ARTICLE IN PRESS

0277-9536/$ - se

doi:10.1016/j.so

�Correspondfax: +4122 791

E-mail addr

fritzen@nus.ed

Social Science & Medicine 65 (2007) 1751–1764

www.elsevier.com/locate/socscimed

What influences government adoption of vaccines in developingcountries? A policy process analysis

Syarifah Liza Muniraa,�, Scott A. Fritzenb

aThe Global Fund to Fight AIDS, Tuberculosis and Malaria, Geneva, SwitzerlandbLKY School of Public Policy, National University of Singapore, Singapore

Available online 17 July 2007

Abstract

This paper proposes a framework for examining the process by which government consideration and adoption of new

vaccines takes place, with specific reference to developing country settings. The cases of early Hepatitis B vaccine adoption

in Taiwan and Thailand are used to explore the relevance of explanatory factors identified in the literature as well as the

need to go beyond a variable-centric focus by highlighting the role of policy context and process in determining the pace

and extent of adoption. The cases suggest the feasibility and importance of modeling ‘causal diversity’—the complex set of

necessary and sufficient conditions leading to particular decisional outcomes—in a broad range of country contexts. A

better understanding of the lenses through which government decision-makers filter information, and of the arenas in

which critical decisions are shaped and taken, may assist both analysts (in predicting institutionalization of new vaccines)

and advocates (in crafting targeted strategies to accelerate their diffusion).

r 2007 Elsevier Ltd. All rights reserved.

Keywords: Hepatitis B vaccine; Vaccine introduction; Policy analysis; Developing countries; Taiwan; Thailand

Introduction

The 20th century saw tremendous advances invaccine technology and, in the last two decadesprimarily, its broadened application to healthproblems afflicting the poor throughout the world(World Bank, 1993). The expanded program forimmunization (EPI) package of vaccines alone isestimated to have saved approximately 20 millionlives in the past two decades (UNICEF website,2006). Yet the process of diffusion of these vaccines

e front matter r 2007 Elsevier Ltd. All rights reserved

cscimed.2007.05.054

ing author. Tel.: +4122 791 82 97,

17 01.

esses: lizamunira@yahoo.com (S.L. Munira),

u.sg (S.A. Fritzen).

has been far from even, as countries with differentpriorities and access to resources adopt and sustainvaccines to different extents. Differences in mortal-ity and morbidity rates across even similarlysituated poor countries were quite pronounced bythe late 1980s, due largely to the different extent towhich these countries had adopted and successfullyimplemented the ‘basic’ EPI package of vaccines(Wang, 2002; Wilkinson, 1992).

What determines the pace at which countriesmove to incorporate emerging vaccines with poten-tially great positive health impacts into theirvaccination programs? Given the dynamism char-acterizing vaccine development, this question ishighly relevant and important. A better under-standing of the lenses through which government

.

ARTICLE IN PRESSS.L. Munira, S.A. Fritzen / Social Science & Medicine 65 (2007) 1751–17641752

decision-makers filter information, and of thearenas in which critical decisions are shaped andtaken, may help us anticipate the diffusion of newvaccines, and help advocates craft targeted strate-gies to influence their adoption (Batson, 1998;Vaccine Alliance, 2006; Walt & Gilson, 1994).

This paper examines the process by which adop-tion of one important vaccine—that for HepatitisB–took place in two middle-income countries,Taiwan and Thailand. These countries were in thefirst wave of countries to adopt the vaccine into theirmass immunization schedules, several years after thevaccine became available in 1984, but before theWHO’s endorsement in 1992. It asks, what werethe factors that led to early adoption in these cases?Were these countries highly idiosyncratic, or werethere important common factors at work? Beyondshedding light on the two cases, the purpose of thispaper is to present a framework that can be used as astarting point for the more systematic modeling ofthe decision-making dynamics underlying vaccineadoption in developing countries. We have labeledthis a ‘policy process’ approaches because of thesalience of the sequencing and policy context inwhich critical decisions are made.

Beyond ‘cost’ and ‘political will’ in explaining uptake:

the case of Hepatitis B

Government decision-points in vaccine development

The introduction and expanding use of vaccinesfollow a typical pattern (Vandersmissen, 2001).

Needs

Early Demand

MatureDemand

General Use

Private Mark

% Doses s

Private Mar

Publi

PM

Private Market

Fig. 1. Traditional vaccine market developme

During a period of early demand, the vaccine islaunched in the private market of industrializedcountries, with low quantities and high prices. It issubsequently integrated into the public healthpolicies of industrialized markets. As quantitiesgrow, a multi-tiered price structure appears, withsupplies to the public market at a lower price thansales to the private market. The loss of thecustomers who previously were part of the privatemarket of industrialized countries is partly compen-sated for by new introduction of the vaccine in theprivate markets of developing countries, usually atintermediate prices. Finally, the vaccine becomesgenerally used, with massive purchases at low pricesin the public markets of developing countries,directly or through international procurementagencies such as The Pan American Health Orga-nization (PAHO) and The United Nations Chil-dren’s Fund (UNICEF).

The premise of this paper is that the pace of theprocess shown in Fig. 1 is shaped by decisions thatoccur in the interface between the private sectormarket and government adoption into standardizedpackages of immunizations, such as EPI. Thefactors underlying government decisions in this areaare complex, and in general under theorized; but thetwo most often cited considerations are vaccine cost(Batson, 2005; Mahoney & Maynard, 1999) and,as a residual category, political will (Catford, 2006).It is generally feared that even vaccines withfavorable cost–benefit ratios at high prices will notgo into general use until prices drop considerably, aprocess that may take up to 20 years or more

et

old

ket

c Market

ublicarket

(Limited in poorer countries)Moderate: establishing

evaluative criteria

Critical: decision-point regarding diffusion

Government Role

nt (adapted from Vandermissen, 2001).

ARTICLE IN PRESSS.L. Munira, S.A. Fritzen / Social Science & Medicine 65 (2007) 1751–1764 1753

(Vandersmissen, 2001). Where high-income coun-tries can potentially benefit from the vaccine, theywill generally lead the wave of adoption, while invery low-income countries, adoption may be sub-jected to the binding constraint of price (McGuire,2003), awaiting foreign funding and initiatives. Thismay help explain the limited interest in the literaturegiven generally to the government decision-makingprocess in the presence of this ostensibly bindingconstraint. Middle-income countries are potentiallythe most unpredictable in terms of where they willfall in any sequence. With their greater (and oftenincreasing) discretionary expenditure as comparedwith low-income countries, and the often growingsophistication of health policymaking networks,adoption may occur at a more intermediate pointin this process.

