Online Program

Polypharmacy as a Predictor for Hospitalization in a National Longitudinal Study of Middle-aged Americans (307812)

Keywords: polypharmacy, hospitalization, propensity score matching, MIDUS

The detriments of polypharmacy include adverse drug events and impaired functional status which can lead to poorer health outcomes. Using the Midlife in the United States (MIDUS) dataset looking at middle-aged Americans, this study looks at whether polypharmacy predicts the frequency of hospitalizations. Polypharmacy was defined as taking more than one type of medication. We matched using a nearest neighbor propensity score method and ran a Poisson regression model with polypharmacy as a predictor for hospitalization. Our matched sample includes 2738 respondents who reported on the number of prescription medications and number of hospitalizations in two separate waves. Controlling for age, race, education, smoking status, and comorbidities such as cancer, diabetes, depression and heart disease, respondents who are on polypharmacy have an 164% (p=0.0001) increased risk of hospitalization compared to those not on polypharmacy. These results suggest that polypharmacy is a marker for hospitalization even with our conservative definition. A systematic medication review may be able to address medication appropriateness and improve polypharmacy health outcomes.