Background: Tuberculosis (TB) remains a critical health concern for people living with HIV (PHIV), significantly contributing to high mortality rates in hospitalized individuals. Improving TB diagnostic approaches is essential to enhance patient outcomes in this vulnerable population.
Methods: A cluster randomised trial was conducted at Zomba Central Hospital in Malawi to assess the impact of enhanced TB diagnostics on TB treatment initiation among adult PHIV admitted to the hospital. The study compared “enhanced TB diagnostics” (urine lipoarabinomannan (LAM) antigen tests – SILVAMP-LAM and Determine-LAM, and digital chest X-ray with computer-aided diagnosis (dCXR-CAD)) plus usual care, against “usual care” alone. The primary outcome was TB treatment initiation during hospital admission. Secondary outcomes included 56-day mortality and TB diagnosis within 24 hours.
Findings: The study involved 415 adult PHIV, with a high proportion (90.8%) already on antiretroviral therapy (ART) upon admission. The enhanced diagnostic arm utilized digital chest X-ray with CAD (CAD4TBv6), revealing a median CAD4TBv6 score of 60. Urine LAM tests also showed positivity in a subset of patients. Significantly, TB treatment initiation was higher in the enhanced diagnostics arm (22%) compared to usual care (12%) (Risk Ratio 1.92, 95% CI 1.20-3.08). However, the study found no significant difference in 56-day mortality, rapid TB diagnosis within 24 hours, or undiagnosed TB at discharge between the two arms.
Interpretation: The implementation of enhanced TB diagnostics, including digital chest X-ray with CAD and urine LAM tests, led to a notable increase in TB treatment initiation among hospitalized PHIV in Malawi. This increase appears to be primarily driven by the use of Determine-LAM urine tests. While digital chest X-ray CAD (CAD4TB) was part of the enhanced diagnostic package, the study suggests Determine-LAM had a more pronounced impact on treatment initiation rates. Interestingly, the enhanced diagnostic approach did not translate to improved short-term mortality outcomes. This highlights the ongoing challenge of high inpatient mortality among adults with HIV and TB, even with improved diagnostic strategies. Further research is needed to understand the discrepancies between Determine-LAM and SILVAMP-LAM urine tests and to develop more effective interventions to reduce mortality in this high-risk population.
Conclusion: Digital chest X-ray with CAD, alongside urine LAM tests, can play a valuable role in enhancing tuberculosis diagnosis and increasing treatment initiation for hospitalized people living with HIV. However, these improved diagnostic methods alone are insufficient to address the alarmingly high mortality rates in this patient group, underscoring the need for comprehensive strategies that go beyond initial TB diagnosis to improve overall outcomes.