The case of Hepatitis B in middle-income countries

Cost has certainly played a role in the diffusionprocess of the Hepatitis B vaccine, commercially

1989 1991 1993 1995

1992 1994

1982

1984 1990

2

HebB vaccine

becomes available

WHO’s targ

countries w

carrier of 8

WHO endorsed

recommendation

:

: middle income countries (eligible for Vaccine Fund) that introduced HB with GAVI support

: low income countries

: low income countries that introduced HB with GAVI support

: higher middle income countries

: high income countries

lower middle income countries

Fig. 2. Hepatitis B vaccine uptake by country status, 1982–2004 (da

available since 1982 and globally recommendedsince 1991. Unlike other vaccines used in nationalEPIs (immunization programs), which have alwaysbeen in the public domain, this vaccine was bothcontrolled by active patents and introduced into thepublic sector at a fairly early stage in its productionlifecycle. At the time of the recommendation forglobal introduction, the vaccine price had droppedfrom over US$50 to roughly US$3 per dose, yeteven at this level the full series (three doses) of HBvaccination cost more than 10 times all the otherEPI vaccines combined, putting it out of reach forpublic use (Batson, 1998). At present, The GAVIAlliance continues to support countries introducingnew and underused vaccines such as Hepatitis Bvaccine, as even the current price of US$1.5 canstrain health budgets in low-income countries(WHO, 2006; GAVI Alliance, 2006).

Fig. 2 shows the progress of Hepatitis B vaccineuptake in different countries since it first becameavailable in 1982. Each star in the chart represents acountry that has included the Hepatitis B vaccine

1997 1999 20032001

1996 1998 2000 2002 2004

et for

ith a HB

% or greater

WHO’s target for all

countries

ta from WHO and the World Development Indicators 2004).

ARTICLE IN PRESSS.L. Munira, S.A. Fritzen / Social Science & Medicine 65 (2007) 1751–17641754

into its national immunization program, listedaccording to the year of vaccine adoption. Theevidence is consistent with a hypothesis that cost is acritical factor: the median time to adoption of thevaccine is 11 years for high-income countries, 15years for middle-income countries and 20 years forlow-income countries, respectively.

Yet clearly cost serves poorly as the sole variablepredicting uptake. Some affluent countries onlyadopted the vaccine well after 1997, the WHO’starget for universal adoption. And a number ofcountries at the lower income brackets decided earlyon to integrate the new vaccine. Indeed, severalmiddle-income countries were among the earliestadopters of the vaccine, integrating the antigen intotheir existing national immunization programs.

The Hepatitis B vaccine is thus a good example ofthe uneven diffusion process of new vaccines—withand without donor support—in developing coun-tries. This unevenness exerts a cost in terms of liveslost to preventable disease; and the difficulty inpredicting uptake and the mistaken assumption(usually implicit) that cost is the sole importantfactor—may well lead to missed opportunities forthe targeting of advocacy efforts. The question is,what factors will influence this process? Does theliterature help us to predict where and how it mayhappen?

Methods and empirical strategy

To further unpack the factors at work in theHepatitis B case, we reviewed a broad swathe ofliterature on vaccines in which factors predictinguptake are mentioned directly or implicitly. Wefound that this literature, as described below, suffersfrom two basic shortcomings. First, it generally hasnot considered systematically what comparativemethodologists call ‘‘causal complexity’’ (Ragin,1987, 2000), which simply means that rather than aparticular explanatory variable having a consistentimpact on the dependent variable of interest—suchas government adoption of a vaccine—this impactmay vary depending on the configuration of othervariables in the case context. There is thus noconsistent formula that will predict adoption;rather, mapping different ‘causal pathways’ (Frit-zen, 2000)—i.e. different combinations of explana-tory factors that lead to a similar outcome mayprove more fruitful for the purpose of drawing outthe implications of limited, comparative case studyevidence for future cases.

A second shortcoming of the literature lies in itsneglect of issues of policy process and context; thus,we draw selectively on some core theories in thepublic policy literature in our delineation of keyrelevant variables in the framework below. We alsohighlight political context in the conclusion of thepaper.

The framework is applied to data on the Taiwanand Thailand cases that were developed from twosources. The first is a thorough review of publishedsources—some of them quite extensive and de-tailed—providing a descriptive account of adoptionin the two countries. These were compared andcontrasted with nine expert interviews conducted, inperson and via email, from among governmentofficials, medical association members, and thepharmaceutical sector in the two countries. Theintention was to view the adoption process frommultiple stakeholder viewpoints so as to improvedata quality and achieve a more grounded analysisof the stakeholder dynamics that are at the core ofour conceptual framework.5

Hypothesized factors predicting uptake

Despite its importance, this question of how topredict vaccine uptake remains only lightly re-searched in the literature on the role of vaccines inthe public health systems of developing countries.The focus in the literature has instead been on issuessuch as system capacity and sustainability (Salis-bury, Beverley, & Miller, 2002), socioeconomicdeterminants of individual access to vaccines,vaccine cost-effectiveness (Glennerster, Kremer, &Williams, 2006; Stanton, 1994) and financingstrategies (McGuire, 2003)—all of which havegenerated kernels of speculation about the factorsthat may be at work in vaccine adoption. Whilethere have been few attempts to assess the relativeimportance of different factors in particular cases,three categories of factors can be elaborated on inindividual cases in order to assess the combination offactors leading to a particular outcome: vaccinecharacteristics, health systems characteristics, andpolicy process and context.

Vaccine value characteristics

Vaccine-related factors that can influence uptakeinclude the disease burden targeted, safety andperformance, and cost-effectiveness. Disease burden

ARTICLE IN PRESSS.L. Munira, S.A. Fritzen / Social Science & Medicine 65 (2007) 1751–1764 1755

has been consistently mentioned by policymakers incountries to be the number one factor in settingpriorities for vaccines to be introduced intoimmunization programs; the higher the burden,the more attractive a potential addition to theimmunization regime of the country would be(World Bank, 1993).

