Retrospective Ethiopian hospital study finds no detectable five-year TB trend

A five-year retrospective analysis of routine Xpert MTB/RIF records from Woldia Comprehensive Specialized Hospital in northeastern Ethiopia found no statistically significant temporal trend in the detection of Mycobacterium tuberculosis (MTB) or rifampicin resistance between 2020 and 2024. MTB was detected in 504 records, or 11.6%. Among 498 MTB-detected records with determinate rifampicin-resistance results, 28, or 5.6%, showed rifampicin resistance.

Neither outcome had a significant annual difference or linear trend across the study period. The study, published in BMC Infectious Diseases , examined laboratory records generated from January 2020 through December 2024. The investigators used descriptive statistics and logistic regression to assess MTB detection, rifampicin resistance and associated factors.

Chi-square tests evaluated annual proportions, while continuous-year logistic regression assessed whether calendar year was associated with either outcome. A case-mix-adjusted sensitivity analysis examined temporal changes in MTB detection after accounting for age, sex, previous TB treatment, specimen type and referring facility. Firth-penalized logistic regression was used to assess the robustness of associations with rifampicin resistance in the presence of sparse resistant cases.

The records represented a hospital testing population rather than a population-based sample, so the results describe facility-level testing patterns and cannot directly establish trends in community TB incidence or drug resistance. Calendar year was not significantly associated with MTB detection in the continuous-year model. The estimated odds ratio was 1.007 per year, with a 95% confidence interval of 0.946 to 1.071.

Year was likewise not significantly associated with rifampicin resistance, for which the odds ratio was 1.089 and the 95% confidence interval was 0.848 to 1.399. Adjustment for the measured case-mix factors did not change the finding for MTB detection: the adjusted odds ratio per year was 1.025, with a 95% confidence interval of 0.957 to 1.097 and a p value of 0.481. The lack of a statistically detectable trend does not demonstrate that the underlying burden or resistance level remained constant.

Previous TB treatment had strong associations with both outcomes in the adjusted analyses. Compared with people without a reported history of prior TB treatment, those previously treated had an adjusted odds ratio of 13.99 for MTB detection, with a 95% confidence interval of 7.85 to 24.95. For rifampicin resistance, the corresponding adjusted odds ratio was 49.72, with a 95% confidence interval of 18.60 to 132.90.

These retrospective estimates represent associations among the tested records rather than causal effects or direct measures of individual risk. Male sex was also associated with rifampicin resistance, with an adjusted odds ratio of 4.59 and a 95% confidence interval of 1.32 to 16.00. HIV positivity was associated with MTB detection in univariable analysis, yielding a crude odds ratio of 7.03 and a 95% confidence interval of 1.99 to 24.77, but extensive missing HIV data constrained the interpretation of that result.

Only 28 rifampicin-resistance events were identified, contributing to wide confidence intervals, particularly for the association between previous TB treatment and resistance. The findings are also limited to one hospital, its recorded testing population and the five-year period from 2020 through 2024. A null temporal association in these records does not establish the absence of trends in other settings or at other times.

Overall, the analysis identified MTB in approximately one in nine tested records and rifampicin resistance in 5.6% of MTB-detected records with determinate results. Previous TB treatment was the factor most strongly associated with both outcomes, while the study detected no significant year-to-year shift in either MTB detection or Xpert-detected rifampicin resistance during the study period.

Subject of Research: Temporal patterns in tuberculosis detection and rifampicin resistance among Xpert-tested records at a northeastern Ethiopian hospital

Article Title: Trends in Mycobacterium tuberculosis detection and rifampicin resistance using Xpert MTB/RIF assay at Woldia comprehensive specialized hospital, Northeastern Ethiopia: a 5-year retrospective study

Article References: Tamrat, E., Kassanew, B., Mollaw, N. G., Girma, T., Asfaw, T., Habtamu, T., Jemal, A., Gashaw, M., Bazezew, A., Damtie, A. A., Gedfie, S., Gashaw, Y., Asmare, Z., Biset, S., & Reta, M. A. (2026). Trends in Mycobacterium tuberculosis detection and rifampicin resistance using Xpert MTB/RIF assay at Woldia comprehensive specialized hospital, Northeastern Ethiopia: a 5-year retrospective study. BMC Infectious Diseases. https://doi.org/10.1186/s12879-026-14670-z

Image Credits: AI Generated

DOI: 10.1186/s12879-026-14670-z

Keywords: tuberculosis, rifampicin resistance, Xpert MTB/RIF, Ethiopia, retrospective study, hospital records, Mycobacterium tuberculosis, diagnostic testing, Trends, Mycobacterium, detection, rifampicin