Introduction: This article reports the results and the lessons learned from implementing the decentralized approach to tuberculosis (TB) detection and treatment, embedded with Human Immunodeficiency Virus (HIV) co-infection in health district. The objective was to increase the TB screening indicators in the district using the common ways for offering care to patients in health district.

Methods: Conducted from August 2006 to July 2007, this large-scale intervention using Non-experimental study Designs has implemented a decentralized approach for fighting against TB in Orodara Health District (OHD), Burkina Faso. Pretest-posttest design has been used for quantitative part using indicators in one hand, and postests-only design for the qualitative part in other hand. In the pretest-posttest design, the TB indicators from years before 2006 (from 2002 to 2005) were used as earlier measurement observations allowing examining changes over time. The decentralized approach was incorporated into the annual planning of the OHD. For the quantitative study design, indicators used were those from National TB Program in Burkina Faso: TB detection rate, incidence density of TB per 100,000 inhabitants per year, and HIV prevalence in incident TB cases with positive smears. Data entry and analysis employed Microsoft Access and Excel software. For the qualitative, in-depth interview was used in which a total of 16 persons have been interviewed. Discussions were tape-recorded and transcribed verbatim for analysis using the computer-based qualitative software program named QSR NVIVO

Results: There were a total of 99,259 outpatient visits during the study period: the7,345 patients (7.43%) presented with cough. Of the 7,345 patient having cough, 503 cases (6.8%) were declared chronic coughing. These 503 patients were screened for TB, including 35.59% whose coughing had lasted 10 to 15 days. We observed an increase in a measured variable was observed. The TB detection rate and incidence-density rate based on positive smears were 16.11% (11.00% in 2005) and 10.42 per 100,000 inhabitants per year (6.88 per 100,000 inhabitants in 2005), respectively. There were 29 patients positive for TB: 41.37% of these had cough lasting 10 to 15 days, 10.34% were also positive for HIV, and 68.97% were from rural areas. Health workers and patients reported satisfaction with the intervention. It was found that implementing a decentralized approach to TB prevention in rural areas is plausible and effective under some conditions: considering that health district system is functional; carefully designing the intervention for TB case management; setting up and implementing of decentralized approach including strong monitoring; and taking into account the all financing, community and volunteer involvement, evaluation of the cost savings from integrating specific donor funding, and being supported by regional and central levels including National TB program.

Conclusion: The study has shown that TB detection rate can be increased by implementing a decentralized approach to primary care. When carefully implemented, a decentralized approach is a suitable approach to TB and HIV prevention in rural and inaccessible settings.