J Ambulatory Care Manage Vol. 32, No. 2, pp. 150–171 Copyright c 2009 Wolters Kluwer Health | Lippincott Williams & Wilkins The Impact of Primary Healthcare on Population Health in Low- and Middle-Income Countries James Macinko, PhD; Barbara Starfield, MD, MPH; Temitope Erinosho, PhD Abstract: This article assesses 36 peer-reviewed studies of the impact of primary healthcare (PHC) on health outcomes in low- and middle-income countries. Studies were abstracted and assessed according to where they took place, the research design used, target population, primary care measures, and overall conclusions. Results indicate that the bulk of evidence for PHC effective- ness is focused on infant and child health, but there is also evidence of the positive role PHC has on population health over time. Although the peer-reviewed literature is lacking in rigorous experimental studies, a small number of relatively well-designed observational studies and the con- sistency of findings generally support the contention that an integrated approach to primary care can improve health. A few large-scale experiences also help identify elements of good practice. The review concludes with several recommendations for future studies, including a focus on better con- ceptualizing and measuring PHC, further investigation into the advantages of comprehensive over selective PHC, need for experimental or quasi-experimental research designs that allow testing of the independent effect of primary care on outcomes over time, and a more detailed conceptual framework guiding overall evaluation design that places limits on the parameters under consider- ation and describes relationships among different levels and types of data likely to be collected in the evaluation process. Key words: developing country, evaluation, population health, primary healthcare T HE WORLD HEALTH ORGANIZATION (WHO) formalized its commitment to pri- mary healthcare (PHC) in 1978, when it was identified as central to the achievement of the goal of “Health for All”and as a key instrument for improving health throughout the world (WHO, 1978). In the decades following Alma Ata, many low- and middle-income countries have un- Author Affiliations: Department of Nutrition, Food Studies, and Public Health, New York University, New York (Drs Macinko and Erinosho); and Department of Health Policy and Management, The Johns Hopkins Medical Institutions, Baltimore, Maryland (Dr Starfield). Dr Erinosho is now with Health Promotion Research Branch, National Cancer Institute, Rockville, Maryland. Corresponding Author: James Macinko, PhD, Depart- ment of Nutrition, Food Studies, and Public Health, New York University, 35 W 4th St, 12th Floor, New York, NY 10012 ([email protected]). dergone dramatic changes, including democ- ratization, economic liberalization in an in- creasingly globalized world, redefining the role of the state, and reforming their health and social services systems. Health reforms, in particular, have aimed at streamlining health- care financing and decentralizing authority for planning and implementation. There is increasing evidence that not all of these re- forms have strengthened PHC, nor have they uniformly contributed to improving health or equity in its distribution (Infante & de Mata, 2000; Mackintosh, 2000; Varatharajan & Thankappan, 2004). In many high-income countries, various at- tributes of primary care have been shown to exert a positive influence on health costs, appropriateness of care, and outcomes for most of the major health indicators (Bindman et al., 1996; Forrest & Starfield, 1996, 1998; Starfield, 1998; Starfield et al., 2005a, 2005b). 150
22
Embed
Macinko J, Starfield B, Erinosho T. The impact of primary health care
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
The Impact of PrimaryHealthcare on PopulationHealth in Low- andMiddle-Income Countries
James Macinko, PhD; Barbara Starfield, MD, MPH;Temitope Erinosho, PhD
Abstract: This article assesses 36 peer-reviewed studies of the impact of primary healthcare (PHC)on health outcomes in low- and middle-income countries. Studies were abstracted and assessedaccording to where they took place, the research design used, target population, primary caremeasures, and overall conclusions. Results indicate that the bulk of evidence for PHC effective-ness is focused on infant and child health, but there is also evidence of the positive role PHChas on population health over time. Although the peer-reviewed literature is lacking in rigorousexperimental studies, a small number of relatively well-designed observational studies and the con-sistency of findings generally support the contention that an integrated approach to primary carecan improve health. A few large-scale experiences also help identify elements of good practice. Thereview concludes with several recommendations for future studies, including a focus on better con-ceptualizing and measuring PHC, further investigation into the advantages of comprehensive overselective PHC, need for experimental or quasi-experimental research designs that allow testing ofthe independent effect of primary care on outcomes over time, and a more detailed conceptualframework guiding overall evaluation design that places limits on the parameters under consider-ation and describes relationships among different levels and types of data likely to be collected inthe evaluation process. Key words: developing country, evaluation, population health, primaryhealthcare
THE WORLD HEALTH ORGANIZATION(WHO) formalized its commitment to pri-
mary healthcare (PHC) in 1978, when it wasidentified as central to the achievement of thegoal of “Health for All”and as a key instrumentfor improving health throughout the world(WHO, 1978).
In the decades following Alma Ata, manylow- and middle-income countries have un-
Author Affiliations: Department of Nutrition, FoodStudies, and Public Health, New York University,New York (Drs Macinko and Erinosho); andDepartment of Health Policy and Management, TheJohns Hopkins Medical Institutions, Baltimore,Maryland (Dr Starfield). Dr Erinosho is now withHealth Promotion Research Branch, NationalCancer Institute, Rockville, Maryland.
