Decision-makingTowards evidence-informedpolicy-making in human resourcesfor <strong>health</strong>: <strong>the</strong> state of researchArticle by Manuel M Dayrit (pictured), Mario Rober<strong>to</strong> Dal Poz,Hugo Mercer <strong>and</strong> Carmen DoleaHealth systems in many countries are failing <strong>to</strong> respond<strong>to</strong> <strong>the</strong> <strong>health</strong> needs of <strong>the</strong>ir population. Part of <strong>the</strong>reason for this is <strong>the</strong> current crisis in <strong>the</strong> <strong>health</strong>workforce, which is expressed in severe shortages,imbalances <strong>and</strong> a poor knowledge base on effectiveness ofinterventions. In order <strong>to</strong> make <strong>health</strong> systems moreresponsive <strong>to</strong> current <strong>and</strong> emerging <strong>health</strong> needs, countriesmust streng<strong>the</strong>n <strong>the</strong>ir <strong>health</strong> workforce. Research can helppolicy-makers pose <strong>and</strong> find answers <strong>to</strong> <strong>the</strong> critical questionsregarding <strong>the</strong> status of <strong>the</strong>ir workforce, its level ofperformance <strong>and</strong> <strong>the</strong> problems <strong>health</strong> workers face.Need for more evaluation studies <strong>to</strong> knowwhat works <strong>and</strong> what does notResearch is required – not only <strong>to</strong> give policy-makers a betterunderst<strong>and</strong>ing of <strong>the</strong> present situation of <strong>the</strong>ir country’s<strong>health</strong> workforce, but also <strong>to</strong> discover which policyinterventions work or do not work. What can we say about<strong>the</strong> status of research in human resources for <strong>health</strong>?First, <strong>the</strong> knowledge base in human resources for <strong>health</strong>development is weak <strong>and</strong> uneven overall, compared <strong>to</strong> o<strong>the</strong>rdomains of <strong>health</strong> systems research, such as <strong>health</strong> financingor <strong>health</strong> sec<strong>to</strong>r reform, even though efforts <strong>to</strong> identifypriorities in <strong>the</strong> <strong>health</strong> workforce research agenda havealready started 1,3 . Given that close <strong>to</strong> half of <strong>health</strong>expenditure is spent on <strong>the</strong> <strong>health</strong> workforce, it seemsincredible that <strong>the</strong>re is so little research investment or solidevidence in this area. Moreover, <strong>the</strong> existing knowledge baseis largely skewed <strong>to</strong>wards high-income countries, medicaldoc<strong>to</strong>rs <strong>and</strong> descriptive reports, as opposed <strong>to</strong> interventionstudies or best practice assessments 1,2 .An examination of <strong>the</strong> Cochrane Collaboration systematicreviews identified only 12 reviews on <strong>to</strong>pics related <strong>to</strong> humanresources for <strong>health</strong>, most of <strong>the</strong>m in <strong>the</strong> domain of <strong>health</strong>workforce management 4 . A more detailed analysis points outthat not only is <strong>the</strong>re a very limited number of systematicTitle of Cochranesystematic reviewResearch questionNumber of studies(initial/final)Total numberof subjectsResultsSubstitution of doc<strong>to</strong>rsby nurses in primary careTo investigate <strong>the</strong> impact ofnurses working assubstitutes for primary caredoc<strong>to</strong>rs on:✜ <strong>health</strong> outcomes✜ process of care✜ resource use✜ costs4253 articles initially25 articles, relating <strong>to</strong> 16studies, met inclusioncriterian/a✜ No difference in quality of care <strong>and</strong> <strong>health</strong> outcomesbetween appropriately trained nurses <strong>and</strong> doc<strong>to</strong>rs.✜ Nurses tend <strong>to</strong> provide more <strong>health</strong> advice <strong>and</strong> achievehigher levels of patient satisfaction, compared withdoc<strong>to</strong>rs.✜ Even though using nurses may save salary costs, nursesmay order more tests <strong>and</strong> use o<strong>the</strong>r services, which maydecrease <strong>the</strong> cost savings of using nurses instead ofdoc<strong>to</strong>rs.Lay <strong>health</strong> workers (LHW)in primary <strong>and</strong>community <strong>health</strong> careTo assess <strong>the</strong> effects ofLHW interventions inprimary <strong>and</strong> community<strong>health</strong> care on <strong>health</strong> carebehaviours, patients’<strong>health</strong> <strong>and</strong> well-being, <strong>and</strong>patients’ satisfaction withcare8637 abstracts initially400 potentially eligible43 eventually included210 110consumers✜ LHWs show promising benefits in promotingimmunization uptake <strong>and</strong> improving outcomes for acuterespira<strong>to</strong>ry infections <strong>and</strong> malaria, when compared <strong>to</strong>usual care.