Trends, challenges and opportunities in tuberculosis control in rural Ethiopia : Epidemiological and operational studies in a resource-constrained setting

Many people still die of tuberculosis (TB). One-third of the world’s population is infected with M. tuberculosis, and the poor suffer most. More than 95% of TB cases and deaths are in developing countries and TB is closely linked to poverty. The prevalence of TB increases globally, mainly because of the worsening HIV pandemic. Emerging drug-resistant TB poses another challenge to efforts to control TB. In 1991, the World Health Organization (WHO) introduced a comprehensive approach to TB control, eventually called DOTS- directly observed treatment, short course. Later WHO declared TB a global emergency. Ethiopia is one of the 22 TB high-burden countries in the world. DOTS was adopted in the first half of 1990s in a few pilot sites, and later expanded. As a result, in 2003, the TB control programme used DOTS in 95% of the public health institutions in the country. This thesis investigates some elements of DOTS and how these were carried out. The studies focus on epidemiological trends, operational challenges and opportunities to improve TB case finding and treatment outcomes in rural Ethiopia. The aim of the thesis is to assess trend in TB control efforts, estimate the burden of TB, address operational challenges and explore alternative approaches to improve TB case detection and treatment outcomes in rural Ethiopia. The studies were conducted in Hadiya in southern Ethiopia. We used cross-sectional, longitudinal observational and intervention study designs. The studies were conducted in rural communities and in public health institutions. Most of the papers focus on smear-positive TB, the most infectious form of TB. The study findings show that after five years, the DOTS population coverage reached 75%. Simultaneously, case notification and treatment outcomes of TB patients improved. Between 1994 and 2000, the treatment success for smear-positive TB rose from 38% to 73%. With a decline in treatment failure and default rates, the steadily increasing trend in treatment success suggests expanding DOTS led to a significant improvement in treatment outcomes. Trends in case notification and treatment outcomes represent proxy indicators for programme performance. However, to better evaluate the impact of a TB programme, we need both baseline and follow-up data on disease prevalence and incidence. Considering the shortage of resources, we used a simple and less expensive method to estimate the prevalence of smear-positive TB in a rural community. The results show that for every two case of smear-positive TB on anti-TB treatment, there was one undiagnosed infectious case in the community. Such a method of estimating TB burden in a population may bridge the information gap on the extent of TB in resourceconstrained settings where case-notification data are incomplete and more sophisticated approaches of estimating incidence and prevalence are not possible. An intervention study on case finding through a village outreach programme showed the intervention was effective in improving the speed of case detection for smearpositive TB. Though not statistically significant, our study shows a higher case notification rate in the intervention communities compared with the control communities. Patients in both groups had comparable treatment outcomes. This case finding method may be relevant for the new health extension programme in the country that gives due emphasis to community-based approaches. The effectiveness of such an intervention and its cost-effectiveness warrant further investigation. Improved case detection has a meaning only when the detected cases successfully complete the treatment. This study explored possible causes predicting treatment noncompletion among smear-positive pulmonary tuberculosis patients. One in five patients did not complete the treatment and the limiting factor was access to treatment. As most treatment interruption occurred during the continuation phase of TB treatment, clinicians in the TB programme should hold follow-up discussions with patients to ensure treatment compliance in this phase. Building diagnostic competence is a precondition to improved case finding. Evaluation of routine sputum microscopy for acid-fast bacilli at the diagnostic laboratories over three-year period revealed a declining trend in false readings and 97% overall agreement between the readings at the diagnostic laboratories and that in the reference laboratory. Unfortunately, the number of laboratories taking part in such quality assessment scheme declined, signalling a need to revitalise and scale-up the quality assessment service in the region