lable at ScienceDirect

Social Science & Medicine 102 (2014) 190e200

Contents lists avai

Social Science & Medicine

journal homepage: www.elsevier .com/locate/socscimed

Health sector priority setting at meso-level in lower and middleincome countries: Lessons learned, available options and suggestedsteps

David B. Hipgrave a,b,*, Katarzyna Bolsewicz Alderman c, Ian Anderson d,Eliana Jimenez Soto c

aNossal Institute for Global Health, University of Melbourne, Victoria 3010, AustraliabUNICEF, 3 UN Plaza, New York 10017, USAc School of Population Health, University of Queensland, Herston, AustraliadCentre for International Health, Curtin University, Perth 6854, Australia

a r t i c l e i n f o

Article history:Available online 10 December 2013

Keywords:Priority-settingPlanningBudgetingLower- and middle-income countries(LMICs)Meso-levelHealth systemsResource allocationEvidence-based policy

* Corresponding author. Permanent address: Box 103184, Australia.

E-mail addresses: dhipgrave@gmail.com, dhipgravk.alderman@uq.edu.au (K.B. Alderman), ian.and(I. Anderson), e.jimenez@sph.uq.edu.au (E.J. Soto).

0277-9536/$ e see front matter � 2013 Elsevier Ltd.http://dx.doi.org/10.1016/j.socscimed.2013.11.056

a b s t r a c t

Setting priority for health programming and budget allocation is an important issue, but there is littleconsensus on related processes. It is particularly relevant in low resource settings and at province- anddistrict- or “meso-level”, where contextual influences may be greater, information scarce and capacitylower. Although recent changes in disease epidemiology and health financing suggest even greater needto allocate resources effectively, the literature is relatively silent on evidence-based priority-setting inlow and middle income countries (LMICs). We conducted a comprehensive review of the peer-reviewedand grey literature on health resource priority-setting in LMICs, focussing on meso-level and theevidence-based priority-setting processes (PSPs) piloted or suggested there. Our objective was to assessPSPs according to whether they have influenced resource allocation and impacted the outcome in-dicators prioritised. An exhaustive search of the peer-reviewed and grey literature published in the lastdecade yielded 57 background articles and 75 reports related to priority-setting at meso-level in LMICs.Although proponents of certain PSPs still advocate their use, other experts instead suggest broader el-ements to guide priority-setting. We conclude that currently no process can be confidently recom-mended for such settings. We also assessed the common reasons for failure at all levels of priority-settingand concluded further that local authorities should additionally consider contextual and systems limi-tations likely to prevent a satisfactory process and outcomes, particularly at meso-level. Recent literatureproposes a list of related attributes and warning signs, and facilitated our preparation of a simpledecision-tree or roadmap to help determine whether or not health systems issues should be improved inparallel to support for needed priority-setting; what elements of the PSP need improving; monitoring,and evaluation. Health priority-setting at meso-level in LMICs can involve common processes, but willoften require additional attention to local health systems.

� 2013 Elsevier Ltd. All rights reserved.

Introduction

Health needs always exceed the available resources, so priority-setting is a key element in health resource allocation. It is tradi-tionally undertaken by governments responding to market failuresin health care, and to support public goods like immunisation.

49, Brighton Rd. RPA, Elwood

e@unicef.org (D.B. Hipgrave),erson.economics@gmail.com

All rights reserved.

However, in both developed anddeveloping countries the process ofsetting priority for public spending in health has been perenniallydifficult, and the subject of considerable debate. Prudent govern-ments take priority-setting seriously because the resources at theirdisposal e budget, staff time, equipment and facilities e areprecious, and all have alternative uses inside and outside the healthsector. Ideally, governments should collaborate with other stake-holders during the priority setting process (PSP), including popu-lation representatives, local interest groups and developmentpartners, to determine how best to utilise available resources. Suchinclusive priority-setting has been recommended for decades(Navarro, 1969; Paalman, Bekedam, Hawken, & Nyheim, 1998), but

D.B. Hipgrave et al. / Social Science & Medicine 102 (2014) 190e200 191

can be very difficult and affected by context, often resulting infunding choices influenced primarily by history, or “grand-fathering”. Another key issue is the difference between the macro-level priority-setting that occurs at national level and has been thesubject of much research and comment, and the more program-focused priority-setting that occurs at meso-level, on which farless has been written, and which is the subject of this paper.

In decentralised systems the focus of national or macro-levelpriority-setting in health is usually which interventions may befinanced with public money, while the difficult task of deciding themix of programs, resources and strategies for delivering in-terventions is usually undertaken by meso-level authorities (e.g.provinces, states or districts). Ideally, the PSPs at each level arelinked, and allocations reflect the needs and preferences of allstakeholders in a well-described, cascading and participatory pro-cess. The outcome would meet the efficiency goal of health econ-omists, the effectiveness goal of clinicians and be legitimate andreasonable according to relevant policies and cross-sectoral inputs.Moreover, the outcome would be equitable and just, and the pro-cess itself would be accepted by all (McDonald & Ollerenshaw,2011; Sibbald, Singer, Upshur, & Martin, 2009). In practice,priority-setting seems difficult at any level and the links betweenthe levels have not been well described, particularly in lower- andmiddle-income countries (LMICs).

Effective priority-setting is probably even more important nowas populations increase, expectations of good health rise, technicalsolutions to health problems expand and yet resources becomeincreasingly stretched. This is particularly the case for many LMICsnegotiating the epidemiologic transition and the so-called doubleburden of disease (Abegunde, Mathers, Adam, Ortegon, & Strong,2007), and especially for meso-level authorities considering solu-tions for a new constellation of issues. Money wasted on a failedPSP or misguided allocations could have been spent on alternativeprocesses or interventions. Indeed, the problems identified inpriority-setting at macro-level aremost likely accentuated at meso-level, especially in LMICs where limitations to effective priority-setting are likely to be greater. In the increasing number of LMICswith decentralised health systems, these limitations may evenoutweigh the benefits of greater local experience and account-ability among local managers (compared tomanagers in centralisedsystems). Accordingly, LMIC authorities should benefit from a re-view of others’ experiences and suggestions on how to proceedwith health priority-setting at sub-national levels.

We sought to assess the evidence on processes available to guidemeso-level LMIC health authorities considering strategies for scale-up of accepted health interventions. We therefore conducted acomprehensive review of studies describing meso-level PSPs, theirimpact on resource allocation and related lessons from the field.Given the dearth of reports from meso-level, we also included re-view articles on macro-level PSPs. We first report our review of theliterature here. Drawingon this reviewof processes and experiencesand additionally on the perspectives of experts, particularly thoserelated to what is feasible in LMICs, a roadmap for approachingmeso-level health priority-setting in such contexts is proposed.

Method

This research was undertaken during 2012 in the context ofwork to develop evidence-based recommendations on how todevelop and use investment scenarios to take forward the UnitedNations Secretary General’s Global Strategy for Women’s and Chil-dren’s Health in LMICs of the Asia-Pacific region (Jimenez-Soto,Alderman, Hipgrave, Firth, & Anderson, 2012). Our objective wasthus to critically review formal processes for priority-setting inLMICs from a policy perspective.

Search for relevant literature

The first step in our review of the evidence involved brain-storming on key resources and the establishment of limitations.This yielded a list of resources, mostly grey literature, and wasfollowed by a series of searches in formal literature. Since the focuswas on investment scenarios, we agreed that the review should berestricted to evidence for resource allocation and PSPs, and thusexcluded studies related to broader areas such as evidence-basedplanning or policy-setting not specifically related to resource allo-cation. This decision was reinforced by the fact that when weinitially searched for recent literature using broad terms such as(“planning” OR “budgeting” AND “health care”) we found over128,000 references. Even when narrowed down by relevant cate-gories (e.g. health care sciences services), over 35,000 studies werefound and very few of the first 300 appeared relevant. Weaccordingly restricted our search to articles published in the lastdecade using the following key words: “resource allocation” AND“health care” AND “policy” OR “priority-setting” AND “health care”AND “policy”. We did not restrict the search to LMICs. Web of Sci-ence and Econlit databases were searched for relevant peer-reviewed articles. The Web of Science search was narrowed downby categories (“health policy services” or “economics” or “planningdevelopment” or “public administration”) and subject areas(“health care sciences services” or “public administration” or“mathematical methods in social sciences”). The search yielded 239references, many of which were relevant. Upon further discussionthe key word ‘policy’ was dropped to expand the search, yielding874 references. The Econlit database search produced 351 refer-ences, most of which overlapped those from the Web of Science orwere deemed irrelevant. The Cochrane database was searched forsystematic reviews on the subject but none were found. Finally thekey search terms were also entered into the Google search engine;no relevant new articles were identified. The literature search wasconducted during FebruaryeMarch, 2012. Ethics approval was notrequired for this research, which did not involve human subjects.

