Estimating private demand for cholera vaccines

Along with my work on the Nile, the other major research effort undertaken during my graduate studies was related to estimating the private benefits of cholera vaccines, and then conducting policy analyses by combining these data with assessments of vaccine costs. The benefit valuation work used a pair of nonmarket valuation methodologies commonly used by environmental economists but rarely applied to vaccines – the contingent valuation method (a stated preference approach) and the travel cost method (a revealed preference approach). I led the fieldwork in Beira, Mozambique, one of the global hotspots for cholera, while colleagues conducted similar studies (and several for typhoid as well) in locations such as Jakarta Indonesia, Kolkata India, Dhaka Bangladesh, and Hue Vietnam, among others.

Cholera vaccine demand site Beira Mozambique Figure 2

Cholera vaccine demand site

Main research collaborators:

  • Marcelino Lucas (Government of Mozambique)
  • John Clemens, Brian Maskery (IVI)
  • Dale Whittington; Don Lauria; Melissa McMahon (UNC-Chapel Hill)
  • Joe Cook (University of Washington)
  • Christy Poulos (RTI)

Related Publications:

Jeuland, M.; M. Lucas; J. Clemens; D. Whittington (2009). “Estimating the private benefits of vaccination against cholera in Beira, Mozambique: a travel cost approach.” Journal of Development Economics 91: 310-322.

Abstract: This paper reports the results of the first study that estimates households’ private demand for cholera vaccines using the travel cost method.We take advantage of an unusual natural experiment. In January 2004, more than 41,000 residents from various locations in Beira, Mozambique received two doses of oral cholera vaccine free of charge during the first vaccination trial to test its effectiveness in an endemic cholera zone of Africa. About 30,000 people participated from outside the target zone, resulting in long queues and an average waiting time of about 85 min per dose. We estimated travel cost models of the revealed demand for cholera vaccines among households informed of the trial using information collected in in-person interviews conducted during the summer of 2005. To explore households’ participation in the trial, we used standard and zero-inflated household count models for all household members and dichotomous choice models for the head of the household. Our analysis shows that the quantity of vaccines obtained by households and the likelihood of participation decreased as travel cost—in time and transport expenses—rose. Our best estimates of per capita willingness to pay for the two required doses of cholera vaccine are about 0.85 USD. These estimates are sensitive to the assumed value of time spent acquiring vaccines.

Jeuland, M.; J. Cook; J. Clemens; C. Poulos; D. Whittington; DOMI Cholera Economics Study Group (2009). “Incorporating herd protection into cost-effectiveness calculations of new generation oral cholera vaccines: a multi-site analysis.” Value in Health 12 (6), 899-908.

Objectives: We evaluated the cost-effectiveness of a low-cost cholera vaccine licensed and used in Vietnam, using recently collected data from four developing countries where cholera is endemic. Our analysis incorporated new findings on vaccine herd protective effects. Methods: Using data from Matlab, Bangladesh, Kolkata, India, North Jakarta, Indonesia, and Beira, Mozambique, we calculated the net public cost per disability-adjusted life year avoided for three immunization strategies: 1) school-based vaccination of children 5 to 14 years of age; 2) school-based vaccination of school children plus use of the schools to vaccinate children aged 1 to 4 years; and 3) community-based vaccination of persons aged 1 year and older. Results: We determined cost-effectiveness when vaccine herd protection was or was not considered, and compared this with commonly accepted cutoffs of gross domestic product (GDP) per person to classify interventions as cost-effective or very-cost effective. Without including herd protective effects, deployment of this vaccine would be cost-effective only in school-based programs in Kolkata and Beira. In contrast, after considering vaccine herd protection, all three programs were judged very cost-effective in Kolkata and Beira. Because these cost-effectiveness calculations include herd protection, the results are dependent on assumed vaccination coverage rates. Conclusions: Ignoring the indirect effects of cholera vaccination has led to underestimation of the cost-effectiveness of vaccination programs with oral cholera vaccines. Once these effects are included, use of the oral killed whole cell vaccine in programs to control endemic cholera meets the per capita GDP criterion in several developing country settings.

Jeuland, M.; M. Lucas; J. Clemens; D. Whittington (2009). “A Cost Benefit Analysis of Vaccination Programs in Beira, Mozambique.” World Bank Economic Review 23 (2):235-267.