Corresponding Author: James Macinko, PhD, Depart-ment of Nutrition, Food Studies, and Public Health, NewYork University, 35 W 4th St, 12th Floor, New York, NY10012 ([email protected]).
dergone dramatic changes, including democ-ratization, economic liberalization in an in-creasingly globalized world, redefining therole of the state, and reforming their healthand social services systems. Health reforms, inparticular, have aimed at streamlining health-care financing and decentralizing authorityfor planning and implementation. There isincreasing evidence that not all of these re-forms have strengthened PHC, nor have theyuniformly contributed to improving healthor equity in its distribution (Infante & deMata, 2000; Mackintosh, 2000; Varatharajan &Thankappan, 2004).
In many high-income countries, various at-tributes of primary care have been shownto exert a positive influence on health costs,appropriateness of care, and outcomes formost of the major health indicators (Bindmanet al., 1996; Forrest & Starfield, 1996, 1998;Starfield, 1998; Starfield et al., 2005a, 2005b).
150
Primary Healthcare in Low- and Middle-Income Countries 151
There is also evidence that countries charac-terized by a strong primary care orientationhave better and more equitable health out-comes than those systems that are orientedtoward specialty care (Macinko et al., 2003;Starfield, 1996; Starfield & Shi, 2002). Nev-ertheless, there is considerable debate abouthow effective PHC has been in improvingpopulation health in low- and middle-incomecountries (Filmer et al., 2000; Lewis et al.,2004).
The 30-year anniversary of the Alma Atameeting, the changing health challenges inthe developing world, and the widespreaddissatisfaction with the status quo have gen-erated interest in a renewed and reinvigo-rated approach to health systems develop-ment based on PHC (“Margaret Chan putsprimary health care,” 2008; Pan AmericanHealth Organization, 2005; WHO, 2008). Toaid in this process, the present review as-sesses the peer-reviewed literature for evi-dence of the effectiveness of previous PHCexperiences with the goal of identifyinglessons learned and providing suggestions forstrengthening the PHC evidence base.
METHODS
The literature review was conductedby searching the US National Library ofMedicine’s PubMed database, the CochraneDatabase of Systematic Reviews, and the In-ternet (via Google) for articles that containedthe phrases “primary care”or “primary healthcare” along with the terms “evaluation” or“impact” in either the title or the abstract.Several journals that publish on healthcarein the developing world were also handsearched. All articles were then culled toidentify additional references. This processrevealed more than 10 000 potential articlesas of July 2008.
From the large potential pool of articles,we excluded all commentaries and non–peer-reviewed works and all articles related to Eu-ropean or other Organization for EconomicCooperation and Development countries. Ab-stracts and study designs were then reviewedto identify articles that addressed the evalua-
tion of PHC programs, systems, and servicesand to exclude articles that (1) did not explic-itly define the scope of the PHC intervention;(2) evaluated only one component of selectiveprimary care services (eg, immunization, oralrehydration therapy); or (3) did not includedata on changes in health outcomes attributedto the PHC intervention. Overall, 36 key arti-cles were retrieved and abstracted.
The Appendix contains a synthesis of themain objectives, study designs, outcomes,PHC measures, and results of the reviewed ar-ticles. We adopt the term “selective”to charac-terize interventions directed at selected indi-vidual health conditions (such as control of di-arrheal diseases) and “integrated” to describeapproaches that are more directed at health ingeneral. In the presentation of results, we dis-tinguish between PHC tasks or services (ie, di-rected at a specific health problem) and PHCfunctions (ie, directed at assuming the mainrole of PHC within health systems, regardlessof the specific health problem).
RESULTS
Figure 1 shows the distribution of new ar-ticles by year on the topic of PHC, which hasincreased each year and, after a relatively sta-ble period from 1995 to 2003, now averagesabout 500 new articles per year.
Table 1 shows characteristics of the 36abstracted studies. Geographically, they arefairly evenly distributed: slightly more than athird are from Africa, about a third are fromLatin America and the Caribbean, a quarterfrom Asia, with the remaining representingmultiple regions. In terms of study design,most (45%) use a pre- and postinterventioncross-sectional design with controls or com-parison groups, about 14% use a case-controldesign, another 11% use multivariable longitu-dinal analyses of ecological data, 1 study usesan experimental design, and 1 uses a cohortapproach. All remaining studies employeda variety of observational designs withoutcontrols.
In terms of outcomes, more than three-quarters of the studies focused on infant orunder-5 mortality, with the remainder dealing
152 JOURNAL OF AMBULATORY CARE MANAGEMENT/APRIL–JUNE 2009
Figure 1. The number of new articles published in PubMed with “primary health care” in the abstract ortitle, by year (1975–2007).