Selection of papers for inclusion

All identified titles and/or abstracts were reviewed by two of theauthors (KA and EJ) using pre-determined inclusion criteria. Arti-cles not specifically related to priority-setting and resource allo-cation in the health sector were excluded (e.g. those only describingprocesses of planning or development of clinical services). Inreviewing the abstracts it became apparent that in addition to ar-ticles describing individual approaches to priority-setting forhealth, there were others describing common elements of thatprocess. The former “approaches” articles, were included if theydescribed in detail or reviewed a systematic approach to evidence-based health priority-setting. Because there were few reportsfocussing specifically on sub-national priority-setting, we alsoincluded articles that reviewed approaches to priority-setting atmacro-level in several countries. However, we excluded individualcountry studies of macro-level approaches (such as health tech-nology assessments). We also excluded papers which focused onlyon the statistical techniques underlying some of the approaches,such as the use of cost-effectiveness analysis (CEA). The latter“common elements” or “background” articles were included only iftheir focus was on the policy implications of a particular element ofpriority-setting.

Using the above strategy a list of 75 “approaches” articles and 57“background” articles was compiled. All 132 articles were read infull and abstracted using one of two templates prepared by theauthors (available upon request). The “approaches” abstractiontemplate included categories such as its objectives; content areas;

Table 1Papers and reports describing an approach to health priority-setting included in theanalysis (N ¼ 75), and their publication date.

Number, type and focus of “approach”papers or reports selected for analysis

Spread across countries bysocio-economic typology(High-income countries (HICs)or low- or middle-incomecountries (LMICs))

ReviewsReviews of approaches, N ¼ 5 HICs: 3 reviews

(Kenny & Joffres, 2008)(Vuorenkoski, Toiviainen, &Hemminki, 2008)(Mitton et al., 2009)LMICs: 2 reviews(Rudan et al., 2010)(Youngkong, Kapiriri, &Baltussen, 2009)

Reviews of priority-setting approaches(PSPs) at macro-level, N ¼ 9

HICs: 7 reviews(Hoffmann & Graf von derSchulenburg, 2000)(Shepherd et al., 2007)(Sabik & Lie, 2008)(Eddama & Coast, 2009)(Stafinski et al., 2011)(Golan, Hansen, Kaplan, & Tal,2011)(Glassman et al., 2012)LMICs: 2 reviews(Doherty et al., 2004)(Iglesias, Drummond, & Rovira,2005)

Individual studiesPapers focussing on or using

Accountability for Reasonableness(A4R) to assess priority-setting,N ¼ 14

HICs: 6 studies(Martin, Giacomini, & Singer,2002)(Hasman & Holm, 2005)(Jansson, 2007)(Kapiriri, Norheim, & Martin,2007)(Sinclair et al., 2008)(Waldau, Lindholm, & Wiechel,2010)LMICs: 8 studies(Daniels et al., 2005)(Kapiriri & Martin, 2006)(Mshana et al., 2007)(Byskov et al., 2009)(Maluka et al., 2010)(Kapiriri & Martin, 2010)(Maluka et al., 2011a)(Maluka et al., 2011b)

Papers focussing on multi-criteriadecision analysis (MCDA), N ¼ 7

HICs: 3 studies(Baltussen & Niessen, 2006)(Baltussen et al., 2010)(Airoldi & Morton, 2011)LMICs: 4 studies(Baltussen, 2006)(Baltussen et al., 2006)(Baltussen et al., 2007)(Jehu-Appiah et al., 2008)

Papers focussing on public budgetingand marginal analysis (PBMA), N ¼ 8

HICs: 8 studies(Mitton & Donaldson, 2002)(Mitton & Donaldson, 2003b)(Mitton & Donaldson, 2003c)(Mitton & Donaldson, 2003a)(Mitton & Donaldson, 2004)(Dionne et al., 2009)(Mitton et al., 2011)(Tsourapas & Frew, 2011)

Papers focussing on use of amultidisciplinary approach thatincludes PBMA, N ¼ 4

HICs: 4 studies(Gibson et al., 2006)(Patten, Mitton, & Donaldson,2006)(Urquhart, Mitton, & Peacock,2008)(Peacock et al., 2009)

D.B. Hipgrave et al. / Social Science & Medicine 102 (2014) 190e200192

scientific method/methods of analysis; the process by which re-sources, data and claimed outputs are or should be used indecision-making on health investment; the context of the priori-tisation; documented impact on plans/budgets/resource allocationand action; outputs produced; location(s) where the approach wasused and lessons learned. To draw a robust picture of an identifiedapproach (such as program budgeting and marginal analysis(PBMA), multi-criteria decision analysis (MCDA) etc.), relevant in-formation from a number of publications related to that approachwas consolidated into a single abstraction sheet. The abstractiontemplate for “background” articles was simpler and focused onlessons learned or concepts that could be used to develop commonapproaches to priority-setting for health resource allocation.

The abstraction process was preceded by a pilot stage afterwhich the “approaches” template was revised. The articles werethen divided among three of the authors (DH, KA and EJ) andabstracted. A sample of fivewas first abstracted by all three authors,subsequent to which these authors’ notes on the process werecompared and difficulties discussed. Based on this discussion a setof guidelines for abstraction was devised to guide the process.

We first reviewed the “approaches” articles and developedsummaries of three established PSPs (PBMA, MCDA and account-ability for reasonableness (A4R)) that have been used or suggestedfor priority-setting at meso-level in developed countries or LMICs,and several other less established processes that have been putforward. In doing this, we also considered reviews of PSPs at meso-and macro-level. The subsequent additional review of the “back-ground” articles also assisted us to draw conclusions on the chal-lenges to meso-level priority-setting in LMICs.

Results

Table 1 summarizes the 75 reports describing an approach topriority setting (“approaches” articles) that met our inclusioncriteria. It divides them according to the named approach describedand whether they described processes in high-income countries(HICs) or LMICs.

Reviews of priority-setting at macro-level

Regarding macro-level priority-setting for health, the literaturefocuses mostly on processes undertaken in developed nations;several reviews of such activities have been published (Kenny &Joffres, 2008; Sabik & Lie, 2008; Stafinski, Menon, Philippon, &McCabe, 2011; Vuorenkoski, Toiviainen, & Hemminki, 2008). Theoverarching conclusion was that even in high-income settingswhere participatory, accountable and rational approaches to healthpriority-setting should be achievable, the process and outcomes ofsuch exercises have been unsatisfactory. For example, reviews ofpriority-setting in the introduction of health technology in 20developed countries (Stafinski et al., 2011) and of drugs in fourdeveloped countries (Vuorenkoski et al., 2008) revealed manyproblems with the processes used. The other reviews concludedthat PSPs based on principles were too vague, and called for threesets of values to govern health-sector priority-setting: proceduralvalues (clarity of process to limit the influence of contextual issueslike public perception and political influence), substantive values(the priority-setting criteria and principles to be followed), andterminal values (the goals of the process) (Kenny & Joffres, 2008;Sabik & Lie, 2008).

Macro-level priority-setting in resource-poor settings has beenthe subject of much less reporting. One review assessed priority-setting in 12 LMICs, mostly at national level, over 1999e2008(Youngkong, Kapiriri, & Baltussen, 2009). The reviewed reportsmostly sought the best criteria for priority-setting and included

Table 1 (continued )

Number, type and focus of “approach”papers or reports selected for analysis

Spread across countries bysocio-economic typology(High-income countries (HICs)or low- or middle-incomecountries (LMICs))

Paper focussing on use of amultidisciplinary approach N ¼ 1

LMICs: 1 study(González-Pier et al., 2006)

Papers focussing on the Business Caseapproach N ¼ 6

HICs: 6 papers(Reiter et al., 2007)(Western AustralianDepartment Of Health, year notstated)(Department for InternationalDevelopment, 2010)(Department for InternationalDevelopment, 2011)(Department for InternationalDevelopment, 2012)(Flanagan & Nicholls, 2007)

Papers focussing on use of the LivesSaved Tool, N ¼ 6

Overview(Winfrey, McKinnon, & Stover,2011)LMICs: 5 papers(Bryce et al., 2010)(Hazel et al., 2010)(Ricca et al., 2011)(Larsen, Friberg, & Eisle, 2011)(Acuin et al., 2011)

Papers focussing on the InvestmentCase approach, N ¼ 6

LMICs: 6 paperss(Jimenez-Soto et al., 2009)(Trisnantoro, Widiati, &Kurniawan, 2011)(Varghese et al., 2011)(Aldaba et al., 2011)(Morgan et al., 2011)(Jimenez-Soto, La Vincenteet al., 2012)

Papers focussing on a Balance Sheetapproach, N ¼ 1

LMICs: 1 paper(Makundi, Kapiriri, & Norheim,2007)

Combined normative-empiricalapproach, N ¼ 2

HICs: 2 reports(Singer, Martin, Giacomini, &Purdy, 2000)(Martin & Singer, 2003)

Papers focussing on public participationas a key approach, N ¼ 2

LMICs: 2 papers(Kapiriri, Norheim, &Heggenhougen, 2003)(O’Meara et al., 2011)

Papers recommending a mix ofqualitative and quantitativeapproaches, N ¼ 2

HICs: 1 paper(Smith, Mitton, & Peacock,2009)LMICs: 2 papers(Nelson et al., 2005)

