Background
The Khorana Score is a validated risk score for predicting 6‐month incidence of cancer‐associated venous thromboembolism (CAT) among patients starting chemotherapy. Venous thromboembolism (VTE) risk factors important in the general population, including age, sex, prior VTE, and hospitalization, are not included in this score, their association with VTE in cancer patients is unknown.
Objective
To examine risk factors for CAT and the impact of incorporating longitudinal hospitalization into risk assessment.
Methods
Risk factors were recorded among patients starting chemotherapy at a single institution from 2012‐14. Hospitalization and time‐periods after hospitalization were assessed as time‐varying covariates. Logistic regression was used to determine factors related to 6‐month CAT risk (the Khorana Score endpoint). Proportional hazard models were used for risk factor identification throughout the 3‐year observation period.
Results
Among 1,583 patients starting chemotherapy (mean age 60, 48% male), 187 developed CAT (11.8%) with 129 (69%) cases occurring within 6 months of starting chemotherapy. In the 6‐month analysis, no additional risk factors were associated with CAT. In the 3‐year analysis, male sex (HR 1.57, 95%CI 1.21, 2.07), prior VTE (HR 2.12, 95%CI 1.41, 3.18), and hospitalization (HR 2.69, 95%CI 1.92, 3.75) were associated with increased hazard of CAT, adjusting for risk factors in the Khorana score.
Conclusions
When time‐to‐event data were incorporated into CAT risk assessment, male sex, prior VTE, and hospitalization were important risk factors. These data highlight the need to consider dynamic methods for assessing VTE risk in cancer patients, with particular attention to the period around hospitalizations.