Prevalence and factors associated with hematological adverse events in RR-TB patients on linezolid-based regimens in Uganda: a multicenter retrospective cohort study

Abstract

Abstract
Background
Linezolid (LZD) is a cornerstone of rifampicin-resistant tuberculosis (RR-TB) treatment, yet its hematological adverse events (AEs), particularly anemia, leukopenia, and thrombocytopenia, remain understudied in resource-limited settings like Uganda. With a high RR-TB burden and comorbidities such as HIV and malnutrition, understanding the prevalence and risk factors of these hematological AEs is critical to improving treatment adherence and outcomes.
Aim
This study aimed to determine the prevalence of hematological AEs and identify factors associated with hematological AEs among RR-TB patients receiving LZD-based regimens in Uganda.
Methodology
A multicenter retrospective cohort study was conducted using medical records of RR-TB patients treated with LZD at four Ugandan hospitals (Mbale RRH, Iganga General Hospital, Moroto RRH, and St. Kizito Matany Hospital) from 2020 to 2024. Hematological AEs were defined as anemia (Hb < 12 g/dL in women or < 13 g/dL in men), thrombocytopenia (platelets < 150 × 10³/µL), and leukopenia (WBC < 3.7 × 10⁹/L). Sociodemographic, clinical, and treatment-related data were extracted. A modified Poisson regression model with robust standard errors was used to estimate adjusted prevalence ratios (aPR) and 95% confidence intervals (CI).
Results
Among 412 patients, 62.9% (259/412) had at least one hematological AE at baseline or during follow-up, and 36.4% (150/412) had hematological AEs first detected after LZD initiation. Among patients with available baseline CBC, 40.0% (98/245) developed incident hematological AEs. First detected anemia, thrombocytopenia, and leukopenia occurred in 21.4% (88/412), 14.8% (61/412), and 13.8% (57/412) of patients, respectively. Rural residence (aPR 1.60, 95% CI 1.11–2.33) and divorced or widowed marital status (aPR 4.10, 95% CI 1.58–10.77) were independently associated with hematological AEs. Patients with hematological AEs had lower treatment success (83.0% vs. 93.5%) and higher loss to follow-up (15.1% vs. 5.9%) than those without AEs.
Conclusion
Hematological AEs were frequent among RR-TB patients receiving LZD in Uganda and were associated with reduced treatment success and increased loss to follow-up, while mortality remained low. Rural residence and divorced or widowed marital status were independently associated with these events, highlighting the need for targeted monitoring in higher-risk groups.