Association of comorbidities, symptoms, and laboratory markers with previous tuberculosis among geneXpert-confirmed patients: a retrospective analysis

Tuberculosis remains a major public health problem in high-burden settings. Patients with previously treated disease may differ clinically from those presenting with a first episode, but this has seldom been examined in microbiologically confirmed cohorts. We assessed characteristics associated with a history of previous tuberculosis among GeneXpert-confirmed patients. Retrospective observational study at a tertiary care hospital in Karachi, Pakistan. Records of all patients with a positive GeneXpert MTB/RIF result over 12 months were reviewed after the study period closed, yielding 162 consecutive patients. Previous tuberculosis was defined as a documented, bacteriologically confirmed episode preceding the index result. Comparisons were restricted to the 71 patients in whom this status was documented. Analyses were performed on a complete-case basis, with denominators reported throughout. Fisher’s exact test, the Mann–Whitney U test, and binary logistic regression, unadjusted and adjusted for age and sex, were used. The mean age of the 162 patients was 40.8 ± 20.1 years, and 88 of 161 (54.7%) were male. Of the 71 patients in whom previous-tuberculosis status was documented, 11 (15.5%) had previously treated disease. Chronic kidney disease was more frequent among previously treated patients (3 of 11, 27.3% versus 3 of 60, 5.0%; p = 0.044; odds ratio [OR] 7.12, 95% confidence interval [CI] 1.22–41.54), and median serum creatinine was higher (0.95 versus 0.70 mg/dL; p = 0.030). No other factor reached significance, including diabetes mellitus (OR 2.11, 95% CI 0.57–7.84), hypertension (OR 2.40, 95% CI 0.65–8.83), ischemic heart disease (OR 2.79, 95% CI 0.59–13.04) and smoking (OR 3.11, 95% CI 0.50–19.54). Mean age did not differ significantly (52.7 ± 20.6 versus 42.1 ± 21.5 years; p = 0.142). After adjustment for age and sex, the chronic kidney disease association was attenuated and no longer significant (OR 5.54, 95% CI 0.89–34.29; p = 0.066). All confidence intervals were wide. Chronic kidney disease and raised serum creatinine were the only characteristics associated with previously treated tuberculosis, and the chronic kidney disease association did not persist after adjustment. These findings are exploratory and constrained by the small previously treated group, but support routine assessment of renal function in patients presenting with tuberculosis.