Cost-Effectiveness of Interventions against Tuberculosis
Some questions about investing in TB control are broad and strategic (for example, should money be spent on the control of TB rather than on the control of some other condition?); others are specific and technical (for example, which laboratory diagnostic procedures should be used?). On whatever level the question is posed, cost-effectiveness analysis (CEA) has become a prominent method for evaluating and choosing among different health interventions.
Background
Between 1980 and 2004, 32 studies of the cost-effectiveness of TB control were published from the low- and middle-income countries considered by the Disease Control Priorities Project (table 16.1; online annex 3 summarizes the 32 studies that have been published according to the country and year of publication, the question being addressed, the strategies compared, the subjects and costs considered, the effectiveness of measures used, whether or not transmission is considered, and the main results and conclusions). Almost all of these studies (28, or 88 percent) have concerned ways of finding, diagnosing, and treating patients with active TB, and most (18, or 56 percent) have been done in eight countries in Sub-Saharan Africa (Floyd 2003). Three studies (all in Sub-Saharan Africa) have investigated TLTI, and one study in Indonesia has examined BCG vaccination. The principal findings are that short-course chemotherapy for active TB is a comparatively cost-effective intervention and one of the most cost-effective of all health interventions. TB patients can be treated more cheaply and conveniently outside hospitals on an ambulatory basis, by health staff or with the help of family and community members, without compromising the success of treatment. Supplementary methods, such as standardized second-line drug treatment for MDR-TB, appear to be affordable and cost-effective in some settings.
[Table .]
What does not emerge from this compilation of data is a comprehensive overview of the value for money provided by current and potential interventions against TB in all major regions of the world, expressed using a common measure of effectiveness and based on a consistent approach to the evaluation of transmission. (The returns on investment in infectious disease control include the immediate benefits to individuals treated—for example, those vaccinated or given drug therapy—plus the longer-term benefits gained by preventing secondary cases through reduced transmission.) Little work has been done in China, India, and other large countries in Asia, even though Asia carries the largest burden of TB, and only limited information is available for Europe and Central Asia, Latin America and the Caribbean, and the Middle East and North Africa. Of the 32 studies, only 10 used a measure of effectiveness that allows comparison with other diseases (table 16.2), and only 9 attempted to include an estimate of the benefits gained from reduced transmission (table 16.1). The benefits from reduced transmission are usually assessed through mathematical modeling (using computer simulations) for a given epidemiological situation, an approach that produces specific solutions for each setting rather than results that are generally applicable. In addition, although the benefits from prevented transmission are lower when TB is endemic, existing studies do not make a clear distinction between the cost-effectiveness of interventions in epidemic (outbreak) and endemic situations.
[Table .]
Methods
In this study, a general analytical framework was used to evaluate the total costs and total effects (defined as cases prevented, deaths averted, and DALYs gained) of the principal interventions against TB across six regions of the world (see online annexes 4-7 for further details). A dynamic infectious disease model (online annex 4) was used to derive general formulas for calculating the cost-effectiveness of interventions to control endemic (online annex 5) and epidemic (online annex 6) TB in a wide variety of settings. The formulas are approximate, but they are simple and able to provide insights into the strategies that give value for money under a wide variety of epidemiological circumstances. The model was then supplied with cost and efficacy data (online annex 7) for each of the six World Bank regions for four main groups of interventions:
-
immunization with BCG (proportion of infants, m, assumed to be protected against severe, noninfectious childhood TB only), or a new vaccine that prevents infection and progression to pulmonary and extrapulmonary TB in children and adults
-
isoniazid treatment of latent TB infection (TLTI, given at per capita rate Rho), for people infected with M. tuberculosis, with or without HIV coinfection and with or without the use of radiography to exclude patients with active disease
-
short-course chemotherapy, delivered as a component of the DOTS strategy, for smear-positive or smear-negative pulmonary disease and extrapulmonary disease (with a combination of drugs given at per capita rate Tau), and for patients infected with HIV, with or without supporting anti-retroviral therapy
-
treatment for MDR-TB using a standardized regimen including first- and second-line drugs or using individualized regimens of first- and second-line drugs that are tailored to each patient's drug susceptibility pattern.
Costs were considered from a health system or provider perspective. They were calculated by combining estimates of the quantities of resources required for each intervention (per patient or per person treated) with the unit prices of those resources (in 2001 U.S. dollars) using the cost categories and unit prices defined in the Disease Control Priorities costing guidelines.