Table 1. Characteristics of studies reviewed(N = 36)
Number
Domain of studies
Geographic region
Latin America and the
Caribbean Sub-Saharan
12
Africa 16
Asia 8
Other (or multiple regions) 1
Study design
Experimental or
quasi-experimental
5
Prospective study with
control group
1
Repeated (pre/post)
cross-sectional design
with control
16
Case-control 5
Repeated (pre/post)
cross-sectional design
without control
5
Systematic literature review 1
Observation/qualitative
study/single cross-section
3
Main outcome studied
Infant or under-5 mortality 28
Other (child) 1
Other (adult) 7
with maternal mortality, life expectancy, all-cause mortality, and cause-specific mortalityin adults. All but 3 studies measured PHC ex-posure by residence in a geographic area inwhich PHC services were being delivered.The other 3 assessed individual use of spe-cific PHC services. Nearly all studies point to apositive impact of the PHC intervention stud-ied: only 5 articles show no improvement at-tributable to PHC.
The magnitude of impact also varied con-siderably. Reductions in infant and under-5mortality attributed to PHC averaged morethan 40% and varied from 0 to as high as71%, with interventions lasting from 2 yearsto more than 10 years.
Studies on specific PHC tasks
Several studies analyzed the association ofspecific primary care tasks with health out-comes. Moore et al. (2003) conducted a panelstudy of 22 Latin American countries overthe period from 1990 to 1998. The studyfound that the most important contributorto lower under-5 mortality was women’s lit-eracy, followed by vaccination coverage anduse of oral rehydration therapy. A similar anal-ysis conducted by using Demographic andHealth Surveys from 5 East African countriesestimated that nearly three-quarters of the
Primary Healthcare in Low- and Middle-Income Countries 153
attributable risks for mortality in childrenyounger than 1 year might be amenable topreventive services, including antenatal care,immunizations, fertility regulation, and use ofpotable water (Brockerhoff & Derose, 1996).In both studies, the extent to which these ser-vices were part of an integrated PHC systemor the result of an effort targeted only at majorcauses of infant and child mortality is unclear.
Dugbatey (1999) assessed the relationshipbetween a set of “Health for All” policies(health education, nutrition, water and san-itation, and maternal/child health services)and health outcomes at the national level in4 African countries in the 1990s. Througha comparative case study design, the authorshowed that PHC-sensitive conditions (suchas infant mortality) were improved in the2 countries with more comprehensive PHCpolicies (Botswana and Zimbabwe), as op-posed to those with a less coherent set of PHCpolicies (Ghana and Cote d’Ivoire), in spite ofthe latter having higher gross national productper capita (Dugbatey, 1999).
The Bellagio Child Survival Study Groupconcluded that nearly 10 million child deathsworldwide could be averted by tasks or ser-vices including combined use of oral rehy-dration therapy (Victora et al., 2000), immu-nization (England et al., 2001), micronutrientsupplementation, promotion of exclusivebreast-feeding (Arifeen et al., 2001), and oth-ers, all but one of which (neonatal intensivecare) would be expected to be delivered bya PHC system. This estimate is supported byanother study (Berman, 2000) that estimatedthat about 62% of all disability adjusted lifeyears (lost in the adult and child populationof developing countries) would be amenableto primary care services (termed “ambulatoryhealthcare” by Berman). Access to primarycare appears to be particularly important inAfrica; some authors suggest that up to 80%of child deaths occur at home, without thechild having any contact with the health sys-tem (Oluwole et al., 2000)
Integrated management of childhood ill-ness (IMCI) reflects a horizontal primary careapproach in the sense that it combines sev-eral specific interventions. An evaluation of
IMCI programs in Brazil, Peru, Uganda, Egypt,and Tanzania showed that although the ap-proach was in many cases more compre-hensive and effective than individual verticalinterventions, poor access, low levels of uti-lization, and structural weaknesses in healthsystems limit its impact on population health(Bryce et al., 2003; WHO, 2004) A recent re-view suggested that a more comprehensiveapproach to PHC and health systems devel-opment will be required for strategies such asthe IMCI program to flourish (Freedman et al.,2005).
Rohde et al. (2008) identify 13 coun-tries that have implemented comprehen-sive PHC (Thailand, Turkey, Vietnam, Brazil,Sri Lanka, El Salvador, Tunisia, DominicanRepublic, Iran, Kazakhstan, Turkmenistan,China, and Cuba). Their analysis suggests thatthese countries experienced important healthgains and that in comparison with coun-tries having more selective PHC approaches,health improvements—particularly for condi-tions that require sustained and coordinatedcare—seem to “depend on progression tocomprehensive primary health care with a re-liable referral system linking to functioning fa-cilities” (Rohde et al., 2008, p. 958).
Studies of specific primary
care programs
The Navrongo experiment in Ghana wasthe only experimental study identified. In it,villages received 1 of 4 different interven-tions: professional community nurses; vol-untary community health workers (CHWs);a combination of both; and nothing (con-trol). In the nurse-only intervention areas,under-5 mortality fell by 14% during 5 yearsof program implementation, compared withthat before the intervention period (Penceet al., 2007; Phillips et al., 2006). In thevolunteer-only villages, under-5 mortality in-creased by 14%. The professional nurse inter-vention added approximately $2 per capita tothe $6.80 per capita budgeted for PHC ser-vices. Note that the study used a “plausibil-ity” rather than a “probability” design, mean-ing that treatments were not truly randomlyassigned.
