.11

Back to contents1 - Method andData returns1 – Method and data returnsExpert groupA multidisciplinary group of experts comprisingconsultants from emergency medicine, intensive care,anaesthesia, cardiology, nursing, resuscitation officers,representatives from the Resuscitation Council (UK), a layrepresentative and a scientific advisor contributed to thedesign of the study.AimTo describe variability and identify remediable factorsin the process of care of adult patients who receiveresuscitation in hospital, including factors which mayaffect the decision to initiate the resuscitation attempt,the outcome and the quality of care following theresuscitation attempt, and antecedents in the preceding48 hours that may have offered opportunities forintervention to prevent cardiac arrest.ObjectivesBased on the issues raised by the expert group, theobjectives of this study were to collect information onthe following:1. The organisational structures and governance inplace to provide resuscitation, including training andthe uptake of training by members of staff.2. The structures in place to identify patients at riskof a cardiac arrest, and so identify opportunities tointervene.3. Outcome following resuscitation.4. DNAR/DNACPR status of patients who have hada cardiac arrest and describe the appropriatenessof resuscitation with regard to each patient whoreceived CPR.5. The process of the resuscitation attempt, and sodifferentiate between the organisational structuresin place to provide resuscitation, and what actuallyhappens.6. The quality of care in the 48 hours prior to cardiacarrest.7. The quality of care in the post-resuscitation period.PopulationAll adult patients who had a cardiac arrest, triggeringeither a call to the resuscitation team (or equivalent) via2222 (or the completion of an audit form subsequent tothe resuscitation attempt) that led to the delivery of chestcompressions or defibrillation during the 14 day studyperiod: 1st-14th November 2010 inclusive.Hospital participationNational Health Service hospitals in England, Wales andNorthern Ireland were expected to participate, as well ashospitals in the independent sector and public hospitalsin the Isle of Man, Guernsey and Jersey. Hospitals thattreated only children were not required to participate inthe study.Within each hospital, a named contact, referred to asthe NCEPOD Local Reporter, acted as a link betweenNCEPOD and the hospital staff, facilitating caseidentification, dissemination of questionnaires and datacollation.13

1 - Method andData returnsAdvisor groupA multidisciplinary group of Advisors was recruited toreview the case notes and associated questionnaires. Thegroup of Advisors comprised clinicians from the followingspecialties: emergency medicine, anaesthesia, critical care,cardiology/general medicine, and resuscitation officers.All questionnaires and case notes were anonymised bythe non-clinical staff at NCEPOD. All patient, clinicianand hospital identifiers were removed. Neither ClinicalCo-ordinators at NCEPOD, nor the Advisors had accessto such identifiers.After being anonymised each case was reviewed byone Advisor within a multidisciplinary group. At regularintervals throughout the meeting, the chair allowed aperiod of discussion for each Advisor to summarise theircases and ask for opinions from other specialties or raiseaspects of a case for discussion.The grading system below was used by the Advisors tograde the overall care each patient received.Good practice – a standard that you would accept foryourself, your trainees and your institutionRoom for improvement – aspects of clinical care thatcould have been betterRoom for improvement – aspects of organisationalcare that could have been betterRoom for improvement – aspects of both clinical andorganisational care that could have been betterLess than satisfactory – several aspects of clinical and/or organisational care that were well below satisfactoryInsufficient information submitted to assess the qualityof careQuality and confidentialityEach case was given a unique NCEPOD number so thatcases could not easily be linked to a hospital.The data from all questionnaires received wereelectronically scanned into a preset database. Priorto any analysis taking place, the data were cleaned toensure that there were no duplicate records and thaterroneous data had not been entered during scanning.Any fields in an individual record that contained spuriousdata that could not be validated were removed.Data analysisThe qualitative data collected from the Advisors’ opinionsand free text answers in the clinician questionnaireswere coded, where applicable, according to content toallow quantitative analysis. The data were reviewed byNCEPOD Clinical Co-ordinators and Clinical Researchersto identify the nature and frequency of recurring themes.Case studies have also been used throughout the reportto illustrate particular themes.All data were analysed using Microsoft Access and Excelby the research staff at NCEPOD.The findings of the report were reviewed by the ExpertGroup, Advisors and the NCEPOD Steering Group priorto publication.Study sample denominator data by chapterWithin this report the denominator used in the analysismay change for each chapter and occasionally withineach chapter. This is because data have been taken fromdifferent sources depending on the analysis required. Forexample in some cases the data presented will be a totalfrom a question taken from the clinician questionnaireonly, whereas some analyses may have required datafrom the clinician questionnaire and the Advisors’ viewtaken from the case notes.15

1 - Method andData returnsIn total 526 sets of case notes were assessed by theAdvisors. The remainder of the returned case noteextracts (54 sets) were too incomplete for assessment.The number of clinician questionnaires included in thestudy analysis is 585, the number of resuscitation formsincluded is 787 and organisational data were collectedfrom 460 hospitals, with 383 of these sites having adedicated resuscitation team on-site and completing thefull questionnaire. A further 200 hospitals returned dataon what happens in the event of a cardiac arrest.Data returnsFigure 1.1 shows the case returns for the study.Cases identifiedfor inclusion in thepeer review842Excluded cases103 *did not meetstudy criteriaIncluded cases739787 resuscitationforms returned739 clinicalquestionnairessent out739 case noterequestssent out98 not matchedto identifiedcases585 clinicalquestionnairesreturned526 casesreviewedby advisors689 formsmatchedto cases509 cases withclinicalquestionnaire andcase notesreturned512 cases withresuscitationform reviewed byadvisor564 cases withresuscitationform and clinicalquestionnaireFigure 1.1 Data returned16

Back to contents2 - OrganisationalData2 – Organisational dataThis chapter aims to provide an overview of theavailability of certain key facilities, policies and clinicalpathways that would be relevant to the care of patientswho have an in-hospital cardiac arrest. Organisationaldata were collected from participating hospitals thathad an on-site resuscitation team. Those withouta resuscitation team on-site were asked to provideinformation on what happens in the event of a patientsuffering a cardiac arrest.Table 2.1 gives details of the hospitals that participated.The ‘other’ group included community hospitals (159),rehabilitation centres (11) and peripheral site of trust (8).Table 2.1 Participating hospital typesHospital typeDistrict general hospital 500 beds 57University teaching hospital 62Private hospital 132Tertiary specialist unit 44Other type of hospital 185Total 593nTable 2.2 shows the type of hospital by whether or notit had a designated resuscitation team for adult patientson-site (hospitals with an on-site resuscitation teamcompleted the whole organisational questionnaire).Those without a resuscitation team tended to be privatehospitals and ‘other’ were identified as treatment centres,specialist hospitals and small community hospitals andother facilities where it would be difficult to provide onsiteresuscitation teams (192/200).Table 2.3 Response to a cardiac arrest where there was noon-site resuscitation teamResponse to a cardiac arrest n %Basic life support and 999 call 160 82.1Summon cardiac arrest team fromanother site 13 6.7Other (e.g. bleep holder) 22 11.2Subtotal 195Not answered 5Total 200Table 2.3 shows what action hospitals without an on-siteresuscitation team took in the event of a cardiac arrest.In most hospitals without a resuscitation team the actionwas to initiate basic life support and summon assistancevia the ambulance service.Table 2.2 Hospital type and whether they had an on-site resuscitation teamHospital had a dedicated on-site resuscitation teamHospital type Yes No Not answered TotalDistrict general hospital 500 beds 55 2 0 57University teaching hospital 60 2 0 62Private hospital 117 14 1 132Tertiary specialist unit 26 15 3 44Other 16 163 6 185Total 383 200 10 59317

2 - OrganisationalDataMost hospitals with a resuscitation team had only onesuch team (323/369 – 88.5%). However, the remainderhad multiple teams and 11 hospitals reported thatthey had four or more teams. It may well be that wheremultiple teams exist, the teams had extended roles andfunctions as it is unlikely that there are sufficient cardiacarrests to account for this number.Table 2.4 Type of outreach teams availableOutreach teams n %Critical care outreach 195 87.1Medical emergency team 52 23.2Rapid response team 22 9.8Answers may be multiple (n/224; not answered in 159)In addition to designated teams to respond to cardiacarrests, hospitals were asked if they had any of theoutreach teams detailed in Table 2.4. Of the hospitalsfrom which a response was received 224 indicatedthat they had an outreach team of some sort and somehospitals indicated that they had more than one type.There is possibly some overlap in the descriptors used inTable 2.4. However 159 hospitals (44.1%) did not indicatethat they had any form of outreach team. This is a keyelement of a systematic response to acute illness andraises the possibility that practice is not in keeping withNICE guidance on this topic. 9Table 2.5 shows the number of 2222 calls (or equivalent)to the resuscitation team between 1/1/09 and 31/12/09.Ninety-four hospitals had 100 or less calls in that 12month period, 137 hospitals had between 100 – 500calls and 16 hospitals had more than 500 calls. It mustbe remembered that many hospitals will use the samenumber (2222) to summon urgent help for patients whomay not have had a cardiac arrest but are consideredacutely unwell, and for other reasons, for examplesecurity breaches. Therefore not all these calls mayrepresent cardiac arrests, this is demonstrated in Figure2.1 which highlights the percentage of calls that wereactual arrests. In hospitals where the number of calls isvery low, opportunities for training will be reduced, andensuring that skills and knowledge are retained is animportant consideration. It was notable that 136 hospitalsdid not answer this question – raising the possibility thatdata on cardiac arrests were not collected.Number of hospitals3025201510500%1-10%11-20%21-30%31-40%41-50%51-60%61-70%71-80%81-90%91-100%Percentage of callsFigure 2.1 Percentage of 2222 calls reported between 1/1/09-31/12/09 that representedresuscitation attempts (n= 172, No 2222 calls or CPR attempts recorded in22 hospitals, data missing in 189 hospitals)18

2 - OrganisationalDataTable 2.5 Number of 2222 calls to the resuscitation teambetween 1/1/09 and 31/12/09Number of 2222 calls n %0 22 8.91-10 32 13.011-20 15 6.121-50 17 6.951-100 8 3.2101-200 35 14.2201-300 48 19.4301-400 32 13.0401-500 22 8.9501-1000 13 5.3>1000 3 1.2Subtotal 247Not answered 136Total 383Table 2.6 Number of actual CPR attempts between 1/1/09and 31/12/09Number of CPR attempts during a year n %0 53 22.91-10 49 21.211-20 7 3.021-50 10 4.351-100 27 11.7101-200 58 25.1201-300 19 8.2301-400 7 3.0401-500 1

2 - OrganisationalDataOf the hospitals from which a response to this questionwas received, 60/261 stated that a resuscitation officernever attended any cardiac arrest calls (23%) and afurther 79/261 hospitals responded that resuscitationofficers attended less than 20% of cardiac arrest calls(30%). It is hard to interpret these figures as the role ofresuscitation officers has evolved quite rapidly in somehospitals. Resuscitation officers now have a crucial role intraining, risk management, data collection and audit, buttheir value in demonstrating clinical skills and leadershipduring actual cardiac arrests should not be neglected.Table 2.7 Estimated or actual number of calls attended by aresuscitation officerEstimate/actual n %Estimate 153 59.3The actual number 105 40.7Subtotal 258Not answered 3Total 261Where hospitals provided details on the percentage ofcalls attended by a resuscitation officer it was asked ifthis was an estimated or actual number. Table 2.7 showsthat it was an actual number in 40% of hospitals, anestimate in 60% and unanswered in 125 hospitals. Againthis does not give assurance that data on cardiac arrestsare collected well in all hospitals.Training in management of a cardiac arrest is importantand hospitals should ensure that their staff are trainedappropriately to undertake this task. Table 2.8 showsif the hospital had any record of the resuscitationcompetencies of the resuscitation team.Overall 93/367 hospitals did not record the resuscitationcompetencies of their staff (25%). There was somevariation by type of hospital and private hospitals had thehighest recording rate of competency of staff (103/113:91%). It is difficult to understand how hospitals areconfident that staff have the appropriate skills if there wasno record of competencies in one in four hospitals.Table 2.8 Hospital had a record of the resuscitation competencies of their staffHospital type Yes No Unknown Subtotal Not answered TotalDistrict general hospital 500 beds 33 15 5 53 2 55University teaching hospital 33 24 0 57 3 60Private hospital 103 6 4 113 4 117Tertiary specialist unit 18 7 1 26 0 26Other 10 5 1 16 0 16Total 257 93 17 367 16 383Table 2.9 shows the response to the question ‘whatnumber do you call to summon the resuscitation team?’by type of hospital.20

2 - OrganisationalDataTable 2.9 Resuscitation team call number by type of hospitalHospital type 2222 Other Subtotal Not answered TotalDistrict general hospital 500 beds 52 2 54 1 55University teaching hospital 58 2 60 0 60Private hospital 66 39 105 12 117Tertiary specialist unit 25 1 26 0 26Other 13 2 15 1 16Total 317 52 369 14 383Of the hospital that responded 317/369 (86%) used 2222as the standard number for summoning the cardiac arrestteam. Of the 52 hospitals that did not use 2222, 39 wereprivate hospitals (75%). Where 2222 was not used 21hospitals used a bleep or cardiac arrest alarm and 24used an alternative telephone number (6666 was themost commonly used: 12 hospitals). It is recommendedthat 2222 should be the standard number used tosummon a cardiac arrest team 15 and in 14% of hospitalsthis had not been implemented. Whilst standardisationof the number used to summon the cardiac arrest teamis sensible it must be remembered that once the patienthas had a cardiac arrest the outlook is poor and mostopportunities to improve outcome are in the pre-cardiacarrest phase. 16EquipmentTable 2.10 Type of defibrillators usedDefibrillator type n %Automated external defibrillators exclusively 34 9.0Shock advisory defibrillators exclusively 28 7.4Manual defibrillators exclusively 3

2 - OrganisationalDataTable 2.12 Hospital had a policy that there was at least one trained member of staff able to perform basic life support and useAED and/or manual defibrillators on each wardHospital type Yes No Unknown Subtotal Not answered TotalDistrict general hospital 500 beds 39 12 2 53 2 55University teaching hospital 44 14 1 59 1 60Private hospital 114 2 0 116 1 117Tertiary specialist unit 21 3 2 26 0 26Other 14 1 1 16 0 16Total 308 59 7 374 9 383The specific question asked was ‘Is it policy that 24hours/day, 7days/week, there is at least one trainedmember of staff able to perform basic life support anduse an AED and/or manual defibrillators on each ward?’As can be seen 308/374 hospitals (82%) confirmed thatthis was the case. Private hospitals more commonlyanswered ‘yes’ than NHS hospitals – 57/59 cases wherethis was not policy were NHS hospitals.All hospitals that responded stated that there wasstandardised provision of drugs specifically for use inthe event of a cardiac arrest. And all hospitals apartfrom three stated that they had a policy for standardisedemergency equipment (trolley) contents for use inemergency situations.Table 2.13 Frequency that emergency trolleys were checkedFrequency n %After every resuscitation attempt 190 50.0Every shift 46 12.1Every 24 hours 313 82.4Once a week 21 5.5Less frequently 10 2.6Table 2.13 shows that only 190 hospitals checked theemergency trolley after every resuscitation attempt.Equipment problems, with regard to both function andavailability, are frequently reported incidents duringresuscitation 14 and if trolleys are not checked andrestocked after every use then these are more likelyto happen.Policies and documentationTables 2.14 and 2.15 show information on the use of earlywarning systems and linkage of early warning systems toescalation protocols.Table 2.14 Early warning system usedEarly warning system was used n %Yes 376 98.9No 4 1.1Subtotal 380Not Answered 3Total 383Answers may be multiple (n/380; not answered in 3)22

2 - OrganisationalDatacollected immediate survival data following resuscitationattempts but many fewer collected any longer termdata; only 33% collected data about survival at 24 hoursafter cardiac arrest and only 57% collected data abouthospital survival (Table 2.21). Whilst it is reassuring tosee that resuscitation activities are audited, perhapshospital survival and functional outcome of survivors(which is considered later in this report), should be givenmore importance when measuring the effectiveness ofresuscitation.Table 2.24 Goals were set to reduce the number of cardiacarrests occurring in-hospitalGoals were set n %Yes 80 23.8No 256 76.2Subtotal 336Not answered 47Total 383Table 2.22 Hospital had a resuscitation committeeResuscitation committee n %Yes 337 88.9No 42 11.1Subtotal 379Not answered 4Total 383Table 2.23 Frequency the resuscitation committee metFrequency n %Once per year 7 2.1Twice per year 36 11.0More frequently 283 86.8Subtotal 326Not answered 11Total 337Tables 2.22 and 2.23 provide data on hospitalresuscitation committees. It can be seen that 42hospitals did not have a resuscitation committee(36/42 were private hospitals) although this is a keygovernance committee in most organisations. However,where they did exist it appears that most met more thantwice per year.Quality improvement work has highlighted the importanceof goal-setting as a stimulus to change but only 80hospitals had set any goals to reduce the number ofcardiac arrests.Where goals have been set the range of improvementtargets was high. Table 2.25 shows these data.Table 2.25 Target perentage reduction in the number ofcardiac arrestsTarget percentageTotal5-10% 2511-20% 321-30% 830-50% 1351-100% 6Subtotal 55Not answered 25Total 80The questionnaire asked if a local goal for reducing thenumber of cardiac arrests leading to a resuscitationattempt had been set (Table 2.24).25

Back to contents3 – Patient population, initial assessment and first consultant review3 - Patient population,initial assessment andfirst consultantreviewPatient populationThis section describes the patient demographicsand characteristics on admission to provide anunderstanding of the patient population and context ofthe study. Figure 3.1 shows the age range of patientsincluded within the study.A clinician questionnaire was returned for 585 cases.The median age for the sample included was 77 years(inter-quartile range 68-84). Forty-six percent of thesample was female (272/585).Comorbidities were reported commonly within the studypopulation. Table 3.1 shows the comorbidities reportedon the clinician questionnaire.Table 3.1 Chronic disease comorbiditiesComorbidities n SubtotalCardiovascular 341 524Respiratory 170 491Renal 133 483Immunosuppression 50 456Liver insufficiency 34 451Answers may be multipleFigure 3.2 shows body mass index (BMI) for thepopulation where both weight and height weredocumented (110 patients). Just over one third ofthe patients was considered to be within normal BMIrange (18.5-25), one third was considered to be in theoverweight BMI range (25-30) and just over one quarterwas considered to be obese (BMI >30). This reflects theBMI range across the current UK population. 19Number of patients120FemaleMale100806040200

3 - Patient population,initial assessment andfirst consultantreviewPercentage of patients4035302520151050

3 - Patient population,initial assessment andfirst consultantreviewPercentage of patients302520151050Bowels (572)Bladder (438)Grooming (447)Dressing (445)Transfer (471)Toilet use (441)Mobility (434)Stairs (353)Bathing (335)Domain assessedFigure 3.3. Barthel Index of Activities of Daily living: Percentage of patients scoringzero for each domain (i.e. much help required). Denominator for eachdomain is shown in brackets.Table 3.3 provides detail on the location of patients priorto hospital admission. The majority of patients were livingin their own home prior to hospital admission. Threepercent of the study population (19/585) were transferredfrom a nursing home.Table 3.4 shows the time of initial admission to hospital.These data were available for 506/585 cases. Mostpatients arrived during daytime (08:00 to 17:59) andless than one fifth arrived in the overnight period(00:00-07:59)Table 3.3 Patients’ location prior to hospital addmissionLocation total %Own home 472 80.7Another hospital 43 7.4Nursing home 19 3.2Other 26 4.4Unknown 25 4.3Total 585Table 3.4 Time of admission to hospitalTime n %00:00-07:59 91 15.608:00-17:59 270 46.218:00-23:59 145 24.8Unknown 79 13.5Total 585Data were not collected on level of independence or needfor support to maintain home living, other than the detailof the Barthel Index already shown.29

3 - Patient population,initial assessment andfirst consultantreviewPercentage of patients706050403020100Emergencyadmissionvia emergencydepartmentEmergencyadmissionvia GPPlannedadmissionPathway of admissionTransferredanotherhospitalReferralfrom GPOtherFigure 3.5 Pathway of admission (n=578, not answered in 7)Initial assessmentConsiderable detail was requested and analysis made onthe admission process to hospital. This was based on thepremise that rapid, complete and accurate assessmentof patients was likely to lead to prompt, appropriate andcomprehensive care. Conversely, if the initial directionof care was incorrect then it was more likely thatsubsequent management could be adversely affected.Table 3.6 shows where the data for this analysis wasobtained from.Table 3.6 Source of data used to assess admission processSource n %Clerking on admission to ward 239 47.0Emergency department assessment 132 26.0Both 137 27.0Subtotal 508Not answered 18Total 526If both emergency department assessment andclerking on admission to a ward were available then theassessment considered both, and used the best dataavailable (for example the most senior grade of either theemergency department doctor or ward doctor was used).For three quarters of the patients the ward assessmentwas used, either alone or in conjunction with theemergency department assessment.Table 3.7 Time of the initial assessmentTime total %00:00-07:59 100 23.108:00-17:59 201 46.418:00-23:59 132 30.5Subtotal 433Unknown 93Total 526Table 3.7 shows the time of initial assessment. Thesedata are similar to time of hospital admission althoughslightly fewer were assessed during daytime and moreassessed out of hours when compared to the admissiontime, suggesting that patients admitted during the dayhave some delays in assessment.31

3 - Patient population,initial assessment andfirst consultantreviewThe grade of clinician making the initial assessmentis shown in Table 3.8 and the definitions of grade areincluded in Appendix 5.Table 3.8 Grade of clinician undertaking the initialassessmentGrade of clinician n %Consultant 11 3.0Staff grade/associate specialist 3

