Predicting when T2D patients may experience a cardiovascular event
A new model for predicting the risk of cardiovascular (CV) events could improve care for older patients with type 2 diabetes, according to new data published in the Journal of Managed Care + Specialty Pharmacy.
“The CV risk prediction model used in this study, specifically developed and trained in a Medicare population with type 2 diabetes, using administrative claims, fills a gap in the current literature by providing a model that identifies risk of disease in this population, which is much different from the general at-risk population," wrote lead author Eleanor O Caplan, PharmD, with Humana Healthcare Research, and colleagues.
The model was utilized to identify patients at a low, moderate or high risk of a CV event.
Data came from patients included in the Humana Research Database from January 2011 to December 2018. Patients with type 2 diabetes were identified based on diagnoses and/or medications from 2012 to 2013.
A total of 362,791 patients were included. While 248,142 patients made up the training dataset, another 106,346 patients made up the test dataset. Fifty-one percent of patients were women, and the median patient age was 74 years old.
CV events the team focused on included inpatient hospitalizations for MI, ischemic stroke, unstable angina or heart failure and any revascularization diagnoses or procedures.
Overall, 20.9% of patients had at least one composite CV episode during the mean follow-up of 4.1 years.
After the model had categorized each patient based on their risk, the team found that approximately 11% of low-risk patients, 27% of moderate-risk patients and 51% of high-risk patients experienced the study's primary outcome over the course of the study.
"These results were generally consistent across the training, test, and holdout datasets," they wrote.
Caplan et al. noted that their model shows potential as a way to identify patients with type 2 diabetes who need a focused intervention. However, they added, additional adjustments and validation of the model may be necessary to achieve optimal performance and utility across other patient populations in the future.
Read the full study here.