Background: The majority of states have implemented Prescription Drug Monitoring Programs (PDMPs) in an effort to curb opioid prescribing in high-risk patients; however, the effectiveness of these programs at hospital discharge is not well established. Studies in the post-operative setting have demonstrated mixed results in opioid prescribing after requiring PDMP queries.1-2 Little is known about the impact of PDMPs on opioid discharge prescribing outside of the surgical setting. The objective of this study was to analyze the impact of a PDMP integrated with the electronic health record (EHR) on opioid prescribing habits at discharge among patients discharged from medicine services.
Methods: Patients who were discharged from medicine units and received a prescription for oxycodone immediate release were retrospectively compared during a three month period before PDMP implementation and after implementation. Demographic and admission data were collected from the electronic medical record. Demographic and admission data of the pre- PDMP and post- PDMP implementation groups were described using descriptive statistics for frequency. The associations between patients’ characteristics and PDMP implementation were evaluated using a chi-square test. Relative risk (RR) and 95% confidence limits were determined to measure the strength of association between PDMP implementation and length of opioid therapy (≤3 days or >3 days). Multivariate analysis was performed using a log-binomial regression to evaluate the association between PDMP implementation and other patient variables. The variance inflation factor (VIF) was used to identify multicollinearity. A negative binomial regression model was used to identify relationships between length of therapy and independent factors. Analyses were performed with SAS version 9.4 (SAS Institute, Cary, NC).
Results: 378 patients were included in the study with 194 patients in the pre- intervention group and 184 patients in the post- intervention group. 77% (291/378) of patients were 3 days (72% compared to 57% in the pre- PDMP group; p=0.002). Patients without a reported opioid exposure prior to admission received a shorter course of oxycodone therapy post- intervention [RR=1.46, 95% CI 1.11-1.93; p=0.007, and the days of therapy were reduced by 29% [95% CI -43, -11; p=0.003] Patients with no history of cancer or sickle cell disease received shorter courses of therapy [RR=1.38, 95% CI 1.02-1.88; p=0.034]; however, the PDMP did not impact this finding.
Conclusions: In this study, a PDMP did not impact oxycodone prescribing at discharge except in patients without a history of opioid use prior to admission. This finding may have been influenced by the significant differences in past medical history and hospital course between the pre and post intervention patients. Additional research is needed to establish PDMPs effects on controlled substance prescribing at discharge among adult medicine patients.