154 JOURNAL OF AMBULATORY CARE MANAGEMENT/APRIL–JUNE 2009
In the Gambia, Hill et al. (2000) comparedPHC and non-PHC communities over a 15-year period. Primary care was assessed by thepresence of a community health nurse as asupervisor to village health workers (VHWs)and traditional birth attendants; presence ofan expanded program of immunization and abasic package of maternal and child health ser-vices; and community participation in health-care provision, priority setting, and fund rais-ing. The study found that although child mor-tality declined in intervention and controlvillages, the decline was generally steeperin PHC villages (Hill et al., 2000). As a typeof natural experiment, the authors reportthat once PHC services were stopped in thevillages (because of lack of funds), the trendreversed and infant mortality increased tolevels higher than those in control (non-PHC)villages.
Velema et al. (1991) studied access and lon-gitudinality of primary care services to a pop-ulation of about 13 000 in Benin. In a matchedcase-control study, 2 factors were partic-ularly influential in predicting death risk:measles vaccination before their first birthday(Odds ratio [OR] = 0.4) and regular con-tact with VHWs (OR = 0.36). The authorsconcluded that regular contact with VHWs,which is consistent with the person-focusedcare over time function of primary care, im-proves the likelihood of child survival.
In Haiti, the activities of an integrated lo-cal health system, based on a PHC model(the Hopital Albert Schweitzer or HAS), wereassociated with infant and under-5 mortalitythat are about half of those in other areaswith similar income levels (Perry et al., 2007).This was accomplished through a decades-long partnership with local communities. Interms of resources, HAS had fewer physiciansand fewer hospital beds per capita than didthe rest of Haiti but more nurses, CHWs, andother outreach and support staff. The HASsystem costs about $19 per capita, includ-ing community development initiatives (Perryet al., 2006).
Other studies of PHC programs using pre-and postintervention measures and control orcomparison groups include the following:
• A large NGO-delivered PHC program(focused on maternal and child healthservices) targeted about 340 000 poorhouseholds in Bangladesh. Services weredelivered through trained family healthvisitors and included regular householdvisits, illustrating the importance of a fam-ily focus in PHC. After 5 to 6 years,program areas experienced a 52% reduc-tion in infant mortality and a 49% re-duction in under-5 mortality, larger de-creases than those experienced in controlareas (Mercer et al., 2004). Another studyin Matlab, Bangladesh, showed the im-pact of a community-based PHC approachemploying supervised and trained VHWs(with referrals to health centers staffedby healthcare professionals) on loweringunder-5 mortality from acute lower respi-ratory tract infections by 32% in 2 years(Fauveau et al., 1992). Prior studies inMatlab had also documented reductionsin under-5 and maternal mortality due todifferent PHC interventions (Chen et al.,1983; Fauveau et al., 1991).
• In a cohort study conducted inPondicherry, India, provision of a broadrange of PHC services, including homevisits by PHC nurses in 12 villages (totalpopulation of about 16 000), decreasedinfant and child mortality by more than65% (Dutt & Srinivasa, 1997). Anothersmaller-scale Indian study assessed theeffects of VHW provision of primaryand maternity care and health educationto pregnant women and grandmothers,resulting in reductions in neonatal mor-tality by 62% and infant mortality by 71%,as compared with preintervention levels(Bang et al., 1999).
• In Liberia (Becker et al., 1993) and Zaire(Chahnazarian et al., 1993; Taylor et al.,1993), a more selective PHC approachwas attributed to reductions in under-5mortality by as much as 28% over a 5-yearperiod, an improvement that was signifi-cantly greater than that reported in com-parison areas.
• A study using 2 waves of nationally repre-sentative surveys in Indonesia found that,
Primary Healthcare in Low- and Middle-Income Countries 155
while holding other village- and maternal-level variables constant, the addition of amaternity clinic and a physician to a vil-lage was found to decrease the odds ofinfant death (relative to an infant born be-fore the clinic existed) by about 15% and1.7%, respectively (Frankenberg, 1995).
• In Bolivia, a comprehensive community-based PHC program (delivered by paidnurses and community volunteers withsome physician support) serving a pop-ulation of about 15 000 successfully re-duced under-5 mortality by more than52% over a 5- to 6-year period, as com-pared with control areas (Perry et al.,1998, 2003). Costs for the program wereestimated at about $10 per person.
• In Pakistan, a case-control study of chil-dren who had diarrhea or acute respira-tory tract infections showed that the useof a traditional healer (as opposed to atrained VHW) raised the odds of a child’sdeath by a factor of 14 (OR = 14.5; 95%confidence interval [CI] = 4.23–49.8),and frequent changing of providers (ie,lack of continuity with a PHC provider)raised the odds of death 8 times (OR = 8;95% CI = 2.22–28.8) (D’Souza & Bryant,1999).
Studies of countrywide PHC experiences
There have been only a few studies that di-rectly test the hypothesis that health systemsbased on a strong PHC orientation (basedon PHC principles) lead to better overallindicators.
By 1985, Costa Rica’s life expectancy hadreached 74 years, and infant mortality de-clined from 60 per 1000 in 1970 to 19 per1000, levels comparable with those in moredeveloped countries. Explanations for thisrapid progress include the development of auniversal social security system and a mul-tidimensional approach to health improve-ment, which included expanding PHC ser-vices, investing in education and sanitation,and improving access to secondary and ter-tiary healthcare services (Haines & Avery,1982; Klijzing & Taylor, 1982; Rosero-Bixby,
1986). PHC improvements beginning in the1970s were estimated to have reduced in-fant mortality by between 40% and 75%(Rosero-Bixby, 1991).
