Survival outcomes in chemotherapy patients using a multi-institutional risk model. A retrospective analysis

Background: Patients with cancer who are receiving chemotherapy have a high risk of complications due to the disease and from adverse events due to therapy. The Chen Model, a proprietary risk stratification tool (multivariable analysis) developed by Mercy Health, St. Louis, identifies chemotherapy patients at risk for adverse outcomes. The Chen Model stratifies patients into high risk (score>=90), moderate high risk (80 to 89), and low risk groups (<80). From our multi-institutional analysis, we aim to identify the 6- and 12-month mortality rates among adult chemotherapy patients (non-leukemia, age >18) stratified by the Chen model. Methods: We conducted a retrospective analysis of 18,854 chemotherapy patients treated between July 2023 and August 2025 at the Mercy Hospital system, which includes multiple hospital sites and oncology practices across three states (Missouri, Arkansas, and Oklahoma). Patients were assigned to cohorts based on their maximum Chen model score during the 7-day post-chemotherapy period. A total of 2139 patients were in the high-risk group, 2449 in moderate high-risk, and 14266 in the low-risk group. Kaplan-Meier estimates and Cox proportional hazards models were used to assess 6- and 12-month mortality, with hazard ratios (HR) and 95% confidence intervals (CI) calculated. Results: Six-month mortality rates were 39.96% (95% CI: 37.85%-42.14%) for the high-risk group, 22.59% (95% CI: 20.91%-24.40%) for the moderate high-risk group, and 11.65% (95% CI: 11.12%-12.21%) for the low-risk group. Twelve-month mortality rates were 51.63% (95% CI: 49.31%-54.00%), 31.44% (95% CI: 29.36%-33.62%), and 19.54% (95% CI: 18.84%-20.26%), respectively. Cox models showed significantly elevated hazard of death for high-risk (HR=4.4 at 6 months, HR=3.7 at 12 months; p<0.005) and moderate high-risk (HR=1.5 at 6 months, HR=1.4 at 12 months; p<0.005) groups compared to the low-risk group. Further analysis on composite palliative care and hospice services utilization was identified to be 23.2% in high risk and 15.6% in moderate high-risk groups. Conclusions: The Chen model effectively stratifies chemotherapy patients by mortality risk, with high-risk patients facing substantially worse outcomes at 6 and 12 months. We also identified a lower palliative care utility in the higher-risk groups. These findings support the model's utility in identifying patients who may benefit from intensified monitoring and early involvement of supportive services, particularly palliative care services. Further studies are needed to validate these results and assess the impact of targeted interventions.