The First National Tuberculosis Prevalence Survey in the United Republic of Tanzania\ud Final Report

Tanzania is classified as one of the 22 high burden countries for tuberculosis (TB). It was the first country in the world to use the now standard Direct Observed Treatment Short Course to treat tuberculosis. The National Tuberculosis and Leprosy Programme (NTLP) was established in 1977. Although the routine TB surveillance data are consistent over the years, there are still areas of uncertainty, which makes that these routine data cannot be translated easily into an approximation of TB incidence as an indicator for the burden of disease. This lack of information on the true burden of TB disease in the country justified the conduct of a national TB prevalence survey to provide the muchneeded context in which all other available data can be re-assessed. The survey was designed as a nation-wide population-based survey in the adult population, in which districts were randomly selected, followed by a random selection of a single ward (denoted as cluster) within each district. A set number of participants in each ward was invited to participate in the survey. Participants were screened for being suspect of having TB by a simple symptom questionnaire and a chest X-ray (CXR). Identified TB-suspects were requested to submit three sputum specimens, of which two were assessed by microscopy in a field laboratory and the third was transported to the CTRL for culture. The prevalence of bacteriological confirmed TB was 295 per 100,000 adult populations. Prevalence was higher in mainland Tanzania compared to Zanzibar, rural compared to urban populations, men compared to women, older compared to younger participants and in participants with lower compared to higher socio-economic position. The prevalence of HIV-infection in identified TB cases was 6.8%. Case Detection of new smear-positive adult TB patients was estimated to be between 42 and 54%. The majority of identified TB cases were 54 years or older, indicating a shifting epidemic from young HIV-infected patients. The survey was conducted to high standards as acknowledged by external monitors from KNCV Tuberculosis Foundation (provided Technical Consultant from the start of the study design) and the WHO Task Force on Impact Measurements. The detailed protocol and SOPs facilitated the implementation of the survey. The support by the Ministry of Health and Social Welfare, the close cooperation with community leaders and local NTLP staff, and the assistance of Community Health Workers, facilitated the implementation of the survey. The survey was conducted in the adult population only, which makes it impossible to assess the burden of childhood TB. Data analysis was hampered by missing data due to recording errors and misplacement of survey records, especially for the central laboratory. However, formal imputation analyses to account for this situation did not change the conclusions of the survey. The prevalence of bacteriological TB in the adult population of Tanzania is higher than expected; the case detection of new smear-positive adults is markedly lower than previously reported. There is an urgent need to assess patient identification and the conduct of laboratory procedures in the diagnostic centres. This can be achieved by intensifying supportive supervision in the country which has been decreased in frequency and intensity during the last few years