3 - Patient population,initial assessment andfirst consultantreviewHistory and examinationAn accurate and complete history is important to ensurea full understanding of the acute presentation and providethe context of previous conditions, treatments andfunctional status. The Advisors considered this aspect ofthe admission process in detail and the number of caseswhere elements of the history were not obtained or wereincomplete are shown in Figure 3.6. The denominator isless than the 526 cases due to lack of data to assess in anumber of cases.It can be seen that there was a substantial number ofcases where elements of the history were not obtained.The major missing elements were drug history, socialhistory and information on activities of daily living. In only164 cases were all elements of the history obtained. Insome cases it may have been appropriate to limit theinitial history or the circumstances may have not allowedall the elements of the history to be obtained. In additionsome of this information may be contained in otherparts of the medical record; for example information onactivities of daily living may have been recorded in thenursing notes. For this reason the Advisors were asked togive an overall impression of the adequacy of the historyin the context of that particular case. This is shown inTable 3.10.Table 3.10 Adequacy of the past medical history taken -Advisors’ opinionAdequate history was taken n %Yes 419 85.7No 70 14.3Subtotal 489Unknown 37Total 526Percentage of patients10090YesNoIncomplete80706050403020100Presentingcomplaint(501)History ofpresentingcomplaint(496)Past medicalhistory(493)Domain coveredDrug history(485)Social history(487)Activities ofdaily Living(481)Figure 3.6 Elements of medical history covered by initial assessment(the denominator for each domain is shown in brackets).33

3 - Patient population,initial assessment andfirst consultantreviewWhen assessed in the context of each particular case theAdvisors were of the opinion that in almost 15% of casesan adequate history had not been obtained during theinitial admission process that could be determined fromreviewing the medical or nursing notes.Similarly, assessment of elements of physical examinationwas performed by the Advisors. Table 3.11 shows asummary of these data.Table 3.11 Assessment of elements of physical examinationdetermined by the AdvisorsSystem assessed n %Cardiovascular system 448 93.5Respiratory system 432 90.2Gastrointestinal system 362 75.6Central nervous system 290 60.5Genitourinary system 120 25.1None 17 3.5Answers may be multiple (n/479; not answered in 13 andinsufficient data in 34)It can be seen that there was a substantial number ofcases where physical examination of particular systemswas not performed. As with history-taking it may beappropriate not to carry out a full systematic examinationin all cases; the clinical scenario may lead to a focusedclinical examination. To account for this, Advisors wereasked to consider whether the clinical examination wasadequate at first contact (in the context of each particularcase). Table 3.12 shows that Advisors considered clinicalexamination incomplete in one in four cases.Professional standards for admission and assessment areavailable. 22 It appears, in the opinion of the Advisors, thatthe practice found in this patient population fell short ofthese standards in many cases.Outputs from initial assessmentThe key outputs from the initial assessment processare: diagnosis or differential diagnosis, investigationplan, treatment plan and monitoring plan. The Advisors’opinion on whether these elements had been performedand documented or not are shown in Figure 3.7.These data show that only 73% of patients had arecorded diagnosis or differential diagnosis, 84% had aplan of investigations, only 77% had a treatment plan,and only 29% had a monitoring plan. It would appearthat these are important deficits and could representan obstacle to timely and effective patient treatment.Furthermore the lack of a monitoring plan increasesthe chance of unrecognised deterioration and lack ofescalation in the event of deterioration.It may well be that some areas or hospitals have a defaultmonitoring plan and that this explains the low level ofdirection from admitting medical staff in this domain inthis sample of cases. However, it would still be goodpractice for the admitting doctor to set some parameters(e.g. initial frequency or type of observations) to ensure asafe baseline and little evidence of this could be found.Table 3.12 Completeness of clinical examination at the firstcontact - Advisors’ opinionComplete clinical examination n %Yes 362 75.6No 117 24.4Subtotal 479Unknown 47Total 52634

3 - Patient population,initial assessment andfirst consultantreviewPercentage of patients9080YesNoIncomplete706050403020100Diagnosis or DifferentialDiagnosis (493)Investigation Plan(491)Monitoring Plan(444)Treatment Plan(487)Figure 3.7 Outputs from initial assessment (denominator for each questionis shown in brackets)History and examination should allow an assessment ofseverity to be made by the admitting doctor. Table 3.13shows the Advisors’ opinion on whether the admittingdoctor appreciated the severity of the situation.Table 3.13 Severity of condition recognised by the admittingdoctor - Advisors’ opinionSeverity of the situation wasappreciated n %Yes 342 82.2No 74 17.8Subtotal 416Unknown 110Total 526Table 3.14 shows the grade of clinician who performed theinitial assessment in those 74 cases where the Advisorsbelieved the severity of the situation was not appreciated.Table 3.14 Grade of clinician making the initial assessmentwhere severity of the situation was not recognised -Advisors’ opinionGrade of cliniciantotalConsultant 2Staff grade 1Senior specialist trainee 3Junior specialist trainee 4Basic grade 34Specialist nurse practitioner 1Other registered nurse 3Other 2Subtotal 50Not answered 24Total 7435

3 - Patient population,initial assessment andfirst consultantreviewAs basic grade doctors performed the majority of initialassessments it is unremarkable that they assessed themajority of the cases in Table 3.14. It is also recognisedthat basic grade doctors may not have the skill orexperience to make a correct assessment of complexand challenging situations. It may be that they arebeing asked to assess and provide initial treatment forpatients when they do not have the competency todo so. This raises the issue of training, to ensure thatdoctors are suitably skilled for the tasks they are requiredto undertake, and suitably supervised, to ensure thatdelivery of tasks is adequate, that staff are supportedand that patient safety is maintained.Part of good care is escalating the care of the patientto more senior members of the team when necessary.Table 3.15 shows the Advisors’ opinion of this domain. Inalmost one in six cases the Advisors were of the opinionthat escalation of care to more senior team members didnot occur in a timely manner.Table 3.15 Timely escalation of care - Advisors’ opinionTimely escalation of care n %Yes 286 82.4No 61 17.6Subtotal 347Escalation not required 59Unknown 120Total 526There are many possible reasons why escalation did nothappen in a timely manner. Working patterns, availabilityof staff and workload may have been contributing causes.However, one major constraint was lack of appreciationof the situation and lack of awareness that escalation ofcare may be required. Table 3.16 gives some detail onthis aspect.Table 3.16 displays escalation by appreciation of theseverity of the situation. In 23 cases the doctor involveddid not appear to appreciate the severity of the situationand did not escalate in a timely manner in the opinion ofthe Advisors.Improved training and supervision are required to addressthis problem as it is difficult to criticise the decision not toescalate the care if the situation is not fully appreciated.In 31 cases the doctor involved did appreciate theseverity of the situation but still did not escalate thecare of the patient to a more senior doctor; this wasapproximately one in ten. This study did not collectdata to describe the reasons behind this problem butanecdotally the view of the clinicians involved with thisstudy recognised culture or unwillingness to ask for help,workload, competing demands and lack of support assome of the possible factors.Case study 1 gives an example of where escalation wouldhave been appropriate.Table 3.16 Timely escalation against appreciation of the severity of illness by the clerking doctor - Advisors’ opinionAppreciation of the severity of situationTimely escalation of care Yes No Unknown TotalYes 216 31 39 286No 31 23 7 61Subtotal 247 54 46 347Escalation not required 44 4 11 59Unknown 51 16 53 120Total 342 74 110 52636

3 - Patient population,initial assessment andfirst consultantreviewCase study 1Patient management until first consultant reviewor 24 hours if no consultant reviewAn elderly patient was admitted to a medicalassessment unit because of shortness of breath.The patient had a long past medical historyincluding life-long smoking, diabetes, ischaemicheart disease, previous coronary artery surgery,heart failure and chronic kidney disease. Thepatient was assessed promptly by an FY2doctor who made a differential diagnosis of heartfailure or chest infection and started treatmentwith antibiotics and increased diuretics. At thetime the patient was distressed and unable tospeak, oxygen saturations were 84% on highflowoxygen, respiratory rate was 32 breaths perminute, blood pressure was 85/45 mmHg, pulserate 140 beats per minute (atrial fibrillation) andarterial blood gasses showed a compensatedmetabolic acidosis. There was no record ofescalation to more senior doctors.Six hours after admission to the medicalassessment unit the patient had a PEA cardiacarrest and despite prompt CPR that continued for15 minutes the patient could not be resuscitated.The patient had not been reviewed by any seniordoctors prior to this.Advisors raised concerns about recognition ofseverity of situation and escalation to more seniordoctors. They also raised concern that there wasno intervention to treat rapid atrial fibrillation. TheAdvisors considered that more senior involvementmay have lead to a referral for higher level ofcare and also that CPR status may have beenconsidered.The data presented in this section relate to aspects ofpatient management after the initial assessment and upto the time of first consultant review (or first 24 hours ifconsultant review was not evident).Table 3.17 shows that a diagnosis or differential diagnosiswas recorded in 9 out of 10 cases.Table 3.17 Differential diagnosis was made and recordedduring the initial reviewDiagnosis or differential diagnosiswas made n %Yes 442 91.1No 43 8.9Subtotal 485Unknown 41Total 526In the opinion of the Advisors the correct diagnosis (or thedifferential diagnosis list included the correct diagnosis)was recorded in 9 out of 10 of these cases. (Table 3.18)Table 3.18 Correct diagnosis was included - Advisors’opinionCorrect diagnosis n %Yes 359 89.3No 43 10.7Subtotal 402Unknown 40Total 44237

3 - Patient population,initial assessment andfirst consultantreviewTable 3.19 shows the Advisors’ opinion of treatment planin this phase of care (initial assessment to initial consultantreview, or first 24 hours if no consultant review identified).Table 3.19 Appropriateness of the treatment plan - Advisors’opinionTreatment plan was reasonable n %Yes 409 83.5No 81 16.5Subtotal 490Insufficient data 36Total 526Advisors were unable to give an opinion on thereasonableness of treatment plan in 36 patients. Howeverin those with sufficient data it was thought that 16.5% ofcases did not have a reasonable treatment plan relatingto their condition.Matters raised by the Advisors as being related totreatment provided to the patient were categorised intoappropriateness of treatment and timeliness of treatment.These Advisor opinions are shown in Table 3.20.Table 3.20 Timely treatment against appropriate treatment -Advisors’ opinionTreatmentwastimelyYes No TotalYes 353 28 381No 44 34 78Total 397 62 459Not answered in 67 casesTreatmentwasappropriateOnly 353 patients out of the 459 with data for fullassessment had both appropriate and prompt therapy(77% of cases). There was inappropriate treatment in62/459 cases (14%) and delays in treatment in 78/459cases (17%). Case study 2 shows an example of this.Case study 2A very elderly patient was admitted to hospitalfollowing “collapse”. On admission their heart ratewas recorded at 35 beats per minute and an ECGshowed complete heart block. Blood pressurewas low and the patient was unable to sit up inbed due to postural hypotension and dizziness.The patient was admitted to the coronary careunit. There was no treatment plan to correct thebradycardia. Consultant review (24 hours later)confirmed complete heart block and noted a planto insert a permanent pacemaker. Over the next36 hours the patient’s heart rate was documentedregularly between 30 and 35 beats per minute. Anasystolic cardiac arrest occurred three days afteradmission to hospital and before the permanentpacemaker had been inserted.This case appeared to highlight both unacceptabledelays and inadequate treatment. A temporarypacemaker could have been inserted or permanentpacing expedited and this may have avoided thecardiac arrest. The Advisors considered that thiswas a potentially avoidable death.Table 3.21 Resuscitation status recordedDecision about CPR status was recorded n %Yes 44 10.1No 391 89.9Subtotal 435Insufficient data 91Total 526Table 3.21 provides data on resuscitation status decisionsduring the admission period: for clarity that was theperiod of initial admission, assessment and planning upto the first consultant review.38

3 - Patient population,initial assessment andfirst consultantreviewWhether or not the consultant considered the CPR statuswill be looked at later under the ‘first consultant review’.These data included patients in whom it could beassessed that a decision had been made that eitherCPR would be attempted and also patients in whomCPR would not be attempted (DNACPR). It must beremembered that this group of patients all ended uphaving a resuscitation attempt.Case study 3An elderly patient was admitted to hospital from anursing home with abdominal pain and vomiting.Past medical history included diabetes, atrialfibrillation, ischaemic heart disease and dementia.The patient was very dependent on help withactivities of daily living. On assessment the patientwas noted to be very frail. Blood pressure wasunrecordable and the patient was unrouseable.Biochemistry revealed severe renal impairment(urea 32 mmol/l, creatinine 507 microm/l), aprofound metabolic acidosis (pH 7.05) and raisedlactate (12 mmol/l). The patient was reviewedjointly by ST2 doctors from medicine and surgerywho decided that at that time the patient was notstable enough to have CT of the abdomen but thatischaemic bowel was the most likely diagnosis.The plan was to commence fluid resuscitationand re-assess. It was noted that the outcome waslikely to be poor but no decision about CPR statuswas documented. More senior doctors were notconsulted. The patient had a cardiac arrest 4hours later and underwent 10 minutes of CPR.This was unsuccessful.The Advisors raised concerns that there was a lackof appreciation of the severity and urgency in thiscase and that escalation to senior doctors shouldhave taken place and CPR status considered.Advisors were asked if, from the information available tothem in the questionnaires and case notes, they believedthat the actual actions regarding DNACPR status onadmission were appropriate; these data are shown inTable 3.22.Table 3.22 Actions regarding DNACPR status wereappropriate - Advisors’ opinionAppropriate DNACPR status n %Yes 237 63.2No 138 36.8Subtotal 375Insufficient data 151Total 526In 138 cases it was felt that the action was inappropriate,and in 89 of these cases it was felt that a DNACPRdecision should have been made (but had not beenmade). It is well recognised by all doctors that it can bechallenging to make decisions about appropriatenessof CPR within a short time period after admission butprocedures should be in place to do so when such adecision is indicated.It must be noted that more generally, many patientsdo have DNACPR decisions made and many deathsin hospital occur without CPR attempts. This is goodpractice where death is an inevitable outcome and CPRwill not work. This study only assessed care in patientswho underwent a CPR attempt and so did not measure,or assess, this level of good practice.One point raised during this study was the lack ofevidence of an explicit decision about resuscitation statusin the majority of patients in the study. Table 3.21 showsthat there was no explicit decision and documentationof resuscitation status in 90% of cases on admission.This lack of clarity is a potential source of confusion.The position that CPR should be attempted as a defaultmay lead to inappropriate CPR attempts. Perhaps moreimportantly this becomes the default position withouttruly examining if CPR will work or is indicated.39

3 - Patient population,initial assessment andfirst consultantreviewShould the intent to, as well as not to, undertake anintervention of this magnitude be clearly recorded inpatients who are acutely unwell, and a system used sothat all members of the team are aware of this decision?The requirement to record this information will encourageconsideration, discussion where appropriate and clarityon the intent and likely outcome should the patient have acardiac arrest. It can also be used as an audit standard toassess practice against.Case study 4A very elderly patient was admitted to hospitalafter collapsing at home. On admission they hada GCS of 9. A CT scan of their brain revealedextensive subarachnoid haemorrhage and thepatient was admitted to a ward for ongoing care.Over the next 24 hours the patient’s conditiondeteriorated and they became more obtunded.At the request of family members a chaplainvisited to offer comfort. Sixteen hours later thepatient had a cardiac arrest and the resuscitationteam was summoned. CPR was initiated andcontinued for approximately 10 minutes. No returnof circulation was achieved and death was certified.It appeared that death was the expected outcomein this case but a plan for what to do in the event ofcardiac arrest had not be written down.The Advisors considered that this CPR attemptappeared inappropriate and was an undignifiedprocess at the end of life.Table 3.23 provides data on the Advisors’ globalassessment of whether there were deficiencies presentin the initial assessment and treatment phase. As canbe seen it was felt that there were deficiencies in 47%of cases. As discussed before, it is more likely thattreatment will be compromised if the initial phase is eitherdelayed or directs attention in the wrong direction.Table 3.23 Deficiencies in initial assessment - Advisors’opinionDeficiencies in the initialassessment n %Yes 230 47.6No 253 52.4Subtotal 483Insufficient data 43Total 526Deficiencies were present in all domains, as shown inTable 3.24.Table 3.24 Areas of deficiencies in care - Advisors’ opinionDeficiencies n %Decision making with regards to CPR status 107 48.0Examination 85 38.1Treatment plan 79 35.4Diagnosis 76 34.1Recognition of severity of illness 69 30.9Seniority of doctor 68 30.5History taking 60 26.9Monitoring 66 29.6Investigation 66 29.6Answers may be multiple (n/223; not answered in 7)Many of these deficiencies were in basic elements ofmedical practice. This raises the more general questionas to whether the current structure and process for theinitial assessment and treatment of emergency patientsis fit for purpose.40

3 - Patient population,initial assessment andfirst consultantreviewThe Academy of Medical Royal Colleges recentlypublished a report titled ‘The benefits of consultantdelivered care’ 23 This concludes that there are realevidence based benefits to moving to a system ofconsultant-delivered care. Principle benefits includerapid and appropriate decision making and improvedoutcomes. This document is very pertinent to thecare of the acutely unwell patient cohort examinedin this report.Case study 5A middle aged patient was admitted to hospitalwith severe chest pain. The patient was known tohave a thoraco-abdominal aortic aneurysm andsurgical or radiological intervention had previouslybeen ruled out following multidisciplinary teamassessment. The patient had chronic lungdisease and home oxygen therapy had beenprescribed. The patient was hypotensive andappeared pale. Fluids, oxygen and analgesiawere prescribed whilst basic investigations werestarted. The impression was that this was aleaking aneurysm. Little analgesia was given dueto concerns over respiratory depression and thepatient continued to complain of severe pain. Fourhours after admission to the surgical assessmentunit the patient had a cardiac arrest. CPR wascommenced and continued for 20 minutes untilthe on-call surgical consultant arrived. CPR wasstopped after assessment of the situation at thattime and death confirmed.This case highlighted the need to assess likelyoutcomes in acutely ill patients and ensure thatCPR is not commenced when it will not work. TheAdvisors commented that earlier consultant reviewmay have facilitated better care. This is a challengein all clinical care settings.Table 3.25 shows the initial care location grouped byAdvisor opinion of level of care 24 (see definitions of levelof care in Appendix 6).Table 3.25 Initial location where care was providedLevel of care n %Level 1 care 402 83.2Level 2 care 62 12.8Level 3 care 19 3.9Subtotal 483Unknown 43Total 526Table 3.26 Actual level of care provided assessed byAdvisors’ opinion of where the patient should have goneLevel 1 careLevel of careLevel 2 careUnable to answerLevel 3 careAdvisors’opinion ofrequired levelof careLevel 1 care 355 0 1 9 365Level 2 care 35 61 0 13 109Level 3 care 2 0 18 1 21Subtotal 392 61 19 23 495Unknown 10 1 0 20 31Total 402 62 19 43 526TotalIt appears that patients who received Level 2 or Level 3care received an appropriate level of care. However therewere concerns raised relating to the group who receivedLevel 1 care. In almost one in ten cases of this group itwas agreed by the Advisors that a higher level of careshould have been provided.41

3 - Patient population,initial assessment andfirst consultantreviewFirst consultant reviewThe first consultant review of a newly-admitted patientis a vital process. Royal Colleges have highlightedthe importance of consultant review to ensure goodtreatment and have produced guidelines and standardsto promote early consultant involvement. 25,26 . Morerecently the Academy of Medical Royal Colleges hashighlighted the benefits of, and need to move towards,consultant delivered care. 23 Not only is the first consultantreview a vital safety net for the patient; it is also anessential training opportunity for junior doctors and otherhealth professionals.Table 3.27 shows if the first consultant review could beidentified from the case notes provided.Table 3.27 First consultant review was recorded in thecase notesFirst consultant review recorded n %Yes 277 53.2No 244 46.8Subtotal 521Insufficient data to assess 5Total 526The first consultant review could only be identified in277 cases (53%). Where consultant review could notbe identified it was not clear how often patients hadnot been seen and how often the review had not beenrecorded clearly enough to permit its recognition. Onlythe first 24 hours of case notes were requested inaddition to the notes for the 48 hours preceding cardiacarrest. If consultant review did not take place within 24hours of admission then the Advisors would not have thedocumentation to find it.Figure 3.8 shows time to first consultant review. Thiscould only be determined in 198 of the 277 cases whereconsultant review could be identified.In 48% of cases, when times could be identified, firstconsultant review occurred more than 12 hours afteradmission. This does not adhere to professionalguidelines 27,28 and appears to be a deficiency in theprovision of appropriate care.Percentage of patients60504030201000-≤1 >1-≤6 >6-≤12 >12Time from admission (hours)Figure 3.8 Timing of first consultant review (n=198, data missing in 79)42

3 - Patient population,initial assessment andfirst consultantreviewAdvisors were asked to consider if the first consultantreview was timely. Table 3.28 shows this data.Table 3.28 Timely consultant review - Advisors’ opinionTimely consultant review n %Yes 212 82.5No 45 17.5Subtotal 257Unknown 20Total 277Of the 45 cases in which the Advisors considered thetiming of consultant review to be inappropriate, in 25cases review occurred more than 12 hours. In threecases, the review took place in less than 12 hours.There was a difference between the number of patientsreviewed later than 12 hours after admission and thenumber in whom the Advisors felt that consultant reviewwas not within an appropriate time period. Whilst thismay give reassurance that longer delay to consultantreview may be acceptable sometimes it must beremembered that early consultant review is a professionalstandard and it provides support to the more juniormembers of the team. As shown earlier, there weredeficiencies in junior doctors’ appreciation of severityand urgency of the situation and lack of escalation wasa problem. Early consultant review is an essential safetynet and whilst in some cases it may not be required itneeds to be applied consistently to ensure that patientsin whom it is essential are not missed.Table 3.29 shows the Advisor opinion of timeliness ofconsultant review by time of day. Patients admittedduring the daytime were considered not to have had atimely review in 19% of cases compared with 9% forpatients admitted in the evening and overnight period.Data were not collected to explore the reason for thisdifference, although during the case review process itwas clear that competing work commitments (e.g.clinic attendance) was one possible reason for thisreduced timeliness during the normal working day.Table 3.29 Consultant review was within an appropriatetimeframe for the patients’ condition by time of admissionConsultant review withinappropriate timeframeTime of admission Yes No Unknown Total00:00-07:59 33 4 5 4208:00-17:59 105 27 9 14118:00-23:59 54 6 3 63Subtotal 192 37 17 246Not answered 20 8 3 31Total 212 45 20 277Table 3.30 Changes in management of care followingconsultant reviewChanges made in: n %Investigations 100 39.1Monitoring 29 11.3Diagnosis 34 13.3Other 82 32.0No evidence of change 83 32.4Answers may be multiple (n/256; not answered in 21)Table 3.30 shows the changes were made followingconsultant review. In two thirds of patients changesto their management were made following consultantreview. The diagnosis was changed in 13% of cases andfurther investigations requested in 39% of cases. Theimpact of consultant review on patient care appears to besubstantial. This demonstrates the need for, and benefitof, early review by a consultant.43

