4 months of rifampin compared with 9 months of isoniazid for the management of latent tuberculosis infection: A meta-analysis and cost-effectiveness study that focuses on compliance and liver toxicity

Background. One-third of the world's population is infected with tuberculosis, and 9 months of isoniazid monotherapy is the treatment of choice for latent tuberculosis infection. However, this approach has been associated with hepatotoxicity and poor compliance. A shorter (4-month) rifampin regimen has been evaluated in recent clinical trials. Methods. We performed a meta-analysis of the published studies to compare compliance, toxicity, and costeffectiveness between the 2 strategies. Pooled effects were calculated as risk ratios (RRs) by means of randomeffects and fixed-effects models. Results. Pooled data from 3586 patients suggested that 4-month rifampin therapy was associated with a significant reduction in the risk of noncompletion (RR for random-effects model, 0.53; 95% confidence interval [CI], 0.44-0.63). Noncompletion rates were lower among patients who received 4-month rifampin therapy (range, 8.6%-28.4%), compared with noncompletion rates among patients who received 9-month isoniazid therapy (range, 24.1%-47.4%). Also, rates of hepatotoxicity (defined as grade 3 or 4 liver failure leading to drug discontinuation) were lower for patients who received 4-month rifampin therapy (range, 0%-0.7%), compared with the corresponding rates for patients who received 9-month isoniazid therapy (range, 1.4%-5.2%), and rifampin was associated with significant reduction in the risk of hepatotoxicity (RR for fixed-effects model, 0.12; 95% CI, 0.050.30). Notably, with the data from our meta-analysis, we calculated that the 4-month rifampin strategy is also cost-effective and results in $213 savings per patient treated ($90/patient when doctor fees are not included). Conclusions. The improved compliance, safety, and cost associated with the 4-month rifampin therapy suggest that the efficacy of this approach needs to be evaluated in detail. An extended posttreatment follow-up in future studies will clarify the unresolved issue of tuberculosis reactivation rates.