Vaccine safety and performance is a second factor.Most countries have mentioned safety, low rates ofside effects and vaccine performance as key criteriain considering introduction of a new vaccine.Decision-makers in one survey (DeRoeck, Clemens,Nyamete, & Mahoney, 2005) required evidence ofvaccine safety and effectiveness in the local popula-tion before considering use in the public sector, evenfor vaccines already licensed locally and used in theprivate sector. Minimum acceptable performancelevels for a number of countries for vaccines such astyphoid fever, shigellosis and cholera included aminimum protection length of 5 years and efficacyrates ranging from 75% to 85%; in some countries,the latter figure increased to 90% for governmentuse. (Hepatitis B vaccine compares relatively favor-ably in this regard, with an efficacy level of 75–95%against chronic infection, and duration of immunityof 15 years after the primary series of injections(WHO, 2002).)

A final factor is cost-effectiveness. The economicimpacts of a disease and potential cost savingseffectuated by vaccine coverage have featuredprominently in discussions about the need forenteric vaccines (Griffiths, Hutton, & Pascoal,2005). Evidence of cost-effectiveness of vaccines—as opposed to treatment—has been cited byinformants in Thailand, India, Bangladesh, Paki-stan, and Vietnam, as critical to convincinggovernment decision-makers to finance publicsector use of a new vaccine (DeRoeck et al.,2005).

Taken together, these factors suggest that a newvaccine is more likely to be adopted when theburden for the relevant disease is high; decision-makers are convinced the vaccine meets acceptablestandards of efficacy and safety; and when cost-effectiveness is assessed to be high.

Health system characteristics

A second category of explanatory factors pre-dicting vaccine uptake focuses on how the proposedvaccine fits into health delivery systems of acountry. One variable that has attracted muchinterest by analysts is programmatic feasibility,

broadly signifying the level of ease with which avaccine can be added to existing service deliverynetworks (Dasgupta & Priya, 2002; Duma, 1995).Programmatic feasibility may be affected by a rangeof issues such as: the timing of immunizationschedules, including whether a birth dose is recom-mended; impact on cold chain capacity; the propermix of single/multi-dose vials; the number ofinjections per visit; vaccine security; whether itmay negatively affect the uptake of other antigens inthe existing program; possibility of local vaccineproduction (Woodle, 2000); and cost (Wittet, 2001).In addition, personnel requirements necessary todelivery a new vaccine influence potential cost of theprogram immensely.

As noted above, vaccine price has been men-tioned as one of the most important factors ininfluencing decisions to introduce a new vaccine,often second only to the issue of disease burden(Mahoney & Maynard, 1999). Yet at least onestudy, focusing exclusively on middle-income coun-tries, has found that vaccine prices may not be themajor driving factor, especially when it comes tovaccines that are in combination form; it points outthat many of the poorer countries among themiddle-income countries included in the study havebeen paying premium prices (Milstien, Munira, &McKinney, 2003). Price may best be considered interms of both affordability, i.e. as a percentage ofnational health budgets—and in conjunction withother factors such as perceived impact and policycontext.

Policy actors, process and context

As noted above, few papers examine the processby which decisions to incorporate vaccines intonational health systems are made, despite theobvious relevance of this literature to explainingor predicting adoption outcomes. Instead, ‘politicalwill’ is often deployed as a residual concept maskinga host of potential factors at work; to help unpackand operationalize the notion of ‘political will’ as anexplanatory factor (Catford, 2006), this paper drawson some well-established approaches in the policyprocess literature, including the Sabatier andJenkins-Smith’s Advocacy Coalition framework(1999), Kingdon’s framework for policy adoption(1995), and Grindle and Thomas’ framework forcontextualizing specifically developing country pol-icy environments (1991). Concepts are borrowedvery selectively in this paper, and the concluding

ARTICLE IN PRESSS.L. Munira, S.A. Fritzen / Social Science & Medicine 65 (2007) 1751–17641756

section notes additional areas where the policyliterature may be fruitfully exploited for under-standing vaccine adoption.

Actor interests and interactions. A variety ofactors, both inside and outside a predictable core,must interact in ways to influence the adoption of avaccine (Walt & Gilson, 1994). The first, mostobvious actor is what policy analysts sometimesrefer to as ‘sovereigns’ (Jenkins-Smith & Sabatier,1994), the government authorities governing thehealth sector who ultimately determine strategiesand allocate scarce resources. Other groups areappropriately seen as attempting to influence them.These include the second group, the scientific/medical grouping, which must form a consensuson the medical value and level of scientific evidenceassociated with a new vaccine, and advocate for it inpolicymaking circles. For this group, the availabilityof scientifically presented evidence regarding vac-cine value characteristics, including the existence ofpilot studies and trials, are likely to be critical.Formal recommendations by bodies such as theWHO are often critical in both reflecting andshaping medical community consensus, and ininfluencing governmental decisions.

Third, the role of the public and non-healthinterest groups is also important to analyze. Thepublic, via the media or non-governmental associa-tions, may or may not push politicians andphysicians to take action to address a perceivedhealth problem via a vaccine. Public support mightbe gauged by looking at the visibility of the mediatreatment of the vaccine issue over time. Local and/or international drug manufacturers who maybenefit from vaccine adoption may be mobilized tosupport adoption. As a result, the feasibility of localproduction of a vaccine would be an importantconsideration in determining the political forcespushing for adoption (Clemens, 2003; Muraskin,1996; Woodle, 2000).

A fourth type of actor is the ‘policy entrepreneur’(Kingdon, 1995), typically either government offi-cials or prominent researchers who decide to ‘own’the vaccine issue by pushing consistently andcreatively for the articulation of specific proposalstimed to the policy adoption cycle (Duma, 1995).

Decision-making process and context. Mention ofpolicy entrepreneurs leads to the issue of processand context. One prominent approach to explainingthe circumstances surrounding policy adoptionfocuses on the convergence of three factors or‘‘streams’’ onto a defined decision-point, at which

time a ‘policy window’ is said to open (Kingdon,1995):

�

Government officials fix attention onto a parti-cular problem, from among a huge and shiftingnumber of alternative problem choices, that mayarise on their agenda through various ‘‘focusingevents’’ such as crises, disasters or the availabilityof a new, high-profile study (p. 112). � Specific, actionable proposals or policies are

‘‘generated, debated, redrafted, and acceptedfor serious consideration’’ (p. 143); and

� The politics of an issue, which may be influenced

by ‘‘swings of national mood, election results,changes of administration, changes of ideologicalor partisan distributions in Congress, and inter-est group pressure campaigns’’, for instance,promote proposals to high-agenda status, wherethey may or may not be acted on (p. 162).