Paper providing a framework onpriority-setting at local level, N ¼ 1

HIC: 1 paper(McDonald & Ollerenshaw,2011)

Diamond model, N ¼ 1 HIC: 1 study(Lu, Huang, & Chiang, 2011)

D.B. Hipgrave et al. / Social Science & Medicine 102 (2014) 190e200 193

multiple stakeholders through group discussions or interviews.Nonetheless, eight countries relied on the literature to select saidcriteria rather than establishing them locally, and used a wide va-riety of methods to select them and which interventions should beprioritised, and how these selections should be presented (in rankorder, in lists or in prose). The review concluded that fair andlegitimate priority-setting in LMICs should use a range of criteria(health and non-health) to account for context, involve multiplestakeholders and use quantitative and qualitative techniques toaccount for contextual variables (ethical and cultural consider-ations, approaches to complex interventions etc.) that cannot easilybemeasured. Contextualisation, while reducing the generalizabilityof conclusions, was also acknowledged by experts involved inMexico’s successful health care reform (González-Pier et al., 2006),

and has been recommended to augment cost-effectiveness analysis(CEA) (Hoedemaekers & Dekkers, 2003; Smith, Mitton, & Peacock,2009), including in a low-resource setting (Yothasamut, Tantivess,& Teerawattananon, 2009). To provide clarity, it was alsoconcluded that PSPs should aim to rank or at least rule-in or rule-out interventions (Youngkong et al., 2009).

Another recent review of national-level priority-setting forhealth service packages and health technologies in LMICs(Glassman et al., 2012) noted that designing a framework forpriority-setting in such countries is particularly difficult. Whilesuggesting principles and process elements, the authors acknowl-edged that most LMICs lack institutional mechanisms, valid healthstatistics and costings to compare the various “priorities”, evaluatepolitico-economic constraints and engage a broad range of stake-holders. Many are also heavily influenced by donors onwhat shouldbe publicly funded, and political influences that outweigh theclaims of lower-ranked stakeholders.

From these published reviews of macro-level priority-setting inmany developed and developing countries, one may conclude thatcontext is more important than process, and none of the processesreported or recommended could easily be replicated across con-texts or for different purposes in the health sector.

Experience with recognised PSPs

We found three comprehensive PSPs which acknowledge therole of context (political, social, financial, cultural etc.), and whichhave been put forward as candidates to guide priority-setting in awide rangeof settings, including LMICs.Web-table 1 describes thesePSPs (A4R, MCDA and PBMA), each of which have been piloted andprospectively assessed, two of them in both developed and devel-oping countries and two atmeso-level.While two are in regular use,it is evident from the following summaries that none provides aprocess yet able to be recommended at any level in LMICs.

i. Accountability for reasonableness (A4R)

A4R (Daniels & Sabin, 1998) was developed to provide an ethicalfoundation for priority-setting, making it acceptable to all stake-holders (Peacock, Mitton, Bate, McCoy, & Donaldson, 2009). Whileit does not prescribe a process as such, it prescribes four key con-ditions in priority-setting (Web-table 1) , and has been used toevaluate health priority-setting at macro-, meso- andmicro-level indeveloped countries and LMICs (Kapiriri & Martin, 2006; Kapiriri,Norheim, & Martin, 2007) and for other health-related purposes(Jansson, 2007; Sinclair et al., 2008).

Evaluations of A4R in both HICs and LMICs focus onwhether theprocess is regarded as fair by stakeholders (Mshana et al., 2007;Waldau, Lindholm, & Wiechel, 2010), rather than being clear andreplicable, or on the outcomes. Moreover, in evaluating A4R inKenya, Tanzania and Zambia (Byskov et al., 2009; Mshana et al.,2007) researchers even found that it could not be implementedaccording to its four key conditions (Maluka et al., 2010; Maluka,Kamuzora et al., 2011). They and others in HICs found A4R to beconceptually abstract, difficult to implement and susceptible tocontextual limitations (Hasman & Holm, 2005; Jansson, 2007;Maluka, Kamuzora et al., 2011); others emphasise the importanceof adapting A4R-driven PSPs according to the key actors and theirinterpersonal relationships; the institutional setting (culture,informal rules, routines) and the wider context (Maluka, Hurtiget al., 2011; Sinclair et al., 2008). While primarily noted in HICs,these limitations are also relevant to priority-setting at meso-levelin LMICs.

In summary, A4R provides a set of standards against which tomeasure fairness and equity in priority-setting, but despite being

D.B. Hipgrave et al. / Social Science & Medicine 102 (2014) 190e200194

the foundation for attempts to improve priority-setting in severalLMICs, does not provide a practicable framework for implementingpriority-setting in developing countries.

ii. Multi-criteria decision analysis (MCDA)

MCDA is claimed to maximize consensus in the allocation ofresources, and was applied to health priority-setting after a longhistory of usage in other disciplines (Baltussen & Niessen, 2006). Asthe name implies, MCDA simultaneously rates interventions or al-locations in the health sector against multiple criteria. It has beenutilised for population-level (Baltussen, 2006; Baltussen, Stolk,Chisholm, & Aikins, 2006; Baltussen et al., 2007; Jehu-Appiahet al., 2008) or local priority-setting (Airoldi & Morton, 2011) inboth HICs and LMICs, but most reports describe macro-level pro-cesses. Some claim that MCDA is the recommended evaluationprocess for a PBMA exercise (Airoldi & Morton, 2011).

Data requirements for MCDA may be quantitative and/or qual-itative, depending on the area of focus. They may include a CEA ofvarious options, district accounts analysis, equity effectiveness ormarginal budgeting for bottlenecks (MBB) (Baltussen & Niessen,2006). MCDA can account for special interests and local context,but can only rank interventions according to quantifiable criteria;in LMICs it has been shown to require a secondary, deliberativeprocess to account for non-quantifiable issues (ethical, political,cultural etc.) (González-Pier et al., 2006; Youngkong et al., 2009).This reduces the ability to generalize an MCDA process beyond theimmediate context, but adds to its accountability and transparency(Baltussen, Youngkong, Paolucci, & Niessen, 2010).

MCDA has been piloted in several LMICs at macro-level and waswell-received by government authorities involved in resourceallocation and selection of interventions (Baltussen et al., 2006;Baltussen et al., 2007; Jehu-Appiah et al., 2008). It was also usedto augment CEA and PBMA at meso-level in the UK (Airoldi &Morton, 2011). However, there is little documented evidence thatMCDA has routinely influenced resource allocation. In theory,MCDA provides a replicable, transparent process that could workwell in LMICs, but more experience is needed on whether itadequately accounts for context, and how easily it can be applied,especially at meso-level.

iii. Program budgeting and marginal analysis (PBMA)

PBMA is a systematic and explicit approach to prioritisingresource allocation and/or disinvestment in the context of currentlybudgeted activities. It assesses the cost and benefits of proposedbudget changes, using a clear and replicable method and structuredinvolvement of stakeholders (through an advisory panel). It pro-duces clear outputs and evaluation is built-in. Moreover, some inHICs have experimented with augmenting PBMA, such as usingMCDA to help decide the locally relevant criteria and constructmeasurement scales (Airoldi & Morton, 2011; Peacock et al., 2009),or adding participatory research to facilitate PBMA decision-making (Patten, Mitton, & Donaldson, 2006). Others have alignedthe PBMA process with the four A4R principles, to increase itsperceived fairness and legitimacy (Gibson, Mitton, Martin,Donaldson, & Singer, 2006; Urquhart, Mitton, & Peacock, 2008).

PBMA has been used extensively in developed countries(particularly Canada, the UK, Australia and New Zealand) to guidepriority-setting at micro-, macro- and meso-levels, and its influ-ence has been well documented and evaluated (Dionne, Mitton,Smith, & Donaldson, 2009; Mitton, Dionne, Damji, Campbell, &Bryan, 2011; Mitton & Donaldson, 2002, 2003a, b, 2004;Tsourapas & Frew, 2011). A recent evaluation found that itimproved stakeholders’ understanding of priority-setting itself

(52% of cases), implementation of priority-setting recommenda-tions (65%) and actual disinvestment or resource allocation (48%)(Tsourapas & Frew, 2011). However, only 22% of local authoritiesadopted PBMA for later use and this evaluation was undertaken indeveloped countries. Our review did not identify reports of PBMAin LMICs.

Other approaches not yet piloted or evaluated

Certain other approaches to health priority-setting (includingCEA, disease burden impact analysis, equity analysis etc.) have beendescribed as both mutually exclusive and collectively incomplete,and also unable to account for different contexts or perspectives(Hoedemaekers & Dekkers, 2003; Rudan et al., 2010; Smith et al.,2009; Yothasamut et al., 2009). Most reports on these and similarapproaches were excluded from our search, as they focused ontechnical implementation, not their use in policy or priority-setting. The ability to use them is also very limited at meso-level(particularly in LMICs), where data is often scarce and decisionsare more of a managerial nature (Greener & Powell, 2003) or aremore influenced by local context than technical approaches(Reichenbach, 2002).

