Are vaccination programmes delivered by lay health workers cost-effective? A systematic review

In 1978, the Alma-Ata Conference put forward the goal of 'Health for all by the year 2000' and declared primary health care (PHC) the vehicle through which this goal was to be achieved [1]. As a result, PHC service delivery programmes using community or lay health workers (LHWs), a cadre of health worker that was often comprised of ordinary people with minimal health training, were established in many low- and middle-income countries (LMICs) and also became more widespread in high-income settings [2]. However, a combination of factors throughout the developing world in the 1980s, such as economic recession, political and policy changes, population growth, poor governance, and inadequate health systems, led to reduced investments in primary health care, including in LHW programmes [2, 3]. Today, a key challenge of health systems in many countries is the need to develop and strengthen human resources to deliver essential interventions [4, 5]. This has been a key factor in rekindling interest in the use of LHWs [6, 7].

In 2005 Lewin et al. [8] published a Cochrane systematic review examining the global evidence from randomized controlled trials (RCTs) on the effects of LHWs programmes, as compared to usual primary and community health care. This review indicated promising benefits, in comparison with usual care, for LHW interventions in the areas of vaccine promotion; breastfeeding promotion and treatment for selected infectious diseases. However, these results were based only on a limited number of studies. For example, the review identified only three RCTs examining the effectiveness of LHW programmes in improving vaccination uptake. An update of the original review by Lewin et al. [8] to identify and synthesize the results of more recent studies on LHW programmes is being undertaken. An interim report on the updated review identified six trials of vaccination promotion by LHWs [9].

With its focus on RCTs of effectiveness, the original review [8] did not explore factors influencing the costs and cost-effectiveness of LHWs in delivering health services such as vaccinations. Taking intervention costs and effectiveness considerations into account is important for policy decisions and concerns around the affordability of resource inputs for health worker programmes. For governments and funding agencies, the question of whether an intervention is more or less cost-effective compared to alternative interventions, as well as whether there are sufficient funds to pay for the intervention, are factors that influence decision-making. Part of the growing interest in LHW programmes is related to the perception that they are cheaper than those that use professional health staff. However, a health programme is defined as affordable only if each individual or organization financially contributing to the programme is willing and able to contribute to financing its operation on the scale envisioned in the programme design [10]. A greater problem in health programming, from the perspective of those funding these initiatives, is the widespread failure to analyze the future recurrent cost implications of a proposed investment programme and to assess whether these costs will be affordable given available financing sources [10].

These considerations have practical implications for economic evaluations of health worker programmes, and specifically LHW programmes. Generally, conventional economic evaluations, particularly cost-effectiveness analysis, focus narrowly on health outcomes, and do not take into account the role of human-made institutions in shaping economic behaviour. Nor do current economic evaluation methods capture social non-health benefits, such as community empowerment and higher social capital, which may have positive or negative values, and are related to programme-induced changes in the wider community [2]. Through their overly reductionist perspective, conventional economic evaluations of LHW programmes are ill-equipped to deal with institutional changes [11], such as changes in local governance or differences in social values, which are especially important at the community-level. Institutional economics, alternatively, considers the social norms and networks which govern individual and group behaviour and are an important dimension to consider when looking at the cost-effectiveness of LHW programmes. For example, the training of programme staff and other activities that are seen as institution-building, with benefit flows beyond the duration of the programme, are treated as a resource input when valuating outcomes. However, within an institutional economics framework, they may also be considered an intermediate output, with its entire cost subject to amortization as per capital costs [11].

Two non-systematic reviews have indicated the general dearth of cost-effectiveness data on LHW programmes [2, 12]. Similarly, three systematic reviews focussing on LMICs, one on the effects and costs of expanding immunisation strategies [13], the other a systematic review of the grey literature on strategies for increasing coverage of routine immunisations [14], and the third a review of published and grey literature on routine immunisation [15], demonstrated the paucity of cost-effectiveness data on strategies to expand the coverage of vaccination services in developing countries. What continues to be missing, however, is a targeted review of the costs and cost-effectiveness of involving LHWs in vaccination programmes. As part of a wider study on LHW programmes for vaccination uptake in low- and middle-income countries (LAYVAC), a systematic review of the costs and cost-effectiveness of using LHWs to promote or deliver vaccinations was conducted.

The overall aim of this paper was to review the costs and cost-effectiveness of vaccination programme interventions involving LHWs. This paper sought to:

There were 2616 records identified. Eighty-four of these studies were considered potentially eligible for inclusion and full text articles were then retrieved. Five additional studies were known to the authors or identified from hand-searching references of key studies and reviews once the full-text articles were retrieved, giving a total of 89 articles. Three studies fully met the inclusion criteria of an economic evaluation of a vaccination programme involving LHWs, while an additional 11 were retained as they included some cost data associated with a vaccination programme involving LHWs. Four authors were contacted for papers on the basis of their conference abstracts; however, the papers were not available for inclusion in this study. All included studies were published in English or Spanish language journals. The results of the search are shown in Figure 1 (QUORUM flow chart).