3 - Patient population,initial assessment andfirst consultantreviewTable 3.31 CPR status was consideredCPR status was considered n %Yes 31 13.2No 203 86.8Subtotal 234Unknown 43Total 277Table 3.31 shows consideration of CPR status in thosepatients in whom consultant review could be identified.In 13% of cases was a CPR decision considered. In viewof the comorbidities, functional impairment and potentialfatality of this patient cohort it may be suggested that agreater number of patients should have had a DNACPRdecision at consultant review, as this figure varies littlefrom the consideration given during the inital assessment(Table 3.21).However, it is possible that the issue of CPR may havebeen considered but no documentation was made ofthis fact, as the default position would be to leave thepatient for CPR. Lack of transparent decision makingand recording of decisions about CPR status in acutelyill patients is an obstacle to good patient care.Table 3.32 shows the Advisors’ opinion of the quality ofthe admission process. The period considered was fromadmission until first consultant review or if no consultantreview could be identified then the first 24 hours afteradmission.Case study 6A middle-aged patient was admitted to hospitalwith an infective exacerbation of chronic lungdisease. This was the fourth admission within theprevious 12 months. At home, the patient washousebound and unable to walk more than 10-15metres due to breathlessness. The admissionprocess and initial treatment were excellent andconfirmed at consultant review which occurredwithin 12 hours of admission to hospital. Atthis review, after discussion with the patient, itwas agreed that care would not be escalatedto tracheal intubation and ventilation should thepatient fail to respond to treatment. CPR statuswas not discussed or documented. The patienthad a cardiac arrest 48 hours after hospitaladmission and underwent a 25 minute period ofunsuccessful CPR.CPR was unlikely to work in this case and, inthe opinion of the Advisors, a DNACPR decisionshould have been made and documented. WhilstDNACPR in the event of a cardiac arrest doesnot stop the provision of other active treatmentmeasures to prevent deterioration, it appearedthat there may have been a concern that makinga DNACPR decision would result in less than fulltreatment and contribute to poor outcome.Table 3.32 Quality of admission process - Advisors’ opinionRating n %Good 145 51.2Adequate 110 38.9Poor 28 9.9Subtotal 283Insufficient data to assess 243Total 526Quality of the admission process was rated as good inhalf the cases where it could be assessed. This leavesa considerable number of cases where there was scopefor improvement in this aspect of initial patient care.However, it is also remarkable that in half the cases theinformation recorded in the case notes was insufficient tomake an overall assessment.44

3 - Patient population,initial assessment andfirst consultantreviewKey FindingsRecommendationsAn adequate history was not recorded in 70/489 cases(14%) and clinical examination was incomplete at firstcontact in 117/479 cases (24%).Appreciation of the severity of the situation was lacking in74/416 (18%).Timely escalation to more senior doctors was lacking in61/347 (18%).Initial assessment (up to first consultant review or first 24hours if consultant review could not be identified) wasconsidered to be deficient in 230/483 (48%) cases.Deficiencies were present in many domains but by far thegreatest number of concerns was raised about decisionsregarding CPR status (107 cases).Decisions about CPR status were documented in theadmission notes in 44/435 cases (10%). This is despite thehigh incidence of chronic disease and almost one in fourcases being expected to be rapidly fatal on admission.Advisors were of the opinion that a further 89 patientsshould have had a DNACPR decision made in this initialphase of their treatment.In the opinion of the Advisors 37/392 patients admitted toLevel 1 care should have received Level 2 or 3 care (9%).First consultant review could be identified only in 277/521cases (53%) and time to first consultant review could bedetermined only in 198/521 cases (38%).Where time to first consultant review could be identified itwas more than 12 hours in 95/198 cases (48%).Consultant review was considered inappropriately late in45/257 cases (18%).CPR status was considered in only 31/234 cases at firstconsultant review (13%).Standards of clerking/examination and recording thereofshould be improved. Each hospital should ensure thatthe detail required in clerking and examination is explicitand communicated to doctors-in-training as part ofthe induction process. A regular (6-monthly) audit ofperformance against these agreed standards should beperformed and reported through the governance structureof the organisation. (Medical Directors and all Doctors)Hospitals must ensure appropriate supervision fordoctors-in-training. Delays in escalation to more seniordoctors due to lack of recognition of severity of illness bydoctors in training are unacceptable and place patients atrisk. (Medical Directors)Each Trust/hospital must provide sufficient critical carecapacity or pathways of care to meet the needs of itspopulation. (Chief Executives)Each entry in a patient’s case notes must contain date,time, location of patient and name and grade of staff andtheir contact details. It must also contain information onthe most senior team member present during that patientcontact (name and grade). (All health Care Professionals)As previously recommended by NCEPOD and the RCP,all acute admissions must be reviewed at consultant levelwithin 12 hours of admission. Earlier consultant reviewmay be required and arrangements should be in place toensure that this is available. A regular (6-monthly) audit ofperformance against this standard should be performedand reported through the governance structure of theorganisation. (Medical Directors and Consultants)CPR status must be considered and recorded for all acuteadmissions, ideally during the initial admission process anddefinitely at the initial consultant review when an explicitdecision should be made, and clearly documented (forCPR or DNACPR). When, during the initial admission, CPRis considered as inappropriate, consultant involvementmust occur at that time. (All Doctors)45

Back to contents4 - Care before thecardiac arrest4 – Care before the cardiac arrestThis chapter provides details of patient care in the48-hour period leading up to the cardiac arrest andattempted CPR, the data are taken from the clinicianquestionnaire.Figure 4.1 shows data on patient location at the time ofattempted CPR. Approximately half of patients were oneither medical or surgical wards and a quarter were oneither coronary care unit or Level 2 care.The clinician caring for the patient was asked in thequestionnaire if this was the correct location for thepatient. Table 4.1 shows these data.Table 4.1 Appropriate ward for the care needed by the patientAppropriate ward n %Yes 521 92.2No 44 7.8Subtotal 565Unknown 20Total 585This question was not answered in 20 cases but in theremainder it was the opinion of the responsible clinicianthat the patient was on the correct ward in 92% of cases.n = 45n = 97n = 226Medical WardSurgical WardLevel 2/HDUCCUOthern = 110n = 106Figure 4.1. Type of ward the patient was on at the time of cardiac arrest(n=584, not answered in one)47

4 - Care before thecardiac arrestTable 4.2 shows that of the 44 patients who the cliniciansfelt were in the wrong location 24 patients should havebeen cared for in CCU/Level 2 or Level 3 care area.Table 4.2 Location where the patients should have beencared for - Responsible clinicians’ opinionLocationTotalLevel 3 care 3Level 2 care 13Coronary care unit 8Surgical ward 2Medical ward 7Other 10Subtotal 43Not answered 1Total 44The duration between hospital admission anddeterioration to cardiac arrest is important in the contextof this study. Figure 4.2 shows these data.Thirty-two percent of patients had been in hospitalfor less than one day prior to their cardiac arrest. Thisgroup may be the most challenging in terms of ensuringsenior review, treatment-planning and decision-makingregarding CPR in the event of a cardiac arrest. Howeverthis was a minority of patients in this study and 68% ofthe cases included had a hospital stay of longer thanone day prior to cardiac arrest. Of the whole group29% had been in hospital for longer than one week. Itwould appear that in most cases there is opportunity toconsider, decide on, and discuss (where appropriate)CPR in the event of a cardiac arrest.It is possible that some patients are admitted to hospitaland subsequently develop an unrelated acute problemand have an unanticipated cardiac arrest. In thesepatients, despite an in-patient stay of several days, theremay be little warning of cardiac arrest. However, thesepatients are likely to be the minority as it is known thatmost patients who have an in-patient cardiac arrest havewarning signs and it is a predictable event not caused byprimary cardiac disease. 4Number of patients2001801601401201008060402007Duration of hospital stay (days)48Figure 4.2 Duration of hospital stay prior to cardiac arrest(n=583, not answered in 2 cases)

4 - Care before thecardiac arrestNumber of patients60504030201000-4 4-8 8-12 12-16 16-20 20-24Duration of hospital stay (hours)Figure 4.3 Duration of hospital stay in those that stayed less than 24 hours(n=146, not answered in 43)Figure 4.3 shows duration of hospital stay prior to cardiacarrest in the group with a length of stay of less than 24hours. Actual time in hospital prior to cardiac arrest couldbe calculated in 146 of the 189 patients. Of this group99 had a stay of less than 12 hours and 47 a stay oflonger than 12 hours. This highlights the need to ensurerapid consultant review for acutely unwell patients. Thetime from admission to death was short in a substantialnumber and this group of patients should have hadaccess to consultant review.The clinician caring for the patient at time of cardiacarrest was asked if the patient was on an end of lifecare pathway. Table 4.3 shows that one percent of thepatients in this study (7/578), who underwent CPR,were on an end of life care pathway. It is not clear if thecare pathways did not include DNACPR direction or ifthis was due to lack of documentation or handover ofinformation. However it is difficult to conceive a situationwhere attempted CPR in the setting of an end of lifecare pathway can be clinically appropriate or in the bestinterests of the patient. Although this was a small numberof patients in this study this poor practice was a concern.Table 4.3 Patients were on an end of life care pathwayEnd of life care pathway n %Yes 7 1.2No 566 97.9Unknown 5

4 - Care before thecardiac arrestphysiological monitoring plan is needed to ensure earlyrecognition of deterioration and facilitate appropriateintervention and decision-making. Previous NCEPODwork has highlighted this as an area of weakness 2 andboth NCEPOD and NICE 9 have made recommendationsthat there should be a clearly documented physiologicalmonitoring plan. The deficit in this area was substantial inthis sample of patients.Percentage of patients1009080ObservationsrequestedObservationsnot requested706050403020100Pulse (378)Blood pressure(380)Respiratory rate(373)Urine output (346)Figure 4.4. Observations that were requested/not requested during the 48 hoursprior to cardiac arrest (the denominator for each domain is shown in brackets)Table 4.4 Number of patients by requested frequency of observationsFluid balance(341)Central venouspressure (347)Oxygen saturation(379)Parameter measured hourly Four Other Not Number ofhourly specified patientsfor whomobservationsrequestedPulse 33 36 50 39 158Blood pressure 34 36 52 47 169Respiratory rate 34 36 46 36 152Urine 30 16 22 27 95Fluid balance 15 6 30 32 83Central venous pressure 2 0 0 3 5Blood oxygen saturation level (SpO2) 29 35 47 43 154Other 10 6 16 12 4450

4 - Care before thecardiac arrestTable 4.4 provides further details on the physiologicalmonitoring plan in the group of patients whereobservations had been requested. In most cases nomonitoring plan was noted. Where it was noted, itcould be seen that in the majority of plans recorded thefrequency of observations that had been requested; itmust be noted that no frequency was stated in between20 and 40% of cases (depending on the parameterconsidered).In the organisational chapter the use of early warningsystems and escalation protocols were presented; 99%of hospitals stated that they used an early warningsystem and in 98% they were linked to escalationprotocols (Tables 2.14 and 2.15). The findings from thesedata should be viewed in this context.Case study 7A major purpose of a physiological monitoring planis to ensure that deterioration is recognised and thatappropriate escalation of care happens. A key elementof the effectiveness of monitoring is clarity on whento escalate care. Advisors were asked to examine theextracts of the medical record available to them andassess if there were instructions to the nursing staff as towhen to alert the medical staff in the event of deterioration.Table 4.5 shows the results of this assessment.Table 4.5 Instructions to nurses about when to alert medicalstaff that a patient was deteriorating was recorded in thecase notesInstructions recorded n %Yes 85 21.0No 320 79.0Subtotal 405Insufficient data to assess 121Total 526In only 85/405 (21%) cases with enough information forassessment was there evidence of instructions regardingcriteria for escalation. These findings highlighted aconcern that the system will fail to respond appropriatelywhen patients deteriorate. One possible reason for theseapparent deficiencies is that there was organisationalsystem for setting out physiological observations andtriggering an appropriate response without the need forthis to be written down in each individual set of records.An elderly patient was admitted to hospital dueto pain from abdominal distension secondary toascites. The cause of ascites was known to bemetastatic colonic carcinoma and all therapeuticoptions had been explored. The patient was onan end of life care pathway and understood thatthey were nearing the end of life. Paracentesiswas performed to ease the symptoms of pain andbreathlessness. Forty-eight hours after hospitaladmission the patient had a PEA cardiac arrest.The cardiac arrest team was summoned andCPR started promptly. After 10 minutes of CPRthere was a return of circulation and spontaneousrespiratory effort, however the patient remainedobtunded and unresponsive. After discussionwith the consultant in charge it was decided thatfurther investigation or escalation of care was notappropriate. The patient survived for a further 36hours but never regained consciousness.It is not clear why CPR was performed in a patientwho was on an end of life care pathway and wasnearing the end of life. The Advisors consideredthis very poor practice.51

4 - Care before thecardiac arrestAdvisors were asked if the patient was monitored using asystem that could be recognised as a ‘track and trigger’chart. Table 4.6 shows these findings.Table 4.6 Track and trigger monitoring system usedTrack and trigger used n %Yes 282 78.8No 76 21.2Subtotal 358Insufficient data 168Total 526This could be answered in only 358 cases. In 282 of thosecases the Advisors could identify a track and triggermonitoring chart. This does not completely fit with theresponses received from hospitals indicating that almostall hospitals used this type of chart. However, it wasencouraging that these charts were widely used as theirpurpose is to recognise warning signs of physiologicalinstability early and ensure an appropriate response.Table 4.7 shows how patients were assessed against thepresence or absence of the criteria that were given to theAdvisors to assess each case. These were adapted frommedical emergency team calling criteria 29 and have beenused previously in NCEPOD work in this area. 2There were substantial proportions of patients who metthese criteria for significant physiological derangement(or concern). The most prevalent problems werehypoxia, hypotension and tachypnoea. These criteria areconsidered ‘red flags’ for patients at risk and in most trackand trigger systems would trigger a patient review. 29-31Table 4.8 shows that when all markers of physiologicalinstability were considered (from Table 4.7) 322 patientshad at least one marker present, 122 patients had nomarkers present and there was insufficient data to assessthis in 82 patients.In discussions with the Advisor group it was suggestedthat the criteria used were far from the normal range andthat more subtle criteria could be used to assessTable 4.7 Patient assessmentsCriteria reached in 48 hoursprior to cardiac arrest Yes % No % Subtotal Insufficient data Not answeredRespiratory rate 30/min 86 23.6 279 76.4 365 124 37Oxygen saturation

4 - Care before thecardiac arrestTable 4.8 Number of markers of physiological instabilityNumber of markers recorded n %At least one 322 72.5None 122 27.5Subtotal 444Insufficient data to assess 82Total 526patients at risk. Even with these extreme criteria itappears that there may be opportunities to recognisemany of these patients prior to cardiac arrest and tointervene in over 70% of cases.It also worth noting that over one quarter of patients didnot have any of the predefined markers of physiologicalinstability. Whilst it may be that the criteria used were notsufficiently sensitive it is also clear that for some patientsa cardiac arrest may have no preceding signs. A goodexample is a patient with acute coronary syndrome whohas a sudden VF arrest with no preceding abnormalphysiological markers.The duration of physiological instability is another keyfactor. If the duration is long then there appeared to begreater opportunity for intervention. Figure 4.5 shows theduration of physiological instability in all patients whoexhibited at least one marker of instability.Sixty two percent of these patients had physiologicalinstability for longer than six hours, 47% longer than12 hours and 20% longer than 24 hours. It appearsthat there was often a substantial time interval betweenonset of physiological instability and cardiac arrest.This provided potential opportunity to intervene andinfluence patient outcomes if warning signs had beenrecognised and acted upon promptly. This interventionmay be new treatment to halt deterioration and improveoutcome or it may be recognition that new treatmentsare available or appropriate and that CPR status shouldbe considered.Percentage of patients40353025201510500-6 6-12 12-24 24-36 36-48 >48Time prior to cardiac arrest (hours)Figure 4.5 First appearance of markers of physiological instability prior to cardiac arrest(n=190, not answered in 132)53

4 - Care before thecardiac arrestPercentage of patients6024 hours in hospital504030201000-6 6-12 >12Duration of physiological instability (hours)Figure 4.6 Duration of physiological instability for those patients in hospital eitherless than or longer than 24 hours (n= 179, not answered in 101)Patients who had a shorter interval between admissionand cardiac arrest tended to have shorter periods ofphysiological instability before their arrest. Figure 4.6show the duration of physiological instability prior tocardiac arrest in those patients who had a cardiac arrestwithin 24 hours of hospital admission or within a timeframe greater than 24 hours post admission.Case study 8A elderly patient was admitted to hospital as anemergency because of breathlessness. After initialassessment it was suspected that this was due tocommunity acquired pneumonia. Appropriate treatmentwas commenced, and confirmed at consultant review,which took place within six hours of admission. At thattime the patient was tachypnoeic (respiratory rate 20breaths per minute), tachycardic (pulse 110 beats perminute, sinus rhythm) and febrile (temperature 38.2 o C).After a further two hours in the medical assessmentunit the patient was transferred to an acute medicalward for ongoing inpatient treatment with IV antibiotics,oxygen and IV fluids. Physiological observations werecarried out initially on a four hourly basis. Over the nexttwelve hours these documented a rising respiratoryrate (to 32 breaths per minute), rising pulse rate (to120 beats per minute) and hypotension (systolic bloodpressure 80 mmHg). In that time the patient wasreviewed twice by an FY2 doctor. Additional IV fluidswere prescribed but no further action was taken. Thefrequency of observations was increased to hourly, dueto nursing concerns. Eight hours later an ST1 doctorreviewed the patient at the request of the nursing staffon the ward. Blood pressure was lower (systolic 75mmHg) and the patient was less rouseable. Furtherfluid was prescribed and IV antibiotics were changed.There was no request for more senior review or referralto other teams, such as critical care. Four hours laterthe patient had a PEA cardiac arrest and CPR wasunsuccessful. The last recorded observations were:BP 70/35, Pulse 130/min, Respiratory rate 32/min,Saturation – 85% (on 40% oxygen).This case illustrates the antecedent factors to cardiacarrest and lack of appropriate action in the face ofsignificant abnormalities. The Advisors considered thatthis cardiac arrest may have been avoided if escalationto more senior doctors and earlier intervention(haemodynamic and respiratory support) had occurred.54

4 - Care before thecardiac arrestOf those patients that had a cardiac arrest within 24hours of admission, almost half of this group hadabnormalities present for longer than six hours andalmost a third for longer than 12 hours prior to cardiacarrest.Almost three quarters of the group who had been inhospital for at least 24 hours prior to cardiac arrest hadabnormalities present for longer than six hours and morethan 50% had abnormalities for longer than 12 hoursprior to cardiac arrest.In the 526 cases examined by the Advisors, 2368individual reviews or patient contacts by doctors or nurseswere identified in the 48 hour period prior to cardiac arrest.Figure 4.7 shows the number of patients who underwenteach number of reviews for patients who were in hospitalfor less than or longer than 24 hours prior to cardiac arrest.Lack of details in the case notes did not allow these datato be collected in 135 cases. In the remainder it can beseen that many patients had frequent reviews. Indeed160 patients had more than five reviews in the 48 hourperiod prior to cardiac arrest.It is inevitable that patients who have a shorter length ofstay prior to cardiac arrest will tend to have had fewerreviews than those with longer hospital stays prior toarrest. Most patients who had a cardiac arrest within 24hours of hospital admission had between one and threereviews. However, even within this group 18 patients hadmore than five reviews prior to their cardiac arrest. In thegroup who were in hospital for longer than 24 hours priorto cardiac arrest 142 patients had more than five reviewsprior to cardiac arrest. What is clear from these findingsis that there was a substantial amount of clinician/patientcontact in the period leading up to cardiac arrest.Number of patients60≤ 24 hours≥ 24 hours504030201001 2 3 4 5 6 7 8 9 10+Number of reviewsFigure 4.7. Number of reviews in patients in hospital for less than or longerthan 24 hours (n=391, not answered in 135)55

4 - Care before thecardiac arrestInformation on the grade/professional groups whoundertook the patient reviews was collected and is shownin Figure 4.8.Figure 4.8 shows that of the 2368 reviews where gradewas available basic grade doctors were responsible for24% of reviews and registered nurses for 33% of reviews.Figure 4.9 details the contribution of each grade/grouping to the individual patient reviews. For exampleof the 462 patients in column one undergoing the firstreview in the 48 hours prior to cardiac arrest, consultantswere responsible for 13% of these reviews and basicgrade doctors for 28%. Of the 151 patients in columnseven undergoing the 7th review prior to cardiac arrestconsultants were responsible for 8% of these reviewsand basic grade doctors for 20%.An essential part of any ‘track and trigger’ system is areliable and escalated response to problems or concerns.These patients who had a cardiac arrest had manyreviews but continued to deteriorate and have a cardiacarrest. The findings in Figure 4.9 appear to demonstratea lack of escalation to senior staff. As the number ofreviews increase it might be expected that the proportionof cases seen by more senior staff would increase butthis was not the case. In the group of patients who hadeight reviews prior to cardiac arrest only 5% of the eighthreview were by consultants or senior specialist traineeswhilst 57% were by basic grade doctors or nursing staff.It is remarkable that escalation to more senior doctorsdid not occur despite abnormal physiological signsand frequent patient contact. More senior involvementmay allow treatment or escalation in level of care to beinitiated that may stop further deterioration and improveoutcome. Alternatively this intervention may allowdecisions about CPR status and appropriate levels ofcare be made prior to cardiac arrest.In addition to the objective criteria used to assessphysiological instability Advisors were asked if in theiropinion there were ‘warning signs’ or markers that thepatient was at risk of deterioration and cardiac arrest.Percentage of patients35302520151050ConsultantSeniorspecialisttraineeJuniorspecialisttraineeBasic grade Nurse Other UnknownGrade of clinicianFigure 4.8 Summary of the grade of clinician reviewing patients during the 48 hoursprior to cardiac arrest (n=2368)56