The context of policy-making plays a critical rolein determining the roles, and relative strength, of theactors noted above. To borrow a pattern reported inother sector contexts, in periods in which theprominence of the issue is rising following anepidemic, decision-making is more likely to bedominated by senior political figures concernedwith the political impacts and perceptions, whereasin ‘bureaucratic-politics-as-usual’ times, technicalpersonnel concerned with some mixture of thetechnical costs and benefits, and the bureaucraticturf issues, are more likely to play a prominent role(Grindle & Thomas, 1991). In addition, the weightgiven to the scientific evidence, and the evaluationof the political and technical risks involved invaccine adoption, may be influenced by thesequence of international adoption. This is espe-cially true at the bookends of the sequence: the cuesthat countries at the early phases of internationaladoption will be sensitive to are likely to be differentfrom those at the end, for instance (Mahoney &Maynard, 1999).

The above considerations orient our examinationof contextual and process-oriented variables affect-ing the governmental adoption of vaccines. We nowturn to such an analysis as applied to two middle-income countries, Taiwan and Thailand. It is basedon three sources: extensive interviews with a rangeof governmental, non-governmental actors andinternational actors in 2005; a review of secondarydocumentation from the countries’ Ministries of

ARTICLE IN PRESSS.L. Munira, S.A. Fritzen / Social Science & Medicine 65 (2007) 1751–1764 1757

Health; and published accounts of adoption in thesecontexts.

Results

Taiwan

Taiwan in the early 1980s was considered a highendemic country for liver cancer; nearly half of allmale and one in every seven female in Taiwan werelikely to develop cirrhosis or hepatoma (Republic ofChina (ROC) Department of Health (2004)).Studies conducted on the population in the 1960sand 1970s revealed the association between thesediseases and the hepatitis B virus (HBV) infection(Huang & Lin, 2000). Evidence showed thatover 9000 people annually died of hepatocellularcarcinoma, liver cirrhosis or chronic liver disease,while more than 80% of hepatocellular carcinomacases and liver cirrhosis in Taiwan were related tochronic HBV infection. Scientists then discoveredthat the HBV infection in Taiwan occurred early inlife and at very high rates; and that perinatalinfection led to chronic liver diseases in 90% ofthose affected, emphasizing that the mother-to-infant transmission was the main route of HBVtransmission in Taiwan. These facts convincedhealth advocates of the need to reduce the diseaseburden and prevent the further spread of this deadlyvirus.

In 1984, Taiwan became the first country in theworld to adopt the vaccine into its nationalimmunization program, only 2 years after it becamecommercially available—a major step for a countrythat had only in recent years reached the status ofan industrialized state. The policy decision andprocess that lead to it, however, was certainly noteffortless.

At the time, several groups of scientists in Taiwanhad closely observed and studied the spread of theHepatitis B virus and its related diseases. Theyjoined forces to lobby the government by puttingforward the following conclusions (personal com-munication with M. H. Chang):

�

that there was a high prevalence of Hepatitis Bcarriers in Taiwan, and that 90% of these carrierswere likely to suffer from liver cirrhosis; � that the liver cancer caused by the Hepatitis B

virus has proven to be mainly transmittedthrough vertical infection, from mother to child,directly during but not prior to parturition;

�

that the transmission/infection needed to beinterrupted during the first 24 h of life by givingthe vaccine to new-born infants.

The assessment of the researchers was that theadoption of Hepatitis B vaccine would depend onstrong political support from senior policymakersgiven the highly centralized, hierarchical decision-making process of the government (and the healthdepartment itself). The President of Taiwan at thattime, Y.S. Sun, was in the process of articulating apolicy of advanced health care as part of a generalpush in the area of science and technology, with avision to tapping into and adopting leading globaltechnologies in a range of fields. The president’svision was influenced by, and associated with, J.T.Lee, who in 1984 held the title of Minister withoutportfolio. These scientists thus targeted Lee in theiradvocacy efforts; and gaining his support, thegovernment decided to form a committee in mid-1984 under the Department of Health to studypolicy responses, sponsor further scientific research,and conduct a pilot immunization program forHepatitis B.

Lee proved an essential supporter of the process(personal communication with D. S. Chen). Onerole he played was to defend the vaccine-pilotingphase from various attacks, such as opposition toconducting trials targeted at infants. Lee hosted aninternational Hepatitis conference in Taipei result-ing in a unanimous conclusion that a clinical trial ofthe Hepatitis B vaccine was indispensable forTaiwan, and that the focus of such trials shouldbe on children at high risk.

Since the vaccine had just been launched, theprice offered in the international market to date hadaveraged US$12 per dose, more than the cost of thebasic EPI vaccines combined. There were twovaccines qualified and ready in the market, onefrom Merck and the other from Pasteur. TheDepartment of Health decided to purchase fromthe latter, which had agreed to give Taiwan licenseto produce the vaccines locally—an importantpolitical consideration that figured in policy debatessurrounding early introduction; and which agreedto the price of US$4 per dose in part to use Taiwanas an example for other countries to follow. To fundthe vaccine, the government in Taiwan simplyabsorbed the cost within an already expandinghealth budget, which from 1980 to 1990 increasedfrom 3.3% to 4.2% of Taiwan’s GDP (a trend thathas continued into the present decade (Republic of

ARTICLE IN PRESSS.L. Munira, S.A. Fritzen / Social Science & Medicine 65 (2007) 1751–17641758

China (ROC) Department of Health, 1997). For thefirst 2 years, vaccines were purchased throughPasteur’s parent company; by the third year, localproduction was enabled, a factor which, togetherwith the favorable results of the initial targetedphase, helped convince the government to expandprovision of the vaccine, at no user cost, to allinfants in 1986 (personal communication with D.S.Chen).