Given the small number of full economic evaluations identified, the following section provides a short description of each. All costs were reported in US dollars (except where noted) and are reproduced here as originally stated in the respective studies (see Table 1 and Additional file 2).

Deuson et al. [17] assessed the value for money of a community-based Hepatitis B vaccination catch-up project for 4384 Asian American children in Philadelphia, USA, implicitly compared with usual care. Staff in the community-based organizations acted as LHWs through educating parents about the hepatitis B vaccination and visited homes of children due for a vaccine dose. Costs per child, per dose, and per completed series were $64, $119, and $537, respectively while the cost per discounted year of life saved was $11 525.

San Sebastian et al. [18] compared the costs and outcomes of two different vaccination strategies for children under five years of age between 1993 and 1995. The District Hospital (DH) strategy was centrally planned and managed by the DH and fully vaccinated five children, resulting in a cost of $777.60 per vaccinated child. The community health worker (CHW) strategy was planned and implemented in conjunction with the CHW Association and fully vaccinated 113 children at a cost of $32 per child.

Weaver et al. [19] conducted an economic evaluation of a community-based outreach initiative to promote pneumococcal and influenza vaccines for people aged over 65 years, compared with no outreach. The authors found that the cost per quality-adjusted life year (QALY) gained was $35 486 for the combined outreach initiative, $53 547 per QALY for the pneumococcal vaccine and $130 908 per QALY for the influenza vaccine. The cost-effectiveness ratio of the intervention targeted to people who had never received the influenza vaccine the previous year was $11 771 per QALY.

The remaining studies did not fulfil the definition of a full economic evaluation but contained some data on the vaccination- and human resource-related costs of vaccination programmes. Of these, four studies looked at LHWs delivering vaccinations only [20‐23], five studies evaluated LHWs to promote vaccinations [24‐28] (including canvassing, publicizing and persuading people to get vaccinated), and two studies reported using LHWs for both promotion and vaccination [29, 30]. Comparing costs in any meaningful way was difficult due to the differences in outcome reporting. More in-depth descriptions of these studies can be found in Additional file 3.

Despite keeping the inclusion criteria broad and general for sensitivity purposes, and despite systematically searching a large number of databases, there was a dearth of published economic evaluations of LHWs in vaccination programmes. Recently published studies point to the potential expansion of LHW involvement in vaccine delivery, especially related to the latest vaccine-related technological innovations, such as thermostable vaccines [33] and Uniject devices [23]. Combined with the emerging trend of adding more services to immunisation campaigns (e.g. vitamin A, insecticide-treated nets, etc.), we may see more studies reporting the use of LHWs in the future.

The results of the three economic evaluations included in the systematic review show that LHWs were more cost-effective options than the comparator, which did not include LHWs. However, given the diversity in the population groups targeted, as well as in the types of interventions and settings, it is difficult to draw generalizations from these studies. For example, Weaver et al. [19] found that targeting interventions to people who had never received the pneumococcal vaccine or who had not received the influenza vaccine in the previous year improved cost-effectiveness, while Deuson et al. [17] focused on increasing coverage of Hepatitis B vaccination for first-generation children of Asian and Pacific Islander descent, aged between two and 13 years.

The inclusion criteria for this review excluded studies not mentioning lay health workers, vaccines or economic evaluations, or terms related to these. Studies were included when they specifically mentioned LHW involvement in vaccination alongside other health services and indicated costs [26, 28]. However, we may have excluded a body of economic evaluation literature concerning the delivery of vaccinations in which LHWs were involved, but packaged with other targeted health services such as family planning interventions. Simmons et al. [34], for example, evaluated the cost-effectiveness of family planning research programmes delivered by LHWs in rural Bangladesh as compared to government programmes; they indicated that vaccines comprised 0.12% of the total programme budget from 1978-1985.

Vaccine delivery by LHWs can be characterized as a complex intervention, whose components usually include behaviours, parameters of behaviours (e.g. frequency, timing) and methods of organizing and delivering those behaviours (e.g. type(s) of practitioner, setting and location); the number of groups or organizational levels targeted by the intervention; and the number and variability of outcomes [35]. To add to the complexity, vaccination programmes are bundled increasingly with other health campaigns, offering a challenge in determining the cost-effectiveness of the immunisation component. For example, a recent cost-effectiveness analysis was conducted of insecticide-treated net (ITN) distribution as part of the 2004 measles vaccination campaign in Togo, with shared costs assumed to be equally attributed between the two health interventions [36]. The results suggested that substantial efficiency gains may be derived from the joint delivery of vaccination campaigns and malaria interventions [36]. Because it is rare for vaccinations or other health services to be delivered in isolation from one another, it is often difficult to determine the indirect costs associated with immunisations in particular. As can be seen by the paucity of full economic evaluations of LHWs and vaccination found in this review, it is also difficult to evaluate the costs associated solely with LHW involvement, mainly due to the interaction of various types of health personnel in service provision. For example, an evaluation of house-to-house versus fixed-site oral polio vaccine delivery strategies in a mass immunisation campaign in Egypt included the costs of physicians, nurses, hygienists, clerks and drivers, in addition to community workers, with differences in personnel costs not only linked to fixed-site versus house-visit, but also linked to urban versus rural areas [30]. Therefore it is difficult, if not impossible, to tease out the contribution of the LHWs.