Other named and unnamed approaches to priority-setting havebeen developed and reviewed (Table 1), but are not currentlyapplied widely or were ad-hoc and have not been replicated. Thewell-established private sector practice of developing a businesscase (BC) has also been suggested by the United Kingdom Depart-ment for International Development (Department for InternationalDevelopment, 2010, 2011) and others (Government of WesternAustralia Department of Health, 2008) to assist decisions onintervention funding. This requires a consistent and transparentappraisal of alternative interventions according to the principles ofeconomy, efficiency, effectiveness and cost-effectiveness, alongwith a clear process and standards. While the BC approach mightset the gold standard for documentation and appraisal, manydeveloping countries would struggle to develop sound BCs for theirhealth priorities, as most of the basic data is missing (ThePartnership for Maternal Newborn and Child Health, 2012). Wefound no documented evaluations of the BC approach in LMICs.

Priority-setting in international child health research was theoriginal focus for a review of several other methodologies in low-resource settings (Rudan et al., 2010). This report describedseveral approaches to guide evidence-based priority-setting,including MBB, use of the Lives Saved Tool (LiST) and the ChildHealth and Nutrition Research Initiative (CHNRI) methodology.(These tools are summarised at http://www.who.int/pmnch/topics/economics/costing_tools/en/index.html.) The review concludedthat the CHNRI approach provides policymakers with an overviewof the strengths and weaknesses of competing investment optionsagainst many criteria, based on expert technical input and isadjustable according to stakeholder perspectives. The principlesestablished could be applied at macro- and meso-level, but no pi-lots have been implemented. The LiST has been evaluated for itscapacity to inform priority-setting at national level in Africa (Bryceet al., 2010). Although it stimulated discussion between parties,none of the three countries in which it was evaluated changedformal targets or plans based on its application. MBB has beenwidely used to estimate the cost of scaling-up interventions inAfrica and Asia by UNICEF and national partners, but we found noevaluations of its sustained impact on routine resource allocation.

Finally, another approach used to aid evidence-based priority-setting specific to maternal and child health (MCH) in several Asia-Pacific countries is the “Investment Case for Scaling up EquitableProgress in MCH” (IC) (Jimenez-Soto, La Vincente et al., 2012). Thisapproach engages policymakers in quantitative problem-solving

Table 2Ten key criteria for any framework on health priority-setting.a

Process Stakeholder engagement The organization’s efforts toidentify and effectively engageinternal and externalstakeholders

Explicit process A transparent process, clear onwho is making the decisions,and how and why they aremade

Clear and transparentinformation management

The process by which theinformation is collected,collated and made available todecision makers

Consideration of values andcontext

The extent to which thestrategic context and values ofthe organization, the staff andother stakeholders areconsidered

Revision or appeals mechanism A formal mechanism forreviewing decisions andaddressing disagreements

Outcome Stakeholder understanding Improved insights into therationale for making decisionsand the organization’s strategy

Shifted priorities and/orreallocated resources

Changes in strategic directions,use of resources (e.g. humanresources or facilities) orreallocation of resources

Decision-making quality For example, appropriate use ofevidence andinstitutionalization of anevidence-based approach forpriority-setting

Stakeholder acceptance andsatisfaction

Stakeholders’ buy-in to andcontentment with the PSP

Positive externalities Peer emulation, or health sectorrecognition, or changes inpolicies as a result of the PSP

a Based on (Sibbald et al., 2009).

D.B. Hipgrave et al. / Social Science & Medicine 102 (2014) 190e200 195

and other exercises to prioritise interventions to equitably addressmaternal and child mortality. Outputs include related country-specific needs and constraints, feasible and cost-effective strate-gies to address them, and ranking of investment scenarios based ontheir estimated costs and impacts onmortality. The IC approach hasbeen introduced at macro-level and at meso-level in selected areasof India (Orissa and Uttar Pradesh) (Varghuese et al., 2011),Indonesia (Trisnantoro, Widiati, & Kurniawan, 2011), Nepal(Morgan et al., 2011) and the Philippines (Kraft, Aldaba, & LaVincente, 2011). So far it has had varied impact on annual plansand budgets in each country. Many strategies required for scale-upof the interventions considered are outside the remit of district-level authorities and involve many partners and complex fundingarrangements. These and other systemic issues might prevent theIC from impacting MCH planning and budgeting at this level(Jimenez-Soto, La Vincente et al., 2012).

Evidence on challenges to priority setting at meso-level in LMICs

The foregoing review of the approaches identified did notenable recommendation of any formal or defined PSP for meso-level in LMICs. Moreover, several reports from LMICs stressed thechallenges and frustrations involved in setting priorities at meso-level in countries that have decentralised (Maluka, Hurtig et al.,2011; O’Meara, Tsofa, Molyneux, Goodman, & McKenzie, 2011). Agood example incorporating most of these challenges described anattempt to establish a system of needs-based resource allocationand budgeting in one province of Pakistan (Green, Ali, Naeem, &Ross, 2000). After listing the perceived health needs and practicalissues involved in the PSP and implementing resource allocationaccordingly, problems typical to such settings appeared. Theseincluded central government refusal to acknowledge districts’ pri-orities; inconsistent and unpredictable financial flows; a rigid andcomplex budget process with multiple and restricted fundingsources; absence of budgeting experience among staff; weak sup-ply and logistic systems; unreliable information systems; politicaland medical profession/special-interest groups demonstratingvested interest in the status quo or particular priorities; low salariesleading to low commitment to the public sector etc. The Pakistanprocess yielded deeply inequitable allocations and an inefficientPSP, and eventually an alternative to needs-based priority-settingwas required. The authors noted that unanticipated resistance tochange at central and local levels, unclear decision space bound-aries along with other contextual influences far outweighed con-cerns about local health needs.

Unclear decision space boundaries such as those documented inPakistan (Green et al., 2000) highlight the lack of independence oflower level authorities from the influence of higher authorities,even in decentralised systems. This absence of coordination downthe levels is noted almost universally in the literature onmeso-levelpriority-setting from LMICs. In decentralised Uganda, meso-levelallocations “reflected macro-level priorities in regard to nation-ally determined priority programs” (Kapiriri et al., 2007, p.90). InKenya, community-based priority-setting was heavily restricted bythe need to prioritise and report on specific health service in-dicators chosen at national level (O’Meara et al., 2011). “The endresult was a work plan that was more reflective of national prior-ities than local priorities” (p.241). In aid-dependent countries, lackof independence is compounded by the influence of donors. InTanzania too, local authorities complained that national planningguidelines and budget ceilings limited their ability to allocate re-sources according to local priorities (Maluka, Kamuzora et al.,2011). However, in this setting, and to a degree in the Pakistancase (Green et al., 2000) it was also noted that national prioritiesare selected with good reason, and imposed restrictions may

actually improve equity and fairness at local level, given the heavyinfluence of certain stakeholders on priority-setting at sub-nationallevel (Maluka, Hurtig et al., 2011), and the lack of capacity in somedistricts (Kapiriri, Norheim, & Heggenhougen, 2003; O’Meara et al.,2011).

The literature also identifies other common challenges to meso-level priority-setting in LMICs. Lack of local data is a major problem,including reliable disaggregated data on disease epidemiology,costs, service uptake and health outcomes. Lack of trust betweenstakeholders, lack of ownership, lack of time available for priority-setting (and the amount of time it can take (Madi et al., 2007)), thedominance of local politics over evidence, discontinuity ofpersonnel, disincentivisation of key stakeholders by the proposedchanges or gaming of the system due to vested interests and lack ofeffective mechanisms to engage the public are all noted (Maluka,Hurtig et al., 2011; Mitton, Smith, Peacock, Evoy, & Abelson, 2009;O’Meara et al., 2011)

Recommended common themes for health priority-setting

Acknowledging the lack of consensus on a “best-practice”framework or normative approach that adequately incorporatescontext, several experts recently utilised mixed empirical methodsto identify ten key elements (five related to process and five tooutcomes) that incorporate the various approaches to priority-setting and definitions of a successful health sector PSP (Sibbaldet al., 2009) (Table 2). Each of the items is claimed to be impor-tant individually but also relates to the others, yielding a “robustand comprehensive definition of successful priority-setting in a

Table 3Seven principles and seven processes for health technology priority-setting inLMICs.a

Principles 1. Ethically sound governance2. Scientific rigour3. Transparency4. Consistency5. Independence from vested interests6. Contestability7. Timeliness and enforceability

Processes 1. Registration and assessment of safety and effectiveness2. Appraisal of what interventions are appropriate3. Cost-effectiveness analysis in the local context4. Budget impact analysis in the local context5. Deliberative processes that engage all relevant stakeholders6. Formal decisions with associated budgets and payment7. An appeal, tracking and evaluation mechanism (preferablywith a time-limit)

a Based on (Glassman et al., 2012).

