From the authors:
- 1Dept of Social and Preventive Medicine, College of Medicine, Inha University, Incheon, 2Division of Pulmonary and Critical Care Medicine, Dept of Internal Medicine and Lung Institute, Seoul National University College of Medicine, Seoul, Republic of Korea.
- J-J. Yim, Division of Pulmonary and Critical Care Medicine, Dept of Internal Medicine and Lung Institute, Seoul National University College of Medicine, 28 Yongon-Dong, Chongno-Gu, Seoul, 110-744, Republic of Korea. E-mail: yimjj{at}snu.ac.uk
We thank A.K. Khurana and P.R. Mohapatra for their interest in our recent work 1. They raised several important issues regarding the setting, results and analysis of our study.
First, they point out that the proportion of extensively drug-resistant tuberculosis (XDR-TB) patients among multidrug-resistant TB (MDR-TB) patients in our cohort was unexpectedly high (21.3%). As only 5.7% of the 1,407 patients with MDR-TB in another Korean cohort had XDR-TB 2, we understand their curiosity regarding our MDR-TB cohort. As described in our paper, our cohort consisted of MDR-TB patients diagnosed and treated at Seoul National University Hospital (Seoul, Republic of Korea), which is the final referral centre in South Korea. Regarding MDR-TB, our institution has not only treated many refractory patients referred from other hospitals for several decades, but has also published research papers on various aspects of MDR-TB 3–6. The relatively high proportion of XDR-TB patients should be understood in this context.
Secondly, A.K. Khurana and P.R. Mohapatra suggest that a description of HIV status in the area in which our cohort was based would help readers to understand the context of our study. Although we excluded MDR-TB patients if they were HIV-infected, we concur that the information on HIV prevalence in the geographic region is valuable. As we mentioned previously, our MDR-TB cohort was referred from hospitals located all over South Korea, for which the predicted HIV prevalence is as low as 0.01% 7.
In addition, A.K. Khurana and P.R. Mohapatra indicate that the success rate (cure or treatment completed) of XDR-TB treatment in our study was surprisingly high (54.8%). However, a recent study on XDR-TB reported a treatment success rate of 60.4% 8. Clinicians' careful selection of an individualised regimen based on drug susceptibility test results and a detailed history of previous treatment, and the patients' firm adherence to the treatment, may have led to improved treatment outcomes, even among XDR-TB patients. Of the 211 patients with MDR-TB in our original cohort, only seven defaulted.
Finally, A.K. Khurana and P.R. Mohapatra argue that the 95% CI for the OR of the effect of streptomycin resistance on an adverse treatment outcome was too large to detect a real difference. However, the CI is not about the association itself but about the precision of the effect estimate, the OR, in our study. Although a wide 95% CI (1.48–98.38) implies that precision was lacking, the large OR (12.05) indicates a very strong association between streptomycin resistance and an adverse treatment outcome. In fact, the p-value for this association was 0.02. In addition, the analysis of a small number of patients with rare diseases inevitably results in a wide confidence interval in most cases. As we mentioned in our paper, additional studies with improved designs may be able to confirm the potentially important role of resistance to injectables, including streptomycin.
Statement of interest
None declared.
- © ERS Journals Ltd