✜ For o<strong>the</strong>r <strong>health</strong> issues, evidence is insufficient <strong>to</strong> justifyrecommendations for policy <strong>and</strong> practice.✜ There is also insufficient evidence <strong>to</strong> assess which LHWtraining or intervention strategies are likely <strong>to</strong> be mosteffective.Audit <strong>and</strong> feedback:effects on professionalpractice <strong>and</strong> <strong>health</strong> careoutcomesAre audit <strong>and</strong> feedbackeffective in improvingprofessional practice <strong>and</strong><strong>health</strong> care outcomes?85 RCTsOnly 10 of <strong>the</strong> 85 includedstudies <strong>to</strong> be of highmethodological quality>3500 <strong>health</strong>professionals✜ Audit <strong>and</strong> feedback can improve professional practice,but <strong>the</strong> effects are variable.✜ When it is effective, <strong>the</strong> effects are generally small <strong>to</strong>moderate.✜ The results of this review do not provide support form<strong>and</strong>a<strong>to</strong>ry or unevaluated use of audit <strong>and</strong> feedback.Table 1: Short description of results of three Cochrane systematic reviews on human resources for <strong>health</strong>160 ✜ Global Forum Update on Research for Health Volume 4
Decision-makingreviews in this field, but <strong>the</strong> final number of studies that met<strong>the</strong> selection criteria is very limited, <strong>to</strong>o, compared <strong>to</strong> <strong>the</strong>initial pool of abstracts. This may mean that ei<strong>the</strong>r <strong>the</strong>selection criteria for systematic reviews were <strong>to</strong>o rigid, or <strong>the</strong>actual number of good quality evaluative studies in <strong>the</strong> fieldof human resources for <strong>health</strong> has been limited (see Table 1for a detailed analysis of three of <strong>the</strong> systematic reviews).Also, most of <strong>the</strong> studies come from developed countries.Second, much of <strong>the</strong> research is descriptive. For example,available studies have pointed <strong>to</strong> global shortages, regionalimbalances <strong>and</strong> increasing migration of <strong>health</strong> workers, aswell as dissatisfaction of <strong>health</strong> workers with <strong>the</strong>ir workingconditions 5-9 . On <strong>the</strong> o<strong>the</strong>r h<strong>and</strong>, many country-basedsituation analyses have pointed <strong>to</strong>wards geographicalimbalances <strong>and</strong> significant understaffing in rural areas. Thereare paradoxical in-country shortages of nurses even wheremany nurses remain unemployed. And <strong>the</strong>re are situationswhere doc<strong>to</strong>rs engage in dual employment 10, 11 .Third, <strong>the</strong> knowledge about effective solutions <strong>to</strong> addresschallenges described is far less developed 5, 12, 13 . And hereinlies a major gap. From a country perspective, policy-makersneed <strong>to</strong> know what solutions are available <strong>to</strong> <strong>the</strong>m. Theywould also like <strong>to</strong> know how much an intervention will cost,how difficult it is <strong>to</strong> implement <strong>and</strong> its likely impact in <strong>the</strong>short run. Knowing what sort of interventions have worked ino<strong>the</strong>r settings would give <strong>the</strong>m an idea of what <strong>the</strong>y can tryin <strong>the</strong>ir own context.Currently, <strong>the</strong>re are very few evaluation studies of effectiveinterventions or impact assessment studies on humanresources for <strong>health</strong>. Also, <strong>the</strong>re are very few sources of bestpractice collections in <strong>health</strong> workforce development. Effortshave started <strong>to</strong> systematize <strong>the</strong> knowledge about effectiveinterventions, but only a very small number of promisingpolicy levers were identified through an extensive search ofrecent systematic reviews 14 .Need for more operations <strong>and</strong> contextspecificresearchWhat is really needed is more operational research <strong>and</strong>context-specific research. Policy-makers in countries may nothave <strong>the</strong> means <strong>to</strong> support sophisticated research such assystematic reviews, but <strong>the</strong>y should certainly have <strong>the</strong>capacity <strong>to</strong> use it. Hence <strong>the</strong>re is a need <strong>to</strong> develop <strong>the</strong> criticalmass of researchers in countries, with <strong>the</strong> appropriate skills <strong>to</strong>do research, use it <strong>and</strong> make its relevance evident <strong>to</strong> policymakers.Also, country-based studies should be encouraged bypolicy-makers, within <strong>the</strong> umbrella of a national strategy of<strong>health</strong> research that should accommodate <strong>the</strong> priorities ofHuman Resources for Health (HRH) research within <strong>the</strong>priorities of overall <strong>health</strong> system