Willingness to pay for prevention and treatment of lymphatic filariasis in Leogane, Haiti

Lymphatic filariasis, a mosquito-transmitted parasitic disease, affects an estimated 120 million people worldwide, causing lymphatic dysfunction in virtually all those infected and recurrent acute adenolymphangitis (ADL), hydrocele, and lymphedema in millions of affected persons. Infection with the parasite results in damage to lymphatic vessels, leading to swelling of the limbs and genitals and painful bacterial infections. As the swelling progresses, it becomes difficult for individuals to carry out their household and job-related activities [1]. The disfigurement caused by the disease can also lead to social stigma and difficulty in fulfilling social roles. The burden of the disease includes medical expenses, lost productivity, value of familial assistance, and the value associated with diminished social function and quality of life [2].

Following the 1997 declaration by the World Health Assembly and the subsequent drug donations by SmithKline Beecham (now GlaxoSmithKline) and Merck & Co. Inc., global and local efforts to eliminate LF as a public health problem were intensified. These efforts are directed toward interrupting transmission of the parasite and providing treatment for people with clinical filarial disease. In selecting and allocating resources, local and national decision makers along with international donors need information on the value that filariasis-endemic communities place on prevention and treatment services. Several studies have estimated the economic value of certain aspects of disease burden, particularly the cost of medical attention and the productivity losses associated with the disease [1, 4–6]. However anthropological studies of communities with endemic LF suggest that the more intangible benefits of preventing and treating filarial disease (e.g. reduced social stigma) may provide a greater contribution to social wellbeing [7]. This study uses a contingent valuation method to estimate the economic value of both the direct and intangible benefits of preventing and treating lymphatic filariasis in a Haitian community. Contingent valuation studies are typically used to characterize the mean or total value placed on an improvement in health within a community, and compare it to the costs of the improvement. Here the approach also is used to examine distribution of household willingness to pay (WTP) for interventions within the community and the expected impact on participation rates, to provide a more complete measure of the benefits of these projects.

Lymphatic filariasis is endemic to parts of Haiti, including the town of Leogane and its surrounding communities. The descriptions of the treatment and prevention programs in the valuation exercises are modeled after projects conducted in the Leogane area. The treatment intervention includes the management of lymphedema and prevention of adenolymphangitis (ADLs) through improved hygiene, skin care, and physical measures, such as elevation and movement. Two prevention strategies can be used to interrupt transmission of LF, either regular use of table salt fortified with diethylcarbamazine (DEC) [8] or mass administration with a 2-drug combination: DEC plus albendazole or ivermectin plus albendazole [9]. Both kill the microfilaria produced by adult parasites, preventing further transmission. For this study DEC-fortified salt was the intervention used in the valuation exercise.

Willingness to pay studies are usually used to estimate the economic value of the burden created by a disease or the benefit of reducing that burden as a result of a specific intervention. Although no studies have directly estimated the medical costs or productivity losses from filariasis in this community, the estimated WTP for LF prevention in this community ($252,000 / year) is likely to exceed actual medical expenses and productivity losses. The WTP values estimated in this study reflect the value of avoided pain, suffering, and reduced social functioning, as well as any expected medical expense savings or lost earnings.

The monetary value of improved health associated with an intervention, when combined with other benefit measures, can be compared to the costs of the intervention in a cost-benefit analysis. Estimating the costs of LF treatment and prevention programs is beyond the scope of this paper. However, the estimated annual value of $12 per person greatly exceeds initial estimates of the annual per capita costs of mass treatment with anti-filarial drugs. Estimates from the first three years of mass drug administration in Leogane, Haiti with DEC and albendazole suggest an annual cost of $1–2 per person [15]. These initial cost estimates are within the wide range of costs ($0.05 – $3) in studies published elsewhere [16, 17]. This suggests that programs to interrupt LF transmission in Haiti are likely to have very high net value, with the monetary value of benefits exceeding costs. Similarly, the mean WTP for lymphedema treatment of $805 per year greatly exceeds the actual annual costs of clinic-based treatment of individuals with lymphedema in this community, estimated at $22–78 per person depending on the stage of disease (S. Kanjilal, personal communication, 2003).

Respondents' knowledge and/or attitudes regarding filariasis affected willingness to pay for treatment and prevention. However the results do not necessarily indicate that improved information through health education will increase participation. It was initially expected that individuals who identified magical powders as a source of lymphedema would be willing to pay less for prevention or treatment interventions which they may see as ineffective. WTP for prevention was not influenced by beliefs about etiology. However WTP for treatment was negatively influenced by both the belief in magical etiology and correct knowledge about the mechanisms of filarial infection. A trend toward higher WTP for prevention was observed among people who perceived themselves to be at higher risk of getting LF. Although the result was not statistically significant, it may suggest that health education efforts that focus on informing individuals about disease risks and the opportunity for prevention and treatment may be more effective than those aimed at improving knowledge about the biology or causes of LF.

Individuals' familiarity with lymphedema or hydrocele (whether they or a family member have one of the conditions) significantly negatively affected WTP for treatment and prevention. This finding is consistent with valuation studies for other health conditions [18]. Having the condition (or having a family member with it) may lead to coping mechanisms and increased acceptance, which in turn reduce an individual's willingness to pay to treat the condition. In the case of treatment, an alternative hypothesis may be that individuals with lymphedema may be more cognizant of the immediate financial constraints on paying for treatment, or they may have felt that the prices asked about were not 'fair', given their assessment or knowledge of the actual costs of the treatment services. Another factor may be that they do not believe that the treatment would be effective. Individuals who already have filarial lymphedema also may consider it too late to prevent it.