4 - Care before thecardiac arrestPercentage of patients100%80%60%40%20%0%1 (n=462)2 (n=401)3 (n=357)4 (n=291)5 (n=240)6 (n=196)7 (n=151)8 (n=118)9 (n=89)10 (n=63)ConsultantSeniorspecialisttraineeJuniorspecialisttraineeBasic grade Nurse Other UnknownReview numberFigure 4.9 Grades of clinician that reviewed patients during 10 reviews in the 48 hoursprior to cardiac arrest (the denominator for each review number is shown in brackets)Table 4.9 Warning signs were apparent that the patient wasdeteriorating - Advisors’ opinionWarning signs were apparent n %Yes 344 74.5No 118 25.5Subtotal 462Insufficient data 64Total 526Table 4.10 Action taken if warning signs were present -Advisors’ opinionThe signs were: Yes % No %Recognised 152 64.1 85 35.9Acted on adequately 104 43.9 133 56.1Communicated toappropriatesenior doctors 106 44.7 131 55.3Answers may be multiple (n/237; not answered in 107)For 75% of the patients in whom sufficient data werereturned, the Advisors stated that there were clearwarning signs present (Table 4.9), and gave an opinionof what was done in response to the warning signs thatpatients were deteriorating (Table 4.10).In 152/237 of these cases the Advisors believed thatthe signs were recognised, in only 44% (104/237) ofcases did the Advisors feel that adequate action wasundertaken and in 45% (106/237) did appropriatecommunication to more senior doctors happen. Theseopinions complement the objective data on duration ofphysiological instability prior to cardiac arrest. Somepatients may continue to deteriorate despite recognitionand intervention and in that instance it would seem harshto criticise the time interval between onset of instabilityand arrest. However it is clear, from Advisor opinion, andthe objective data from the clinicians directly involvedin the patients’ care, that there were problems withrecognition, action and senior involvement. The RoyalCollege of Physicians has recently published an acutecare toolkit to highlight the constraints and challengesfaced and make recommendations to ensure delivery ofhigh quality acute care which should aid this. 3257

4 - Care before thecardiac arrestPercentage of cases6050Poor (%)Adequate (%)Good (%)403020100Organisationalaspects (490)Clincians’ knowledge(478)Appreciation ofurgency (469)Supervision ofjunior staff (468)Seeking of advicefrom seniordoctors (465)Domain assessedFigure 4.10 Advisor grading of clinical aspects of care in 48 hours prior tocardiac arrest (the denominator for each domain are shown in brackets)Percentage of cases6050Poor (%)Adequate (%)Good (%)403020100Airway (458) Breathing (460) Circulation (459) O 2 therapy (445) Patient monitoring(458)Domain assessedFigure 4.11 Advisors grading of aspects of patient management in the 48 hoursprior to cardiac arrest (the denominator for each domain is shown in brackets).59

4 - Care before thecardiac arrestKey FindingsRecommendationsMost patients who had an in-hospital cardiac arrestwere considered to be on the correct ward at the time(521/565; 92%).68% of patients (394/583) had been in hospital for longerthan 24 hours prior to cardiac arrest.7/573 patients who underwent CPR were on an end of lifecare pathway. All seven patients died in hospital.Appreciation of urgency, supervision of junior doctorsand the seeking of advice from senior doctors were rated‘poor’ by Advisors.Physiological instability was noted in 322/444 (73%) ofpatients who subsequently had a cardiac arrest.Advisors considered that warning signs for cardiac arrestwere present in 344/462 (75%) of cases. These warningsigns were recognised poorly, acted on infrequently, andescalated to more senior doctors infrequently.Many patients had multiple reviews in the 48 hour periodprior to cardiac arrest, 160/391 had more than 5 reviews.There was no evidence of escalation to more senior staffin patients who had multiple reviews.NICE Clinical Guideline 50 (Acutely Ill patients in hospital:Recognition of and response to acute illness in adultsin hospital ) is not applied universally. Each hospitalmust ensure that they comply with this NICE guidance.(Medical Directors)For all patients requiring monitoring, there must be clearinstructions as to the type and frequency of observationsrequired. Where ‘track and trigger’ systems are used theinitial frequency of observations should be stated clearlyby the admitting doctor. (All Doctors)Where patients continue to deteriorate after nonconsultantreview there should be escalation of patientcare to a more senior doctor. If this is not done, thereasons for non-escalation must be documented clearlyin the case notes. (All Doctors)Hospitals should undertake a detailed audit of the periodprior to cardiac arrest to examine whether antecedentfactors were present that warned of potential cardiacarrest and what the clinical response to those factorswas. (Medical Directors)A national standard dataset should be developed to auditantecedent factors against.Advisors considered that the cardiac arrest waspredictable in 289/454 (64%) and potentially avoidablein 156/413 (38%) of cases.60

Back to contents5 - Resuscitation status5 – Resuscitation statusIn the previous chapter we commented on when CPRstatus had been considered at the ‘initial assessment’and then again at the ‘first consultant review’.This chapter covers this topic in more depth, detailingconsideration of CPR status and decision making at anypoint prior to the cardiac arrest.The clinician caring for the patient at the time of thecardiac arrest was asked if there was a statement in thenotes detailing the patients’ resuscitation status priorto the cardiac arrest. Table 5.1 shows the results of thisquestion.Table 5.1 CPR status was recorded in the notesCPR status was recorded n %Yes 122 22.1No 430 77.9Subtotal 552Not answered 33Total 585If an explicit decision had been documented the clinicianswere asked what this decision was. Table 5.2 showsthese data.Table 5.2 Explicit CPR decision had been madeCPR status n %For CPR 70 57.4DNACPR 52 42.6Total 122In 22% (122/552) there was a written statement in thenotes making an explicit decision about CPR status(Table 5.1). In 70/122 (57%) the patient was for CPRin the event of a cardiac arrest and in 52/122 (43%)the decision was DNACPR (Table 5.2). This poses twoquestions:1. Why did 52 patients who had a documentedDNACPR decision undergo a CPR attempt?Was this due to poor documentation, poor handover,poor communication or poor systems for holdingthis information? Data were not collected toanswer these questions but we consider that it isnot appropriate that 52 patients underwent CPRwhen a decision had been made that this was notappropriate.2. Why was no explicit decision made in 430/522(78%) of patients?Does this reflect the default position is that allpatients are for CPR unless stated otherwise andif so is this the correct approach in acutely unwellpatients as those sampled for this study? Such animportant decision as DNACPR or for CPR in thisgroup should be explicitly documented. Not doing somay lead to inappropriate treatment and confusionwithin the team caring for the patient. Furthermorethe requirement to document a decision will promptconsideration of CPR status and provide a basis forauditing compliance and decision making. Furtherdetails are shown in Table 5.4.61

5 - Resuscitation statusCase study 10A middle-aged patient with advanced inoperablepancreatic carcinoma was admitted to hospital.A biliary stent had been inserted two monthsprior to provide symptomatic relief from bile ductobstruction and jaundice. This admission wasdue to sepsis, abdominal pain and worseningjaundice. Ultrasound examination showed biliarytree dilatation. Chest x-ray showed consolidationat the right lung base. All treatment options werediscussed with the patient who declined furtherinvasive procedures or escalation of care. Thepatient understood the severity of the situationbut was content to be treated with antibiotics,fluids, oxygen and analgesia. Resuscitationstatus was discussed with the patient by theresponsible consultant and it was agreedthat CPR would not be initiated and this wasdocumented in the notes.The patient had a cardiac arrest and CPRwas initiated, after 10 minutes of resuscitationspontaneous return of circulation was obtained.The patient was intubated, had poor respiratoryeffort and remained obtunded. The patientwas transferred to the intensive care unit. Noconsultants were involved in this process. Thepatient continued to deteriorate and died fourhours later.The Advisors commented that care wasappropriate and good, including a DNACPRdecision. They questioned why CPR was initiateddespite the DNACPR decision and questionedthe lack of consultant involvement in decisionmaking after CPR.In some circumstances DNACPR decisions may involvequality of life considerations. There are circumstanceswhere CPR may work and the patient may survive butconcerns exist about the burden of disease and quality oflife after CPR. In these circumstances it is very importantto enter into sensitive discussions with patients and/or next of kin, to understand their views and to allow anagreed course of action to be followed.The previous paragraphs highlight key differences in therole of CPR, depending on whether or not it will work,and the way in which patients and next of kin might beengaged.Irrespective of these differences this work has revealedthat involvement of patients (or next of kin whereappropriate) in discussions about CPR status was notpracticed commonly. This might be due to the fact thatCPR would not work and therefore discussion wasfelt not to be needed. However, the lack of records toanswer the question of whether the patient and/or nextof kin had been involved raises some questions – lackof good process, lack of documentation or both? TheGeneral Medical Council guidance on recording andcommunicating CPR intentions is very clear about thisand states ‘Any discussions with a patient, or withthose close to them, about whether to attempt CPR,and any decisions made, should be documentedin the patient’s record or advance care plan. If aDNACPR decision is made and there has been nodiscussion with the patient because they indicateda wish to avoid it, or because it was your consideredview that discussion with the patient was notappropriate, you should note this in the patient’srecords’. 33 It does not appear that practice is followingthis guidance in this sample of patients.The treating clinicians were asked the reasons why aDNACPR decision had not been made in the remainderof the patients.63

5 - Resuscitation statusTable 5.5 Duration of hospital stay prior to cardiac arrestDuration14 days 3Subotal 38Not answered/data missing 22Total 60Only a minority of patients in this group had a cardiacarrest within 12 hours of hospital admission and thatmost were in hospital for one or more days. Lack of timeappears to be a poor reason not to have made decisionsabout CPR status in this group.As part of the peer review process the Advisors wereasked if there was evidence of CPR status beingrecorded at any point from admission to cardiac arrest.Table 5.6 CPR status recorded at any point from admissionto cardiac arrest - Advisors’ opinionRecord made n %Yes 62 12.3No 443 87.7Subtotal 505Insufficient data to assess 21Total 526nAdvisors found documentation of CPR status in 62/505cases (12%). Table 5.7 provides details of this decision.As can be seen in 24 patients the decision was DNACPRwhilst in 38 the decision was to perform CPR in the eventof a cardiac arrest. Again it is worth emphasising thatperforming CPR in patients with documented DNACPRdecisions is poor practice.Table 5.7 Recorded decision stated that the patient was forresuscitation - Advisors’ opinionRecorded decisionYes 38No 24Total 62Where an explicit decision had been made anddocumented regarding CPR status the Advisors lookedfor the grade of staff involved. Table 5.8 shows the gradeof staff who made the decision about CPR status.Table 5.8 Grade of clinician who made the CPR statusdecision - Advisors’ opinionGrade of clinicianConsultant 23Staff grade 2Trainee with CCT 0Senior specialist trainee 5Junior specialist trainee 6Basic grade 5Specialist nurse practitioner 0Other registered nurse 0Resuscitation officer 0Other 2Subtotal 43Insufficient data 19Total 62nn65

5 - Resuscitation statusConsultants were noted to have made the decision in23/43 cases. However, basic grades or junior specialisttrainees were noted to have made the decision in 11/43cases. This is not in keeping with good practice orprofessional recommendations.Advisors were asked if there was evidence of discussionof CPR status with the patient or next of kin. Tables 5.9and 5.10 show these data.Table 5.9 CPR status was discussed with the patient -Advisors’ opinionDiscussed with patientYes 11No 29Subtotal 40Insufficient data to assess 22Total 62Table 5.10 CPR status was discussed with the patient’srelatives - Advisors’ opinionDiscussed with relativesYes 22No 16Subtotal 38Insufficient data to assess 24Total 62There was a greater involvement of next of kin rather thanpatients in decision making about CPR status. However,in both responses patient and next of kin the involvementwas low (11/40 patients and 22/38 next of kin).Discussions of decisions about CPR status are anessential part of good end of life care, where patientsare not likely to benefit from CPR or the application ofCPR may not be in their best interests. The followingquote is from the Department of Health’s End of LifeCare Strategy. 36nn“During the development of the end of life strategymany people have identified the lack of opendiscussion between health care staff and thoseapproaching the end of life, as one of the keybarriers to the delivery of good end of life care. Thisrepresents a major challenge. It requires a significantculture shift both amongst the public and withinthe NHS. Clinicians and managers need to acceptthat death does not always represent a failure ofhealthcare and that enabling people to die as well aspossible is one of the core functions of the NHS.”Table 5.11 shows the grade of staff involved in the 11cases where CPR status was discussed with patients.Table 5.11 Grade of clinician involved where CPR status wasdiscussed with patientsGrade of clinicianConsultant 3Staff grade 0Trainee with CCT 0Senior specialist trainee 1Junior specialist trainee 2Basic grade 2Specialist nurse practitioner 0Other registered nurse 0Resuscitation officer 0Other 0Subtotal 8Insufficient data to assess 3Total 11The numbers are small and there are some missing data.However only three cases were discussed by consultantswhilst four cases were discussed by basic grade or juniortrainee doctors. Again this does not appear to fit withgood practice. Similarly in the 22 cases discussed withrelatives only three were by consultants and six by basicgrade or junior trainee doctors (data not shown).n66

5 - Resuscitation statusCase study 11An elderly patient with severe dementia wastransferred from a nursing home to an acute hospitalbed due to an acute confusional state. Over theprevious few months the patient had experiencedsignificant weight loss. It was noted that food intakewas very poor even with help and encouragement.It was felt that the reason for the patient’sconfusional state may be infection, either chestor urinary tract, and antibiotics were started tocover both possibilities. Over the next few days thepatient remained very confused. Due to concernsover poor oral intake a nasogastric tube wasinserted. However this was pulled out several timesand no effective nutrition was delivered. Six daysafter admission the patient was noted to be moreobtunded, had a high respiratory rate (30 breaths perminute) and urine output was very poor. The patientwas reviewed by a CT1 doctor who prescribedfurther fluids and changed the antibiotics. Concernwas expressed in the notes by nursing staff that thepatient was dying and that there should be clarityabout what to do in the event of a cardiac arrest.The patient was reviewed a further two times byjunior medical staff over the next 24 hours. CPRstatus was not considered during those reviews.Shortly after the last review the patient had a cardiacarrest. When the cardiac arrest team arrived theyfound the patient to be in asystole. CPR continuedfor 10 minutes before a decision was taken by aSpecialist Registrar in medicine that this was futileand the CPR attempt stopped. There was no returnof circulation.The Advisors considered that this was an undignifiedprocedure at the end of life. Furthermore theythought that it should have been recognised that thepatient was deteriorating despite active therapy andthat death was a likely outcome. CPR in the contextof this case was felt to be inappropriate.Advisors were asked whether a DNACPR decision shouldhave been made prior to cardiac arrest. They were askedto base this opinion on information gathered about pasthistory, functional status, acute illness, course of illnessand likelihood of survival. Advisors felt able to form anopinion in only 230 cases and it was believed that aDNACPR decision should have been made in 196/230cases and that CPR should have been attempted in only34/230 cases. Clearly in the vast majority of these casesthe Advisors felt that CPR was not indicated.Table 5.12 Advisors’ opinion whether the patient should havehad a DNACPR and whether they did or notAdvisors’ opinion:Patient should have hada DNACPR decisionPatient had DNACPRYes No SubtotalYes 22 174 196No 30 4 34Subtotal 52 178 230Insufficient data/not answered in 296Table 5.12 shows the Advisors’ opinion related to theDNACPR decision. In the 178 cases where the treatingclinician had not made a DNACPR decision the Advisorsstated that in 174 cases this should have been made.However, in the 52 cases where the treating clinicianhad made a DNACPR decision the Advisors disagreedin 30 cases.The Advisors reviewed cases where a DNACPRdecision had been made to consider documentation ofcommunication. There was sufficient information in 23cases to assess this. In 11 cases it was agreed that therewas an effective system for recording this information andin five cases this was felt not to be so. In six cases it wasfelt that communication with the patient was effectiveand in eight cases this was not present. In 10 cases itwas stated that communication with the next of kin waseffective and in five cases this was not so.67

5 - Resuscitation statusKey FindingsRecommendationsCPR status was recorded in only 122/552 (22%) ofpatients. Of these 122 patients, 70 were for CPR and 52had a DNACPR decision.52 patients who received CPR had a documentedDNACPR decision.430/522 patients (78%) had no documentation aboutCPR status.Reasons stated for patients remaining for CPR included:Patient remained for full and active treatment (326/424;77%) and lack of time to discuss or document decision(60/424; 14%)In 196/230 cases where there was sufficient dataAdvisors felt that a DNACPR decision should havebeen made.An effective system for recording all decisions anddiscussions relating to CPR/DNACPR must beestablished, allowing all people who may care for thepatient to be aware of this information. (Medical Directors)Health care professionals as a whole must understandthat patients can remain for active treatment but that inthe event of a cardiac arrest CPR attempts may be futile.Providing active treatment is not a reason not to considerand document what should happen in the event of acardiac arrest. (All Health Care Professionals)The use of ‘ceilings of care’ documentation wouldfacilitate decision making and clarity of intent. There isneed for a national project to lead this work.68

Back to contents6 - Resuscitation attempt6 – Resuscitation attemptThis section of the report covers the period of timeimmediately surrounding the cardiac arrest andCPR attempt. Data were sourced from the clinicianquestionnaires, Advisor assessment forms andresuscitation questionnaires. The latter were completedmainly by the person responsible for leading theresuscitation team. The number of resuscitation formsreturned was higher than the number of clinicianquestionnaires and therefore for some analyses thedenominator is different.Table 6.1 Location of cardiac arrestLocation n %Surgical ward 217 27.8Medical ward 212 27.1Coronary care unit 94 12.0Emergency department 63 8.1Procedure/intervention area 54 6.9Operating room/post-operativeanaesthetic care unit 13 1.7Outpatient area 10 1.3Level 2 care 9 1.2Other 109 14.0Subtotal 781Not answered 6Total 787Table 6.1 provides more detail on the location of cardiacarrest than that presented earlier (taken from the clinicianquestionnaire). The most common locations were stillmedical and surgical wards (54% of cardiac arrestsoccurred in these two locations). It is notable that almostone in three cardiac arrests occurred in areas that wouldbe considered higher level care with more monitoring andhigher nurse to patient ratios (operating room, coronarycare unit, emergency department, intervention area, Level2 care).Table 6.2 Time of cardiac arrestTime n %00:00-07:59 285 36.708:00-17:59 318 41.018:00-23:59 173 22.3Subtotal 776Not answered 11Total 787Table 6.2 shows the time of the cardiac arrest. 318/776(41%) cardiac arrests occurred in normal working hours.It is known that out of hours staffing generally involvesfewer available doctors in training, who are also dealingwith a demanding workload, and less direct input ofconsultant staff. This may have a direct bearing onavailability of resuscitation teams and consultant inputinto decision making regarding the cardiac arrest. As60% of cardiac arrests happened out of hours this is apotential consideration.Table 6.3 shows the grade of the team leader on theresuscitation team for each cardiac arrest call. Consultantleadership was low (67/754: 9%). Perhaps this iswhat most clinicians working in hospital would expectgiven their knowledge of how resuscitation teams areconstituted but such low consultant involvement in anacute life threatening situation could be questioned.In 148/754 (20%) cardiac arrests the team leader was abasic grade or junior specialist trainee. Consideration ofthe composition and leadership of resuscitation teamsneeds to take place.69

6 - Resuscitation attemptTable 6.3 Team leader at the resuscitation attemptTeam leader n %Consultant 67 8.9Staff grade/associate specialist 74 9.8Trainee with CCT 2

6 - Resuscitation attemptNumber of patients250200Primary cardiac diseaseNon-cardiac diseaseUnknown150100500VF VT Asystole PEA Not monitored/unknownCardiac RhythmFigure 6.2 Underlying cause and initial primary rhythm of cardiac arrest(n=730; not answered in 57)Table 6.5 Primary rhythm at cardiac arrestPrimary rhythm n %Ventricular fibrillation 79 11.1Ventricular tachycardia 31 4.4Asystole 227 31.9Pulseless electrical activity 375 52.7Subtotal 712Not monitored/unknown/not answered 75Total 787The primary rhythm is presented in Table 6.5. A minorityof patients had VF/VT as the primary rhythm, almost onethird asystole and half the patients PEA. It is known thatthe primary rhythm has an important association withoutcome.Figure 6.2 shows primary rhythm by cardiac/non-cardiacdisease. Most of the VF/VT cases were patients who hada cardiac arrest secondary to primary cardiac disease.Table 6.6 shows if the cardiac arrest was witnessed and/or monitored. In 480/714 cases where this question wasanswered, the cardiac arrest was witnessed (67%). In307/714 cases where this question was answered, thecardiac arrest was monitored (43%). In 660 cases ananswer was provided to both questions – in this group251 cases were both monitored and witnessed (38%) and200 cases were neither monitored nor witnessed (30%).Table 6.6 Witnessed and /or monitored cardiac arrestMonitoredWitnessed Yes No Subtotal Not answered TotalYes 251 184 435 45 480No 25 200 225 9 234Subtotal 276 384 660 54 714Not answered 31 23 54 19 73Total 307 407 714 73 78771

