Treating multidrug-resistant TB (MDR-TB) remains challenging. However, the determinants and timing of poor outcomes during MDR-TB treatment are still poorly understood. We conducted a retrospective cohort study on all adult MDR-TB patients treated at Persahabatan Hospital, Jakarta, Indonesia, between January 2013 and December 2016. Risk factors for poor outcomes were analysed using Cox regression. Death occurred at a median time of 6 months (IQR 4-14) and loss to follow-up (LTFU) at 7 months (IQR 3-11). In multivariate analysis, advanced age (aHR 2.91, 95% CI 1.21-6.96; = 0.017 for age >60 years), having diabetes mellitus (aHR 2.18, 95% CI 1.25-3.82; = 0.006) and HIV co-infection (aHR 3.73, 95% CI 1.14-12.23; = 0.030) were predictive of poor outcome in the first 7 months of treatment, whereas history of LTFU (patients who were LTFU once: aHR 2.14, 95% CI 1.33-3.47; = 0.002; patients who were LTFU more than once: aHR 3.61, 95% CI 1.68-7.77; = 0.001) and baseline body mass index <18.5 kg/m² (aHR 1.98, 95% CI 1.10-3.56; = 0.022) predicted poor outcome after 7 months of treatment. Different subsets of patients with MDR-TB are at risk of poor outcome at different times during treatment.
Download full-text PDF |
Source |
---|---|
http://dx.doi.org/10.5588/ijtld.21.0351 | DOI Listing |
Enter search terms and have AI summaries delivered each week - change queries or unsubscribe any time!