In the 1990s, additional Costa Ricanreforms sought to improve PHC coverageand efficiency. A quasi-experimental studycomparing 3 groups of districts on the basisof when they adopted PHC reforms foundthat, in addition to other improvements inliving standards, PHC reforms significantlyreduced mortality in both adults and children(Rosero-Bixby, 2004a). For every 5 additionalyears after PHC reforms, child mortalitydeclined by 13% and adult mortality by4%. The proportion of the population withinsufficient access to PHC services declinedby 15% in reformed districts compared withonly a 2% decline in districts that did notundergo reforms. The reforms additionally im-proved equity in access by targeting the leastprivileged population first (Rosero-Bixby,2004b).
Brazil’s family health program (FHP) is nowperhaps the largest community-based PHCsystem in the world. In 2007, the programencompassed more than 27 000 community-based teams responsible for providing careto about 85 million people (Brazilian Min-istry of Health, Department of Primary Care,2006). The FHP is based on an explicit strat-egy to provide all core primary care functions,including first-contact access for each newneed, long-term person-focused care, compre-hensive care for most health needs, coordi-nated care when it must be sought elsewhere,and a focus on the family and the community.These functions are achieved through theprogram’s decentralized organization, elimi-nation of copays for services, incentives to lo-cal government for increasing access to theprogram, and multidisciplinary teams com-posed of a physician and a nurse who deliverclinic-based care along with CHWs who makeregular home visits and perform community-based health-promotion activities (Ministry ofHealth of Brazil, 2003). Costs for the pro-gram (which includes access to pharmaceu-ticals) are estimated at between $25 and $35(Macinko et al., 2007). A panel data analysis
156 JOURNAL OF AMBULATORY CARE MANAGEMENT/APRIL–JUNE 2009
of Brazilian states from 1990 to 2002 showedthat an increase in FHP coverage by 10% wasassociated with an average 4.6% decrease ininfant mortality, controlling for other healthdeterminants, including water supply, livingconditions, doctor and hospital supply, andwomen’s education (Macinko et al., 2006).A follow-up study conducted at the microre-gional level for 1999–2004 showed that the ef-fect of FHP coverage was especially strong forconditions that are known to be sensitive toprimary care (such as postneonatal mortalityand deaths from diarrheal diseases) (Macinkoet al., 2007). Several studies also demon-strated associations between CHWs and lowerinfant mortality in specific Brazilian states(Emond et al., 2002; Svitone et al., 2000).
A few other countrywide observations aresuggestive of the role of PHC in populationhealth improvements, although these studiesdo not explicitly quantify the contribution ofPHC to health improvements or explicitly testthe impact of specific PHC interventions.
Cuba’s universal PHC program uses familyhealth physicians and nurses, who provideuniversal, comprehensive, integrated, and in-tersectorial care to geographically defined ar-eas with a focus on families (Evans et al.,2008; Waitzkin, 1997). Changes in PHC ac-cess, organization, and delivery over the past40 years correspond to about a 40% declinein infant mortality over the same period, evenwhile other indicators such as gross nationalproduct per capita have not substantially in-creased (Riveron Corteguera, 2000). Invest-ments in prevention integrated into PHC mayalso have contributed to the control of car-diovascular diseases, resulting in lower-than-expected mortality and fewer avoidable hos-pitalizations for these and related conditions(Spiegel & Yassi, 2004). Lessons learned fromthe Cuban experience suggest the potentialbenefits of organizing an entire health sys-tem around the PHC approach (Franco et al.,2007).
In Mexico, child mortality declined from 64per 1000 live births in 1980 to 23 per 1000in 2006 (Sepulveda et al., 2006). These re-ductions were consequent to a strategy thatbegan with a number of disease-specific pro-
grams and expanded to a broader strategy thatcombined vertical programs with more com-prehensive PHC and human development ap-proaches, including legislation making accessto maternal and child health services a citi-zen’s right (Frenk et al., 2003). Reyes et al.(1997) also found that in Mexico, primarycare characteristics (such as adequate refer-ral processes, continuity of care, being seenby the same provider, and being attended in apublic facility) had an important, independenteffect on reducing a child’s odds of dying.Similarly, Gutierrez et al. (1999) point to theimportance of access to primary care (as mea-sured by nurse and physician supply) as wellas investments in public health (immunizationand improved water and sanitation) and edu-cation as particularly important for reducinginfant mortality in Mexico.
In Thailand, decreases in under-5 infantmortality occurred after primary care reform,which included developing at least 1 pri-mary care health center for every rural villageby 1990 and a government medical welfarescheme started in 1993. In the correspond-ing decade, under-5 mortality declined by 44%in the poorest population quintile, 41% in thenext poorest quintile, 22% in the third, 23% inthe fourth, and 13% in the wealthiest quintile(Vapattanawong et al., 2007)
In Indonesia, a 20% reduction in infant mor-tality during the early 1990s has been at-tributed to improvements in PHC (Simms &Rowson, 2003). Some evidence for this attri-bution comes from the observation that inthe later 1990s, once primary care spend-ing declined substantially (and hospital spend-ing increased by almost 25%), infant mortal-ity actually increased by 14% in almost everyprovince of the country (Simms & Rowson,2003).