A striking aspect of Taiwan’s case was the lack ofdirect international support to the health sectorgenerally, and specifically for the introduction of theHepatitis B vaccine. Some analysts have suggestedthat this fact actually spurred Taiwan to take itspioneering role, as politicians and health officialsbelieved favorable international commentary couldbenefit them and the country in different ways(personal communication with Mark Kane). SomeTaiwanese officials have asserted their pride at beingthe first country to introduce, and successfullyadopt, the Hepatitis B vaccine into its nationalimmunization program over 10 years before therecommended target from the World Health Orga-nization—the same organization that does not grantmembership to Taiwan (Chan, Lee, & Lo, 2004).

Fig. 3 and its accompanying notes summarize thebroad sequencing of government decision-making inTaiwan; compiled from various expert interviews aswell as literature reviews.

Notes to accompany figure: initiators and role

(1) Scientists conduct clinical trials and studiesconducted on liver cancer prevalence in Taiwan,

DiseaseBurden

PoliticalLeaderScientists

Studies

Scientific Dept.ofHealth

1 2 3

4

Fig. 3. Taiwan’s process prior to introduction (compil

and (2) in coalition lobby government, which (3–4)forms a National Committee to oversee pilotprogram and assess policy options, which (5)conducts pilot program and assesses favorably thepossibility of local production, leading to (6) theformal launch of Taiwan’s national immunizationprogram for Hepatitis B in 1984.

Thailand

Thailand was among the earliest countries toadopt the Hepatitis B Vaccine into its routineimmunization schedule; Thailand’s case in particu-lar is also interesting given that although officially amiddle-income country, its GDP per capita in 1992was, according to one measure, only 19.3% that ofTaiwan (Republic of China (ROC) IndustrialDevelopment, 2006). In 1992, the populationchronic carriage rate of HBV was 8–10% andhepatocellular carcinoma was the most commoncancer-related cause of death. Most virus transmis-sion occurred during childhood and about twothirds of neonates were infected by the age of 3months. Hepatitis B vaccine was integrated intoThailand’s EPI in 1992, supported by a pilot projectwhich demonstrated that immunization reducedchronic HBV carriage in children under 5 years byat least 80%. A national Hepatitis B vaccinationcoverage rate of over 90% was subsequentlyachieved using a monovalent product with the firstdose given at birth (Chunsuttiwat, Biggs, Maynard,Thammapormpilas, & Prasertsawat, 2002).

The process that propelled Hepatitis B to a rela-tively high public policy priority was a collaborative

National Committee

ImmunizationProgram

LocalProduction

PilotStudy

6

5

5

6

6

ed from expert interviews and literature review).

ARTICLE IN PRESSS.L. Munira, S.A. Fritzen / Social Science & Medicine 65 (2007) 1751–1764 1759

effort. It involved at different stages the Ministry ofPublic Health’s (MOPH) Department of Commu-nicable Disease Control (CDC) (spearheaded by Dr.Vinij Assawasena, the Director General of theCDC, and Dr. Piroj Ningsanon, the MOPH’sPermanent Secretary), the Thai Medical Associa-tion, the pharmaceutical industry and the media,with an internationally-supported Task Force even-tually playing an important catalytic role (Mura-skin, 1995; personal communication with SomsakLelekha supplemented by anonymous peer re-viewer). The manner in which this process playedout highlights the shifting influence of scientific andprogrammatic evidence as well as shifts in leadactors driving advocacy efforts.

A critical step in the early deliberative processexploring the possibility of mass vaccination was theformation, in late 1985, of the National Hepatitis BCommittee. It comprised some 20–30 experts fromdifferent institutes, associations and governmentdepartments, to be chaired by the Director Generalof the CDC. Some observers felt that the size of thecommittee retarded initial evidence collection byincreasing the costs of bureaucratic coordination(Muraskin, 1995). For example, internally thedecision-making of the committee was subject tobottlenecks of centralization. Any recommenda-tions emerging from its membership had to bescreened and approved by the committee headbefore they could proceed to a higher level in theMinistry to then be presented to the Minister ofPublic Health, who would then be responsible forcoordinating advocacy efforts (Muraskin, 1995).

The MOPH conducted initial policy discussionsregarding potential mass vaccine in its initial year. Italso facilitated early efforts that included the CDCitself, the medical association, and pharmaceuticalcompanies to engage the media through public talksand interviews to raise awareness of the Hepatitis Bvirus, highlighting the high toll taken by the diseasein Thailand. While these early efforts suggested thatthe Hepatitis B agenda had potential, the Commit-tee’s leadership was convinced that political supportcould ultimately only be clinched if more persuasivescientific and cost-related evidence could be mobi-lized.

Towards this end, the Committee moved slowly—but with hindsight systematically—to build theevidentiary basis needed to win political supportfor an eventual mass immunization program. In1988, the MOPH launched a cost-effectivenessstudy, comparing the option of a mass immuniza-

tion program with the alternative, selective vaccina-tions for high-risk groups. The study found theformer to be both more economical and effective,primarily due to the anticipated difficulties inidentifying and achieving acceptable coverage rateswithin high-risk groups (Van Damme, Kane, &Meheus, 1997). Earlier studies had also estimatedthe Hepatitis B carrier rate in Thailand at as high as10%, with 3% due to perinatal infection (mother-to-child) and the rest involving horizontal infection.Such findings convinced scientists that in order tolower the endemicity it was essential for thegovernment to emphasize prevention through massvaccination for infants. The most persuasive mes-sage to convey to policymakers through theCommittee, they found, was that most liver cancerpatients were carriers, and thus that preventivemeasures would enable them to significantly lowerliver cancer morbidity and mortality rates (Poovor-awan et al., 2001). The fact that the efficacy of thevaccine itself had been proven in the private marketby that time was also highlighted.

Price remained a critical remaining considerationfor both many Committee members and govern-ment officials, and was perceived to be an importantbarrier to the adoption of a mass campaign. Yet thegrowing scientific consensus behind mass immuni-zation noted above coincided with the emergence ofincreased international support for its operationsand financing. Specifically, a PATH-affiliatedteam—the International Task Force for HepatitisB, supported also by the Australian InternationalDevelopment Assistance Bureau (AIDAB), and theThai Red Cross Society—launched a mass immuni-zation pilot project in two provinces, and initiatedthe funds to purchase the vaccine for this purpose.The project, which concluded in 1991, successfullydemonstrated that the Hepatitis B vaccine could beoperationally synchronized with the basic EPIpackage without creating a parallel system, a featdeemed critical in winning the support for nation-wide immunization among several senior Ministryof Public Health officials (Muraskin, 1995; Wittet,2001).

