NP/MD prescribing and medication adherence: Evidence from Medicare

Purpose: Previous research has established that outcomes of nurse practitioners (NPs) and physicians (MDs) are comparable for a wide range of health care services. To this end, expanded roles of NPs are viewed integral to addressing the challenge of delivering primary care services that meet the goals of the “Triple Aim”. Yet, very little is known about the prescribing practices and outcomes of NPs compared to MDs, and whether NPs should be allowed to prescribe independently from physicians continues to be a contentious issue. The purpose of this study was to examine differences in (1) medication adherence for NP and MD beneficiaries and (2) prescribing costs for NP and MD beneficiaries.

Data: We used Medicare Part D administrative claims data for the years 2010-2013 for a random sample of NP and MD beneficiaries. Our analysis examined longitudinal data for 334,800 beneficiaries who had continuous Part D coverage across these years. The Area Resource File (ARF) provided state-level covariates such as income, unemployment rate and number of primary care providers. We also controlled for state scope of practice laws that indicated whether NPs could prescribe independently from physicians.

Methods: Patient-level annual medication adherence rates were calculated (measured as the proportion of covered days per year) based on a National Quality Forum approved measure that calculates adherence in a cohort of patients on antidiabetic, antihypertensive, and statin medications. We employed propensity score matching to address selection bias between patients and NPs/MDs. Ordinary least squares regression with controls for patient demographic information, Elixhauser comorbidities, plan information, out-of-pocket share, and state level covariates was used to estimate the effect of provider type on adherence. Pharmaceutical costs were examined using a generalized linear model.

Findings: Preliminary results found no significant differences in medication adherence rates between NP and MD beneficiaries (.79 vs .78; P <0.06). Pharmaceutical costs were significantly higher for NP beneficiaries compared to MD beneficiaries ($3,471 vs $3,348; P<0.001).

Conclusions: Additional research is needed to understand the reasons for higher pharmaceutical costs among NP beneficiaries, including taking into account inpatient and outpatient spending. Because the findings of this study provide new data on NPs serving a large and growing patient population in the US, results provide valuable information for designing workforce policies and are important for the legal ramifications affecting nurse practitioner prescribing.