6 - Resuscitation attemptTable 6.7 Arrest was witnessed assessed against patient outcomeWitnessedPatient survived to dischargeYes No Subtotal Not Not answered TotalapplicableYes 69 261 330 2 7 339No 11 165 176 0 2 178Subtotal 80 426 506 2 9 517Not answered 5 41 46 0 1 47Total 85 467 552 2 10 564Table 6.8 Arrest was monitored assessed against patient outcomeMonitoredPatient survived to dischargeYes No Subtotal Not Not answered TotalapplicableYes 57 144 201 1 5 207No 22 283 305 1 4 310Subtotal 79 427 506 2 9 517Not answered 6 40 46 0 1 47Total 85 647 552 2 10 564Table 6.7 shows survival data broken down by whether ornot the cardiac arrest was witnessed. Survival to hospitaldischarge was greater in cases where the cardiac arrestwas witnessed.Table 6.8 shows survival data broken down by whetherthe cardiac arrest was monitored. Survival to hospitaldischarge was greater in cases where the cardiac arrestwas monitored.These figures are based on only 564 patients from whomall the required data were available. Overall rates ofsurvival to hospital discharge rates were much lower andare presented in the next chapter.However, there appeared to be an association betweensurvival to hospital discharge and witnessing/monitoringof the patient at time of cardiac arrest. This may berelated to speed of recognition but could equally berelated to other factors. Patients most likely to benefitfrom CPR may be more likely to be cared for in areaswhere staffing and facilities allow direct observation andmonitoring.Of the 110 VT/VF cases in Figure 6.2, 84 were witnessedcardiac arrests.72

6 - Resuscitation attemptTable 6.9 shows competence and action of the witness tothe cardiac arrest in cases where the primary rhythm wasa shockable one (VF/VT).Table 6.9 When witnessed, the witness was competent todefibrillateWitness was competent to defibrillate TotalYes, and they did 58Yes, but they did not 8No 8Subtotal 74Not applicable (defibrillationwas inappropriate) 10Total 84In 66/74 cases the witness was competent to defibrillateand in most of the instances (58/66) the witness diddefibrillate the patient. As time to defibrillation is akey intervention to improve outcome these data arereassuring. However, in almost one in eight cases thewitness was not competent to defibrillate leading to delayin providing appropriate treatment.Table 6.10 shows the time interval from recognition ofcardiac arrest to defibrillation. It should be noted that thisquestion was not answered in 34/110 cases (31%) andthis possibly reflects the difficulty of collecting accuratedata in the emergency situation of a cardiac arrest. Wheredata were provided the majority of patients receiveddefibrillation very quickly. However almost one in five hada delay of longer than three minutes prior to defibrillation.The cases where long delays were noted were few butwere of concern.Table 6.11 shows time from recognition of cardiac arrestto initiation of resuscitation attempt.Table 6.11 Time from cardiac arrest to the resuscitationattemptTime n %Immediately 287 59.11-3 minutes 178 36.64-6 minutes 12 2.57-8 minutes 1

6 - Resuscitation attemptThe vast majority of patients had immediate or very quicktreatment. Whilst the numbers are small it is difficult tounderstand why initiation of resuscitation attempt wasdelayed in the remainder. This is self reported data fromthe resuscitation questionnaire and details of delays werenot requested. All health care staff should be capableof providing basic life support and delays in initiation ofbasic life support should not occur.Duration of CPRFigure 6.3 shows the duration of CPR attempt, asreported on the resuscitation questionnaire. No detailof the duration of CPR attempt was provided in 149/787cases. Figure 6.3 provides more detail on the durationof CPR.There was a high proportion of relatively short CPRattempts; 22% of CPR attempts lasted five minutes orless and 43% lasted for 10 minutes or less.Figure 6.4 shows duration of CPR attempt in cardiac andnon-cardiac causes of cardiac arrest (n=481).The totals are lower than those in Figure 6.3 as the causeof cardiac arrest was not stated in some cases. Thereare some differences in duration of CPR between arrestswith cardiac and non-cardiac causes. Thirty-two percentof CPR attempts where primary cardiac disease was thecause of cardiac arrest lasted for five minutes or less.The corresponding figure for cardiac arrests due to noncardiacdisease was 19%. Furthermore almost half ofcardiac arrests due to primary cardiac disease had CPRfor 10 minutes or less.Table 6.12 shows what interventions the patients receivedduring the CPR attempt.Table 6.12 Interventions applied during CPRInterventions n %Chest compressions 726 94.3Assisted ventilation 586 76.1Adrenaline 579 75.2Tracheal intubation 310 40.3Defibrillation 179 23.2Supraglottic airway device 99 12.9Answers may be multiple (n/770; not answered in 17)Number of patients1601401201008060402002 hrsFigure 6.3 Duration of CPR attempt (n=638; not answered in 149)74

6 - Resuscitation attemptPercentage of patients25Cardiac diseaseNon-cardiac disease201510502 hrsDuration of CPRFigure 6.4 Duration of CPR by cardiac and non-cardiac disease (n=481; not answered in 206)Almost all patients received chest compressions but onlythree quarters received assisted ventilation or adrenaline.Of the patients who received assisted ventilation,310/586 had tracheal intubation (53%) and 99 had use ofa supraglottic airway device (99/586:17%).Patients requiring resuscitation often have an obstructedairway. In these cases, prompt assessment, with controlof the airway and ventilation of the lungs, is essential.Without adequate oxygenation it may be impossible torestore a spontaneous cardiac output. Although for awitnessed cardiac arrest in the vicinity of a defibrillator,attempted defibrillation takes precedence over openingof the airway. That said, in this study many patientswho were ventilated were managed with bag andmask ventilation only; 20 patients apparently receivedCPR for over 30 minutes without tracheal intubation orsupraglottic airway device. This may be less than idealpractice when one considers the availability of alternativedevices and techniques that can provide a more patentairway and reduce gastric distension.It is intuitive that the management of the airway duringcardiac arrest will be influenced by the duration of CPRand Figure 6.5 details the management of the airway byduration of cardiac arrest.Figure 6.5 shows the airway technique for each timebandof CPR duration. The percentages refer topercentage of patients in each time-band.The tracheal tube has generally been considered theoptimal method of managing the airway during cardiacarrest. But there is evidence that, without adequatetraining and experience, the incidence of complications,such as unrecognised oesophageal intubation (6-17%in several studies involving paramedics) is unacceptablyhigh. 37Prolonged attempts at tracheal intubation are harmful;the cessation of chest compressions during this time willcompromise coronary and cerebral perfusion. Severalother airway devices (supraglottic airway devices) may beused for airway management during CPR.The supraglottic airway devices are easier to insertthan a tracheal tube and, unlike tracheal intubation,can generally be inserted without interrupting chestcompressions. 38 There were no data collected in thisstudy which supported the routine use of any specificapproach to airway management during cardiac arrest.75

6 - Resuscitation attemptPercentage of patientsNo Tracheal Intubation/Supraglottic AirwayTracheal IntubationSupraglottic AirwayTracheal Intubation & Supraglottic Airway10090807060504030201002 hrsDuration of CPRFigure 6.5 Duration of CPR and treatment received (n=634; not answered in 153)Table 6.13 shows whether an anaesthetist or intensivistwas part of the resuscitation team.Table 6.13 Presence of and anaesthetist/ intenstivist on theresuscitation teamAnaesthetist/Intensivist on team Total %Yes 486 76.7No 148 23.3Subtotal 634Not answered 153Total 787Data were provided on the resuscitation questionnairein 634 cases. In 486 cases (77%) an anaesthetist orintensivist was part of the resuscitation team. Where ananaesthetist or intensivist is not part of the team theremay be concerns about airway management. This maydepend on the skills and abilities of other members of theteam and this may be influenced by the location of thecardiac arrest.76

6 - Resuscitation attemptTable 6.14 shows the location for those cardiac arrestswhere an anaesthetist or intensivist was not part of theteam.Table 6.14 Location of arrestLocation total %Medical ward 48 32.7Surgical ward 33 22.4Coronary care unit 25 17.0Procedure/intervention area 12 8.2Emergency department 9 6.1Outpatient area 1

6 - Resuscitation attemptCase study 12A middle-aged patient was admitted to hospitalwith a presumptive diagnosis of Pneumocystiscarnii pneumonia. The patient was startedpromptly on appropriate therapy and admittedto a ward a few hours later. The patientdeteriorated rapidly and there was inadequaterecognition or escalation. No further medicalreview took place until the patient had a cardiacarrest 36 hours later.During CPR attempts the patient regurgitated,and gastric contents were noted in the maskthat was being used to provide ventilation. Thesuction unit was not working and this took sometime to resolve. The resuscitation team did notinclude anyone who could intubate the patient atthat time so bag and mask ventilation continued.An anaesthetist arrived 10 minutes later andsecured the airway with a tracheal tube. Return ofcirculation was obtained after 25 minutes of CPRand the patient was transferred to the intensivecare unit. The patient did not recover and died inthe intensive care unit seven days later.The Advisors considered that this case highlightsthree main issues:1. Lack of recognition or intervention inresponse to evidence of acute deterioration2. Lack of functional equipment3. Delay in obtaining rapid and definitive airwayprotection.These findings of problems related to CPR attemptssupport work by the NPSA published in 2007. 14Because the number of cases where resuscitationquestionnaires were returned was greater than thenumber of clinician questionnaires or notes, thedenominator for this response is lower (526 cases).However there were 91 individual problems identifiedby the Advisors giving a crude rate of problems of91/526 (17%). The most frequent problems were airwaymanagement (36/526; 7%), presence of appropriatestaff and equipment problems. Communication andteamwork were highlighted infrequently as problems bythe Advisors, this may reflect the difficulty in assessingthis domain retrospectively from the case notes. Airwaymanagement and equipment were highlighted both byclinicians returning the resuscitation form and Advisorsand attention should be focused on both these areas.It is worth noting again in the context of airwaymanagement concerns that only 53% of patientsundergoing CPR had either a tracheal tube or supraglotticairway. Many patients did not have advance airwaymanagement and this is something that requires furtherstudy and possibly better education.Advisors were asked if problems in the resuscitationprocess could have affected outcome and in 5/67 casesthis was considered to be likely.Cardiac arrest leading to CPR attempt may often beconsidered to be a critical incident. Even if the conduct ofthe CPR event is problem free, the very nature of cardiacarrest merits reporting and recording through a criticalincident reporting system. Table 6.17 shows that in only57 cases (9%) was this the case.Table 6.17 Critical incident reportedCritical incident n %Yes 57 8.7No 596 91.3Subtotal 653Unknown/not answered 134Total 78778

6 - Resuscitation attemptOther systems for recording cardiac arrest and CPRattempts may exist but the use of an organisation’scritical incident reporting system is more likely to allowbetter data collection of the incidence of event andreporting through the clinical governance structure.Whilst the use of critical incident reporting systems willprovide data on the incidence of cardiac arrest it will notprovide more detailed information on the circumstances,conduct and outcomes of cardiac arrest. Each trust/hospital should collect structured information on patientswho have a cardiac arrest. The National Cardiac ArrestAudit collects such data and hospitals are encouraged toparticipate. 3979

6 - Resuscitation attemptKey FindingsRecommendationsMore than half of the cardiac arrests in this studyoccurred on medical/surgical wards (429/781; 55%).458/776 cardiac arrests (59%) occurred ‘out of hours’.Most cardiac arrests where the cause was known weresecondary to non-cardiac disease (356/591; 60%).The initial rhythm was pulseless electrical activity in 53%,asystole in 227/712 (32%) and VF/VT in 110/712 (15%).Almost one in five patients in whom defibrillation wasindicated did not receive a shock within 3 minutes ofrecognition of cardiac arrest.In only 486/634 cases (77%) an anaesthetist or intensivistwas part of the resuscitation team.There were 234 problems identified by the treatingclinicians during the 787 resuscitation attempts. Themost common problems were equipment (7%), airwaymanagement (6%) and team work (4%).Hospitals must arrange services and equipment to ensurethat defibrillation is delivered within three minutes ofcardiac arrest (for shockable rhythms). (Medical Directors)All CPR attempts should be reported through theTrust/Hospital critical incident reporting system. Thisinformation should be reported to the Trust/HospitalBoard on a regular basis. (Medical Directors)Each Trust/Hospital should set a local goal for reductionin cardiac arrests leading to CPR attempts. Progressagainst this goal should be reported to the Trust/HospitalBoard on a regular basis. (Medical Directors)Each hospital should ensure there is an agreed plan forairway management during cardiac arrest. This mayinvolve bag and mask ventilation for cardiac arrests ofshort duration, tracheal intubation if this is within thecompetence of members of the team responding tothe cardiac arrest or greater use of supraglottic airwaydevices as an alternative. (Medical Directors)The Advisors reported problems during the resuscitationattempt in 91/526 cases (17%). Of these, 36/91 wereassociated with airway management.80

Back to contents7 - Period after thecardiac arrest7 – Period after the cardiac arrestOutcome after CPROutcome after CPR can be measured at different timepoints and clarity of definition is important to ensurethat data are comparable. 40 Table 7.1 below shows thenumber of patients who had return of circulation andsurvived the immediate CPR attempt.Table 7.1 Outcome of CPR attemptOutcome n %Death 381 65.6Survival 200 34.4Subtotal 581Not answered 4Total 585Whilst 34% of patients had return of circulation afterCPR this needs to be considered in the context ofoverall benefit. Figure 7.1 shows the number of patientswho survived the CPR attempt and survived to hospitaldischarge.Of the 200 patients who survived the immediate CPRattempt only 85 survived to hospital discharge (14.5%survival to hospital discharge rate: 85/585). This is atthe lower end of, but in keeping with, the literature onhospital survival after cardiac arrest, which shows ratesof hospital survival from 14-20%. 41,1Cardiac arrest can lead to hypoxic-ischaemic cerebralinjury and disability in those who survive from cardiacarrest. It is important to understand to what degree thishas occurred in order to have a fuller picture of outcomeafter cardiac arrest. The “Clinical Performance Score”was initially proposed by Jennett and Bond in The Lancetin 1975 as a way of evaluating the outcomes of braininjury, mainly from head trauma. 42 It has been used todetermine the neurological status of patients after cardiacarrest, notably in the setting of clinical trials of therapeutichypothermia 43,44 and is more commonly termed the‘cerebral performance category’ (CPC).Cardiac Arrest(585)Died(381)Not answered(4)Survived CPRattempt 200/581(34.4%)Not answered/not applicable(7)Survived todischarge 85/581(14.6%)Died(108)Figure 7.1 Survived to discharge after CPR81

7 - Period after thecardiac arrestTable 7.2 shows clinicians’ assessment of cerebralperformance category (CPC) at the time of hospitaldischarge in the group who survived. It is worthremembering that these data were obtained from theclinician questionnaire and that the questionnaire wasfilled out retrospectively with the aid of case notes only.It is unlikely that formal CPC scoring had been performedand recorded and likely that clinicians had to make ajudgment from available notes.Table 7.2 Cerebral performance category (CPC)CPCTotal1. Conscious, alert-normal function 712. Conscious, alert-moderate disability 53. Conscious, severe disability 14. Comatose 1Subtotal 78Not answered 7Total 85In this study only 2/78 patients had severe disability orwere comatose at time of hospital discharge. It is worthnoting that only five patients were classified as CPC 2 –conscious and alert with moderate disability. Whilst it isreassuring that 71 patients were classified as CPC 1 itis possible that the differentiation between CPC 1 and 2was difficult. Furthermore given the above paragraph it ispossible that around half of the individuals with a CPC 1actually had moderate to severe handicaps that were notrecognised and that disability was underestimated.Despite the face validity of the CPC, studies by Roineet al 45 and Hsu et al 46 have previously demonstratedthat there is a low correlation between a score of “2” andfunctional abilities as measured on a well-standardisedscale from critical care, and that nearly half of individualswith a “1” actually have moderate to severe cognitivehandicaps. Assessment of quality of life after cardiacarrest survival is complex and the CPC is a crude tool. 47Survived CPRattempt(200; 34.4%)Not answered orNot applicable(still an inpatient)(7)Survived todischarge(85; 14.6%)Died(108)Discharged home(55; 9.5%)Discharged toother institution forfurther care(30; 5.2%)Figure 7.2 Discharge location82

7 - Period after thecardiac arrestPercentage of patients100SurvivedDied9080706050403020100VF VT Asystole PEA Not monitored/unknown rhythmPrimary cardiac rhythmFigure 7.4 Primary cardiac rhythm and survival to discharge (n=537)Initial rhythm when CPR is commenced is associatedwith different outcomes. Figure 7.4 shows these data.It can be seen that patients who presented withshockable rhythms had higher survival to discharge rates.Only nine patients presenting with asystole survived tohospital discharge.Figure 7.5 shows data for the patients who survived tohospital discharge, according to presenting rhythm andwhether or not the cause of cardiac arrest was cardiacor non-cardiac.Figure 7.6 shows the association between length of hospitalstay prior to cardiac arrest and survival to discharge.Patients who suffered an arrest after longer durationsof hospital stay were less likely to survive to discharge.Patients who had been in hospital for more than four dayshospital had survival rates of only 8%. As discussed earlier,insufficient time to document or discuss DNACPR wasfrequently cited as a constraint to decision making. Thedata from Figure 7.6 shows that that survival to hospitaldischarge was only 12% in patients who had a hospital stayof longer than one day prior to cardiac arrest (44/376): is itreasonable to cite lack of time in this group of patients?Only 2/78 patients (2.5%) with asystole secondary tonon-cardiac causes and only 12/147 (8%) patientswith PEA secondary to non-cardiac causes survived tohospital discharge.84

7 - Period after thecardiac arrestPercentage of patients100Cardiac causeNon-cardiaccause9080706050403020100VF VT Asystole PEA Not monitored/unknownrhythmPrimary cardiac rhythmFigure 7.5 Percentage of patients that survived to discharge and typeof primary rhythm and cause of cardiac arrest (n=424)Percentage of patients100Survived to dischargeDied90807060504030201007Duration of hospital stay (days)Figure 7.6 Duration of hospital stay and survival to discharge (n=551)85

7 - Period after thecardiac arrestTable 7.3 Outcome by time of arrestPatient survived to dischargeTime Yes % No % Subtotal Insufficient data Totalto assess00:00-07:59 13 7.4 163 92.6 176 5 18108:00-17:59 44 20.1 174 79.8 218 6 22418:00-23:59 15 12.5 105 87.5 120 1 121Total 72 14.0 442 86.0 514 12 526Table 7.4 Outcome by day of the weekPatient survived to dischargeDay Yes % No % Subtotal Insufficient data Totalto assessMonday to Friday 57 15.4 314 84.6 371 7 378Saturday and Sunday 15 10.5 128 89.5 143 5 148Total 72 14.0 442 86.0 514 12 526Table 7.5 Outcome by arrests occurring out of hoursPatient survived to dischargeDay Yes % No % Subtotal Insufficient data Totalto assessMonday to Friday in hours 52 19.8 210 80.2 262 2 264Saturday and Sunday/outof hours 20 7.9 232 92.1 252 10 262Total 72 14.0 442 86.0 514 12 526Table 7.3 shows the association between time of day ofcardiac arrest and survival to discharge. The variationis considerable: 20.1% during daytime, 12.5 % in theevening and 7.4% at night. Patients having CPR attemptsduring the day were almost three times more likely tosurvive than those having CPR attempts at night. Thepoorer survival of patients having a cardiac arrest atnight has previously been shown. 48 It is not clear whetherthis is due to problems with provision of appropriatepersonnel and facilities overnight or due to decisionmaking (DNACPR status or post cardiac arrest care) inthis period. However it does highlight a particular issueand individual hospitals should pay attention to their owndata and the pattern of medical care that can be providedin the overnight period.Table 7.4 shows survival to discharge by day of cardiacarrest. The hospital survival of patients who had acardiac arrest at weekends was lower than during theweek. The association of weekend admissions withmortality has been found previously. 49 From Table 7.4 itappears that this was also a finding in patients who havea cardiac arrest over the weekend period.86

7 - Period after thecardiac arrestPercentage of patients100Survived to dischargeDied9080706050403020100Operating room/PACUCoronary care unitEmergency departmentProcedure areaOutpatient areaHigh dependency unitMedical wardSurgical wardOtherLocation of cardiac arrestFigure 7.7 Location of cardiac arrest and survival to discharge (n=547)Table 7.5 presents data on cardiac arrests that occurredbetween the hours of 08:00-18:00 Monday to Friday andcardiac arrests that occurred at weekends and between18:00-07:58 (out of hours).It is clear that there was an association between outcomeand time of cardiac arrest; patients who had a cardiacarrest during the normal working week had more thandouble the hospital survival rate of those patients whohad a cardiac arrest out of hours or at weekends.Figure 7.7 shows the association between location attime of cardiac arrest and survival to hospital discharge. Itdoes appear that cardiac arrests that led to the initiationof CPR were associated with a better outcome if theyoccurred in a highly monitored environment (operatingtheatre/post anaesthetic care unit, procedure/interventionarea, Level 2 care). This may not be surprising. Arreststhat occurred in general wards (medical and surgical)or in the emergency department were associated withvery poor outcomes. These data are potentially complexas many other factors may be involved – disease state,acute presentation, appropriateness of CPR attempt toname a few.It is important to remember that the clinician caring forthe patient considered that the patient was on the correctward at the time of cardiac arrest in 521/565 cases (92%)and therefore had no concerns about appropriateness oflocation for the level of care required.87