Table 4Recommended attributes for health sector priority-setting processes.a

Attribute set Summary of collected attributes

Soundness of the situation analysisinforming strategy

Comprehensive, contextualised andparticipatory analysis of healthdeterminants and health outcometrends, based on data or qualitativeanalysis; measurable, realistic, relevant,resourced and time-bound priorities,objectives and outcomes.

Soundness/inclusiveness ofdevelopment and endorsementprocesses

Open, fair, multi-stakeholderinvolvement in the PSP; selectedpriorities fit well with those of variousinterest groups, relevant non-health-sector agencies, and local political andhigher-level authorities, to assureimplementation and funding

Sound financing and auditprocesses

The prioritised interventions orstrategies are under-written by acomprehensive costing/budget,governed by a sound financialmanagement system and transparentaudit process that aims for “value formoney”

Sound implementation andmanagement

Roles and responsibilities ofimplementing partners are agreed andclear, and are governed by an equitableand realistic resources andimplementation schedule.Procurement, governance andmanagement meet relevant standards,and are adequate for the objectives andproposed interventions.

Inclusion of monitoring, evaluationand review

The monitoring and evaluation (M&E)plan clearly describes output andoutcome impact indicators and agreedtargets and milestones, along with therelated sources of information andtiming. An agreed, regular reviewprocess is in place.

a Based on (Travis & Klingen, 2009).

D.B. Hipgrave et al. / Social Science & Medicine 102 (2014) 190e200196

broad conceptual framework” (p.7). The detailed definition pro-vided for each is noteworthy for its explicit commitment toevolutionary improvement in priority-setting and the link thisimplies to the context, values and stakeholder participation. Thesame team evaluated this approach in a developed-country setting(Sibbald, Gibson, Singer, Upshur, & Martin, 2010) with positiveoutcomes.

Similarly, for priority-setting in selection of health servicepackages and technologies in LMICs, a set of seven principles andseven processes have recently been recommended (Glassman et al.,2012) (Table 3). However, the authors and others (Dickinson,Freeman, Robinson, & Williams, 2011) realistically note the poten-tial obstacles related to capacity, stakeholder understanding of theprocess, fragmentation across agencies, weak governance or cor-ruption, weak information systems and weak links betweenappraisal, decision-making and practice. To this list should be addedobstruction due to political or vested interests and incomplete de-centralisation in LMICs and elsewhere (Green et al., 2000; Maluka,Hurtig et al., 2011; O’Meara et al., 2011; Reichenbach, 2002; Wirtz,Cribb, & Barber, 2003), and also the cost of the PSP itself.

Finally, an attempt to define the key elements and facilitateevaluation of national health strategies has recently been devel-oped as the Joint Assessment Tool of the International HealthPartnership (IHP) (http://www.internationalhealthpartnership.net/en/partners; (Travis & Klingen, 2009)). Although developed toharmonise support for national health planning, the Tool couldequally be useful at meso-level, and is not prescriptive. It facilitatesexamination of the strengths and weaknesses of five sets of attri-butes considered the foundation of any health strategy (Table 4).One particularly helpful component of the Tool guidelines is thehighlighting of signs that warn of problems in the planning process,many of them common to the literature on priority-setting. Ex-amples of these warning signs for each set of attributes are pro-vided in Web-table 2.

Possible roadmap for priority-setting in LMICs

Our main objective was to identify an approach for LMIC gov-ernments to use in health priority-setting at meso-level. The ap-proaches assessed do not provide LMICs with clear guidelines or anevidence base for a process that might meet recommended stan-dards, at any level. Technical approaches such as CEA or disease-burden analysis also do not capture key context-specific factorssuch as the local politics and decision-making culture (Kapiriri &Martin, 2007). More importantly, health priority-setting refersnot only to deciding on and allocating resources, but is intimatelyconnected to the status of national and sub-national health

systems. Indeed, the importance of accounting for systems andother contextual issues in HPS (see “Recommended commonthemes” above) resonates with our own experience (Hipgraveet al., 2012; Jimenez Soto et al., 2013; Jimenez-Soto, La Vincenteet al., 2012; La Vincente et al., 2013). This implies that where majorsystemic obstacles are evident, countries should deal with thosebefore moving to implement costly or complex initiatives. Mean-while, however, health authorities must continue to make de-cisions on health resource allocation.

The absence of a recommendable approach suggests that suchauthorities would benefit from guidance on their assessment andimprovement of priority-setting, in a health systems context. Aroadmap is proposed, based on one originally designed to provideboth a normative and empirical approach to priority-setting atmeso-level (Martin & Singer, 2003), involving research at each stepto support a context-driven process. A variation on this approachhas also been recommended for LMICs (Kapiriri & Martin, 2007). Inthe more systems-focused approach depicted (Fig. 1), the original“describe, evaluate and improve” conceived by Martin and Singer(2003) has been re-ordered into: assessing (describing) thecontext, deciding on whether it is feasible to proceed with a PSP oradditionally focus on improving the systems elements required toassure sustained impact. The original emphasis on foundationalresearch, monitoring and evaluating of the process is retained, withan additional focus on the health outcomes. In addition to thefoundation of these four steps in the earlier consideration of ex-perts, they are also based on the reported experience of priority-setting in LMICs, as follows:

Fig. 1. A possible roadmap for assessing and improving health priority-setting at local or national level.

D.B. Hipgrave et al. / Social Science & Medicine 102 (2014) 190e200 197

- Assess the current prioritization/resource allocation process: thereviewed evidence suggests that improvements in priority-setting may not be realized unless they are preceded by asound analysis of the existing situation. The roadmap includesthe domains suggested by others as key elements of suchanalysis (Gibson et al., 2006; Kapiriri et al., 2007; Martin &Singer, 2003), and the IHP warning signs (Travis & Klingen,2009) can guide assessment of the health system aspects(Web-table 2). A mix of partners from public and non-statesectors, civil society, private providers and developmentagencies should be engaged early in this assessment and sub-sequent policy work. Experience suggests the actual practice ofsetting priorities often deviates fromwritten policy (Green et al.,2000; O’Meara et al., 2011).

- Decide whether improvement in the current PSP is needed and/orviable: there is no simple guideline for this. However, as sum-marised above (see “Evidence on challenges to priority setting atmeso-level in LMICs” above, Table 4 and Web-table 2) commonsystemic factors may act as barriers to an approach that islegitimate and fair, and warning signs of problems may exist.Policymakers should consider improvements to the PSP onlywhen these barriers can be adequately overcome.

- Improve the current PSP: the reviewed “approaches” (seeExperience with recognised PSPs and Table 1) suggest a set oftangible elements of PSPs that should be included in consider-ation of any improvement thereof (Fig. 1). These are required fora systematic approach to health priority-setting at sub-nationallevel. However, they are not intended to be prescriptive andshould be adapted to the individual country situation and thespecific approach adopted.

- Evaluate the process adopted: Very few approaches to healthpriority-setting have been evaluated, and “there is no clearguidance from the literature on how to evaluate the success of apriority-setting exercise” (p.178, Tsourapas & Frew, 2011). Thesuggested common principles, processes and outcomes listed inTables 2 and 3 and discussed above (see Recommendedcommon themes) provide a starting point for any such evalua-tion. Experts have recently attempted to define the content andpracticalities of evaluating priority-setting in LMICs, recom-mending an equal focus on the outcomes and the process,including its efficiency, use of evidence, fairness (using the A4Rcriteria) and the level of public confidence in the process(Kapiriri & Martin, 2010). Use of objectively verifiable indicators

and means of verification for each of the suggested parametersis also recommended.

Discussion

Priority-setting is an important and difficult component of theprocess of health resource allocation, with which health andfinance authorities, community leaders and other stakeholdershave struggled for decades. In seeking to assist LMICs prioritiseallocation of resources to scale up recommended health in-terventions we reviewed the related literature on approaches topriority-setting. Our focus was on meso-level which is increasinglythe setting for health resource allocation as various forms of de-centralisation become established. Compared to reports onpriority-setting in developed countries, relatively little informationwas available, possibly reflecting not only the low priority given toprocess evaluation but also publication bias due to reluctance toreport PSPs with negative outcomes.

Given this scarcity of evidence, the political sensitivity ofresource allocation and the dominance of context, we conclude thatgeneralizable guidelines on priority-setting in LMICs cannot beestablished. In particular, the utility of the three PSPs reviewed indetail (A4R, MCDA and PBMA) seems limited in such settings;although all have promising elements none can be unequivocallyrecommended at this time. The other tools and approachesconsidered, many of them developed specifically to assist priority-setting in LMICs (e.g. LiST, CHNRI, IC, MBB), either lack an actualframework for decision-making, or are not yet supported by evi-dence. Even in developed nations there was limited published ev-idence that PSPs have consistently benefited health resourceallocation or the issues for which they were undertaken, at anylevel.