As expected, household income positively affected estimates of mean WTP for both lymphedema treatment and LF prevention. Willingness-to-pay studies often restrict WTP estimates to be less than reported income, given that households cannot expect to be willing to pay more than their actual income [12]. In this study we did not artificially restrict WTP to be less than reported income. For some households estimated WTP for lymphedema treatment exceeded their reported income. However this may not actually represent a violation of the logical restrictions on WTP. Although households may not realistically be able to pay their entire income to prevent LF or treat lymphedema, there are other resources that contribute to household ability to pay. Respondents may base their WTP on past savings or expected future earnings, rather than their current monthly income (which may be highly variable). In addition, WTP for treatment and prevention may be strongly influenced by resources of the extended family (including individuals living and working abroad). This is supported by the influence of household construction type (as a proxy for wealth) on WTP estimates. Having a house constructed with cement (as opposed to wood or thatch) was positively related to WTP for both prevention and treatment.

In this study we used a 2-dimensional Monte Carlo model to estimate the distribution of WTP, reflecting the fact that some individuals may be willing to pay a great deal to participate in programs, while others have negative WTP. This suggests that even if the interventions were offered free of charge, they would choose not to participate. Negative WTP estimates suggest that some people may see the interventions as ineffective or not worth the individual costs associated with participating in them, such as the time, travel, and other costs. It may also reflect a sense of fatalism about their condition [7]. Their participation would require an incentive or subsidy of some sort, to offset these costs.

Such a distribution of household willingness to pay may have important program implications. For lymphedema treatment, if the cost of the intervention exceeds a household's WTP, then that household may choose not to participate. For example, if the annual cost of treatment were the equivalent of $100, the model estimates that 45% would choose not to seek treatment, if they had lymphedema. For the majority of those who would choose to participate, the value of the service (their WTP) would exceed the actual cost to them, resulting in a net benefit or surplus.

For the prevention intervention, the distribution of WTP has important implications for the ability of the program to achieve transmission interruption. The distribution suggests that as the cost of the intervention to households increases the rate of participation is likely to decline. It is believed that a minimum coverage level must be sustained for several years in order to interrupt transmission, although the precise level is unknown and is likely to differ between settings [19].

For an individual to decide to participate, the value of prevention must be perceived to be at least equal the cost incurred by participation. For a DEC-fortified salt program this could include any additional costs for the fortified salt. Programs that distribute anti-filarial drugs through mass drug distribution also may introduce costs to participating individuals. In some countries cost recovery practices are used for similar programs (e.g. onchocerciasis) to cover program costs. Households may encounter a variety of costs including transportation to distribution points or lost time from work.

Based on Figure 3, if it is assumed that 80% of households need to participate in order to interrupt transmission, the expected cost (or household contribution) for the intervention (e.g. cost recovery fees, transportation costs, or additional cost of DEC-fortified salt) would have to be below the 20^th percentile of the WTP distribution ($5.09/person/yr). Above that value, it is estimated that less than the required 80% would participate. If the actual cost per household were greater than the appropriate threshold WTP, then a subsidy may be required to reach the needed coverage rate. If 90% participation were needed, then the cost to households would have to be less than $1.53/person/yr. Other factors may also affect participation, including illness or pregnancy (contraindications to treatment), fear of adverse drug reactions, and not being available at the time of drug distribution.

Cost recovery policies may provide short-term revenue, but may also reduce participation levels, leading to the need for longer intervention periods to interrupt transmission. Policies that introduce other costs to households, such as the limiting number of drug distribution posts or not providing medication for adverse reactions free of charge, may have a similar effect. Reducing these costs may increase short-term participation, resulting in lower long-term costs. While cost recovery policies will almost always result in reduced participation, the consequences may be particularly important for LF elimination and other programs that require very high levels of participation.

Subsidies, whether financial or in the form of other health services, may be useful in increasing participation, and decreasing long-term program costs. The anthelminthic benefit resulting from co-administration of albendazole with DEC or ivermectin can be viewed as such a subsidy. Efforts to improve participation in these ways or through educational campaigns may be particularly important in specific settings – including poor communities (low income or lower quality housing) and households of individuals with clinical disease (Figure 4).

While WTP estimates capture many of the components of disease burden and program impact, the method may miss other societal benefits from treatment and prevention programs. For example, other individuals who travel into the region might be willing to pay for protection during short visits. For prevention activities, our WTP estimates do not capture the benefit to future generations who would no longer be at risk (only the value that these current residents put on future benefits to others). For treatment, we may not completely capture benefits to other household members who may benefit from another family member's reduced morbidity. Alternative measures of burden of disease and program benefits include health metrics (Disability-Adjusted Life Years, Quality-Adjusted Life Years) and monetary measures of medical costs and lost productivity. Each of these captures different aspects of a disease on individuals and society. These approaches provide burden information which complements WTP estimates of burden.

The WTP estimates generated in this study expand our understanding of the economic impact of the disease in two ways. First, WTP measures provide an estimate of welfare gain that includes changes in morbidity, economic production, and other intangible values such as social function. All of these aspects of welfare are affected by lymphatic filariasis, and should be considered when assessing the burden of the disease. Second, WTP studies help identify factors that, if addressed through social mobilization, subsidies, or other strategies, may improve program participation rates critical to an elimination goal.