ReviewTRIVAC decision-support model for evaluating the cost-effectiveness of Haemophilus influenzae type b, pneumococcal and rotavirus vaccination
Introduction
The Expanded Programme on Immunization (EPI) has already had a major impact on the numbers of deaths and episodes of disease caused by diphtheria, pertussis, tetanus, polio and measles [1]. In more recent years a new generation of vaccines has become available. Two of these (Haemophilus influenzae type b [Hib] vaccine and pneumococcal conjugate vaccine [PCV]) protect against pneumonia, meningitis and other invasive bacterial diseases, and one (rotavirus vaccine [RV]) protects against severe diarrhoea. In the year 2008, diarrhoea and pneumonia were estimated to have caused 28% of all deaths in children aged less than 5 years. With the inclusion of meningitis and other invasive bacterial diseases, this fraction increases to around one-third [2]. Hib vaccine, PCV and RV have the potential to prevent a significant proportion of these deaths and have already been introduced in many countries.
Historically, vaccines have been among the ‘best buys’ for public health programmes, costing a matter of cents per dose [3]. The newer vaccines are more expensive. If the aim is universal coverage, even a relatively modest cost for each child can add up to a large amount for a population, so budgetary constraints and prioritisation emerge as critical issues.
Inevitably government decisions about the introduction of new vaccines will be subject to advocacy and lobbying by interested parties. However, if the maximum public health benefit is to be had from health care spending, prioritisation should be based on how much improvement in health an intervention produces in relation to its cost. Cost-effectiveness depends on factors such as the incidence and severity of the diseases in question, the effectiveness and cost of the vaccine, and to some extent the health care costs avoided by preventing disease. Most of these factors vary considerably between countries, so cost-effectiveness will also vary by country. Ideally then, each country should determine its own priorities using evidence relevant to its own circumstances. There is growing recognition of and support for this principle, both in general and for vaccines in particular [4].
The evidence is not straightforward however. There is a good deal of uncertainty about the scale of the burden of disease and how much of it could be prevented by these new vaccines. There is also uncertainty about the future prices of new vaccines and the extent and duration of funding support from the GAVI Alliance and other donors. Uncertainties of this kind can lead to a view that evidence-based decision-making is ‘too hard’, leading in turn either to decision paralysis or to decisions based on less worthy considerations. This is where decision support models (DSMs) can be helpful, not so much to provide the answer as to provide a framework within which data and assumptions are made explicit, and the implications of choosing different options under different scenarios can be explored. In particular, they allow the implications of the evidence, and the uncertainties about the evidence, to be presented in a transparent and coherent way.
In this paper we describe the structure, methods and assumptions of the TRIVAC decision support model. The model has been designed for use at country level by teams led by the Ministry of Health [5], [6]. The development of the model was funded by the Pan American Health Organization (PAHO) ProVac Initiative (www.paho.org/provac) and GAVIs Hib Initiative (www.hibaction.org). To date, it has been used to perform national and/or sub-national cost-effectiveness evaluations in Albania, Argentina, Belarus, Bolivia, Costa Rica, Ecuador, El Salvador, Guatemala, India, Nicaragua, Pakistan, Panama, Paraguay, Peru and Uzbekistan.
In 2011, the PAHO ProVac Initiative received requests from other regions (e.g. Africa, Eastern Mediterranean, Eastern Europe) to use both the TRIVAC model and the ProVac methodology of which the model is just one part. This methodology involves providing practical training at facilitated regional workshops, forming national teams, providing technical assistance, and supporting presentation of results to National Immunization Technical Advisory Groups (NITAGs) and high-level Ministry of Health authorities [5]. In response to these requests, a ProVac International Working Group (IWG) has been established with funding from the Bill and Melinda Gates Foundation [see Toscano C, same Supplement].
One of the challenges of using a standardised model in a diverse set of countries is that the influence of each parameter may vary across settings. A better understanding of this from the outset should allow national teams with limited time and resources to prioritise their data collection efforts and scenario analyses.
This paper aims to: (i) provide a methodological reference document for national teams and partners working with the TRIVAC model; and (ii) identify the TRIVAC parameters likely to have the greatest influence on the cost-effectiveness of Hib vaccine, PCV and RV, in different epidemiological and economic contexts.
Section snippets
Methods
The TRIVAC model is a spreadsheet software program that calculates incremental cost-effectiveness ratios (ICERs) and other indicators for three childhood vaccines (Hib, PCV and RV) utilising parameters such as demography, disease burden, vaccine costs, vaccine coverage, vaccine efficacy, health service utilisation and costs. Where more than one vaccine is evaluated, the common model framework provides a consistent basis for comparison.
The model is designed for use in low- and middle-income
Results
Fig. 3 shows the relative and absolute influence of each key parameter on the cost-effectiveness of Hib vaccine, PCV and RV respectively, from a government perspective, in each WHO sub-region. The full country breakdown is available in the Appendix. For all three vaccines, relative coverage and the herd effect multiplier were the most influential parameters. Both factors are important because they have a direct impact on the amount of health and economic benefit without affecting the programme
Strengths and weakness of the model
To date, TRIVAC has been used widely for decision support relating to PCV and RV in Latin America, and mainly for Hib vaccine in a few countries elsewhere. The challenge has been to devise a model that produces results that decision makers can believe in and defend, under sometimes hostile scrutiny, and TRIVAC has a number of advantages in this respect:
- 1.
Transparency. National teams led by the Ministry of Health, and ultimately the committees and decision makers they engage with, need to
Conflict of interest
All authors declare no conflict of interest.
Funding
PAHOs ProVac Initiative and GAVIs Hib Initiative.
References (23)
- et al.
Global, regional, and national causes of child mortality in 2008: a systematic analysis
Lancet
(2010) - et al.
Strengthening the technical capacity at country-level to make informed policy decisions on new vaccine introduction: lessons learned by PAHO's ProVac Initiative
Vaccine
(2011) - et al.
Supporting new vaccine introduction decisions: lessons learned from the Hib Initiative experience
Vaccine
(2010) - et al.
Distributional impact of rotavirus vaccination in 25 GAVI countries: estimating disparities in benefits and cost-effectiveness
Vaccine
(2012) - et al.
Timing of children's vaccinations in 45 low-income and middle-income countries: an analysis of survey data
Lancet
(2009) - Appendix G – Impact of vaccines in the 20th & 21st centuries, Centers for Disease Control and Prevention. Epidemiology...
- The Disease Control Priorities Project. Using cost-effectiveness analysis for setting health priorities. Available...
- Glassman A, Chalkidou K. Priority-setting in health. Building institutions for smarter public spending. A report of the...
- The World Bank. How we classify countries. Available from: http://data.worldbank.org/about/country-classifications...
- WHO. List of member states by WHO region and mortality stratum. Available from:...