Abstract: Economic and epidemiological data collected in Beira, Mozambique, are used to conduct this first social cost–benefit analysis for cholera vaccination in Sub-Saharan Africa. The analysis compares the net economic benefits of three immunization strategies with and without user fees: school-based vaccination for school children only (age 5–14), school-based vaccination for all children (age 1–14), and a mass vaccination campaign for all people older than one year. All options assume the use of a low-cost new-generation oral cholera vaccine. The analysis incorporates the latest knowledge of vaccine effectiveness, including new evidence on the positive externality associated with the resulting herd protection (both protection of unvaccinated individuals and enhanced protection among vaccinated individuals arising from vaccination of a portion of the population). It also uses field data for incidence, benefits (private willingness to pay, public cost of illness), and costs (production, shipping, delivery, private travel costs). Taking herd protection into account has important economic implications. For a wide variety of parameters values, vaccination programs in Beira pass a cost–benefit test. Small school-based programs with and without user fees are very likely to provide net benefits. A mass vaccination campaign without user fees would result in the greatest reduction in the disease burden, but the social costs would likely outweigh the benefits, and such a program would require substantial public sector investment. As user fees increase, mass vaccination becomes much more attractive, and the reduction in disease burden remains above 70 percent at relatively low user fees.

Cook, J.; M. Jeuland; B. Maskery; D. Lauria; D. Sur; J. Clemens; D. Whittington (2008). “Using private demand studies to calculate socially optimal vaccine subsidies in developing countries.” Journal of Policy Analysis and Management 28 (1): 6-28.

Abstract: Although it is well known that vaccines against many infectious diseases confer positive economic externalities via indirect protection, analysts have typically ignored possible herd protection effects in policy analyses of vaccination programs. Despite a growing literature on the economic theory of vaccine externalities and several innovative mathematical modeling approaches, there have been almost no empirical applications.
The first objective of the paper is to develop a transparent, accessible economic framework for assessing the private and social economic benefits of vaccination. We also describe how stated preference studies (for example, contingent valuation and choice modeling) can be useful sources of economic data for this analytic framework. We demonstrate socially optimal policies using a graphical approach, starting with a standard textbook depiction of Pigouvian subsidies applied to herd protection from vaccination programs. We also describe nonstandard depictions that highlight some counterintuitive implications of herd protection that we feel are not commonly understood in the applied policy literature.
We illustrate the approach using economic and epidemiological data from two neighborhoods in Kolkata, India. We use recently published epidemiological data on the indirect effects of cholera vaccination in Matlab, Bangladesh (Ali et al., 2005) for fitting a simple mathematical model of how protection changes with vaccine coverage. We use new data on costs and private demand for cholera vaccines in Kolkata, India, and approximate the optimal Pigouvian subsidy. We find that if the optimal subsidy is unknown, selling vaccines at full marginal cost may, under some circumstances, be a preferable second-best option to providing them for free.

Lucas, M.; M. Jeuland; J. Deen; N. Lizaro; M. McMahon; A. Nyamete; A. Barreto; L. von Seidlein; A. Cumbane; F. Songane; D. Whittington (2007). “Private Demand for Cholera Vaccines in Beira, Mozambique.” Vaccine 25 (14): 2599-2609.

Abstract: In the summer of 2005, we interviewed 996 randomly selected respondents in Beira, Mozambique concerning their willingness and ability to pay for cholera vaccine for themselves and for other household members. Respondents were told that two doses of the vaccine would be required 2 weeks apart, and that the cholera vaccine would offer excellent protection against infection for the first year following vaccination, and some protection during the second and third year after a person is vaccinated. This research was carried out in order to learn more about private demand for vaccines in a cholera-endemic area.We asked two types of valuation questions: (1) a discrete-price offer for a vaccine that could be purchased for household members and (2) a payment card designed to assess uncertainty in the respondent’s demand for a vaccine for self-protection.We estimate average household willingness to pay (WTP) for cholera vaccines in Beira to be 2005 US$ 8.45. This estimate of household WTP represents the perceived private economic benefits to a household – six persons on average – of giving all members free cholera vaccines.