This review assessed the efficacy of standard directly observed treatment, short-course 6-month regimens of treatment in preventing the recurrence of tuberculosis. The authors concluded that few studies have assessed recurrence rates of such regimens under routine conditions. Poor reporting of review methodology means that this conclusion is probably, though not certainly, reliable.

To assess the efficacy of standard directly observed treatment, short-course (DOTS) regimens of treatment for 6 months in preventing the recurrence of tuberculosis (TB).

PubMed (from 1950), EMBASE (from 1974) and the Cochrane Controlled Trials Register were searched to December 2006; the search terms were reported. Four relevant journals were handsearched, reference lists were examined and related reports were identified. Only studies published in English were eligible for inclusion.

Studies of any design that evaluated one of the six recommended DOTS regimens for the treatment of pulmonary TB, and reported recurrence rates after successful treatment, were eligible for inclusion. The included studies used different combinations of isoniazid, rifampicin, pyrazinamide and ethambutol. Study definitions of recurrence were accepted. Studies with active and passive follow-up of patients were both eligible for inclusion. The included studies were of both new and previously treated patients. The majority of studies did not test for drug susceptibility. In the included studies, patients were either human immunodeficiency virus (HIV) negative or were not tested for HIV. Follow-up in the included studies ranged from 12 to 60 months. The included studies were randomised controlled trials (RCTs), controlled clinical trials (CCTs), uncontrolled prospective studies and observational studies under programme conditions. The studies were conducted in a wide range of countries; the majority were in developing nations, including four from India. The included studies used differing definitions of treatment success and recurrence.

The authors did not state how the papers were selected for the review, or how many reviewers performed the selection.

The authors assessed the reporting of measures to ensure adherence to initial treatment, and withdrawals from initial treatment, as well as the adequacy of follow-up procedures used to determine recurrence and the percentage of patients excluded from the follow-up analysis. The studies were graded as 'high', 'average' or 'poor' on each criterion.

The authors did not state how many reviewers performed the validity assessment.

For each study, the percentage of patients with recurrent TB was extracted and the 95%confidence intervals calculated using the exact binomial method. Recurrence was classified on the basis of whether new TB had occurred, HIV status, drug-resistance and bacteriological confirmation of treatment success.

The authors did not state how many reviewers performed the data extraction.

The studies were combined in a narrative and differences between the studies were discussed.

Sixteen studies (at least 4,246 participants) were included in the review: 6 RCTs, 3 CCTs, 6 observational studies under programme conditions and one prospective study under programme conditions.

Two studies were rated 'high quality' for TB treatment, seven were rated 'average' and seven were 'poor'. For follow-up, 8 studies were rated 'high quality', three were 'average' and five were 'poor'. Only one study differentiated between reinfection and relapse.

Rates of TB recurrence in the included studies ranged from 0 to 14%. Reported rates of recurrence tended to be highest in studies rated as poor quality for treatment carried out in Uzbekistan, South Africa and India. The authors noted a high level of clinical and methodological heterogeneity across a range of variables; in particular, they noted the more restrictive enrolment criteria of clinical trials compared with the programme conditions documented in observational studies.

Few studies have assessed the ability of DOTS regimens to produce lasting cure for TB patients treated under routine conditions.

The review question and the inclusion criteria were clear, although broad with respect to the study design. The authors searched relevant databases and other sources; however, the decision to limit the review to published studies reported in English might have increased the possibility that some relevant studies were not included in the review. The authors did not report using methods designed to reduce reviewer bias and error in the review process. Their assessment of the studies addressed some aspects of validity, but did not examine some important elements such as randomisation and blinded assessment. In view of the high level of clinical and methodological heterogeneity between the studies, the decision to employ a narrative synthesis appears appropriate. The authors' conclusions are an accurate reflection of the results of the review and, as such, are appropriately cautious. However, given the poor reporting of review methodology and the incomplete nature of the validity assessment, it is difficult to be certain of their reliability.

Practice: The authors did not state any implications for practice.

Research: The authors stated that any new regimens for the treatment of TB should, in addition to normal clinical trials, be assessed under routine programme conditions. They also stated that studies should differentiate between relapse and reinfection, and assess the influence of coinfection with HIV.

Australian National Health and Medical Research Council.

This is a critical abstract of a systematic review that meets the criteria for inclusion on DARE. Each critical abstract contains a brief summary of the review methods, results and conclusions followed by a detailed critical assessment on the reliability of the review and the conclusions drawn.