research.In general, preference should be given <strong>to</strong> research that:✜ develops <strong>and</strong> evolves from purely descriptive studies <strong>to</strong>conceptual research, policy analysis <strong>and</strong> operationsresearch;✜ promotes international <strong>and</strong> comparative research byconsidering multi-site, multi-level <strong>and</strong> multi-countryresearch projects <strong>to</strong> improve comparability <strong>and</strong>transferability of findings, such as <strong>the</strong> African migrationstudy or <strong>the</strong> European nursing exit study;✜ integrates research in<strong>to</strong> ongoing or planned interventions<strong>and</strong> processes such as <strong>health</strong> sec<strong>to</strong>r reforms so thatlessons can be drawn, compared <strong>and</strong> shared from <strong>the</strong>experiences of different countries 1 .Bringing <strong>the</strong> answers in<strong>to</strong> <strong>the</strong> policy-makingprocessOnce research has provided us with <strong>the</strong> answers we need, weshould move <strong>to</strong> <strong>the</strong> next stage: using that knowledge.In <strong>the</strong> field of human resources for <strong>health</strong>, effective actionmust cut across many sec<strong>to</strong>rs. These include, among o<strong>the</strong>rs:<strong>health</strong>, education, labour, civil service, <strong>the</strong> private sec<strong>to</strong>r –which operates many educational institutions – <strong>and</strong> <strong>the</strong>regula<strong>to</strong>ry system. Ga<strong>the</strong>ring all <strong>the</strong> various ac<strong>to</strong>rs,researchers <strong>and</strong> policy-makers from <strong>the</strong> different sec<strong>to</strong>rsaround <strong>the</strong> same table will take some doing. A recentexample from Canada may provide inspiration for similarprocesses <strong>to</strong> be developed in o<strong>the</strong>r settings.In <strong>the</strong> year 2000, <strong>the</strong> Canadian government established anew national programme of Research Chairs <strong>to</strong> attract <strong>and</strong>retain some of <strong>the</strong> best researchers. One of <strong>the</strong> specific aimsof <strong>the</strong> Chairs is <strong>to</strong> promote <strong>the</strong> best possible use of researchresources through strategic institutional planning <strong>and</strong> throughcollaboration among institutions <strong>and</strong> between sec<strong>to</strong>rs. Forexample, in <strong>the</strong> domain of <strong>health</strong> workforce research at <strong>the</strong>University of Toron<strong>to</strong>, research <strong>to</strong>pics are both agreed upon<strong>and</strong> discussed with local <strong>and</strong> national policy-makers. Theprocess of engaging policy-makers in <strong>the</strong> development ofresearch <strong>to</strong>pics, keeping <strong>the</strong>m informed of <strong>the</strong> results ofresearch <strong>and</strong> cultivating champions among policy-makershas helped <strong>to</strong> address some crucial policy issues in nursingworkforce development in Canada (L O’Brien-Pallas, oralpresentation at World Health Organization (WHO), Geneva,January 2006).Regional <strong>and</strong> national observa<strong>to</strong>ries are o<strong>the</strong>r potentialmechanisms for harvesting <strong>and</strong> disseminating newknowledge, provided <strong>the</strong>y effectively engage <strong>the</strong> full range ofstakeholders <strong>and</strong> <strong>the</strong>ir institutions. The experience of <strong>the</strong>Observa<strong>to</strong>ry on Human Resources for Health in <strong>the</strong> region of<strong>the</strong> Americas demonstrates that <strong>the</strong> observa<strong>to</strong>ry can be aneffective mechanism <strong>to</strong> improve information <strong>and</strong> evidence<strong>and</strong> <strong>to</strong> advance <strong>the</strong> advocacy for human resources issues.The recently launched Africa HRH Observa<strong>to</strong>ry is hoped<strong>to</strong> provide a new impetus for research <strong>and</strong> evidence forthat region 15 .The way forward: encourage local production<strong>and</strong> <strong>reach</strong> out <strong>to</strong> external researchIn conclusion, without good research <strong>to</strong> answer <strong>the</strong> policyquestions in <strong>health</strong> workforce development, we will fail<strong>to</strong> find innovative solutions <strong>to</strong> long-festering problems. Policymakersmust <strong>reach</strong> out <strong>to</strong> researchers in <strong>the</strong>ir own countries<strong>and</strong> facilitate <strong>the</strong> development of national research within<strong>the</strong> framework of a national strategy of <strong>health</strong> systemsresearch.Therefore, this policy brief recommends that both policymakers<strong>and</strong> researchers at country level:Global Forum Update on Research for Health Volume 4 ✜ 161