Finally, the 2008 World Health Report onPHC presents numerous case studies of PHCexperiences. Although it does not contain asystematic review of the evidence on the ben-efits of PHC, it reviewed the evidence for thebenefits of PHC components and concludedthat there is an overwhelming justification fora focus on developing and strengthening PHCin all countries (WHO, 2008).
Primary Healthcare in Low- and Middle-Income Countries 157
Studies finding little or no impact of
PHC on health outcomes
In Niger, a prospective study found thatthere was no additional survival advantagefor children in villages with a “village healthteam”present, although the presence of a dis-pensary lowered the odds of death by 32%,as compared with villages with no services(Magnani et al., 1996). In explanation for theapparent lack of an effect, Magnani et al. sug-gest the need for more comprehensive pack-ages of health services than those delivered bythe project, because the single interventionswere possibly offset by continued high levelsof exposure to other unchanged factors.
In the Philippines, the Bohol project pro-vided very low-cost PHC services to a popula-tion of about 400 000 residents for 5 years.The evaluation included pre- and postin-tervention surveys and comparison with acontrol village. The project increased theutilization of some health services but didnot significantly decrease infant mortality(Williamson, 1982). Williamson suggests thatpotential reasons for the lack of an effect in-clude the generally poor quality of health ser-vices provided, a selective rather than a com-prehensive approach to PHC with a strongemphasis on family planning (fertility did de-cline), and overworked and/or inadequatelytrained staff.
In a retrospective study with control com-munities in the Gambia (De Francisco et al.,1994), there were no significant differences(P = .88) in under-5 mortality between vil-lages with VHWs and those without them(35.5/1000 vs 35.8/1000). De Francisco et al.suggest that different health service utiliza-tion patterns (based on the type of child ill-nesses) and preferences for traditional heal-ers may partially explain the lack of effect. Inaddition, there was no indication that theseVHWs were achieving PHC functions, includ-ing provision of good-quality care and refer-ral to trained healthcare professionals whenindicated.
In Brazil, one study found that participa-tion in FHP between 1994 and 1998 did notsignificantly improve child health indicatorsin municipalities with high coverage, as op-
posed to those in municipalities with lowor no coverage. Infant mortality declined by42% and 45.5% in the intervention and con-trol groups, respectively, a nonsignificant dif-ference (Morsch et al., 2001). A possible ex-planation for the lack of an effect might havebeen the inability to control for variables re-lated to the performance of primary care ser-vices, such as the technical quality of care oraccessibility, which vary by municipality.
Finally, in their systematic review of “in-tegration” of primary care in developingcountries, Briggs et al. (2001) discuss an es-sential feature of primary care: the extent towhich it provides a range of services meant toattend to most common healthcare problems.This feature of primary care is more oftentermed “comprehensiveness.” The review ofBriggs et al. (2001) contains only 4 studies,and they conclude that no overall conclusionscan be drawn from their results. As a possibleexplanation for the lack of a conclusivefinding, the authors point to the poor qualityof many of the studies conducted, includingpoor recording of outcomes, inadequaterandomization processes, and control groupsthat were not entirely comparable with exper-imental groups. Moreover, each study definedand measured integration in a differentway.
DISCUSSION
This review of the evidence of the ef-fectiveness of PHC on population health inlow-income countries has shown that sev-eral analyses provide consistent evidenceof the impact of PHC on improved healthoutcomes. Nevertheless, many studies suf-fered from important methodological weak-nesses, including inadequate controls forindividual- or community-level confoundersin multivariable analyses. Reductions in in-fant mortality (the most frequently studiedoutcome) attributed to PHC actions averagedabout 40% and varied from 0% to as highas a 71% over intervention periods rangingbetween 2 and 10 or more years. Costs forcomprehensive PHC programs ranged fromabout $10 to $35 per capita per year.
158 JOURNAL OF AMBULATORY CARE MANAGEMENT/APRIL–JUNE 2009
Despite the apparent consistency of results,analysis of the studies revealed that PHC hasonly rarely been evaluated in a consistent andreproducible way. Rather, it is often only var-ious aspects of health services that are as-sumed to be part of PHC that have been the fo-cus of attention. For example, all but 3 studiesmeasured PHC exposure as residence withina geographic area in which the PHC programor project was implemented. In addition, thedefinition of the PHC program also varied con-siderably, from the mere presence of a VHWin a community to the use of specific ser-vices to the development of an integratednetwork of health and social services in thecommunity. For this reason, there is little thatcan be gleaned regarding the mechanisms bywhich these PHC approaches might achieveimportant primary care functions, such asfirst-contact access, longitudinality, compre-hensiveness, and coordination of care.
The general failure to use an operationalconceptualization of PHC has also made iden-tification of studies about PHC difficult. Forthis reason, the literature probably containsmore evidence than is discoverable from theabstracts or titles of published articles.