Like effectiveness outcomes, the costs of (complex) interventions can be strongly determined by contextual factors; by the exact combination and 'dose' of intervention components; or by the behavioural predispositions of participants or providers. A population's attitude toward health care and interventions, compliance and adherence, utility valuations of health status, and incentives---such as level of co-payment---are also important components that can have a significant impact on cost-effectiveness [37]. The difficulty in generalising or transferring economic evaluation results to other settings arises because we do not know what caused the particular relationship between opportunity costs and outcomes in each instance. As interventions become more complex, it becomes even more difficult to explain how a specific bundle of intervention components (and their associated resource use), provided in a given context, has generated the levels and types of outcomes measured [38, 39].

In his review 'Systematic reviews of economic evaluations: utility or futility?', Anderson argues that it has become increasingly recognised in public health and health promotion that only asking whether an intervention "is effective" has limited value, because effectiveness is more complex and contingent on the specific combination of elements in an intervention, and/or its interaction with different community and organisational contexts [51]. Rather, he argues, it makes much more sense to ask "how and why" an intervention is or is not effective or cost-effective in different circumstances. As noted by Drummond, "there is widespread recognition amongst economists, and possibly amongst decision makers, that whether or not a particular intervention is cost-effective depends on the local situation" [16]. However, a common characteristic of economic evaluation studies in healthcare is that though sensitivity analyses are undertaken to deal with uncertainties in the models, few studies look explicitly at variability between locations [37], let alone attempt to explain how different levels of resources contribute to different levels and combinations of outcomes.

This review highlights the dearth of LHW vaccination strategies that have been evaluated on an economic basis. The very small number of studies identified that evaluated the economic aspects of LHWs promoting or delivering vaccination, as well as the heterogeneity of these studies, makes it difficult to draw conclusions on whether the use of LHWs in vaccination programmes represents good value for the resources invested. The lack of studies is especially surprising given that vaccination is one of the most cost-effective public health interventions [52] and that vaccination comprises a basic component of primary health care and comprises a key part of Millennium Development Goal 4 [53].

It is conceivable that with a larger number of economic evaluations than these three studies, specific characteristics of LHWs in vaccination programmes that could be generalized to help inform decision making would have been identified. The current lack of standardization in the design, analysis and reporting of results from economic evaluations, and substantially different outcomes [54], also lead to a lack of comparability. In this review, outcomes of the included studies were: cost per discounted year of life [17], cost per fully vaccinated child [18] and cost per quality adjusted life year [19]. Though there is a role for peer review to play in upholding and regulating reporting standards for the economic evaluations published [55], as well as in the quality of the studies published, there is also a need for more consistency in adhering to the numerous recommendations and guidelines for conducting economic evaluations [16, 56]. This, in turn, would aid the potential of systematic reviews to provide insights for planning and decision making.

Further research on the costs and cost-effectiveness of LHWs in delivering and promoting vaccinations is needed (Table 3), especially with closer examination of: the links between LHW-roles and strengthened primary-care facilities and first-referral services [3]; potential LHW involvement in long-term human resource planning; better training and supportive supervision [57]; the substitution of nursing and other professional tasks by lay workers (e.g. CHWs, pharmacy assistants) [3]; and cost-effective approaches to determining the allocation of PHC services based on health needs [3]. Concomitant with technological advancements in improving the safety, efficiency and thermostability of vaccinations, assessments should be conducted as to how LHWs may provide an increasingly important role in vaccine delivery at the community level.

Building on these recommendations, it is proposed that this area of research would also benefit strongly from a randomized community trial, or series of trials, comparing the cost-effectiveness of LHWs for vaccination in a range of low-, middle- and high-income settings. It is important to adhere to existing guidelines for the conduct of cost-effectiveness studies and to build on these by using the holistic economic evaluation framework proposed by Jan et al. [11]. This would aid in incorporating aspects of institutionalist economics, which takes into account context-specific norms and values, and better reflects the wider social value of health programmes within a community. Further to this, taking into consideration sustainability issues will help ensure continuing programme responsiveness to community needs, and allow LHWs to maximise their effectiveness in the context in which they are working.