Effectiveness of empiric tuberculosis treatment for presumptive smear-negative pulmonary TB among people living with HIV, a systematic narrative review

INTRODUCTION: Timely diagnosis and prompt initiation of treatment are crucial for the effective management of tuberculosis (TB), particularly in People living with HIV (PLHIV). A delay in diagnosing TB,in this special population, not only postpones the start of appropriate treatment but also increases the risk of poor clinical outcomes and TB transmission. In resource-limited settings, where advanced diagnostic tests such as Xpert-MTB assay and mycobacterium culture may be unavailable, there is a notable lag in diagnosing smear-negative pulmonary TB among PLHIV. Empiric anti-TB treatment is a widely utilized strategy to address this challenge. However, there has been limited research evaluating its impact on mortality. Given this knowledge gap, it becomes imperative to conduct a systematic review that can provide evidence-based management strategies for healthcare professionals working with PLHIV. METHOD: A systematic search strategy was conducted using various databases including MEDLINE, EMBASE, SCOPUS, Cochrane Library, and ClinicalTrials.gov up until May 2023. Observational and interventional studies involving adults with HIV and clinically suspected (unconfirmed) pulmonary TB were included. Due to clinical and methodological heterogeneity, a narrative synthesis was performed. Findings were categorized by the effect of empiric TB treatment on mortality. FINDINGS: The search strategy identified seven eligible studies (including 2,352 patients) meeting the inclusion criteria. Two studies revealed higher long-term mortality rates among PLHIV who received empiric tuberculosis treatment compared to those who did not. However, most of these patients had more advanced HIV disease and a higher likelihood of having TB. A statistically significant increase in mortality was observed in only one of these two studies (adjusted HR: 3.7, 95% CI 92.2-6.3, PAÿ=Aÿ0 .115). On the other hand, two studies suggested that there may be a potential benefit of empiric treatment for those with signs of severe TB resulting in approximately 17% reduction in mortality (ARR: 17%; 95% CI: 5.5%-29%; pAÿ=Aÿ0.004). No mortality benefit was observed among stable patients without signs of severe TB who were empirically treated. A separate study focusing on severely ill HIV patients, showed a numerically lower mortality rate in the group who received empiric treatment compared to the untreated group but not reaching statistical significance. CONCLUSION: The effectiveness of empiric TB treatment for PLHIV remains uncertain. Multiple studies included in this systematic narrative review have yielded conflicting results, indicating the complexity of managing TB-HIV co-infection in resource-limited settings. Current evidence points toward risk-stratification when taking the decision to initiate empiric TB treatment for PLHIV. Empiric treatment may reduce mortality in critically ill patients with danger signs and should be initiated promptly. However, it should be avoided in clinically stable patients where further diagnostic evaluation is preferred

• Population(s)
  • General HIV+ population
• Co-infections
  • Tuberculosis