Implementation and Utilization of Real-time Prescription Benefit Tools Across Three Large Academic Health Systems
Muhammad Ali,Jeremy I Schwartz,Jiangxia Wang,Jing Liu,Niteesh Potu,William Rottier,Peter Dziedzic,Erika Smith,Tanvi Mehta,Nitu Kashyap,Michael J. Fliotsos,Jing Luo,Jeremy Epstein,Jodi B. Segal,Harold Lehmann,Bradley H. Crotty,Fasika A. Woreta
DOI: https://doi.org/10.1101/2024.12.27.24318670
2024-12-30
Abstract:Background: Real-time prescription benefit (RTPB) tools provide clinicians an out-of-pocket cost (OOPC) estimate at the time of prescribing and may help them select medications with lower OOPC for their patients.
Objective: To evaluate how RTPB tools altered medication orders during clinical encounters and assess patient and clinician characteristics associated with the display of RTPB tool and alternative selection.
Design: Retrospective cohort study
Setting: Outpatient encounters at three academic medical centers
Participants: Patients and clinicians
Measurements: Patient characteristics, clinician characteristics and medication alternatives suggested by RTPB tool were compared across the three sites.
Results: From April 2019 to October 2021, 6,562,442 patient encounters occurred between 3,624 clinicians and 1,261,551 unique patients. Medications were prescribed in 2,152,772 (32.8%) encounters, of which RTPB tool retrieved and displayed alternatives in 968,811 (45.0%) encounters. The clinician selected an alternative in 68,731/968,811 (7.1%) of the encounters during which 89,050 medications were prescribed. The unit cost of alternative medications remained the same for most orders (n=41,212; 58.4%), while 18,629 (26.4%) had lower cost and 10,728 (15.2%) alternate orders had higher cost. Clinicians selected a different pharmacy among 39,634 (44.5%) and a different pharmacy type (mail vs retail) among 7,508 (12.7%) of the alternate medication orders, of which most were to mail order 4,680 (62.3%).
Limitations: We could not assess the role of pharmacy benefit manager coverage, cash-based alternative pricing, and impact of prior authorizations which may be assessed during future investigations.
Conclusion: Alternate prescribing after implementation of RTPB tools was low across the three institutions. Unit cost of the medication did not often change. Most changes reflected pharmacy choice. Further studies are needed to assess the barriers in adoption of RTPB tools.