7 - Period after thecardiac arrestTable 7.6 Advisors’ opinion of appropriateness of whether the patient should have had a DNACPR decision by location.Should have had a DNACPRLocation of cardiac arrest Yes No Subtotal Insufficient data Totalto assessOperating room/postanaesthetic care unit 2 0 2 5 7Coronary care unit 14 2 16 49 65Emergency department 16 8 24 20 44Procedure area 5 3 8 25 33Outpatient area 2 1 3 4 7Level 2 care 1 1 2 3 5Medical ward 68 4 72 72 144Surgical ward 59 9 68 69 137Other 22 5 27 39 66Subtotal 189 33 222 286 508Not answered 2 1 3 1 4Total 191 34 225 287 512The Advisors’ opinion of DNACPR, combined withthe responsible clinicians’ view of appropriateness oflocation, reinforces the literature that points to cardiacarrests occurring on general wards being associated withpoor outcomes and shows that our Advisors believedthat CPR attempts in the patients in these locations werevery often not indicated. Clinicians who are responsiblefor patient care in medical and surgical wards shouldunderstand the potentially poor outcome for their patientsshould they have a cardiac arrest and, if the patients’location is believed to be appropriate, be prepared toquestion the appropriateness of initiating CPR should thepatient have a cardiac arrest.In Chapter 5 it was shown that 52 patients who theclinician had documented that DNACPR was appropriateunderwent CPR. Figure 7.8 shows the outcome for thesepatients.Survived todischarge 2(3.8%)Survived cardiacarrest 21(40.4%)DNACPR decisionby responsibleclinician 52Died19Died31Figure 7.8 Patients who had a DNACPR decision butstill underwent CPRTwenty-one patients survived the resuscitation attemptand 19 of those event survivors died later in their hospitalstay. Two patients survived to hospital discharge.88

7 - Period after thecardiac arrestCare after a cardiac arrestSurvivors of CPR require rapid and systematic care ifchances of survival and good neurological outcome areto be maximised. Often this starts with some simpleinvestigations to understand better the cause of thecardiac arrest, to determine the degree of physiologicalderangement and to allow planning of subsequent care.Table 7.7 shows the number of patients who had a rangeof simple investigations performed in the immediate postarrest period. It is noted that substantial numbers ofpatients did not have these investigations performed.appropriate. Table 7.8 shows that in the post arrestperiod 84/191 (44%) were made DNACPR in the event ofa further cardiac arrest. As discussed earlier it is possiblethat full and active therapy would still be delivered tothese patients and so a DNACPR decision does not ruleout elements of post cardiac arrest care. However it doesraise the likelihood that many patients may have beenmanaged with a more palliative approach. Table 7.8 alsoshows that all but 11 of the patients with a DNACPRdecision died in hospital which raises again the questionof whether some of these patients should have had aDNACPR decision in the first instance.Table 7.7 Investigations performed in the immediate postarrest periodInvestigations performed n %12 lead ECG 122 79.7Full blood count 111 72.5Urea and electolytes 110 71.9Chest x-ray 89 58.2Arterial blood gasses 103 67.3Answers may be multiple (n/153; not answered in 55)One reason for apparent ‘deficiencies’ in post cardiacarrest care may be that patients were considered to bedying and that active therapy was no longer consideredIt is recognised that interventional cardiology has asignificant role to play in patients who have a cardiacarrest secondary to myocardial ischaemia. In thosepatients who survived the cardiac arrest, Advisors wereasked to consider if the cause of the arrest was likely tobe due to myocardial ischaemia. Advisors consideredthat this was likely in 100/190 cases with sufficientinformation to make an assessment.In those 100 patients with likely myocardial ischaemiaas a cause for the cardiac arrest data were extractedfrom the case notes to assess if cardiology input wasrequested, angiography considered or angiography/intervention undertaken. Figure 7.9 shows the resultsof this analysis.Table 7.8 Outcome when a DNACPR order was made after CPRPatient survived to dischargeFollowing CPR, DNACPRorder was made Yes No Subtotal Not applicable Not answered TotalYes 11 72 83 0 1 84No 70 33 103 1 3 107Subtotal 81 105 186 1 4 191Unknown 4 3 7 0 2 9Total 85 108 193 1 6 20089

7 - Period after thecardiac arrestNumber of patients70YesNo6050403020100Discussed withcardiology?Coronory angiographyconsidered?Angiography/interventionperformed?Input from cardiologyFigure 7.9 Cardiology input in patients with CVS aetiology (n=100)It is well recognised that post-cardiac-arrest patientswith ST elevation myocardial infarction (STEMI) shouldundergo early coronary angiography and percutaneouscoronary intervention (PCI). 50,51 However, chest pain and/or ST elevation are relatively poor predictors of acutecoronary occlusion in these patients. 52 For this reasonthis intervention should be considered in all post-cardiacarrestpatients who are suspected of having coronaryartery disease as the cause of their arrest. 52-54From the findings in this study it appears that there was arelatively low use of PCI after cardiac arrest. This may bedue to an over-estimate, by the Advisors, of cases whereit was thought that myocardial ischaemia may be a factor.However, if patients are to have maximal chance of goodoutcomes the importance of PCI after cardiac arrest mustbe understood and arrangements put in place to ensurethis can be delivered reliably when required.Quality of care immediately following CPRAdvisors were asked to grade the quality of clinical carein the immediate (up to first hour) post arrest period andalso decision making in this immediate period. Figure7.10 shows these opinions.90Whilst it is reassuring that such high numbers of patientswere thought to have received good care and gooddecision making in the immediate post arrest period,there does appear to be opportunities to improve. Wheredecision making was considered less than good thereasons for this are given in Table 7.9.Table 7.9 Reasons for poor decision makingReasonsTotalSpeed of decision making 8Clarity about care required 22Seniority of decision making 17Other 9Answers may be multiple (n/49)It is a theme running through this report that lack of clarityabout the care required (both before and after cardiacarrest) and seniority of decision making to ensure themost appropriate treatment are deficiencies. Addressingboth of these will bring significant improvements to thisgroup of patients.

7 - Period after thecardiac arrestPercentage of patients8070GoodAdequatePoor6050403020100Advisor opinion: clinical careAdvisor opinion: decision makingFigure 7.10 Advisor opinion on quality of care immediately following CPR(Clinical care n=177; Decision making n=189)Table 7.10 Appropriate location after the arrest -Advisors’ opinionAppropriate location after thecardiac arrest n %Yes 194 95.1No 10 4.9Subtotal 204Insufficient data to assess 4Total 208Table 7.11 Referred to critical care after the arrestReferred to critical care Total %Yes 109 64.1No 61 35.9Subtotal 170Insufficient data to assess 38Total 208Advisor opinion of post cardiac arrest location is givenin Table 7.10. In most cases the Advisors felt that thelocation was appropriate to the needs of the patient. In10 cases this was not so and of those cases it was statedthat five patients should have received a higher level ofcare and in four cases there was no benefit to the higherlevel of care provided.Critical careIt has been estimated that there are approximately30,000 treated out of hospital cardiac arrests and 20,000treated in-hospital cardiac arrests each year in the UKand approximately 6350 of these (12.7%) will be admittedto an intensive care unit. 55 It has been established thatinterventions in the post-resuscitation period have asignificant influence on the ultimate outcome. 53,56Table 7.11 shows that 109 patients were referred tocritical care for consideration of admission aftercardiac arrest.91

7 - Period after thecardiac arrestTable 7.12 Grade of clinician who referred the patientto critical careGrade of clinicianTotalConsultant 17Staff grade 2Senior specialist trainee 39Junior specialist trainee 23Basic grade 6Other 1Subtotal 88Insufficient data to assess 21Total 109It was not possible to ascertain the grade of referringclinician in 21/109 cases. However in the remaining 88cases, only 17 cases were referred by consultants (19%)(Table 7.12). Earlier in this study data demonstrated thatthere were often concerns about decision making (bothin speed and clarity of decision making) and this relativelylow level of consultant involvement in post cardiac arrestcare decision making appeared poor.Table 7.13 Outcome of referral to critical careOutcome n %Admitted 75 68.8Not admitted 34 31.2Total 109Table 7.13 shows that of the patients referred to criticalcare 34/109 were declined admission and 75 patientswere admitted.Table 7.14 Grade of clinician making the decision to admitthe patient to intensive careGrade ofNotclinician Admitted admitted TotalConsultant 27 17 44Senior specialist trainee 19 7 26Junior specialist trainee 6 1 7Subtotal 52 25 77Insufficient data to assess 23 9 32Total 75 34 109Table 7.14 shows the grade of staff involved in thedecision on admission to critical care. Just over half ofthe decisions were made by consultants (44/77; 57%).Although the numbers are quite small it is interesting tonote that the proportion of referrals admitted to criticalcare was lower for consultants than senior trainees withjunior trainees admitting the highest proportion. It maybe that consultants are involved with the more complexpatients and that this explains the difference. However,senior decision making is key to ensuring that the correctcare is provided and this is very important in acutesituations such as those surrounding a cardiac arrest.Earlier NCEPOD work has highlighted the importance ofconsultant involvement in decisions to admit to criticalcare and this message needs to be reinforced. 2Table 7.15 Reason for no admission to critical careReasonTotalNo need for admission, patient wouldrecover with lower level care 3No need for admission, patient expected to die 26No critical care beds, patient would havebeen admitted but no facility 1Other 4Total 3492

7 - Period after thecardiac arrestWhere patients were not admitted to critical care theAdvisors were asked to classify why this was. In themajority of cases this was due to futility as the patientwas expected to die, although in 32/113 cases it wasfelt that recovery with current level of care was to beexpected (Table 7.17).Table 7.18 Appropriate admission to critical care - Advisors’opinionAppropriate admission94totalYes 78No 5Subtotal 83Insufficient data 1Total 84Advisor opinion of appropriateness of a Level 3 admissionin those patients who were admitted is shown in Table7.18. It was felt that most admissions were appropriatebut in a few cases it was felt that the patient was likely todie irrespective of critical care admission.Table 7.19 Treatment limitation decisions for the patientsadmitted to critical careTreatment limitationtotalNot for ventilation 11Not for renal replacement therapy 8Not for inotropic support 5Not for escalation of care above currentlevel of organ support 28Other 7Answers may be multiple (n/41)Table 7.20 Opinion of appropriate organ supportAppropriate organ support n %Yes 105 95.5No 5 4.5Subtotal 110Insufficient data to assess 98Total 208Treatment limitation can be an appropriate way ofensuring that treatments that will bring no ultimatebenefit are not provided and that the dying processis not prolonged. Advisor opinion was that all but fivepatients had appropriate organ support provided. In onepatient it was felt that respiratory support should havebeen provided and in two patients it was felt that renalsupport should have been provided and in one patientcardiovascular support was deficient. It appears thatthe treatment limitation decisions in Table 7.19 wereappropriate. Only nine patients were actively cooled inthe post arrest period. Table 7.21 shows the number ofpatients (admitted to critical care) with obtunded cerebralfunction after cardiac arrest.Table 7.21 Obtunded cerebral functionObtunded cerebral functiontotalYes 49No 21Subtotal 70Insufficient data to assess 14Total 84There is good evidence that cooling in the immediatepost arrest period improves outcome in out of hospitalVF arrests 43,57 and there is some lower level evidence(observational data only) indicating that inducedhypothermia may be beneficial after cardiac arrest fromnon-shockable rhythms and after in-hospital cardiacarrest. However, this is controversial. The data from thisstudy show that cooling was infrequent, not even in thepresence of obtunded neurological function.

7 - Period after thecardiac arrestOf the 84 patients admitted to critical care 28 survivedto hospital discharge. Table 7.22 shows the numberof patients in whom life-sustaining therapies werewithdrawn during the stay in critical care.Table 7.22 Life-sustaining therapies were withdrawnTherapies withdrawn n %Yes 38 48.7No 40 51.3Subtotal 78Insufficient data to assess 6Total 84Life sustaining therapies were withdrawn in 38 casesand in the opinion of the Advisors this was the correctdecision in all cases (35) where they could form anopinion. In 13 of these 38 cases life sustaining therapieswere withdrawn on the basis of predicted poorneurological outcome.Organ donation is a possible option to consider inpatients in intensive care when active life sustainingtherapies are being withdrawn. Within this study itappeared that it was only considered in six of the 33cases where it could have been an option. There aremany issues that may preclude organ donation and wedid not collect data on these. However it is important tohighlight the potential for organ donation and ensure thatthis is considered as a usual part of the dying process inall potential organ donors. 33,59Table 7.23 Organ donation was considered (in ICU patientswhere life-sustaining therapy was withdrawn)Organ donation was consideredTotalYes 6No 19Subtotal 25Insufficient data to assess 13Total 38In the 13 patients who had life sustaining therapieswithdrawn on the basis of poor neurological outcome thisdecision was made within 24 hours of cardiac arrest infour cases and within 48 hours of cardiac arrest in sevenpatients. Prediction of neurological outcome is unlikelyto be reliable until at least three days after cardiac arrest;if therapeutic hypothermia has been used it may extendthis time to five to six days after cardiac arrest. 5895

7 - Period after thecardiac arrestKey FindingsRecommendationsSurvival to discharge after in-hospital cardiac arrest was14.6% (85/581).Only 9/165 (5.5%) patients who had an arrest in asystolesurvived to hospital discharge.Survival to hospital discharge decreased as length ofhospital stay prior to cardiac arrest increased. Only14/554 (9.1%) of patients who had a cardiac arrestafter an inpatient stay of >7 days survived to hospitaldischarge.Survival to discharge after a cardiac arrest at night wasmuch lower than after a cardiac arrest during the day time(13/176; 7.4% v 44/218; 20.1%).In the post arrest period 84/191 (44.0%) patients had aDNACPR decision made.Location of post arrest care was judged to be appropriatein 95% of cases.Each hospital should audit all CPR attempts and assesswhat proportion of patients should have had a DNACPRdecision in place prior to the arrest and should not haveundergone CPR, rather than have the decision made afterthe first arrest. This will improve patient care by avoidingundignified and potentially harmful CPR attempts duringthe dying process. (Medical Directors)Consultant input is required in the immediate post arrestperiod to ensure that decision making is appropriateand that the correct interventions are undertaken.(Consultants)Coronary angiography and PCI should be considered inall cardiac arrest survivors where the cause of cardiacarrest is likely to be primary myocardial ischaemia.(Consultants)Organ donation should be considered in every casewhere life sustaining therapies are being withdrawn.(Consultants)Many patients were expected to die post cardiac arrestand were not admitted to critical care (66/113).It was considered that in 100 patients who had returnof circulation, the cause of the cardiac arrest primarymyocardial. Only 30 of these 100 patients had coronaryangiography, and PCI where appropriate, in the postcardiac arrest phase.Organ support in critical care was judged to beappropriate in most cases. (105/110; 96%).Life sustaining therapies were withdrawn in 38 cases.Organ donation was considered in six of those cases.96

Back to contents8 - Overall assessmentof care8 – Overall assessment of careThe Advisors considered the overall care of patients andgraded it as shown in Figure 8.1.Number of patients180160140120100806040200Good PracticeRoom forimprovement -clinicalRoom forimprovement -organisationalRoom forimprovement -bothLess thansatisfactoryInsufficientdataFigure 8.1 Overall quality of care - Advisors’ opinionThe care of patients was considered to be good in only29% (154) of patients assessed in this study. There wasroom for improvement in the clinical care of 132 (25%) ofpatients, room for improvement in the organisational careof 70 (13%) of patients and room for improvement in bothin 115 (22%) of patients. Overall the Advisor’s believedthat care was less than satisfactory in 33 (6%) patients.From Table 8.1 overleaf it is clear that there are bigdifferences when the data are displayed in this way.Of the group who survived 61% were judged to havereceived good care as opposed to 26% of the group whodied. This analysis is potentially subject to have criticismthat it was done in the knowledge of the outcome of thepatients and this may have influenced the judgement.However it appears that patients who died in hospitalwere believed to have less than good care in three out offour cases and that most of the deficiencies in care wereassociated with the ‘room for improvement in clinicalcare’ (including decision making and CPR status).Clinicians returning the clinical questionnaire were askedif there were any factors that, if changed, could haveaffected the outcome positively. Table 8.2 shows thatthere were 71 positive responses to this question andcategories the responses.97

8 - Overall assessmentof careTable 8.1 Advisors’ opinion of quality of care for those patients who survived to hospital discharge (Yes) and those who died inhospital (No).Patient survived to dischargeQuality of care Yes % No % Insufficient Not Totaldata to assess answeredGood practice 42 60.9 109 25.6 2 1 154Room for Improvement-clinical 10 14.5 121 28.5 0 1 132Room for improvement- organisational 9 13.0 60 14.1 0 1 70Room for improvement - both 8 11.6 105 24.7 1 1 115Less than satisfactory 0 0.0 30 7.1 1 2 33Subtotal 69 425 4 6 504Insufficient data to assess 3 17 2 0 22Total 72 442 6 6 526Table 8.2 Action that may have improved outcome ifsomething had been done differently - Clinician caring forthe patients’ opinionActionEarlier treatment of problem 14DNACPR decision 13Better monitoring 12Escalation to higher level of care 5Early warning score acted on 4Correction of wrong diagnosis made 3Escalation to consultant 2Administration of treatment as statedby the consultant 1Correction of wrong treatment 1Other 16Total 71nWhere care was considered to be less than goodthe Advisors were asked if they considered that thedeficiencies may have contributed to death. Table 8.3shows the response.Table 8.3 Less than good care contributed to death -Advisors’ opinionLess than good care contributedto the patients’ death total %Yes 81 31.9No 173 68.1Subtotal 254Insufficient data to assess 75Total 316There was sufficient information to answer this questionin 254/316 cases. In 81/254 cases the Advisors felt thatdeficiencies in care may have contributed to the death ofthe patient (32% of cases rated as less than good care,with sufficient information to form an opinion).98

8 - Overall assessmentof careSummaryIn summary, the care of patients who had an in-hospitalcardiac arrest was less than good in seven out of 10cases. Deficiencies were noted in the admission process,consultant involvement, decision making about CPRstatus, recognition of severity of illness and markersof risk, appreciation of urgency and requirement toescalate to more senior doctors. It appeared that theseaspects of care prior cardiac arrest provide the biggestopportunities to improve patient outcome. The report alsohighlights areas for improvement in both the resuscitationattempt and post cardiac arrest care. However it must beemphasised that the outcome for patients once they havehad a cardiac arrest is very poor and the focus must beon prevention in the first instance.The data from this study seems to give an overall pictureof unreliability in the recognition of the deterioratingpatient; failure to respond to deterioration reliably andfailure to engage senior doctors to direct interventioneither to prevent further deterioration or facilitateDNACPR decisions. This report therefore raises two mainchallenges to all health care professionals:1. To ensure rapid and consistent recognition andmanagement of acute illness in order to maximisepatients’ chance of recovery.2. To ensure that decision making about CPR is appliedconsistently, communicated effectively and thatCPR is performed only on patients who are likely tobenefit from it.99

1 Introduction100

Back to contentsReferencesReferences1. Meaney PA, Nadkarni VM, Kern KB, Indik JH,Halperin HR, Berg RA. Rhythms and outcomesof adult in-hospital cardiac arrest. Crit Care Med2010;38:101-82. National Confidential Enquiry into Patient Outcomeand Death. An acute problem? London: NCEPOD;2005.3. Kause J, Smith G, Prytherch D, Parr M, FlabourisA, Hillman K. A comparison of antecedents tocardiac arrests, deaths and emergency intensivecare admissions in Australia and New Zealand,and the United Kingdom--the ACADEMIA study.Resuscitation 2004;62:275-82.4. Schein RM, Hazday N, Pena M, Ruben BH,Sprung CL. Clinical antecedents to in-hospitalcardiopulmonary arrest. Chest 1990; 98(6):1388-13925. McQuillan P, Pilkington S, Allan A et al. Confidentialinquiry into quality of care before admission tointensive care . BMJ 1998; 316: 1853-1858.6. McGloin H, Adam SK, Singer M. Unexpected deathsand referrals to intensive care of patients on generalwards. Are some cases potentially avoidable?J R Coll Physicians Lond 1999;33(3):255-259.7. Seward E, Greig E, Preston S, et al. A confidentialstudy of deaths after emergency medical admission:issues relating to quality of care. Clin Med 2003;3(5):425-434.8. Franklin C, Mathew J. Developing strategiesto prevent in-hospital cardiac arrest: analysingresponses of physicians and nurses in the hoursbefore the event. Crit Care Med 1994; 22(2):244-247).9. NICE CG 50 (2007): http://www.nice.org.uk/nicemedia/pdf/CG50FullGuidance.pdf10. Bellomo R, Goldsmith D, Uchino S, BuckmasterJ, Hart GK, Opdam H, Silvester W, Doolan L,Gutteridge G. A prospective before-and-after trialof a medical emergency team. Med J Aust 2003;179(6):283-287.11. Chan PS, Khalid A, Longmore LS, Berg RA,Kosiborod M, Spertus JA. Hospital wide code ratesand mortality before and after implementation of arapid response team. JAMA 2008;300:2506-13.12. Hillman K, Chen J, Cretikos M, et al. Introductionof the medical emergency team (MET) system:a cluster-randomised controlled trial. Lancet2005;365:2091-7.13. Harrison DA, Gao H, Welch CA, Rowan KM. Theeffects of critical care outreach services before andafter critical care: a matched-cohort analysis. J CritCare 2010;25:196-204.14. The fifth report from the Patient Safety Observatory.Safer care for the acutely ill patient: learning fromserious incidents. NPSA 200715. National Patient Safety Alert 02. Establishing astandard crash call telephone number in hospitals.National Patient Safety Agency; London:200416. Findlay GP, Stallard NJ. Standardising crash callnumbers. A minor issue compared with need tochange concept of cardiac arrest team. BMJ. 1996.Nov 30;313(7069):1404101

