Abstract
BACKGROUND: In Ethiopia individual report indicated nearly 30% of incident cases of tuberculosis remained undiagnosed. Therefore, this systematic review and meta-analysis was aimed to determine the pooled prevalence rate of undiagnosed smear positive pulmonary tuberculosis (PTB) using community based studies published in Ethiopia. METHODS: MEDLINE/PubMed, 'Cochrane' library, and Google scholar databases were searched, and reference list of studies on tuberculosis in Ethiopia were reviewed. We used table to present descriptive information of original studies and quantitative results were presented in forest plots. The Cochrane Q test and I(2) test statistic were used to test heterogeneity across studies. The Pooled prevalence and point estimates of undiagnosed smear positive PTB were computed by a random effects model. RESULTS: From the nine studies included in the analysis, the pooled prevalence rate and point estimate of undiagnosed smear positive PTB was 0.11%(95% CI, 0.06-013%, p < 0.001) and 79.8/100,000(95% CI; 56.3-112.8) respectively. Pooled prevalence rate and point estimate of bacteriologically confirmed PTB were 0.17%(95%CI; 0.13-0.22%, P < 0.001) and 191/100000(95% CI; 141.3-258) respectively. The ratio of active to passive case detection was 2.3(95% CI, 0.42-4.1). Pooled prevalence rate of presumptive PTB was 2.7%(95% CI; 1.3-5.3%). CONCLUSIONS: The analysis revealed that the magnitude of undiagnosed smear positive PTB cases in the community is high in Ethiopia. This indicated the ongoing transmission of tuberculosis in community due to missed infectious cases. Active tuberculosis finding in the community should be strengthened in Ethiopia. TRIAL REGISTRATION: 140611.