In their retrospective study, Hwang et al. 1 found the prevalence of extensively drug-resistant tuberculosis (XDR-TB) to be 21% amongst multidrug-resistant (MDR) patients. However, the available literature indicates that it is around 10% 2. The XDR-TB data of Hwang et al. 1 was based on the drug susceptibility test for ofloxacin only; it was not performed for other commonly used quinolones. The data would have been higher if drug susceptibility to other commonly used quinolones could have been carried out. The proportion of XDR-TB amongst these patients would therefore have been higher.
XDR-TB has been reported with high HIV prevalence 3. It would therefore have been more appropriate if Hwang et al. 1 had discussed the correlation with HIV and TB co-infection in the geographical areas in which the study was carried out.
Treatment success rates of XDR-TB worldwide are 3–13% and depend on the number of drugs to which the bacilli are resistant 4. However, Hwang et al. 1 found the success rate of treatment of XDR-TB patients to be >50% (23 out of 42 patients). Their success rate was comparable to the standard success rate of MDR-TB (rather than XDR-TB). It would be appreciated if the authors could discuss the treatment regimes/drugs used in XDR-TB patients, which are likely to boost confidence in treating MDR-TB and may be useful in formulating the future guidelines.
By and large, an odds ratio is a measure of an association between categorical responses; it is important in epidemiology because it represents a relative estimate of risk when no direct risk estimate is possible. A confidence interval is a range of plausible values that account for uncertainty in a statistical estimate; a wide interval implies poor precision, suggesting findings are compatible with a wide range of effect sizes. Further studies are needed to reach a conclusion, as the figures of Hwang et al. 1 (adjusted OR 12.05, 95% CI of 1.48–98.38) do not convincingly show a real difference between groups. Hence, we cannot be confident enough to attribute streptomycin resistance to the adverse treatment outcomes in XDR-TB patients.
Statement of interest
None declared.
- © ERS Journals Ltd
