Abstract
BACKGROUND: Visceral leishmaniasis (VL) is highly prevalent in the Indian state of Bihar and, without proper diagnosis and treatment, is associated with high fatality. However, lack of efficient reporting mechanism had been an impediment in estimating the burden of mortality and its antecedents among symptomatic VL cases. The objectives of the current study were to generate a reliable estimate of symptomatic VL caseload and mortality in Bihar, as well as to identify the epidemiologic and health infrastructure-related predictors of VL mortality. METHODOLOGY AND PRINCIPAL FINDINGS: Using an elaborate index case tracing method, we attempted to locate all symptomatic VL patients in eight districts of Bihar. Interviews and medical-record-reviews were conducted with cases (or next-of-kin for the dead) meeting the eligibility criteria. The information collected during the interviews included socio-demographic characteristics, onset of disease symptoms, place of diagnosis, pre- and post-diagnosis treatment history, type and duration of drugs received. In total, we analyzed data on 4925 VL patients-59% were male and 68% were less than 30 years old. There were 158 (3.2%) deaths and the incidence rate of mortality was 3.2/100 person-years. In the adjusted Cox-proportional-hazards analysis, treatment at public facility [Adjusted Hazard Ratio (AHR) = 0.61; 95% CI = 0.43-0.86], shorter (≤30 days) diagnostic delay [AHR = 0.62, 95% CI = 0.43-0.92], and treatment completion [AHR = 0.03, 95% CI = 0.02-0.05] emerged as significant negative predictors of mortality. CONCLUSION: Mortality reduction efforts in Bihar should focus on improving access to early diagnosis, quality treatment and treatment-adherence measures, with special emphasis on marginalized communities.