The Task Force also helped the governmentnegotiate prices and assisted them in selecting themost affordable alternatives. The Task Force’sinitial enthusiasm for local production—whichproved unsuccessful in the end—contributed to astronger negotiating position for the governmentvis-a-vis international suppliers, helping to pushthe price down significantly to US$1 per dose

ARTICLE IN PRESSS.L. Munira, S.A. Fritzen / Social Science & Medicine 65 (2007) 1751–17641760

(Muraskin, 1995). With a choice between twovaccine alternative types—plasma derived andrecombinant—Thailand’s Ministry of Public Healthopted for the lower-priced plasma-derived vaccine,produced by the Korean Green Cross Corporation(KGCC)—to the reported consternation of largerprivate pharmaceutical companies. Interestingly,once the government had decided to implement anation-wide immunization program for Hepatitis B,the Ministry’s aggregate health budget was in-creased specifically to fund the program (personalcommunication with Somsak Lolekha). In theseways, the International Task Force served as anexternal catalyst among the many members of theNational Hepatitis B Committee (Muraskin, 1995).

Thus, the initial impetus for mass immunizationthat had emerged from the National Committee inthe mid-1980s was finally adopted as a nationalpolicy that includes the Hepatitis B vaccine intoThailand’s routine immunization program in 1992,some 7–8 years after its initial public debate. It wasclear that the 4-year pilot project’s success was acritical factor that convinced the government thatthe vaccine adoption was indeed feasible, even withits mid-term results (Chunsuttiwat et al., 1997).Thailand’s case in turn proved to be a strong spur tothe World Health Organization in issuing its 1992recommendation for universal Hepatitis B vaccina-tion in developing countries (Poovorawan et al.,2001). Fig. 4 summarizes the broad sequencing ofgovernment decision-making in Thailand.

The General Public

Scientists

Studies ScientificDisease

Burden

MinistryofHealth

Media

PharmaceuticalCompanies

1

1

1

Fig. 4. Thailand’s process prior to introduction (comp

Notes: initiators and roles

(1) Scientists, supported by media and pharma-ceutical companies present scientific evidence andsuccessfully capture public attention, which leads(2) Ministry of Public Health to prioritize thevaccine and establish (3) a National Committee toreview options, assisted by (4) an International TaskForce, which performs multiple support functions,leading to (5) government adoption of the vaccineinto routine EPI schedule.

Conclusions

The case of the introduction of Hepatitis B inthese two middle-income countries can be used toreflect both on the broad set of determinants ofgovernment vaccine adoption and on the methodsnecessary to gain a deeper understanding of theprocess underlying adoption. On the most basiclevel, we can examine the similarities and differencesin the policy process by which the two countriesreached the same ultimate outcome (early adop-tion). In each country, fixed characteristics (such asthe disease burden being experienced around thetime of adoption) combined with core and non-coreactors to frame decision-points for the use ofscientific evidence. Table 1 lays out in summaryfashion the presence and absence of some of thefactors predicted to be facilitative of vaccine

National Committee

ImmunizationProgram

LocalProduction

Pilot

Study

Int’l TF on HB (PATH/CDC)

Vaccine Price

32

4

4

5

5

iled from expert interviews and literature review).

ARTICLE IN PRESS

Table 1

Hypothesized factors influencing government decision to introduce vaccine into immunization program

No. Factors Taiwan Thailand

1 High disease burden Present Present

2 Programmatic Feasibility Present Present

3 Pilot studies Present Present

4 Scientific evidence Present Present

5 Important role played by:

a. Role of the medical association Present Present

b. Local manufacturers Present Not present

c. International support Not present Present

d. Role of media Not present Present

6 Sensitivity to price Not present Present

7 Policy entrepreneurs Present Not present

8 Other countries already using the vaccine? Not present Present

S.L. Munira, S.A. Fritzen / Social Science & Medicine 65 (2007) 1751–1764 1761

adoption posited in section three above, and theirpresence and absence in the two country cases.

The case studies suggest some important com-monalities in the countries’ experiences. Both shareda high disease burden from the Hepatitis B virus,which set the stage for the articulation of politicaland/or social demand for its prevention. In bothcountries, the demonstration of programmaticfeasibility through pilot projects was an importantimpetus to winning support from technical elites inthe ministries. And central to both was the proactiverole that medical associations played, which usedthe growing scientific evidence and epidemiologicalstudies to advocate for adoption of the new vaccinein policy and indeed political circles. This combinedwith an already strong and well-performing EPI andcentralized health system (particularly in Taiwan),were all supporting factors in the process.

But striking differences were evident as well. Themost important differences lay in the differentconstellation of those actors outside the core circleof government officials and the medical associa-tions, including the local and international pharma-ceutical industry, international donors, and themedia. International support—critical to Thailand,absent entirely in Taiwan—was a most obviouscritical difference. The potential for local manufac-turing was an important supporting factor in thedecision-making calculus in Taiwan, but in Thai-land played only an indirect role in negotiationsover price. Sensitivity to vaccine price differed too.Price was a critical constraint in Thailand untilinternational support and a far lower marketedprice came into play, whereas Taiwan was relativelyprice insensitive to a degree that is not explainedsimply with reference to its higher income. Taiwan

presented evidence of a critical role being played by‘policy entrepreneurs’, while in Thailand no highlyvisible individuals drove proposals towards adecision.

Thus Thailand and Taiwan present two relatedbut distinct pathways to the same ultimate outcomeof early adoption. What light do these two casesshed on broader issue of the determinants of vaccineadoption in developing countries? There are poten-tially huge implications to understanding theprocess better, since the success of efforts toincorporate new vaccines into the regimen clearlyhinge on the ability to persuade policymakers toinvest both in the initial uptake and long terminstitutionalization of vaccines (Hardon & Blume,2005; Mrazek & Mossialos, 2003; Tangcharoen-sathien et al., 2001). We have known for some timethat such factors as political will, resources andinstitutional capacity are important determinants ofthe success of the EPI effort (Catford, 2006;Clemens, 2003). But the ability to predict andsystematically explore these factors still eludes us.

This study is no more than a first step towardsreopening the question of policy processes under-lying vaccine adoption as a field of inquirywarranting greater attention. In moving towards asustained research agenda in this area, two broaderaspects of our findings warrant attention.

