Identifying classes of veterans with multiple risk factors.

Abstract

As researchers examine the efficacy of interventions that simultaneously target more than 1 symptom, it is important to identify ways to help guide research and program development. This study used electronic medical record data to examine the covariation of multiple risk factors regularly assessed among primary care patients. It also examined the health care utilization of those patients identifying where the health care system came in contact with them to help identify the ideal locations these interventions may be most often used. We obtained data for six risk factors, as well as the number of primary care, mental health, and emergency department visits, from Veteran patients with a primary care visit. There were three main groups of primary care patients, identified using latent class analysis and regression. Although the smallest group, the "High Treatment Need" group, had an increased probability of screening positive for all four risk factors, the post-traumatic stress disorder screen was a significant discriminator of this group from the others. Results show that this group had the greatest number of encounters in all health care locations suggesting significant opportunities for intervention. However, future research is needed to examine the current interventions offered and potential avenues where risk factors may be addressed simultaneously.