References17. 2010 Resuscitation Guidelines. ResuscitationCouncil (UK) 2010 htttp://www.resus.org.uk/pages/DNARrstd.htm18. http://www.dh.gov.uk/prod_consum_dh/groups/dh_digitalassets/@dh/@en/documents/digitalasset/dh_4012184.pdf19. Statistics on obesity, physical activity and diet:England, 2010 The Health and social careinformation centre. NHS 2010)20. McCabe WR, Jackson GG: Gram negativebacteremia: I. Etiology and ecology. Arch Intern Med1962, 110:845-847.21. Mahoney FI, Barthel D. “Functional evaluation: TheBarthel Index.” Maryland State Medical Journal1965;14:56-61.22. Academy of Medical Royal Colleges. A clinician’sguide to record standards – Part 2: standards forthe structure and content of medical records andcommunications when patients are admitted tohospital. London: NHS, 2009)23. The benefits of consultant delivered care. TheAcademy of Medical Royal Colleges 201224. The Intensive Care Society. Levels of critical care foradult patients. London: ICS, 2009.25. Acute medical care: The right person, in the rightsetting – first time. Report of the acute medicinetaskforce. Royal College of Physicians, 2007.26. Emergency surgery: standards for unscheduledsurgical care. Guidance for providers,commissioners and service planners. Royal Collegeof Surgeons of England 2011. Guidelines for theprovision of anaesthetic services. Royal College ofAnaesthetists 200927. Consultant physicians working with patients. Theduties, responsibilities and practice of physicians inmedicine. Royal College of Physicians 201128. Standards for consultant staffing of intensive careunits. Intensive Care Society 200629. Lee A, Bishop G, Hillman K, Daffum K. The MedicalEmergency Team. Anaesth Intensive Care 1995;23(2):183-18630. Gao H, McDonnell A, Harrison DA et al. Systematicreview and evaluation of physiological track andtrigger warning systems for identifying at riskpatients on the ward. Intensive Care Medicine2007;33:667-67931. Standardising the assessment of acute illnessseverity in the NHS: recommendations for a NHSearly warning score (NEWS). Report of a workingparty. Royal College of Physicians (working title –due for pub late 2011)32. Acute care toolkit2. High-quality acute care. RoyalCollege of Physicians 201133. Treatment and care towards the end of life: goodpractice in decision making. General MedicalCouncil 201034. Fritz Z, Fuld J, Haydock S, Palmer C. Interpretationand intent: a study of the (mis)understanding ofDNAR orders in a teaching hospital. Resuscitation.2010 Sep;81(9):1138-41.35. Decisions relating to cardiopulmonary resuscitationA joint statement from the British MedicalAssociation, the Resuscitation Council (UK) and theRoyal College of Nursing 200736. End of Life Care Strategy Promoting high qualitycare for all adults at the end of life July 2008.Department of Health www.dh.gov.uk/publications102

References37. Nolan JP, Soar J. Airway techniques and ventilationstrategies. Curr Opin Crit Care 2008;14:279-8638. Gatward JJ, Thomas MJ, Nolan JP, Cook TM.Effect of chest compressions on the time taken toinsert airway devices in a manikin. Br J Anaesth2008;100:351-639. National Cardiac Arrest Audithttps://www.icnarc.org40. Cardiac arrest and cardiopulmonary resuscitationoutcome reports: update and simplification ofthe Utstein templates for resuscitation registries.A statement for healthcare professionals from atask force of the international liaison committee onresuscitation Resuscitation. 2004 Dec;63(3):233-4941. Sandroni C et al. In-hospital cardiac arrest:incidence, prognosis and possible measuresto improve survival. Intensive Care Medicine2007;33:237-4542. Jennett, B. and M. Bond, Assessment of outcomeafter severe brain damage. Lancet, 1975. 1(7905): p.480-4.43. Bernard, S.A., et al., Treatment of comatosesurvivors of out-of-hospital cardiac arrest withinduced hypothermia. N Engl J Med, 2002. 346(8): p.557-63.44. Mild therapeutic hypothermia to improve theneurologic outcome after cardiac arrest. N Engl JMed, 2002. 346(8): p. 549-56.45. Roine, R.O., et al., Nimodipine after resuscitationfrom out-of-hospital ventricular fibrillation. Aplacebo-controlled, double-blind, randomized trial.Jama, 1990. 264(24): p. 3171-746. Hsu, J.W., C.D. Madsen, and M.L. Callaham,Quality-of-life and formal functional testing ofsurvivors of out-of-hospital cardiac arrest correlatespoorly with traditional neurologic outcome scales.Ann Emerg Med, 1996. 28(6): p. 597-605.47. Elliott VJ, Rodgers DL, Brett SJ. Systematicreview of quality of life and other patient-centredoutcomes after cardiac arrest survival. Resuscitation2011;82:247-25648. Survival from in-hospital cardiac arrest during nightsand weekends. Peberdy MA, Ornato JP, Larkin GL,Braithwaite RS, Kashner TM, Carey SM, Meaney,PA, Cen L, Nadkarni VM, Praestgaard AH, Berg RA;National Registry of Cardiopulmonary ResuscitationInvestigators. JAMA. 2008 Feb 20;299(7):785-92.49. Inside your hospital. Dr Foster Hospital Guide 2001-2011 http://drfosterintelligence.co.uk/wp-content/uploads/2011/11/Hospital_Guide_2011.pdf50. Gorjup V, Radsel P, Kocjancic ST, Erzen D, NocM. Acute ST-elevation myocardial infarctionafter successful cardiopulmonary resuscitation.Resuscitation. 2007;72:379-385.51. Garot P, Lefevre T, Eltchaninoff H, et al. Six-monthoutcome of emergency percutaneous coronaryintervention in resuscitated patients after cardiacarrest complicating ST-elevation myocardialinfarction. Circulation. 2007;115:1354-136252. Spaulding CM, Joly LM, Rosenberg A, et al.Immediate coronary angiography in survivorsof out-of-hospital cardiac arrest. N Engl J Med1997;336:1629-3353. Sunde K, Pytte M, Jacobsen D, et al. Implementationof a standardised treatment protocol for postresuscitation care after out-of-hospital cardiacarrest. Resuscitation 2007;73:29-39103

References54. Deakin CD, Nolan JP, Soar J, et al. EuropeanResuscitation Council Guidelines for Resuscitation2010. Section 4. Adult Advanced Life Support.Resuscitation 2010;81.55. Nolan JP, Laver SR, Welch CA, et al. Outcomefollowing admission to UK intensive care units aftercardiac arrest: a secondary analysis of the ICNARCCase Mix Programme Database. Anaesthesia 2007;62: 1207-16.56. Herlitz J, Engdahl J, Svensson L, et al. Majordifferences in 1-month survival between hospitalsin Sweden among initial survivors of out-of-hospitalcardiac arrest. Resuscitation 2006; 70: 404-9.57. Hypothermia After Cardiac Arrest Study Group. Mildtherapeutic hypothermia to improve the neurologicoutcome after cardiac arrest. N Engl J Med 2002;346: 549-56.58. Oddo M. Rossetti AO, Current Opinion in CriticalCare 2011;17:254-259. Standards for themanagement of patients after cardiac arrest. ICS200859. NICE Clinical Guideline 135 (2011) www.nice.org.uk/nicemedia/live/13628/57502/57502.pdf104

Back to contentsGlossaryGlossary2222 calls The most common emergency number called to alert medical staff to a patient havinga cardiac arrestADL (see also Appendix 4 -Barthel Index)Activities of Daily LivingALS trainedThe individual has completed the Resuscitation Council (UK) Advanced Life Support(ALS) course and holds a Resuscitation Council (UK) ALS Provider certificate,which is valid for 4 years The ALS course is a standardised national course teachingevidence-based resuscitation guidelines and skills to healthcare professionalsincluding the knowledge and skills required to:- Recognise and treat the deterioratingpatient using a structured ABCDE approach; Treat cardiac and/or respiratory arrest,including starting CPR, manual defibrillation, life threatening arrhythmias, and postresuscitation care; Care for the deteriorating patient or patient in cardiac and/orrespiratory arrest in special circumstances such as anaphylaxis, and pregnancy; Leada team, work as a team member, and use structured communication skills includinggiving an effective handoverAscitesIs excess fluid in the space between the tissues lining the abdomen andabdominal organsAsystoleA state of no electrical activity in the heartBasic life supportThis refers to maintaining airway patency and supporting breathing and the circulationBMIBody Mass IndexBPBlood PressureCardiac arrestCardiac arrest is the cessation of cardiac mechanical activity as confirmed by theabsence of signs of circulation. For the purposes of this study- receiving chestcompressions or defibrillationCCUCoronary Care UnitCPCCerebral Performance Category - This is a measure of functional outcome aftercardiac arrestCPRCardiopulmonary ResuscitationCTComputed TomographyDefibrillator - AEDAutomated External DefibrillatorAn automated external defibrillator is a defibrillator that analyses the heart rhythm,determines whether a shock is appropriate and provides audio prompts to theoperator. When prompted the operator pushes a button to deliver a shock tothe patient.Defibrillator - Manual Purely manual defibrillators do not incorporate rhythm analysis software - the operatormust interpret the rhythm, determine whether a shock is appropriate and, if so, chargethe defibrillator and deliver the shock105

GlossaryDefibrillator - Shock advisoryDNACPR/DNAREarly warning scoreECGEDGCSGPIVMonitored cardiac arrestNHSOutreach teamPCIPEAResuscitation teamSTEMITrack and triggerVFVTWitnessed cardiac arrestA shock advisory defibrillator can operate in either manual or AED modes - thepreferred mode is selected by the operatorDo Not Attempt Cardiopulmonary Resuscitation/Do Not Attempt ResuscitationA simple physiological scoring system that can be calculated at the patient’s bedside,using parameters which are measured in the majority of unwell patients. It is calculatedfor a patient using five simple physiological parameters: mental response, pulse rate,systolic blood pressure, respiratory rate and temperature. Points are allocated todeviations from the normal range in each parameter, and an overall score is thencalculatedElectrocardiographEmergency DepartmentGlasgow Coma ScaleGeneral PractitionerIntravenousThe patients physiological parameters are being monitored at the time of the arrestNational Health ServiceAlso known as a ’Medical Emergency Team’ or ‘Rapid Response Team’. Their purposeis to provide immediate care to patients on the ward who show signs of physiologicalinstability or clinical deterioration. They provide intervention to prevent, rather thantreat, cardiorespiratory arrestPercutanous Coronary InterventionPulseless electrical activityA team that is activated in response to a cardiopulmonary arrest. Ideally, the teamshould include at least two doctors with current training in advanced life support.The exact composition of the team will vary between institutions, but overall theteam must have the following skills: Airway interventions, includingtrachealintubation; Intravenous cannulation, including central venous access; Defibrillation(advisory and manual) and cardioversion; Drug administration; Ability to undertakeadvanced resuscitation skills (e.g. external cardiac pacing, Skills required for postresuscitationcareST elevation Myocardial Infarction occurs when a coronary artery becomestotally blockedTrack & Trigger system is used to calculate a patient’s physiological score, and adesignated trigger level is agreed; when this is reached, nursing staff alert aclinician. Other calling criteria, based upon routine observations, are activatedwhen one or more variables reaches an extreme value outside the normal range.Ventricular FibrillationVentricular TachycardiaA witnessed cardiac arrest is one that is seen or heard by another person106

Back to contentsAppendicesAppendicesAppendix 1 DNACPR - Example formReprinted with thanks: Resuscitation Council UK107

AppendicesBack to contentsAppendix 2Definitions of comorbiditiesChronic diseaseRespiratoryRenalImmuno- suppressionCardio- vascularLiver InsufficiencyDefinitionIncluding: Chronic pulmonary disease resulting in severe exercise restrictions(e.g. unable to perform household duties or climb stairs); Documented chronic hypoxia;Hypercapnia; Secondary polycythemia; Severe pulmonary hypertension (>40 mmHg);Ventilator dependency.Receiving chronic dialysis.The patient has received therapy that suppresses resistance to infection (e.g.immunosuppression, chemotherapy, radiation, long-term or recent high dose steroids.)or has a disease that is sufficiently advanced to suppress resistance to infection (e.g.Leukaemia, Lymphoma, AIDS).New York Heart Association Functional Classification - Class IV: Severe limitations.Experiences symptoms even while at rest.Including: Biopsy-proven cirrhosis; Documented portal hypertension; Episodes of pastupper GI bleeding attributed to portal hypertension; Prior episodes of hepatic failure/encephalopathy/coma.Appendix 3McCabe ClassificationCategory 1Category 2Category 3Non-fatal diseases (e.g. diabetes,genitourinary, gastrointestinal or obstetric conditions)Ultimately fatal (diseases estimated tobecome fatal within 4 years e.g. metastatic carcinomas,cirrhosis, chronic renal disease)Rapidly fatal (e.g. acute leukemia)108

AppendicesBack to contentsAppendix 4Barthel IndexThe Barthel Index consists of 10 items that measure aperson’s daily functioning specifically the activities ofdaily living (ADL) and mobility. The assessment can beused to determine a baseline level of functioning andcan be used to monitor improvement in activities of dailyliving over time. The items are weighted according to ascheme developed by the authors. The person receivesa score based on whether they have received help whiledoing the task. The scores for each of the items aresummed to create a total score. The higher the score themore “independent” the person. Independence means thatthe person needs no assistance at any part of the task.Bowels2 Continent (for preceding week)1 Occasional accident (once a week or less)0 Any worse grade of incontinenceGrooming1 Independent washing face, combing hair, shavingand cleaning teeth (when implements provided)0 Help neededBladder2 Continent (for preceding week) or able to manageany device (e.g. catheter and bag)without help1 Occasional accident (once a day or less), orcatheterised and needs help with device0 Any worse grade of incontinenceTransfer3 Needs no help2 Needs minor help, verbal or physical: can transferwith one person easily, or needs supervision1 Needs major help: two people or one strong/trainedperson, but can sit unaided0 Cannot sit: needs skilled lift by two people (or hoist)Not able to assessToilet Use2 Able to get on and off toilet or commode, undressand dress sufficiently, and wipe self without physicalor verbal help1 Needs some help, can wipe self and do some of therest with minimal help only0 Needs more help than thisDressing2 Independent putting on all clothes, includingfasteners, zips, etc. (clothes may be adapted)1 Needs some help but can do at least half0 Needs more help than thisMobility3 May use aid (stick or frame, etc., not wheelchair)2 Needs help of one person, verbal or physical,including help standing up1 Independent in wheelchair, including able tonegotiate doors and corners unaided0 Needs more help than thisStairs2 Independent up and down, and can carry anynecessary walking aid1 Needs help verbal or physical, or help carrying aid0 UnableBathing1 Able to get in and out of bath or shower, wash selfwithout help (may use aids)0 UnableMahoney Fl, Barthel DW:Functional evaluation: the Barthel Index.Md State Med J 14:2, 1965109

AppendicesBack to contentsAppendix 5Grade of clinicians/health care professionalsAppendix 6Levels of careConsultantStaff Grade or Associate SpecialistTrainee with CCTSenior Specialist Trainee (SpR3+ or ST5)Junior Specialist Trainee (SpR1, SpR2 , ST3 and ST4)Basic Grade (ST1, ST2, FY1, FY2 or CTs)Specialist Nurse PractitionerOther Registered NurseResuscitation OfficerLevel 1 care - ward careLevel 2 care - high dependency unit (HDU)A specialist unit in a hospital, where patients requiringa high level of specialist intervention are cared for. Highdependency unit care is appropriate for: patients needingsupport for a single failing organ, but excluding thoseneeding advanced respiratory support; patients who canbenefit from more detailed observation than can be safelyprovided on a general ward; patients no longer needingintensive care, but not yet well enough to be returned toa general ward; or postoperative patients who need closemonitoring for longer than a few hours, i.e. the periodnormally spent in a recovery area.Level 3 care - An intensive care unit (ICU/ITU)A specialist area to which patients are admitted fortreatment of actual or impending organ failure, especiallywhen mechanical ventilation is necessary.110

AppendicesBack to contentsAppendix 7Ceilings of care - an example documentReprinted with thanks: Gloucester Hospitals NHS Foundation Trust111

AppendicesBack to contentsAppendix 8The role and structure of NCEPODThe National Confidential Enquiry into Patient Outcomeand Death (NCEPOD) is an independent body to whicha corporate commitment has been made by the Medicaland Surgical Colleges, Associations and Facultiesrelated to its area of activity. Each of these bodiesnominates members on to NCEPOD’s Steering Group.Steering Group as at 1st June 2012Dr I WilsonMr F SmithDr C MannDr S BridgmanProfessor R MahajanDr A BatchelorVacancyMs M McElligottDr E MorrisMrs M WishartDr I DoughtyDr R DowdleProfessor T HendraDr S McPhersonMr R LamontMr M BircherMr D MitchellDr M OsbornMs S PanizzoMrs M WangAssociation of Anaesthetists ofGreat Britain and IrelandAssociation of Surgeons of GreatBritain & IrelandCollege of Emergency MedicineFaculty of Public Health MedicineRoyal College of AnaesthetistsRoyal College of AnaesthetistsRoyal College of GeneralPractitionersRoyal College of NursingRoyal College of Obstetriciansand GynaecologistsRoyal College ofOphthalmologistsRoyal College of Paediatrics andChild HealthRoyal College of PhysiciansRoyal College of PhysiciansRoyal College of RadiologistsRoyal College of Surgeons ofEnglandRoyal College of Surgeons ofEnglandFaculty of Dental Surgery, RoyalCollege of Surgeons of EnglandRoyal College of PathologistsPatient RepresentativePatient RepresentativeObserversMrs J MooneyDr R HunterHealthcare Quality inPartnership (HQIP)Coroners’ Society of England andWales112

AppendicesNCEPOD is a company, limited by guarantee (Companynumber: 3019382) and a registered charity (Charitynumber: 1075588), managed by Trustees.TrusteesMr B Leigh - ChairmanDr D Justins - Honorary TreasurerProfessor M BrittonProfessor J H ShepherdProfessor L ReganProfessor R EndacottCompany Secretary - Dr M MasonSupporting organisationsThe organisations that provided funding to cover the costof this study:Heathcare Quality in Partnership on behalf of theDepartment of Health in England, the Welsh AssemblyGovernment and the Department of Health, SocialServices and Public Safety (Northern Ireland) also theStates of Guernsey Board of Health and States ofJersey, Health and Social Services. Funding is for theClinical Outcome Review Programme into Medicaland Surgical Care.Clinical Co-ordinatorsThe Steering Group appoint a Lead Clinical Co-ordinatorfor a defined tenure. In addition there are seven ClinicalCo-ordinators who work on each study. All Co-ordinatorsare engaged in active academic/clinical practice (in theNHS) during their term of office.Lead ClinicalCo-ordinatorClinicalCo-ordinatorsDr G P Findlay (Intensive Care)Dr M Juniper (Medicine)Dr K Wilkinson (Anaesthesia)Dr A P L Goodwin (Anaesthesia)Mr I C Martin (Surgery)Professor M J Gough (Surgery)Professor S B Lucas (Pathology)Aspen Healthcare LtdBMI HealthcareBUPA CromwellClassic HospitalsEast Kent Medical Services LtdFairfield Independent HospitalHCA InternationalHospital of St John and St ElizabethIsle of Man Health and Social Security DepartmentKing Edward VII’s Hospital Sister AgnesNew Victoria HospitalNuffield HealthRamsay Health Care UKSpire Health CareSt Anthony’s HospitalSt Joseph’s HospitalThe Horder CentreThe Hospital Management TrustThe London ClinicUlster Independent Clinic113

AppendicesBack to contentsAppendix 9Participation114Valid reason for non-return ofcase notesCase notes receivedValid reason for non-return ofquestionnaireQuestionnaires returnedNumber of cases identified for inclusionNumber of resuscitation forms receivedNumber of hospitals that returnedspreadsheet data or informed of null-returnNumber of hospitals returning otherorganisational data*Number of completed organisationalquestionnaires receivedNumber of hospitalsTrust NameAbertawe Bro Morgannwg University Health Board 11 9 0 8 11 11 7 1 7 0Aintree Hospitals NHS Foundation Trust 1 1 0 1 0 3 2 1 2 0Airedale NHS Trust 1 0 0 0 2 2 2 0 1 0Aneurin Bevan Local Health Board 9 8 1 7 11 13 5 4 5 3Anglian Community Enterprise 2 0 0 2 0 0 0 0 0 0Ashford & St Peter’s Hospital NHS Trust 2 2 0 1 6 5 4 0 5 0Aspen Healthcare 3 2 0 2 0 0 0 0 0 0Barking, Havering & Redbridge University HospitalsNHS Trust 2 2 0 0 3 0 0 0 0 0Barnet and Chase Farm Hospitals NHS Trust 2 2 0 0 7 7 4 0 2 0Barnsley Hospital NHS Foundation Trust 1 1 0 0 2 2 1 1 1 0Barts and The London NHS Trust 4 4 0 3 3 2 2 1 2 0Basildon & Thurrock University HospitalsNHS FoundationTrust 1 1 0 0 7 8 6 0 8 0Bedford Hospital NHS Trust 1 1 0 0 1 1 1 0 1 0Belfast Health and Social Care Trust 4 4 0 1 5 8 8 1 8 0Benenden Hospital 1 1 0 0 1 1 1 0 1 0Berkshire Healthcare NHS Foundation Trust 3 0 0 0 0 0 0 0 0 0Betsi Cadwaladr University Local Health Board 17 4 0 15 12 6 3 2 4 0Birmingham Community Healthcare NHS Trust 2 0 0 0 1 0 0 0 0 0Birmingham Women’s Healthcare NHS Trust 1 1 0 1 0 0 0 0 0 0Blackpool, Fylde and Wyre HospitalsNHS Foundation Trust 2 2 0 1 2 4 4 0 4 0BMI Healthcare 52 36 3 46 2 2 0 0 0 0

