Patterns, causes, and predictors of inpatient mortality among cancer patients in a tertiary center in south India: A 10-year retrospective analysis.

e24049 Background: Inpatient mortality among cancer patients reflects a complex interaction between advanced disease, treatment toxicity, and health-system factors. However, systematic data describing the timing, causes, and determinants of inpatient oncology deaths from low- and middle-income countries remain limited. We analyzed a 10-year cohort of inpatient cancer deaths to characterize mortality patterns and identify factors associated with place and cause of death. Methods: All adult inpatient oncology deaths between 2016 and 2025 were analyzed. Demographic, disease, treatment, and admission variables were collected. Causes of death were categorized as cancer-related, treatment-related, or other. Early death was defined as death within 48 hours of admission. Variables underwent univariate screening, followed by multivariable logistic regression to identify predictors of ICU death and multinomial regression with likelihood ratio testing to assess contributors to cause of death. Results: Among 426 inpatient deaths, median age was 56.68 years; 64.5% were male and 83.8% had solid tumors. At last admission, 78.0% had advanced-stage disease, 71% were metastatic, and 82.8% had poor performance status (ECOG 3–4). Emergency admissions accounted for 76.2%. Altered sensorium was present in 44.8% and 74.17% had documented comorbidities. Overall, 159 patients (37.3%) died in the ICU and 266 (62.4%) died in ward/room settings. Overall, 350 deaths (82.15%) were cancer-related, 27(6.3%) were treatment-related, and 49 (11.5%) were due to other causes. Nearly one-third (29.6%) died within 48 hours of admission. Univariate screening identified no significant predictors of early death, indicating that early in-hospital mortality was multifactorial with no single identifiable baseline predictor. On multivariable analysis, decreasing age (adjusted OR aOR 0.98, 95% CI 0.96–0.997, p = 0.024), and greater number of prior systemic therapy lines (aOR 0.68, 95% CI 0.49–0.93, p = 0.017) were independently associated with ICU death. Conclusions: In this large, decade-long inpatient mortality audit, the majority of deaths were cancer-related and occurred outside the ICU, reflecting appropriate recognition of advanced disease trajectories and end-of-life care needs. Treatment-related mortality was infrequent, underscoring the overall safety of systemic therapy delivery. Nearly one-third of patients died within 48 hours of admission, highlighting the complexity of late-stage hospital presentations. ICU death was independently associated with younger age and greater cumulative exposure to systemic therapy. These findings emphasize the need for earlier goals-of-care discussions, timely integration of palliative care, and structured frameworks to guide escalation decisions, particularly in heavily pretreated patients.