The first concerns methodology. The study shedslight on the methodologies that will be important tothe further development of this research agenda. Aclassic approach to predicting uptake would involvethe construction of databases of cross-nationalvariables predicting uptake—indicators of the kindexplored in section three above—followed byeconometric analysis; the aim of such analysis

ARTICLE IN PRESSS.L. Munira, S.A. Fritzen / Social Science & Medicine 65 (2007) 1751–17641762

would in general be to quantify the incrementalimpact of any of the variables while holding allothers in the equation at their mean value. Whilesuch a variable-centric approach has not yet beenconducted and may be feasible, our findings suggestit is unlikely to cast much light on the critical issuesof policy context, process, and stakeholder config-urations. It is likely that the primary variables atwork are triggered and impact on decisional out-comes in complex combinations with other vari-ables; this is the hallmark of ‘causal diversity’ thatforms a core assumption of the present study, andone borne out by the case analysis. Such causalcombinations that will be highly difficult to modelgive limited sample sizes in cross-country analysis(Parashar, 2005).

More promising, therefore, will be methodologiesthat attempt to model sets of necessary andsufficient conditions indicating distinct causal path-ways towards the same outcome, of the typepioneered by Ragin (1987, 2000). Such ‘qualitativecomparative’ and ‘fuzzy-set’ analysis will require theconstruction of a greater range of qualitative casestudies using comparable categories of analysis(such as those suggested in section three). Thesecan – and will certainly need – to extend beyond themiddle-income country range, and should includecases of late or non-adoption as well as the earlyadopters explored in the present paper.

An important challenge facing this type ofcomparative method will be the establishment andjustification of appropriate boundary conditionsunderlying an assessment of the ‘presence’ or‘absence’ of different factors. Sensitivity and fidelityto the case narratives in this work will be vital, sincewithout grounded comparisons the actual modelingof causal conditions will prove meaningless. Indoing so, it is important to note that the coding ofvariables will be based on the relative rather thanabsolute values for variables in the cases; forexample, the comparative lack of a high-profilepolicy champion in the present study’s Thailandcase does not imply that passionately motivatedindividuals played little role; indeed, their presenceamong other places on the International Task Forcewas indeed very important (Muraskin, 1995). Like-wise, few if any countries will be oblivious to priceconsiderations in vaccine adoption, but countrieswill vary in their degree of price sensitivity, as theydid in the present study.

A second set of considerations concerns thecontested, political context of vaccine adoption

described in this paper’s framework. The adoptionprocess described in our cases underlined theimportance of policy-making environments thatwere in part open to multiple inputs and actors,and at least partly evidence based. But thiscondition is far from typical in a large number ofcountries, particularly in the developing world, dueto weakly institutionalized processes of policy-making as well as power and resource asymmetriesbetween donors bringing much needed finance, onthe one hand, and country health and financeministries on the other. Recognition of the politicalcontext of policy-making underscores the need foreclecticism in conceptual approaches—which willneed to draw on theories of as bureaucratic politics(Brinkerhoff & Crosby, 2002; Grindle & Thomas,1991) as well as policy transfer and learning (Rose,1993) in future studies. It also highlights thenecessity of care in interpreting findings-besidesthose provided by sound epidemiological studies(the latter often critically lacking in developingcountries); those seeking to influence the adoptionprocess will not be able to follow a variable-centered‘cook-book’ of conditions or methods facilitatingadoption. But the emphasis placed in the currentstudy on the dynamics of the policy process, coupledwith its approach to making the modeling ofexplanatory frameworks dependent to a greaterextent on local conditions, may point towards aresearch agenda rich with practical implications.

Disclaimer: The views and opinions expressed in this

article are those of the author and does not reflect the

official position of The Global Fund to Fight AIDS,

Tuberculosis and Malaria.

References

Batson, A. (1998). Sustainable introduction of affordable new

vaccines: The targeting strategy. Vaccine, 16, S93–S98.

Batson, A. (2005). The problems and promise of vaccine markets

in developing countries. Health Affairs, 24(3), 690–694.

Brinkerhoff, D. W., & Crosby, B. L. (2002). Managing policy

reform: Concepts and tools for decision-makers in developing

and transition countries. Bloomfield: Kumarian Press.

Catford, J. (2006). Creating political will: Moving from the

science to the art of health promotion. Health Promotion

International, 21(1), 1–4.

Chan, C. Y., Lee, S. D., & Lo, K. J. (2004). Legend of Hepatitis B

vaccination: The Taiwan experience. Journal of Gastroenter-

ology and Hepatology, 19, 121–126.

Chunsuttiwat, S., Biggs, B. A., Maynard, J. E., Thammapormpi-

las, P., & O-Prasertsawat, M. (2002). Comparative evaluation

ARTICLE IN PRESSS.L. Munira, S.A. Fritzen / Social Science & Medicine 65 (2007) 1751–1764 1763

of a combined DTP–HB vaccine in the EPI in Chiangrai

Province, Thailand. Vaccine, 21, 188–193.

Chunsuttiwat, S., Biggs, B. A., Maynard, J., Thamapalo, S.,

Laoboripat, S., Bovornsin, S., et al. (1997). Integration of

Hepatitis B vaccination into the expanded programme on

immunization in Chonburi and Chiangmai provinces. Thai-

land. Vaccine, 15, 769–774.

Clemens, J. D. (2003). Thinking downstream to accelerate the

introduction of new vaccines for developing countries.

Vaccine, 21, S2/114–S2/115.

Dasgupta, R., & Priya, R. (2002). The sustainability of Hepatitis

B immunization within the universal immunization pro-

gramme in India. Health Policy and Planning, 17(1), 99–105.

DeRoeck, D., Clemens, J. D., Nyamete, A., & Mahoney, R. T.

(2005). Policymakers’ view regarding the introduction of new

generation vaccines against typhoid fever, shigellosis and

cholera in Asia. Vaccine, 23, 2762–2774.

Duma, R. (1995). Establishing a national universal immunization

programme. Vaccine, 13(S1), S58–S60.

Fritzen, S. A. (2000). Decentralization and local government

performance: A comparative approach with application to social

policy reform in Vietnam. Ph.D. dissertation. Princeton

University.

