Figure 3.

Summary of study quality assessment. If duration of follow-up was less than 1 month, studies were classified as having a high risk of bias. If more than 15% of patients were lost to follow-up, studies were classified as having a high risk of bias. If all patients were hospitalised or had a rare form of TB, studies were classified as having a high risk of bias. Unclear classifications indicate that insufficient information was available to assess these areas. If studies had a high risk of bias in any of the aforementioned areas, the study was classified as low quality. TB, tuberculosis.