Experts have also proposed principal elements to include in anyPSP, suggesting that in-country processes should involve tailoringthese elements to local needs and context (Kapiriri & Martin, 2007).However this recommendation may not benefit LMICs where theunavailability of funds, weak information systems, limited oppor-tunities for stakeholder engagement and local political or socio-cultural influences may be more limiting than the processapplied, or the inclusion of principle elements. Indeed, it is ourperspective and the reviewed literature suggests that one group’sten recommended process and outcome elements (Sibbald et al.,2009) (Table 2) and another’s seven principles and seven

D.B. Hipgrave et al. / Social Science & Medicine 102 (2014) 190e200198

processes (Glassman et al., 2012) (Table 3) are relatively discon-nected from the reality of priority-setting in most LMICs.

Given this context and the ongoing requirement for successfulpriority-setting, our review suggests that careful examination ofthe systems and other contextual influences on health resourceallocation in LMICs is more important than the priority-settingprinciples and steps suggested; these are irrelevant if the situa-tion prevents a fair, informed, unbiased and functional process.Moreover, priority-setting is intimately linked to and dependent onfunctioning health systems; it is not possible to implement pro-grams however strongly prioritised unless known problems withinthe health system are rectified.

We acknowledge that deciding whether to focus primarily onpriority-setting and perhaps additionally on the systems and othercontextual issuesmay itself be influenced by context. (Is it politicallyacceptable to favour improving the PSP and possibly systems issuesover funding specific health activities? Will new funds allocated tothe health sector be lost if they are not spent on new programs withobjectively measurable outputs or outcomes? What are the risks ofcurrentpatterns ofhealth spending? Is it feasible to resolve the socio-cultural obstacles to priority-setting?) It will also be influenced bythe scale and typeof problems identified; short-termprioritisationofsystems issuesmay not resolve fundamental problems such as someof those listed (Table 4; Web-table 2). However, in the context ofdiminishing resources and increasing expectations in LMICs, focus-sing on the fundamentals of priority-setting, at the least in parallelwith institutional reforms addressing systems issues (e.g. timelydisbursement of funds)may sustainably improve resource allocationas well as health outcomes in such settings. We therefore infer fromthe literature that where there are deep and unresolved systemicbarriers to effective health priority-setting, these barriers should bedealt within parallel to improving the PSP itself.

The roadmap calls for careful monitoring, and recent priority-setting literature appropriately emphasises process and impactevaluation that is capacity-building, a good investment and feasibleeven in LMICs (Kapiriri & Martin, 2010). The absence of relatedevaluation has been noted in recent reviews of one aspect ofpriority-setting, public engagement (Bolsewicz Alderman,Hipgrave, & Jimenez-Soto, 2013; Mitton et al., 2009). Suchengagement and such evaluation should be included in the designof any priority-setting exercises in future (Mansuri & Rao, 2012). Tothe extent that meso-level priority-setting is heavily contextual,counterfactual or alternative scenarios should be included in suchassessment. We encourage authorities engaged in priority-settingat all levels and in all settings to include careful and preferablyindependent monitoring and evaluation as a routine element oftheir work.

Finally, one rarely acknowledged issue that should impact onpriority-setting in LMICs is the extent to which public providers’practices are entirely for the public good, or to which health ser-vices are provided by the private sector and therefore not directlyinfluenced by government priorities. Priority-setting for allocationof health resources implicitly assumes that the process will have asubstantive impact on health outcomes. However, not only doesthis usually ignore the influence of social and other determinants ofhealth, it is only true to the extent that the public sector dominateshealth service delivery or that governments are able to effectivelyinfluence and regulate private practice. This is clearly not the casein many nations in South and East Asia, where private providersdeliver a large proportion of health services, mostly unregulatedand for profit, not public service (Nishtar, 2010; Peters & Bloom,2012). This is an important area for future research. Suffice tosuggest that in prioritising allocation of resources, the extent towhich both public providers and the private sector are well regu-lated, participate in the PSP and pursue the same priorities should

be considered. Parallel measures to improve this may beappropriate.

We acknowledge limitations to our review, and obviously to thesolutions suggested in the roadmap. Although comprehensive, ourreview was not systematic as no formal assessment of bias wasundertaken. The material reviewed described a mix of approaches,possibly biased towards ones easier to conduct in LMICs. Theseincluded qualitative and semi-quantitative field research, accountsand some participant observation, literature reviews and com-mentaries. While the overall quality of evidence may be low, inmost cases the literature reviewed was written by experiencedacademics with many publications in the area of health priority-setting. We have already mentioned publication bias, which mayequally apply to unreported positive outcomes of processes un-dertaken in a non-academic setting. In addition, many of the arti-cles reviewed had their origin in developed country settings; to theextent that priority-setting in such locations is easier, the situationmay be even worse in LMICs, underscoring even more the impor-tance of the first three steps in the roadmap. In developing theroadmap, we relied more on the background articles than the ap-proaches. However, we concluded that the lack of a recommend-able process was more than offset by the practical and groundedsequence depicted and the range of supporting articles, giving it asolid foundation. Whether or not it is stating the obvious thatcountries should “tidy the house and check with the bank beforecommencing a renovation”, the problem remains that many donorsand national authorities are expecting meso-level decision-makersto introduce interventions or fund activities before it is feasible toprioritise them.

Our review of the literature concludes that priority-setting inLMICs is complex, difficult and currently unsatisfactory from mostobjective viewpoints; it suggests consideration of systems andcontextual issues during the process of priority-setting, particularlyat meso-level. We have applied a realistic approach to an issue thathas attracted much interest and many suggestions, but scantacknowledgement of the lack of evidence. Global health partnersare keen to devise and recommend processes or pathways that canbe feasibly implemented in low-resource settings, to improve ef-ficiency, equity and health outcomes. However, allocating preciousresources to strategies or programs before systems and contextualissues have been effectively resolved risks further waste of theseresources, and failure to improve population health outcomes.

Acknowledgements

The work reported in this paper was funded by WHO throughthe Partnership for Maternal, Newborn and Child Health as part ofan agreement with the University of Queensland.

Appendix A. Supplementary material

Supplementary material related to this article can be foundonline at http://dx.doi.org/10.1016/j.socscimed.2013.11.056.

References

Abegunde, D. O., Mathers, C. D., Adam, T., Ortegon, M., & Strong, K. (2007). Theburden and costs of chronic diseases in low-income and middle-incomecountries. The Lancet, 370, 1929e1938.

Acuin, C. S., Khor, G. L., Liabsuetrakul, T., Achadi, E. L., Htay, T. T., Firestone, R., et al.(2011). Maternal, neonatal, and child health in southeast Asia: towards greaterregional collaboration. The Lancet, 377, 516e525.

Airoldi, M., & Morton, A. (2011). Portfolio decision analysis for population health. InA. Salo, J. Keisler, & A. Morton (Eds.), Portfolio decision analysis: Improvedmethods for resource allocation (pp. 359e381).

Aldaba, A., La Vincente, S., Kraft, A., et al. (2011). The Philippines: Developing an in-vestment case for financing equitable progress towards MDGs 4 and 5 in the Asia

D.B. Hipgrave et al. / Social Science & Medicine 102 (2014) 190e200 199

Pacific Region: Scale-up report (unpublished report). Brisbane: School of Popu-lation Health, The University of Queensland.

Baltussen, R. (2006). Priority setting of public spending in developing countries: donot try to do everything for everybody. Health Policy, 78, 149e156.

Baltussen, R., & Niessen, L. (2006). Priority setting of health interventions: the needfor multi-criteria decision analysis. Cost Effectiveness and Resource Allocation, 4,1e9.

Baltussen, R., Stolk, E., Chisholm, D., & Aikins, M. (2006). Towards a multi-criteriaapproach for priority setting: an application to Ghana. Health Economics, 15,689e696.

Baltussen, R., ten Asbroek, A. H. A., Koolman, X., Shrestha, N., Bhattarai, P., &Niessen, L. W. (2007). Priority setting using multiple criteria: should a lunghealth programme be implemented in Nepal? Health Policy and Planning, 22,178e185.

Baltussen, R., Youngkong, S., Paolucci, F., & Niessen, L. (2010). Multi-criteria decisionanalysis to prioritize health interventions: capitalizing on first experiences.Health Policy, 96, 262e264.

Bolsewicz Alderman, K., Hipgrave, D., & Jimenez-Soto, E. (2013). Public engagementin health priority setting in low- and middle-income countries: current trendsand considerations for policy. PLoS Medicine, 10, e1001495.

Bryce, J., Friberg, I. K., Kraushaar, D., Nsona, H., Afenyadu, G. Y., Nare, N., et al. (2010).LiST as a catalyst in program planning: experiences from Burkina Faso, Ghanaand Malawi. International Journal of Epidemiology, 39(Suppl. 1), i40e47.

Byskov, J., Bloch, P., Blystad, A., et al. (2009). Accountable priority setting for trust inhealth systems e the need for research into a new approach for strengtheningsustainable health action in developing countries. Health Research Policy andSystems, 7.

Daniels, N., Flores, W., Pannarunothai, S., Ndumbe, P., Bryant, J. H., Ngulube, T. J.,et al. (2005). An evidence-based approach to benchmarking the fairness ofhealth-sector reform in developing countries. Bulletin of the World Health Or-ganization, 83, 534e540.