AppendicesBradford District Care Trust 5 0 0 2 0 0 0 0 0 0Bradford Teaching Hospitals NHS Foundation Trust 2 1 1 1 3 3 2 1 2 1Brighton and Sussex University Hospitals NHS Trust 3 3 0 2 4 4 3 1 4 0Buckinghamshire Healthcare NHS Trust 9 2 7 0 5 4 4 0 4 0BUPA Cromwell Hospital 1 1 0 1 1 1 0 0 0 0Burton Hospitals NHS Foundation Trust 1 1 0 0 5 5 2 0 5 0Calderdale & Huddersfield NHS Foundation Trust 2 2 0 1 4 2 2 0 2 0Cambridge University Hospitals NHS Foundation Trust 1 1 0 0 3 3 3 0 3 0Cambridgeshire Community Services NHS Trust 4 0 4 4 0 0 0 0 0 0Cardiff and Vale University Health Board 7 7 0 5 8 8 5 1 7 1Care UK 4 3 0 2 0 0 0 0 0 0Central and North West London NHS Foundation Trust 1 0 0 1 0 0 0 0 0 0Central Lancashire Primary Care Trust 1 0 0 1 0 0 0 0 0 0Central London Community Healthcare NHS Trust 1 0 0 0 0 0 0 0 0 0Central Manchester University HospitalsNHS Foundation Trust 4 4 0 3 5 5 2 2 2 3Chelsea & Westminster Healthcare NHS Trust 1 0 0 0 2 2 0 0 0 0Chesterfield Royal Hospital NHS Foundation Trust 1 1 0 0 3 3 3 0 3 0Circle Health Limited 1 1 0 1 0 0 0 0 0 0City Hospitals Sunderland NHS Foundation Trust 2 2 0 3 8 8 8 0 8 0Clatterbridge Centre for Oncology NHS Trust 1 1 0 1 1 1 1 0 1 0Colchester Hospital University NHS Foundation Trust 2 0 0 1 4 3 0 0 0 0Countess of Chester Hospital NHS Foundation Trust 2 1 0 0 2 2 2 0 2 0County Durham and Darlington NHS Foundation Trust 8 6 0 5 4 4 3 1 2 0Croydon Health Services NHS Trust 1 1 0 0 2 1 0 0 1 0Cumbria Partnership NHS Foundation Trust 9 0 9 9 0 0 0 0 0 0Cwm Taf Local Health Board 5 5 0 3 4 3 3 0 3 0Dartford & Gravesham NHS Trust 1 0 0 0 5 4 3 0 2 0Derby Hospitals NHS Foundation Trust 1 1 0 1 9 9 9 0 9 0Doncaster and Bassetlaw HospitalsNHS Foundation Trust 5 3 0 3 5 5 5 0 3 0Dorset County Hospital NHS Foundation Trust 1 1 0 1 1 1 1 0 1 0Dorset Healthcare University NHS Foundation Trust 11 0 0 0 0 0 0 0 0 0Ealing Hospital NHS Trust 1 1 0 0 3 2 2 0 2 0Ealing Primary Care Trust 2 0 0 1 0 0 0 0 0 0East & North Hertfordshire NHS Trust 3 2 0 0 5 5 4 1 4 1115

Appendices116Valid reason for non-return ofcase notesCase notes receivedValid reason for non-return ofquestionnaireQuestionnaires returnedNumber of cases identified for inclusionNumber of resuscitation forms receivedNumber of hospitals that returnedspreadsheet data or informed of null-returnNumber of hospitals returning otherorganisational data*Number of completed organisationalquestionnaires receivedNumber of hospitalsTrust NameEast Cheshire NHS Trust 2 0 0 3 7 7 6 1 7 0East Kent Hospitals University NHS Foundation Trust 5 3 2 2 14 14 9 1 3 0East Kent Medical Services 1 1 0 1 0 0 0 0 0 0East Lancashire Hospitals NHS Trust 3 3 0 2 1 2 2 0 2 0East Riding of Yorkshire Primary Care Trust 4 0 0 0 0 0 0 0 0 0East Sussex Healthcare NHS Trust 6 5 0 0 4 4 4 0 4 0Enfield Primary Care Trust 1 0 0 1 0 0 0 0 0 0Epsom and St Helier University Hospitals NHS Trust 5 5 0 6 6 6 4 0 6 0Fairfield Independent Hospital 1 1 0 1 0 0 0 0 0 0Frimley Park Hospitals NHS Trust 1 1 0 0 2 2 2 0 2 0Gateshead Health NHS Foundation Trust 2 2 0 1 4 4 4 0 3 0George Eliot Hospital NHS Trust 1 0 0 1 0 0 0 0 0 0Gloucestershire Hospitals NHS Foundation Trust 2 2 0 0 8 10 7 0 3 0Great Western Hospitals NHS Foundation Trust 1 1 0 1 2 2 2 0 2 0Guy’s & St Thomas’ NHS Foundation Trust 3 2 0 0 0 0 0 0 0 0Hampshire Hospitals NHS Foundation Trust 2 2 0 2 2 2 2 0 2 0Harrogate and District NHS Foundation Trust 3 2 0 0 4 4 4 0 3 0HCA International 7 2 0 6 0 0 0 0 0 0Health & Social Services, States of Guernsey 2 2 0 1 3 3 1 0 1 0Health & Surgical Holdings 1 1 0 0 0 0 0 0 0 0Heart of England NHS Foundation Trust 3 3 0 0 0 14 4 2 0 0Heatherwood & Wexham Park HospitalsNHS Foundation Trust 4 0 0 3 6 7 3 4 3 4Herefordshire Primary Care Trust 5 0 0 0 1 0 0 0 0 0

AppendicesHertfordshire Community NHS Trust 5 0 0 0 0 0 0 0 0 0Hillingdon Hospital NHS Trust 1 1 0 0 5 5 4 0 1 0Hinchingbrooke Health Care NHS Trust 1 1 0 0 1 1 0 1 1 0Homerton University Hospital NHS Foundation Trust 1 0 0 1 3 3 1 2 1 2Hospital of St John and St Elizabeth 1 1 0 1 0 0 0 0 0 0Hull and East Yorkshire Hospitals NHS Trust 2 2 0 0 7 7 7 0 7 0Hywel Dda Local Health Board 10 9 1 0 6 6 6 0 6 0Imperial College Healthcare NHS Trust 5 5 0 2 11 10 8 0 8 2Ipswich Hospital NHS Trust 1 1 0 0 5 5 5 0 5 0Isle of Man Department of Health & Social Security 2 1 1 2 0 0 0 0 0 0Isle of Wight NHS Primary Care Trust 1 1 0 0 4 0 0 0 0 0James Paget Healthcare NHS Trust 2 2 0 1 2 2 2 0 2 0Kent Community Health NHS Trust 12 0 6 12 0 0 0 0 0 0Kettering General Hospital NHS Trust 1 1 0 0 4 7 7 0 7 0King Edward VII’s Hospital Sister Agnes 1 1 0 1 0 0 0 0 0 0King’s College Hospital NHS Foundation Trust 1 1 0 0 12 11 8 1 11 0Kingston Hospital NHS Trust 1 1 0 0 4 4 3 1 3 1Lancashire Teaching Hospitals NHS Foundation Trust 2 0 0 2 2 8 1 0 1 0Leeds Teaching Hospitals NHS Trust (The) 6 6 0 5 13 11 5 1 4 1Lewisham Hospital NHS Trust 1 1 0 1 5 3 3 0 3 0Liverpool Community Health 1 0 1 1 0 0 0 0 0 0Liverpool Heart and Chest Hospital NHS Trust 1 1 0 0 3 3 3 0 3 0Liverpool Women’s NHS Foundation Trust 1 0 0 0 0 0 0 0 0 0London Clinic 1 1 0 1 0 0 0 0 0 0Luton and Dunstable Hospital NHS Foundation Trust 1 1 0 1 2 2 2 0 2 0Maidstone and Tunbridge Wells NHS Trust 2 1 0 3 11 10 9 0 10 0Medway NHS Foundation Trust 1 1 0 0 6 6 6 0 6 0Mid Cheshire Hospitals NHS Foundation Trust 1 1 0 1 4 4 4 0 3 0Mid Staffordshire NHS Foundation Trust 2 2 0 2 2 2 2 0 2 0Mid Yorkshire Hospitals NHS Trust 5 5 0 7 18 0 0 0 0 0Mid-Essex Hospital Services NHS Trust 1 1 0 1 0 0 0 0 0 0Milton Keynes Hospital NHS Foundation Trust 1 1 0 1 5 0 0 0 0 0Milton Keynes Primary Care Trust 1 0 0 0 0 0 0 0 0 0Moorfields Eye Hospital NHS Foundation Trust 1 1 0 1 0 0 0 0 0 0New Victoria Hospital 1 1 0 1 0 0 0 0 0 0117

Appendices118Valid reason for non-return ofcase notesCase notes receivedValid reason for non-return ofquestionnaireQuestionnaires returnedNumber of cases identified for inclusionNumber of resuscitation forms receivedNumber of hospitals that returnedspreadsheet data or informed of null-returnNumber of hospitals returning otherorganisational data*Number of completed organisationalquestionnaires receivedNumber of hospitalsTrust NameNewcastle upon Tyne Hospitals NHS Foundation Trust 3 3 0 1 8 6 6 0 6 0Newham University Hospital NHS Trust 1 1 0 0 6 6 1 0 1 0NHS Barnet 2 0 0 2 0 0 0 0 0 0NHS Bath & North East Somerset 2 0 0 0 0 0 0 0 0 0NHS Bedfordshire 1 0 0 0 0 0 0 0 0 0NHS Camden 1 0 0 0 0 0 0 0 0 0NHS Cornwall and Isles of Scilly 16 0 0 0 0 0 0 0 0 0NHS Derbyshire County 12 0 0 0 0 0 0 0 0 0NHS Gloucestershire - Gloucestershire Care Services 11 0 11 8 0 0 0 0 0 0NHS Great Yarmouth and Waveney 4 0 4 4 0 0 0 0 0 0NHS Kingston 2 0 2 2 0 0 0 0 0 0NHS Kirklees 1 0 0 0 0 0 0 0 0 0NHS Leicestershire County and Rutland 4 0 0 0 0 0 0 0 0 0NHS Lincolnshire 4 0 1 3 0 0 0 0 0 0NHS Manchester 1 0 1 1 0 0 0 0 0 0NHS Mid Essex 3 0 0 0 0 0 0 0 0 0NHS Norfolk - Norfolk Community Health & Care 1 0 0 0 0 0 0 0 0 0NHS North Yorkshire and York 9 0 0 9 0 0 0 0 0 0NHS Nottinghamshire County 3 0 0 0 0 0 0 0 0 0NHS Richmond 1 0 0 1 0 0 0 0 0 0NHS South West Essex 4 0 1 0 0 0 0 0 0 0NHS South Yorkshire and Bassetlaw 3 0 0 0 0 0 0 0 0 0NHS Surrey 12 0 0 0 1 0 0 0 0 0NHS West Essex 4 0 4 0 0 0 0 0 0 0

AppendicesNHS West Sussex 5 0 0 0 0 0 0 0 0 0NHS Wiltshire 3 0 0 3 0 0 0 0 0 0NHS Worcestershire 5 0 0 1 0 0 0 0 0 0Norfolk & Norwich University Hospital NHS Trust 2 0 0 3 4 4 3 0 2 0North Bristol NHS Trust 2 2 0 2 4 4 4 0 4 0North Cumbria University Hospitals NHS Trust 2 2 0 0 8 0 0 0 0 0North Middlesex University Hospital NHS Trust 1 1 0 1 0 0 0 0 0 0North Tees and Hartlepool NHS Foundation Trust 2 2 0 0 5 5 5 0 5 0North West London Hospitals NHS Trust 2 2 0 2 8 4 4 0 4 0Northampton General Hospital NHS Trust 1 1 0 0 5 5 5 0 5 0Northamptonshire Teaching Primary Care Trust 4 0 0 0 0 0 0 0 0 0Northern Devon Healthcare NHS Trust 18 1 17 16 3 3 3 0 3 0Northern Health & Social CareTrust 4 2 2 0 7 7 5 2 4 0Northern Lincolnshire & Goole HospitalsNHS Foundation Trust 3 2 1 0 4 6 4 0 2 0Northumbria Healthcare NHS Foundation Trust 9 9 0 1 3 2 2 0 2 0Nottingham University Hospitals NHS Trust 2 2 0 0 8 7 7 0 7 0Nucleus Healthcare 1 1 0 1 0 0 0 0 0 0Nuffield Health 29 22 0 20 1 0 0 0 0 0Outer North East London Community Services 1 0 0 0 0 0 0 0 0 0Oxford University Hospitals NHS Trust 6 1 0 4 10 10 8 1 9 0Oxfordshire Primary Care Trust 7 0 0 1 0 0 0 0 0 0Papworth Hospital NHS Foundation Trust 1 1 0 1 3 2 2 0 2 0Pennine Acute Hospitals NHS Trust (The) 4 4 0 2 11 10 7 0 7 0Peterborough & Stamford HospitalsNHS Foundation Trust 2 2 0 2 3 3 1 2 0 3Plymouth Hospitals NHS Trust 2 2 0 0 8 0 0 0 0 0Poole Hospital NHS Foundation Trust 1 1 0 0 2 2 2 0 0 0Portsmouth Hospitals NHS Trust 1 1 0 0 6 6 5 1 6 0Powys Teaching Local Health Board 10 0 10 10 0 0 0 0 0 0Princess Alexandra Hospital NHS Trust 3 1 2 0 2 2 2 0 2 0Queen Victoria Hospital NHS Foundation Trust 1 1 0 1 1 1 1 0 1 0Ramsay Health Care UK 25 17 0 23 1 1 1 0 1 0Robert Jones and Agnes Hunt OrthopaedicHospital NHS Foundation Trust 1 1 0 1 0 0 0 0 0 0Royal Berkshire NHS Foundation Trust 1 1 0 0 3 3 3 0 3 0119

Appendices120Valid reason for non-return ofcase notesCase notes receivedValid reason for non-return ofquestionnaireQuestionnaires returnedNumber of cases identified for inclusionNumber of resuscitation forms receivedNumber of hospitals that returnedspreadsheet data or informed of null-returnNumber of hospitals returning otherorganisational data*Number of completed organisationalquestionnaires receivedNumber of hospitalsTrust NameRoyal Bolton Hospital NHS Foundation Trust 1 1 0 1 2 2 2 0 2 0Royal Bournemouth and Christchurch HospitalsNHS Trust 2 2 0 1 6 6 6 0 6 0Royal Brompton and Harefield NHS Foundation Trust 2 2 0 2 3 3 3 0 3 0Royal Cornwall Hospitals NHS Trust 3 3 0 5 4 4 4 0 4 0Royal Devon and Exeter NHS Foundation Trust 1 1 0 0 3 3 3 0 3 0Royal Free Hampstead NHS Trust 2 2 0 3 9 9 8 1 8 1Royal Liverpool & Broadgreen University HospitalsNHS Trust 1 0 0 0 3 3 3 0 3 0Royal Marsden NHS Foundation Trust (The) 2 2 0 0 1 0 0 0 0 0Royal National Hospital for Rheumatic DiseasesNHS Foundation Trust 1 0 0 1 0 0 0 0 0 0Royal National Orthopaedic Hospital NHS Trust 1 1 0 1 0 0 0 0 0 0Royal Orthopaedic Hospital NHS Foundation Trust 1 1 0 1 0 0 0 0 0 0Royal Surrey County Hospital NHS Trust 1 1 0 0 3 2 2 0 2 0Royal United Hospital Bath NHS Trust 1 1 0 1 4 4 3 1 3 1Royal Wolverhampton Hospitals NHS Trust (The) 1 0 0 0 6 0 0 0 0 0Salford Royal Hospitals NHS Foundation Trust 1 1 0 0 0 1 1 0 1 0Salisbury NHS FoundationTrust 1 1 0 0 2 2 2 0 2 0Sandwell and West Birmingham Hospitals NHS Trust 3 3 0 5 3 3 3 0 3 0Scarborough and North East Yorkshire Health CareNHS Trust 2 2 0 1 6 6 5 0 3 0

AppendicesSEQOL(Care and Support Ptnrship CommunityInterest Company 1 0 1 0 0 0 0 0 0 0Sheffield Teaching Hospitals NHS Foundation Trust 5 5 0 3 10 9 9 0 9 0Sherwood Forest Hospitals NHS Foundation Trust 2 2 0 2 4 3 3 0 3 0Shrewsbury and Telford Hospitals NHS Trust 2 2 0 0 14 14 12 2 14 0Shropshire County Primary Care Trust 4 0 0 0 0 0 0 0 0 0South Devon Healthcare NHS Foundation Trust 1 1 0 1 1 1 1 0 1 0South Eastern Health & Social Care Trust 4 4 0 7 5 5 4 1 4 0South London Healthcare NHS Trust 4 2 2 0 16 14 9 1 9 0South Staffordshire Primary Care Trust 3 0 0 0 0 0 0 0 0 0South Tees Hospitals NHS Foundation Trust 7 2 0 2 0 5 5 0 5 0South Tyneside NHS Foundation Trust 3 1 2 2 6 6 6 0 3 0South Warwickshire NHS Foundation Trust 6 1 1 1 5 5 5 0 5 0Southampton University Hospitals NHS Trust 4 4 0 0 2 2 1 0 0 0Southend University Hospital NHS Foundation Trust 1 1 0 1 6 6 4 0 3 0Southern Health & Social Care Trust 7 3 1 0 4 0 0 0 0 0Southern Health NHS Foundation Trust 11 1 10 11 0 0 0 0 0 0Southport and Ormskirk Hospitals NHS Trust 2 2 0 0 4 5 4 1 4 1Spire Healthcare 33 29 0 32 0 0 0 0 0 0St Anthony’s Hospital 1 1 0 0 1 1 0 0 0 0St George’s Healthcare NHS Trust 1 1 0 0 6 8 5 0 4 0St Helens and Knowsley Teaching Hospitals NHS Trust 2 2 0 3 5 5 5 0 5 0St Joseph’s Hospital 1 1 0 1 0 0 0 0 0 0Staffordshire & Stoke on Trent Partnership NHS Trust 2 0 0 0 0 0 0 0 0 0States of Jersey Health & Social Services 1 0 0 1 0 0 0 0 0 0Stockport NHS Foundation Trust 1 1 0 0 2 3 3 0 1 0Suffolk Primary Care Trust 5 0 5 5 0 0 0 0 0 0Surrey & Sussex Healthcare NHS Trust 1 0 0 0 9 8 6 0 5 0Tameside & Glossop Primary Care Trust 1 0 1 1 0 0 0 0 0 0Tameside Hospital NHS Foundation Trust 1 1 0 0 4 4 3 0 3 0Taunton & Somerset NHS Foundation Trust 1 1 0 0 5 5 5 0 5 0The Christie NHS Foundation Trust 1 0 0 1 0 0 0 0 0 0The Dudley Group NHS Foundation Trust 1 1 0 0 3 3 3 0 3 0The Horder Centre 1 1 0 1 0 0 0 0 0 0The Hospital Management Trust 3 3 0 2 1 1 0 0 0 0The Queen Elizabeth Hospital King’s Lynn NHS Trust 1 1 0 0 5 5 4 1 4 0121

Appendices122Valid reason for non-return ofcase notesCase notes receivedValid reason for non-return ofquestionnaireQuestionnaires returnedNumber of cases identified for inclusionNumber of resuscitation forms receivedNumber of hospitals that returnedspreadsheet data or informed of null-returnNumber of hospitals returning otherorganisational data*Number of completed organisationalquestionnaires receivedNumber of hospitalsTrust NameThe Rotherham NHS Foundation Trust 1 1 0 1 1 1 1 0 1 0The Walton Centre NHS Foundation Trust 1 1 0 1 0 0 0 0 0 0Torbay Care Trust 11 0 0 9 0 0 0 0 0 0Trafford Healthcare NHS Trust 3 0 0 3 2 2 2 0 2 0UK Specialist Hospitals Ltd 1 0 0 0 0 0 0 0 0 0Ulster Independent Clinic 1 0 1 1 0 0 0 0 0 0United Lincolnshire Hospitals NHS Trust 4 4 0 1 10 10 10 0 10 0Univ. Hospital of South ManchesterNHS Foundation Trust 1 1 0 1 3 3 3 0 3 0University College London HospitalsNHS Foundation Trust 5 3 2 8 3 4 3 0 3 0University Hospital Birmingham NHS Foundation Trust 2 2 0 2 4 5 5 0 5 0University Hospitals Coventry and WarwickshireNHS Trust 2 2 0 3 15 16 16 0 16 0University Hospitals of Bristol NHS Foundation Trust 6 6 0 5 4 4 4 0 4 0University Hospitals of Leicester NHS Trust 3 3 0 0 6 6 6 0 6 0University Hospitals of Morecambe Bay NHS Trust 3 3 0 1 2 5 4 1 4 1Velindre NHS Trust 1 1 0 1 0 0 0 0 0 0Walsall Hospitals NHS Trust 1 0 0 0 4 0 0 0 0 0Warrington & Halton Hospitals NHS Foundation Trust 2 2 0 0 2 0 0 0 0 0Wells Hospital and Hospice Charitable Trust 1 0 0 0 0 0 0 0 0 0West Hertfordshire Hospitals NHS Trust 2 2 0 0 0 0 0 0 0 0

AppendicesWest Middlesex University Hospital NHS Trust 1 1 0 0 5 9 4 0 3 1West Suffolk NHS Foundation Trust 1 1 0 0 2 2 2 0 2 0Western Health & Social Care Trust 4 4 0 4 4 4 4 1 2 0Western Sussex Hospitals NHS Trust 3 3 0 1 14 13 12 0 12 0Weston Area Health Trust 1 1 0 0 2 3 2 1 2 1Whipps Cross University Hospital NHS Trust 1 1 0 0 7 7 3 0 1 0Whittington Health 1 1 0 1 1 1 1 0 1 0Wirral University Teaching HospitalNHS Foundation Trust 4 2 2 0 0 0 0 0 0 0Worcestershire Acute Hospitals 3 3 0 1 12 11 9 0 8 0Wrightington, Wigan & Leigh NHS Foundation Trust 3 3 0 5 7 7 4 3 7 0Wye Valley NHS Trust 1 1 0 1 0 1 1 0 1 0Yeovil District Hospital NHS Foundation Trust 1 1 0 0 1 1 1 0 1 0York Teaching Hospital NHS Foundation Trust 1 1 0 1 4 4 4 0 4 0123

Published June 2012by the National Confidential Enquiryinto Patient Outcome and Death5th Floor125 Wood StreetLondonEC2V 7ANT 0207 600 1893F 0207 600 9013E info@ncepod.org.ukw www.ncepod.org.ukISBN 978-0-9560882-7-7A company limited by guarantee Company no. 3019382Registered charity no. 1075588