Severe infection poses a major threat to cancer patients due to treatment-related immunosuppression. Early identification and timely intervention are essential to reducing infection-related mortality. This study investigated whether daily vital signs recorded during radiotherapy could predict severe infection. We conducted a retrospective case-control study of cancer patients who received radiotherapy between 2016 and 2022. Cases were patients who developed severe infection during radiotherapy, while controls completed radiotherapy without infection. Patients with over a 2-day interval between the last radiotherapy session and infection diagnosis were excluded. Included patients had ≥ 5 recorded daily vital-sign measurements. Collected data included demographics, vital signs, cancer type, treatments, comorbidities, and infection outcomes. Logistic regression and area under the receiver operating characteristic curve (AUROC) were used to assess predictive value. A total of 35 patients with severe infection and 200 controls were analyzed. Head and neck cancer and lung cancer were the most common cancer types in the infection group. Significant predictors included lower last systolic blood pressure (SBP), higher last heart rate (HR), an increased last shock index (SI, calculated as HR/SBP), and greater HR variability across the last 2 to 5 radiotherapy sessions. In multivariable models, last SI and HR variability remained independent predictors, with AUROC values ranging from 0.773 to 0.803. In cancer patients receiving radiotherapy, changes in HR, SBP, SI, and HR variability in the days preceding severe infection may serve as feasible, self-monitored indicators for early detection in outpatient or home settings. However, the relatively small number of infection events (n = 35) may limit the statistical robustness of our findings, and further validation is needed.
Li et al. (2026) studied this question.