Objective: To summarise data on the implementation of the DOTS strategy in China in terms of actual observation and treatment adherence, and to review the effectiveness of quality improvement interventions for tuberculosis (TB) control in China.
Design: We included survey studies that reported data on the implementation of DOTS in China and controlled studies that evaluated TB care in specified communities. We excluded studies outside mainland China, pharmacological intervention trials and reviews.
Results: We included 12 survey studies that reported on the performance of TB control services in China. The pooled analysis showed that more than half of TB patients were treated by self-administration (52%) and that only 20% actually had their treatment observed by health workers. We include 85 intervention studies that evaluated the effect of quality improvement interventions. Treatment observers were family members in 37 studies, and health workers in 20 studies. The pooled odds ratio (OR) for cure was 2.48 (95%CI 1.97-3.11, I(2) = 70.9%, P < 0.001); the pooled OR for treatment completion was 2.87 (95%CI 2.23-3.69, I(2) = 66.3%, P < 0.001). Sensitivity analysis found that the estimated treatment effects in meta-analyses using reported and imputed data were much reduced, but still statistically significant.
Conclusion: The proportion of TB patients whose treatment was strictly observed was much lower than reported by official statistics in China. The treatment completion rate was not optimal, which may be an important reason for the reported increases in drug resistance. Community health personnel have become the main work force for TB control in China.