Publication bias is also likely to have lim-ited the scope of this review. Many suc-cessful (as well as unsuccessful) experienceshave simply not been documented in peer-reviewed journals. As an illustration, we wereunable to retrieve any peer-reviewed articlesthat adequately assessed the impact of PHCon population health in Sri Lanka, China, orVietnam, although each country’s PHC ap-proach has been discussed elsewhere as “suc-cessful”(Bloom, 1998; Fritzen, 2007; Halsteadet al., 1985).
In addition, there is little peer-reviewed evi-dence on the role of PHC on improvements inadult health in low- and middle-income coun-tries, because most published studies have fo-cused only on infant and under-5 mortality.Thus, the potential for PHC to help controladult chronic and infectious diseases in the de-veloping world remains largely unexplored.
The studies that found no effect of PHC onhealth indirectly provide support for a com-prehensive approach to PHC: most involvedinterventions that focused only on selective
PHC tasks. They also point to the importanceof accurately measuring variations in the tech-nical quality of primary care delivered, a topicthat certainly deserves far more attention inthe literature reviewed here.
In view of the limitations of these studies,an agenda for the evaluation of the contribu-tion of PHC tasks and functions to populationhealth would benefit from the following con-siderations.
First, a clear conceptualization of primarycare is needed, including specification ofeach of its component features, for example,first-contact access and use, longitudinality(person-focused care over time), comprehen-siveness (addressing the breadth of commonhealth needs), and coordination (integrationof services with other levels of care).
Second, studies should start with a con-ceptual framework to guide the overall eval-uation, design the characteristics under con-sideration, and describe relationships amongdifferent levels and types of data to be col-lected in the evaluation process. This frame-work should provide a model of how primarycare is conceptualized in relation to biologi-cal, social, and environmental influences onhealth (Starfield, 2001).
Third, future studies require clear speci-fication and measurement of the PHC sys-tem, including specific structural character-istics (input and policy), process (servicedelivery modalities), and relevant healthoutcomes and outputs.
Fourth, as noted throughout the PHC lit-erature (Chen et al., 1993; Hill et al., 2000),there is still an urgent need for more rig-orous research designs that allow testing ofthe independent effect of primary care onoutcomes over time. This should includeindividual- and community/contextual-leveldata derived from longitudinal sources, appro-priate control or comparison groups, and con-trol for relevant individual- and contextual-level covariates. Such evaluations will requirea commitment from donor organizations andnational governments to provide necessary re-sources and to ensure the scientific integrityof the research process.
In the short term, 3 approaches couldbe implemented to aid in providing more
Primary Healthcare in Low- and Middle-Income Countries 159
systematic evaluation of primary care, asfollows:
1. Existing or planned cohort studies couldbegin to incorporate PHC measuresthrough the use of validated instrumentssuch as the Primary Care AssessmentTools (Harzheim et al., 2006; Macinkoet al., 2007; Pasarin et al., 2007).
2. Standardized surveys such as the Demo-graphic and Health Surveys or LivingStandards Measurement Surveys couldinclude modules derived from the Pri-mary Care Assessment Tools along withhealth system variables to identify howand where populations are receiving ef-fective PHC services.
3. Researchers could direct their attentionto countries that are currently undergo-ing reform of their primary care system,thus opening the possibility for analy-sis of natural experiments in which re-formed states or municipalities couldbe compared with otherwise similar re-gions without reformed primary caresystems. Better yet, experimental assign-ment of different PHC approaches couldbe used to help phase in reforms andmore rigorously evaluate their impact(King et al., 2007).
Finally, there is a need to encourage thepublication of evaluations of PHC experi-ences, both successful and unsuccessful, sothat the PHC approach can be guided by awider body of high-quality evidence.
CONCLUSION
The WHO proposal for renewing PHC re-inforces the idea that strengthened health sys-tems should be viewed as a necessary (thoughnot sufficient) condition for meeting interna-tionally agreed-upon development goals suchas those contained in the Millennium Devel-
opment Goals (WHO, 2008). Basing healthsystems more strongly on PHC represents animportant strategy to address emerging healthproblems (Fuster & Voute, 2005), scale upexisting interventions, and effectively com-bat health threats such as HIV/AIDS (Buveet al., 2003), tuberculosis (Mahendradhataet al., 2003), chronic illnesses (Rothman &Wagner, 2003), and others.
These observations are also relevant to therenewal of primary care in the United States.Recently, the American Academy of FamilyPhysicians, the American Academy of Pedi-atrics, the American College of Physicians,and the American Osteopathic Association(2007) united to endorse the “Joint Princi-ples of the Patient-Centered Medical Home,”which describes characteristics of a patient-centered medical home (PCMH) as includinga personal physician, physician-directed med-ical practice, whole-person orientation, coor-dinated and integrated care, quality and safety,enhanced access to care, and payment that“appropriately recognizes the added valueprovided to patients. . ..”Lessons learned fromthe evaluation of PHC in the developing worldmay also have relevance to the assessment ofthe PCMH, as it is apparent that definitionsand tools of measurement should be consis-tent, standardized, and based on evidence ofeffectiveness of primary care components.Without greater attention to these aspects, thePCMH model may fall short of reaching itsgoal of renewing a PHC approach to health-care organization and delivery in the UnitedStates.
