Patient characteristics associated with multiple pharmacy use in the U.S. population: Findings from the Medical Expenditure Panel Survey.

Abstract

Multiple pharmacy use (MPU) is an important safety and quality issue, as it results in fragmented patient care. However, few studies have examined patient characteristics predicting the use of multiple pharmacies, and the findings have been inconsistent. To identify patient characteristics associated with MPU using national data. Data were obtained from the 2011 U.S. Medical Expenditure Panel Survey. The dependent variable was MPU, or the use of more than one pharmacy. The Andersen Behavioral Model of Health Service Use was used to guide the selection of independent variables, which were categorized as predisposing, enabling, and medical need related characteristics. Multivariable logistic regression analysis was conducted to identify the relationships between predisposing, enabling, and need variables and MPU in a hierarchical fashion. Point estimates were weighted to the U.S. non-institutionalized population, and to adjust standard errors to account for the complex survey design. MPU was common, with a national prevalence of 41.3%. Individuals aged 40-64 and adults 65 and older were significantly less likely to use multiple pharmacies as patients aged 18-39 years (40-64 years OR: 0.67, CI: 0.58-0.77; ≥65 years OR: 0.49, CI: 0.41-0.58). Females were significantly more likely to use multiple pharmacies than males (OR: 1.16, CI: 1.05-1.29). Individuals lacking health insurance were more likely to use multiple pharmacies as individuals with private health insurance (OR: 1.42, CI: 1.16-1.73); in contrast, individuals having drug insurance were more likely to use multiple pharmacies (OR: 1.25, CI: 1.06-1.47) relative to those without drug insurance. Any mail order use was the strongest predictor of MPU (OR: 6.94, CI: 5.90-8.18). Pharmacists and other health care providers need to be aware that their patients may be using multiple pharmacies, especially younger patients, those lacking access to health insurance, or those using mail order pharmacies. The findings from this study can be used to identify patients that may need additional monitoring to ensure safe and appropriate drug therapy, and has important implications as health care continues to shift toward performance-based reimbursement and quality ratings.