Glennerster, R., Kremer, M., & Williams, H. (2006). Creating

markets for new vaccines. Innovations, 1(1), 67–79.

Griffiths, U. K., Hutton, G., & Pascoal, E. D. (2005). The cost-

effectiveness of introducing Hepatitis B vaccine into infant

immunization services in Mozambique. Health Policy and

Planning, 20(1), 50–59.

Grindle, M., & Thomas, J. (1991). Public choices and policy

change: The political economy of reform in developing

countries. Baltimore: Johns Hopkins University Press.

Hardon, A., & Blume, S. (2005). Shifts in global immunisation

goals (1984–2004): Unfinished agendas and mixed results.

Social Science & Medicine, 60(2), 345–356.

Huang, K. Y., & Lin, S. R. (2000). Nationwide vaccination:

A success story in Taiwan. Vaccine, 18, S35–S38.

Jenkins-Smith, H., & Sabatier, P. (1994). Evaluating the

advocacy coalition framework. Journal of Public Policy, 14,

175–203.

Kingdon, J. (1995). Agendas, alternatives, and public policies. New

York: HarperCollins College Publishers.

Mahoney, R. T., & Maynard, J. E. (1999). The introduction

of new vaccines into developing countries. Vaccine, 17,

646–652.

McGuire, T. G. (2003). Setting prices for new vaccines (in

advance). International Journal of Health Care Finance and

Economies, 3, 207–274.

Milstien, J., Munira, S. L., & McKinney, S. L. (2003). Issues in

selection of DTwP-based combination vaccines. Vaccine, 21,

1658–1664.

Mrazek, M. F., & Mossialos, E. (2003). Stimulating pharmaceu-

tical research and development for neglected diseases. Health

Policy, 64(1), 75–88.

Muraskin, W. (1995). The war against Hepatitis B, a history of the

International Task Force on Hepatitis B immunization.

Philadelphia: University of Pennsylvania Press.

Muraskin, W. (1996). Origins of the children’s vaccine initiative:

The political foundations. Social Science & Medicine, 42(12),

1721–1734.

Parashar, S. (2005). Moving beyond the mother-child dyad:

Women’s education, child immunization, and the importance

of context in rural India. Social Science & Medicine, 61(5),

989–1000.

Poovorawan, Y., Theamboonlers, A., Vimolket, T., Sinlapar-

atsamee, S., Chaiear, K., Siraprapasiri, T., et al. (2001).

Impact of Hepatitis B immunisation as part of the EPI.

Vaccine, 19, 943–949.

Ragin, C. C. (1987). The comparative method: Moving beyond

qualitative and quantitative strategies. Berkeley: University of

California Press.

Ragin, C. C. (2000). Fuzzy-set social science. Chicago: University

of Chicago Press.

Republic of China (ROC) Department of Health (2004).

Exemplary Hepatitis B Prevention and Treatment Practices

of the Past Century: Two Decades of Success in Taiwan.

Taiwan, Republic of China (DVD).

Republic of China (ROC) Department of Health. (1997). Evalua-

tion report of the national health insurance’s two years

implementation. Available at /www.pacificbridgemedical.com/

publications/html/TaiwanMay98.htmS, accessed March 19,

2006.

Republic of China (ROC) Industrial Development and Invest-

ment website. (2006). /investintaiwan.nat.gov.tw/en/env/

stats/per_capita_gdp.htmlS, accessed May 12, 2006.

Rose, R. (1993). Lesson drawing in public policy: Learning across

time and space. New Jersey: Chatham House.

Sabatier, P. A., & Jenkins-Smith, H. C. (1999). The

advocacy coalition framework: An assessment. In P. A.

Sabatier (Ed.), Theories of the policy process. Boulder:

Westview Press.

Salisbury, D. M., Beverley, P. C. L., & Miller, E. (2002). Vaccine

programmes and policies. British Medical Bulletin, 62,

201–211.

Stanton, J. (1994). What shapes vaccine policy? The case of

Hepatitis B in the UK. Social History of Medicine, 7(3),

427–446.

Tangcharoensathien, V., Phoolcharoen, W., Pitayarangsarit, S.,

Kongsin, S., Kasemsup, V., Tantivess, S., et al. (2001). The

potential demand for an AIDS vaccine in Thailand. Health

Policy, 57(2), 111–139.

UNICEF website. (2006) /www.unicef.org/immunization/

index_coverage.htmlS, accessed May 4, 2006.

Vaccine Alliance website. (2006). /www.vaccinealliance.org/

General_Information/About_alliance/index.phpS, accessed

April 5, 2006.

Van Damme, P., Kane, M., & Meheus, A. (1997). Integration of

Hepatitis B vaccination into national immunization pro-

grammes. British Medical Journal, 314, 1033 (5 April).

Vandersmissen, W. (2001). WHO expectation and industry goals.

Vaccine, 19, 1611–1615.

Walt, G., & Gilson, G. (1994). Reforming the health sector in

developing countries: The central role of policy analysis.

Health Policy and Planning, 9(4), 353–370.

Wang, L. (2002). Determinants of child mortality in LDCs:

Emprical findings from demographic and health surveys.

Health Policy, 65(3), 277–299.

Wilkinson, R. G. (1992). National mortality rates: The impact of

inequality? American Journal of Public Health, 82(8),

1082–1084.

Wittet, S., (2001). Hepatitis B vaccine introduction: Lessons

learned in advocacy, communication, and training. Children

Vaccine Program at PATH Occasional Paper #4, available at

ARTICLE IN PRESSS.L. Munira, S.A. Fritzen / Social Science & Medicine 65 (2007) 1751–17641764

/http://www.childrensvaccine.org/files/CVP_Occ_Paper4.pdfS,accessed January 20, 2006.

Woodle, D. (2000). Vaccine procurement and self-sufficiency in

developing countries. Health Policy and Planning, 15(2),

121–129.

World Bank. (1993). Investing in health: The World Development

Report 1993. Washington, DC: World Bank.

World Health Organization (WHO). (2002). Core information for

the development of immunization policy: 2002 Update. Geneva:

World Health Organization.

World Health Organization (WHO) website. (2006) ‘‘State of the

World Vaccines and Immunization,’’ available at

/www.who.int/vaccines-documents/DocsPDF04/wwwSOWV_

E.pdfS, accessed February 12, 2006.