Daniels, N., & Sabin, J. (1998). The ethics of accountability in managed care reform.Health Affairs (Millwood), 17, 50e64.

Department for International Development. (2010). A DfID practice paper: Writing abusiness case. London, England: United Kingdom Department for InternationalDevelopment.

Department for International Development. (2011). DfID’s approach to value formoney. London, England: United Kingdom Department for InternationalDevelopment.

Department for International Development. (2012). Somalia e 2012 humanitarianbusiness case. London, England: DFID.

Dickinson, H., Freeman, T., Robinson, S., & Williams, I. (2011). Resource scarcity andpriority-setting: from management to leadership in the rationing of healthcare? Public Money & Management, 31, 363e370.

Dionne, F., Mitton, C., Smith, N., & Donaldson, C. (2009). Evaluation of the impact ofprogram budgeting and marginal analysis in Vancouver Island Health Author-ity. Journal of Health Services Research & Policy, 14, 234e242.

Doherty, J., Kamae, I., Lee, K. K., Li, H., Li, S. C., Liu, G. G., et al. (2004). What is nextfor pharmacoeconomics and outcomes research in Asia? Value in Health, 7, 118e132.

Eddama, O., & Coast, J. (2009). Use of economic evaluation in local healthcare decision-making in England: a qualitative investigation. Health Policy, 89,261e270.

Flanagan, J., & Nicholls, P. (2007). Public sector business cases using the five casemodel: A toolkit. Bristol.

Gibson, J., Mitton, C., Martin, D., Donaldson, C., & Singer, P. (2006). Ethics andeconomics: does programme budgeting and marginal analysis contribute to fairpriority setting? Journal of Health Services Research & Policy, 11, 32e37.

Glassman, A., Chalkidou, K., Giedion, U., Teerawattanon, Y., Tunis, S., Bump, J., et al.(2012). Priority-setting institutions in health. Global Heart, 7, 13e34.

Golan, O., Hansen, P., Kaplan, G., & Tal, O. (2011). Health technology prioritization:which criteria for prioritizing new technologies and what are their relativeweights? Health Policy, 102, 126e135.

González-Pier, E., Gutiérrez-Delgado, C., Stevens, G., Barraza-Lloréns, M., Porras-Condey, R., Carvalho, N., et al. (2006). Priority setting for health in-terventions in Mexico’s System of Social Protection in Health. The Lancet,368, 1608e1618.

Government of Western Australia Department of Health. (2008). Business caseguidelines. Western Australia Department of Health.

Green, A., Ali, B., Naeem, A., & Ross, D. (2000). Resource allocation and budgetarymechanisms for decentralized health systems: experiences from Balochistan,Pakistan. Bulletin of the World Health Organization, 78, 1024e1035.

Greener, I., & Powell, J. (2003). Health authorities, priority-setting and resourceallocation: a study in decision-making in new labour’s NHS. Social Policy &Administration, 37, 35e48.

Hasman, A., & Holm, S. (2005). Accountability for reasonableness: opening the blackbox of process. Health Care Analysis, 13, 261e273.

Hazel, E., Gilroy, K., Friberg, I., Black, R. E., Bryce, J., & Jones, G. (2010). Comparingmodelled to measured mortality reductions: applying the Lives Saved Tool toevaluation data from the Accelerated Child Survival Programme in West Africa.International Journal of Epidemiology, 32e39.

Hipgrave, D., Guo, S., Mu, Y., Guo, Y., Yan, F., Scherpbier, R. W., et al. (2012). Chinese-style decentralization and health system reform. PLoS Medicine, 9, 1e4.

Hoedemaekers, R., & Dekkers, W. (2003). Key concepts in health care prioritysetting. Health Care Analysis, 11, 309e323.

Hoffmann, C., & Graf von der Schulenburg, J. M. (2000). The influence of economicevaluation studies on decision making. A European survey. The EUROMETgroup. Health Policy, 52, 179e192.

Iglesias, C. P., Drummond, M. F., & Rovira, J. (2005). Health-care decision-makingprocesses in Latin America: problems and prospects for the use of economicevaluation. International Journal of Technology Assessment in Health Care, 21,1e14.

Jansson, S. (2007). Implementing accountability for reasonableness e the case ofpharmaceutical reimbursement in Sweden. Health Economics Policy and Law, 2,153e171.

Jehu-Appiah, C., Baltussen, R., Acquah, C., Aikins, M., d’Almeida, S. A., Bosu, W. K.,et al. (2008). Balancing equity and efficiency in health priorities in Ghana: theuse of multicriteria decision analysis. Value in Health, 11, 1081e1087.

Jimenez-Soto, E., Alderman, K., Hipgrave, D., Firth, S., & Anderson, I. (2012). Priori-tization of investments in reproductive, women’s and children’s health: Evidence-based recommendations for low and middle income countries in Asia and thePacific e A subnational focus. Work commissioned by WHO Partnership forMaternal, Newborn and Child Health. Brisbane: School of Population Health,University of Queensland.

Jimenez-Soto, E., Firth, S., La Vincente, S., Morgan, A., Bauze, A., Dayal, P., et al.(2009). Developing an investment case for financing equitable progress towardsMDGs 4 and 5 in the Asia Pacific Region phase 1: Strategic report. Brisbane:University of Queensland School of Population Health.

Jimenez-Soto, E., La Vincente, S., Clark, A., Firth, S., Morgan, A., Dettrick, Z., et al.(2012). Developing and costing local strategies to improve maternal and childhealth: the investment case framework. PLoS Medicine, 9, e1001282.

Jimenez Soto, E., La Vincente, S., Clark, A., Firth, S., Morgan, A., Dettrick, Z., et al.(2013). Investment case for improving maternal and child health: results fromfour countries. BMC Public Health, 13, 601.

Kapiriri, L., & Martin, D. K. (2006). Priority setting in developing countrieshealth care institutions: the case of a Ugandan hospital. BMC Health ServicesResearch, 6.

Kapiriri, L., & Martin, D. K. (2007). A strategy to improve priority setting in devel-oping countries. Health Care Analysis, 15, 159e167.

Kapiriri, L., & Martin, D. K. (2010). Successful priority setting in low andmiddle income countries: a framework for evaluation. Health Care Analysis,18, 129e147.

Kapiriri, L., Norheim, O. F., & Heggenhougen, K. (2003). Public participation in healthplanning and priority setting at the district level in Uganda. Health Policy andPlanning, 18, 205e213.

Kapiriri, L., Norheim, O. F., & Martin, D. K. (2007). Priority setting at the micro-,meso- and macro-levels in Canada, Norway and Uganda. Health Policy, 82,78e94.

Kenny, N., & Joffres, C. (2008). An ethical analysis of international health priority-setting. Health Care Analysis, 16, 145e160.

Kraft, A., Aldaba, B., & La Vincente, S. (2011). The Philippines: Developing an invest-ment case for financing equitable progress towards MDGs 4 and 5 in the AsiaPacific Region: Scale-up report. Brisbane: School of Population Health, TheUniversity of Queensland.

La Vincente, S., Aldaba, B., Firth, S., Kraft, A., Jimenez-Soto, E., & Clark, A. (2013).Supporting local planning and budgeting for maternal, neonatal and childhealth in the Philippines. Health Research Policy and Systems, 11, 3.

Larsen, D., Friberg, I., & Eisle, T. (2011). Comparison of Lives Saved Tool model childmortality estimates against measured data from vector control studies in sub-Saharan Africa. BMC Public Health, 11, 34.

Lu, T. H., Huang, Y. T., & Chiang, T. L. (2011). Using the diamond model to prioritize30 causes of death by considering both the level of and inequality in mortality.Health Policy, 103, 63e72.

Madi, B. C., Hussein, J., Hounton, S., D’Ambruoso, L., Achadi, E., &Arhinful, D. K. (2007). Setting priorities for safe motherhood programmeevaluation: a participatory process in three developing countries. HealthPolicy, 83, 94e104.

Makundi, E., Kapiriri, L., & Norheim, O. F. (2007). Combining evidence and values inpriority setting: testing the balance sheet method in a low-income country.BMC Health Services Research. doi:15210.1186/1472-6963-7-152.

Maluka, S., Hurtig, A., Sebastian, M., Shayo, E., Byskov, J., & Kamuzora, P. (2011).Decentralization and health care prioritization process in Tanzania: from na-tional rhetoric to local reality. International Journal of Health Planning andManagement, 26, E55eE73.

Maluka, S., Kamuzora, P., San Sebastian, M., Byskov, J., Ndawi, B., & Hurtig, A. K.(2010). Improving district level health planning and priority setting in Tanzaniathrough implementing accountability for reasonableness framework: percep-tions of stakeholders. BMC Health Services Research, 10.

Maluka, S., Kamuzora, P., SanSebastian, M., Byskov, J., Ndawi, B., Olsen, O., et al.(2011). Implementing accountability for reasonableness framework at districtlevel in Tanzania: a realist evaluation. Implementation Science, 6.

Mansuri, G., & Rao, V. A. (2012). Localizing development: Does participation work?Washington, DC: The World Bank.

