Enhanced Model Looks Beyond Comorbidities to Better Predict Hospitalization Risk

An enhanced risk prediction model that considers not only chronic comorbidities but also health factors not typically accounted for in other models improved risk prediction for emergency department (ED) use and hospitalization among a group of patients in a primary care medical home, according to research in the American Journal of Managed Care.

“We found that age, body mass index, previous healthcare utilization, mental health conditions, anemia, heart failure, epilepsy, hyperlipidemia, warfarin use, and narcotic use all simultaneously contributed to the prediction of hospitalization and ED visits in the next year,” Paul Y Takahashi, MD, MPH, of the Mayo Clinic, and colleagues wrote. “This enhanced model had an area under the curve of 0.711 for predicting hospital and ED use, as compared with 0.667 found in the model utilizing only Minnesota medical tiering.”

To better understand additional factors that could help identify at-risk patients, researchers conducted a retrospective cohort investigation of a paneled primary care population of 84,752 adult patients. Using data from 2010 and 2011, researchers developed their model with data from 42,384 patients and then validated it with data from the remaining 42,368. They also compared its risk prediction results with the Minnesota medical tiering model. Factors included previous hospitalizations, specific high-risk illnesses, mental health disorders, and use of high-risk medications

“In comparing high-risk groups of the same size from both models and excluding patients identified in both models, 47% of the patients in the enhanced model experienced hospitalization compared with 32.5% of patients from the Minnesota medical tiering model,” Dr Takahashi and colleagues wrote. “The enhanced model accounts for factors in a younger population (eg, mental health and previous utilization) that are not detected in the Minnesota medical tiering model.”

Specifically, researchers found that 2 or more prior ED visits and 1 prior hospital stay were strong predictors for ED use or hospitalization in the following year. Both a diagnosis of depression as well as prior substance abuse were also linked with increased risk. Other predictors of risk were lack of insurance, Medicaid coverage, age older than 70, and obesity.

“These findings help to develop new models for determining patients at risk for adverse health outcomes,” Dr Takahashi and colleagues wrote. “They focus on potential future areas of clinical practice and research on social determinants of health.”—Jolynn Tumolo