Risk Prediction Model Identifies Cancer Patients Prone to ER Visits During Clinical Trials
Key Insights
A new risk prediction model identifies advanced cancer (search) patients in clinical trials at high risk for unplanned ER visits and hospital stays.
The model incorporates factors like performance status, coronary artery disease (search), hypertension (search), liver disease (search), and prostate cancer (search) status.
Patients with two or more risk factors had over three times the risk of acute care use, highlighting the need for targeted interventions.
A novel risk prediction model has been developed and validated by investigators from the SWOG Cancer Research Network to pinpoint advanced cancer (search) patients participating in clinical trials who face an elevated risk of unplanned emergency room (ER) visits and hospitalizations. The ability to identify these high-risk individuals could pave the way for interventions aimed at decreasing the necessity for such visits, ultimately enhancing the quality of care and lowering healthcare costs.
The findings were presented by Dawn L. Hershman, MD, MS, at the 2024 ASCO Quality Care Symposium in San Francisco.
Hershman's team analyzed data by linking Medicare claims data to data from six SWOG advanced cancer (search) clinical trials to identify hospital stays or ER visits by enrolled patients. The study revealed that a substantial majority, 67.5 percent, of the 1,397 patients analyzed experienced at least one ER visit or hospital stay within a year of enrolling in a clinical trial.
The researchers divided the patient data into a training set (60 percent) and a test set (40 percent) to develop and validate the risk prediction model, respectively. The model considered 23 baseline factors, including sociodemographic, geographic, clinical, and treatment-related variables, as well as the presence of comorbidities.
The final risk model incorporated four key risk factors: patient performance status, coronary artery disease (search), hypertension (search), and liver disease (search). The model also adjusted for cancer (search) type, specifically prostate cancer (search).
Model Validation and Risk Stratification
Patients in the training set with two or more risk factors demonstrated a threefold increase in the risk of acute care use compared to those with zero or one risk factor. Validation using the test data set confirmed the model's performance.
When considering all patients, those in the highest risk quartile (three or four risk factors) had a 4.23 times greater risk of hospital stays or ER visits compared to those in the lowest risk quartile (zero risk factors).
Implications for Clinical Trial Design
Clinical trial eligibility criteria have historically excluded patients with certain comorbid conditions. Recent efforts to broaden inclusion criteria may lead to a higher proportion of patients at increased risk for acute care utilization.
"Trials have become more inclusive of patients with some comorbid conditions due to the work of ASCO and other organizations," said Joseph M. Unger, PhD, associate professor at Fred Hutch Cancer Center (search). "This could also have the salutary effect of reducing disparities in access to trials for patients of different sociodemographic backgrounds... However, our work also highlights how investigators and trialists should anticipate a higher risk of acute care use."
This research was supported by NIH/NCI/NCORP grant UG1CA189974 and The Hope Foundation for Cancer (search) Research.