As national governments, the WHO, andother international organizations move to re-new their PHC strategies, greater clarity inspecifying PHC in terms that allow for morestandardized measurement and investment inrigorous evaluation of PHC effectiveness andits effects on equity will be essential.
REFERENCES
Afari, E. A., Nkrumah, F. K., Nakana, T., Sakatoku, H., Hori,
H., & Binka, F. (1995). Impact of primary health care
on childhood and mortality in rural Ghana: The Gomoa
experience. Central African Journal of Medicine,41(5), 148–153.
American Academy of Family Physicians, American
160 JOURNAL OF AMBULATORY CARE MANAGEMENT/APRIL–JUNE 2009
Academy of Pediatrics, American College of Physi-
cians, & American Osteopathic Association. (2007).
Joint principles of the patient-centered medicalhome. Patient Centered Primary Care Collabora-
tive. Retrieved November 5, 2008, from http://www.
medicalhomeinfo.org/joint%20Statement.pdf
Arifeen, S., Black, R. E., Antelman, G., Baqui, A., Caulfield,
L., & Becker, S. (2001). Exclusive breastfeeding re-
duces acute respiratory infection and diarrhea deaths
among infants in Dhaka slums. Pediatrics, 108, E67.
Berman, P. (2000). Organization of ambulatory health
care provision in developing countries. A critical de-
terminant of health system performance in developing
countries) Bulletin of the World Health Organization,78(6), 791–802.
Bang, A. T., Bang, R. A., Baitule, S. B., Reddy, M. H., &
Deshmukh, M. D. (1999). Effect of home-based neona-
tal care and management of sepsis on neonatal mor-
tality: Field trial in rural India. The Lancet, 354(9194),
1955–1961.
Bang, A. T., Reddy, H. M., Deshmukh, M. D., Baitule, S.
B., & Bang R. A. (2005). Neonatal and infant mortality
in the ten years (1993 to 2003) of the Gadchiroli field
P. (2001). The evidence base for interventions to re-duce mortality from vaccine-preventable diseases inlow and middle-income countries (CMH Working Pa-
per WG5:10). Commission on Macroeconomics and
Primary Healthcare in Low- and Middle-Income Countries 161
Health. Retrieved January 3, 2004, from http://www.
cmhealth.org/docs/wg5 paper10.pdf
Evans, R. (2008). Thomas McKeown, Meet Fidel Castro:
Physicians, population health and the Cuban paradox.
Healthcare Policy, 3(4), 21–32.
Fauveau, V., Stewart, M. K., Chakraborty, J., & Khan, S.
A. (1992). Impact on mortality of a community-based
programme to control acute lower respiratory tract in-
fections. Bulletin of the World Health Organization,70(1), 109–116.
Fauveau, V., Stewart, M. K., Khan, S. A., & Chakraborty,
J. (1991). Effect on mortality of community-based
maternity-care programme in rural Bangladesh. TheLancet, 338(8776), 1183–1186.
Filmer, D., Hammer, J., & Pritchett, L. (2000). Weak links
in the chain: A diagnosis of health policy in poor coun-
tries. The World Bank Research Observer, 15(2), 199–
224.
Forrest, C. B., & Starfield, B. (1996). The effect of first-
contact care with primary care clinicians on ambula-
tory health care expenditures. Journal of the FamilyPractice, 43(1), 40–48.
Forrest, C. B., & Starfield, B. (1998). Entry into primary
care and continuity: The effects of access. AmericanJournal of the Public Health, 88(9), 1330–1336.
Franco, M., Kennelly, J. F., Cooper, R. S., & Ordunez-
Garcıa, P. (2007). Health in Cuba and the millennium
development goals. Revista Panamericana de SaludPublica, 21(4), 239–250.
Frankenberg, E. (1995). The effects of access to health
care on infant mortality in Indonesia. Health Transi-tion Review, 5(2), 143–163.
Freedman, L. P., Waldman, R. J., de Pinho, H., & Wirth, M.
E. (2005). Who’s got the power? Transforming healthsystems to improve the lives of women and chil-dren (Millennium Project Task Force 4: Child Health
and Maternal Health—Interim Report, UN Millennium
Project). United Nations: New York.
Frenk, J., Sepulveda, J., Gomez-Dantes, O., & Knaul, F.
(2003). Evidence-based health policy: Three genera-
tions of reform in Mexico. The Lancet, 362(9396),
1667–1671.
Fritzen, S. (2007). Legacies of primary health care in an
age of health sector reform: Vietnam’s commune clin-
ics in transition. Social Science & Medicine, 64, 1611–
1623.
Fuster, V., & Voute, J. (2005). MDGs: Chronic diseases
are not on the agenda. The Lancet, 366(9496), 1512–
1514.
Greenwood, B. M., Bradley, A. K., Byass, P., Greenwood,
A. M., Menon, A., Snow, R. W., et al. (1990). Evaluation
of a primary health care programme in The Gambia.
II. Its impact on mortality and morbidity in young chil-
dren. The Journal of Tropical Medicine and Hygiene,93(2), 87–97.