Martin, D., Giacomini, M., & Singer, P. (2002). Fairness, accountability for reason-ableness, and the views of priority setting decision-makers. Health Policy, 61,279e290.

Martin, D., & Singer, P. (2003). A strategy to improve priority setting in health careinstitutions. Health Care Analysis, 11, 59e68.

McDonald, J., & Ollerenshaw, A. (2011). Priority setting in primary health care: aframework for local catchments. Rural Remote Health, 11(2), 1714.

D.B. Hipgrave et al. / Social Science & Medicine 102 (2014) 190e200200

Mitton, C., Dionne, F., Damji, R., Campbell, D., & Bryan, S. (2011). Difficult decisionsin times of constraint: criteria based resource allocation in the VancouverCoastal Health Authority. BMC Health Services Research, 11.

Mitton, C., & Donaldson, C. (2002). Setting priorities in Canadian regional healthauthorities: a survey of key decision makers. Health Policy, 60, 39e58.

Mitton, C., & Donaldson, C. (2003a). Resource allocation in health care: healtheconomics and beyond. Health Care Analysis, 11, 245e257.

Mitton, C., & Donaldson, C. (2003b). Setting priorities and allocating resources inhealth regions: lessons from a project evaluating program budgeting andmarginal analysis (PBMA). Health Policy, 64, 335e348.

Mitton, C., & Donaldson, C. (2003c). Tools of the trade: a comparative analysis ofapproaches to priority setting in healthcare. Health Services ManagementResearch, 16, 96e105.

Mitton, C., & Donaldson, C. (2004). Health care priority setting: principles, practiceand challenges. Cost Effectiveness and Resource Allocation, 2.

Mitton, C., Smith, N., Peacock, S., Evoy, B., & Abelson, J. (2009). Public participationin health care priority setting: A scoping review. Health Policy, 91, 219e228.

Morgan, A., Prasai, Y., Jimenez-Soto, E., et al. (2011). Nepal: Developing an investmentcase for financing equitable progress towards MDGs 4 and 5 in the Asia PacificRegion: Scale-up report. Brisbane: School of Population Health, The University ofQueensland.

Mshana, S., Shemilu, H., Ndawi, B., Momburi, R., Olsen, O. E., Byskov, J., et al. (2007).What do district health planners in Tanzania think about improving prioritysetting using ‘accountability for reasonableness’? BMCHealth Services Research, 7.

Navarro, V. (1969). Planning for the distribution of personal health services. PublicHealth Reports, 84, 573e581.

Nelson, B. D., Dierberg, K., Scepanovic, M., Mitrovic, M., Vuksanovic, M., Milic, L.,et al. (2005). Integrating quantitative and qualitative methodologies for theassessment of health care systems: emergency medicine in post-conflict Serbia.BMC Health Services Research, 5.

Nishtar, S. (2010). The mixed health systems syndrome. Bulletin of the World HealthOrganization, 88, 74e75.

O’Meara, W. P., Tsofa, B., Molyneux, S., Goodman, C., & McKenzie, F. E. (2011).Community and facility-level engagement in planning and budgeting for thegovernment health sector e a district perspective from Kenya. Health Policy, 99,234e243.

Paalman, M., Bekedam, H., Hawken, L., & Nyheim, D. (1998). A critical review ofpriority setting in the health sector: the methodology of the 1993 WorldDevelopment Report. Health Policy and Planning, 13, 13e31.

Patten, S., Mitton, C., & Donaldson, C. (2006). Using participatory action research tobuild a priority setting process in a Canadian Regional Health Authority. SocialScience & Medicine, 63, 1121e1134.

Peacock, S., Mitton, C., Bate, A., McCoy, B., & Donaldson, C. (2009). Overcomingbarriers to priority setting using interdisciplinary methods. Health Policy, 92,124e132.

Peters, D. H., & Bloom, G. (2012). Developing world: bring order to unregulatedhealth markets. Nature, 487, 163e165.

Reichenbach, L. (2002). The politics of priority setting for reproductive health:breast and cervical cancer in Ghana. Reprod Health Matters, 10, 47e58.

Reiter, K., Kilpatrick, K. E., Greene, S. B., Lohr, K. N., & Leatherman, S. (2007). How todevelop a business case for quality. International Journal for Quality in HealthCare, 19, 50e55.

Ricca, J., Prosnitz, D., Perry, H., Edward, A., Morrow, M., Ernst, P., et al. (2011).Comparing estimates of child mortality reduction modelled in LiST withpregnancy history survey data for a community-based NGO project inMozambique. BMC Public Health, 3.

Rudan, I., Kapiriri, L., Tomlinson, M., Balliet, M., Cohen, B., & Chopra, M. (2010).Evidence-Based Priority Setting for Health Care and Research: Tools to SupportPolicy in Maternal, Neonatal, and Child Health in Africa. PLoS Medicine, 7.

Sabik, L. M., & Lie, R. K. (2008). Priority setting in health care: Lessons from theexperiences of eight countries. International Journal for Equity in Health, 7, 4.

Shepherd, J., Briggs, J., Payne, L., Packer, C., Kerridge, L., & Ashton-Key, M. (2007).Setting the future policy agenda for health technology assessment: a specialitymapping approach. International Journal of Technology Assessment in HealthCare, 23, 405e413.

Sibbald, S. L., Gibson, J. L., Singer, P. A., Upshur, R., & Martin, D. K. (2010). Evaluatingpriority setting success in healthcare: a pilot study. BMC Health ServicesResearch, 10.

Sibbald, S. L., Singer, P. A., Upshur, R., & Martin, D. K. (2009). Priority setting: whatconstitutes success? A conceptual framework for successful priority setting.BMC Health Services Research, 9.

Sinclair, S., Hagen, N. A., Chambers, C., Manns, B., Simon, A., & Browman, G. P.(2008). Accounting for reasonableness: exploring the personal internalframework affecting decisions about cancer drug funding. Health Policy, 86,381e390.

Singer, P. A., Martin, D. K., Giacomini, M., & Purdy, L. (2000). Priority setting for newtechnologies in medicine: qualitative case study. BMJ: British Medical Journal,321, 1316e1319.

Smith, N., Mitton, C., & Peacock, S. (2009). Qualitative methodologies in health-carepriority setting research. Health Economics, 18, 1163e1175.

Stafinski, T., Menon, D., Philippon, D. J., & McCabe, C. (2011). Health technologyfunding decision-making processes around the world: the same, yet different.Pharmacoeconomics, 29, 475e495.

The Partnership for Maternal Newborn and Child Health. (2012). Knowledge sum-maries #17: Civil registration and vital statistics. Geneva: World HealthOrganization.

Travis, P., & Klingen, N. (2009). Joint assessment of national health strategies andplans; combined joint assessment tool and guidelines (draft). International HealthPartnership and Inter-Agency Working Groups.

Trisnantoro, L., Widiati, Y., Kurniawan, F., et al. (2011). Indonesia: Developing aninvestment case for financing equitable progress towards MDGs 4 and 5 in the AsiaPacific Region: Scale-up report. Australia: School of Population Health, TheUniversity of Queensland.

Tsourapas, A., & Frew, E. (2011). Evaluating ‘success’ in programme budgeting andmarginal analysis: a literature review. Journal of Health Services Research &Policy, 16, 177e183.

Urquhart, B., Mitton, C., & Peacock, S. (2008). Introducing priority setting andresource allocation in home and community care programs. Journal of HealthServices Research & Policy, 13, 41e45.

Varghese, B., Dayal, P., Panicker, R., et al. (2011). Orissa: Developing an investmentcase for financing equitable progress towards MDGs 4 and 5 in the Asia PacificRegion: Scale-up report (unpublished report). Brisbane: School of PopulationHealth, The University of Queensland.

Vuorenkoski, L., Toiviainen, H., & Hemminki, E. (2008). Decision-making in prioritysetting for medicines e a review of empirical studies. Health Policy, 86, 1e9.

Waldau, S., Lindholm, L., & Wiechel, A. H. (2010). Priority setting in practice: par-ticipants opinions on vertical and horizontal priority setting for reallocation.Health Policy, 96, 245e254.

Western Australian Department Of Health. Western Australia business case guide-lines. Available from http://www.health.wa.gov.au/Publications/business_case_guidelines.pdf. Western Australia Department of Health.

Winfrey, W., McKinnon, R., & Stover, J. (2011). Methods used in the Lives Saved Tool(LiST). BMC Public Health, 11.

Wirtz, V., Cribb, A., & Barber, N. (2003). Understanding the role of “the hiddencurriculum” in resource allocation e the case of the UKNHS. Health CareAnalysis, 11, 295e300.

Yothasamut, J., Tantivess, S., & Teerawattananon, Y. (2009). Using economic evalu-ation in policy decision-making in Asian countries: mission impossible ormission probable? Value in Health, 12, S26eS30.

Youngkong, S., Kapiriri, L., & Baltussen, R. (2009). Setting priorities for health in-terventions in developing countries: a review of empirical studies. TropicalMedicine & International Health, 14, 930e939.