After Health System-Wide Redesign of Default Options within the - - PowerPoint PPT Presentation

Change in Generic Medication Prescribing Rates After Health System-Wide Redesign of Default Options within the Electronic Health Record Mitesh Patel, MD, MBA, MS

June 28, 2016

Assistant Professor of Medicine and Health Care Management, Perelman School of Medicine and The Wharton School, University of Pennsylvania; Staff Physician, Crescenz Veterans Affairs Medical Center, Philadelphia

Medical Decision-Making and the EHR

• Manner in which choices are offered and framed

significantly influences our decisions

– Electronic health record (EHR) is a choice environment – Design of default options within the EHR is often overlooked and its impact on health care behaviors has not been well examined

• Default options are the path of least resistance

– Influence decisions without restricting choice

• Optimal setting when dominant option is clear

– Brand name medications with available generic equivalents

Halpern et al. NEJM. 2007. Patel et al. JGIM. 2012

Generics Utilization Improves Adherence and Outcomes

Choudhry et al. Ann Intern Med. 2016.

• 40% of Medicare beneficiaries with cardiovascular disease were prescribed

brand names that had available generic equivalents

• Medicare spent $1.4 billion more on brand name medications for diabetics

when compared to similar patients at the VA

• Estimated that at least $6 billion in drug spending could be saved by simply

using generics when they are available

Generic Utilization at Penn Medicine

• Ranked in the bottom quartile among peers in generic

prescribing rates

• Education and feedback to clinicians had little impact
• Intervention in the electronic health record

– Default was changed from opt-in to opt-out generic substitution – Implemented across all ambulatory clinics among all specialities within the entire health system in November 2014

Objectives

• To evaluate the effect of this health system-wide

change in default options on physician prescribing behaviors

• To test whether physicians maintained the ability to
• pt-out in cases where the brand name medication

was preferred

– Syndroid and levothyroxine often have different levels of thyroid hormone and switching can require dosage re-titration

Pre-Intervention: Opt-In Process

Brand Generic

Post-Intervention: Opt-Out Process

Study Design

• Data

– Monthly prescription rates from Clarity, an EPIC reporting database – Pre (Jan to Oct 2014) and post (Nov 2014 to June 2015) intervention

• Medications

– New prescriptions associated with 10 common outpatient conditions

• Acid reflux, anxiety/insomnia, bacterial infection, depression, diabetes, hyperlipidemia,

hypertension/CHF, hypothyroid, pain, seizure

– Excluded medications without generic equivalents and combination pills

• Analysis

– Chi-square tests for differences in generic prescribing rates – Difference-in-difference approach using a linear probability model to compare changes over time for synthroid/levothyroxine to other medications with similar baseline generic prescribing rates

Results

Percent of Prescriptions Ordered as Generic

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

Acid Reflux Pain Hypertension/CHF Bacterial Infection Seizure Depression Anxiety/Insomnia Diabetes Hyperlipidemia Hypothyroid

75.3% 98.4%

P < 0.001

Results - Opt-Out Differences

Limitations

• Observational data from one health system
• Data on prescribed, not dispensed medications

– However, Penn Medicine immediately moved from bottom to top quartile based on insurer claims data on medication dispensing rates

Summary

• Changing the default to an opt-out process for generic

substitution led to a significant and immediate increase in generic prescribing

• Low-cost intervention that did not disrupt clinician

workflow and preserved their ability to easily opt-out when necessary

• Further opportunities should be considered to

systematically design and test interventions that change EHR choice architecture to nudge clinician behavior

Acknowledgements

• Collaborators

– Susan Day, MD, MPH – Scott Halpern, MD, PhD – Bill Hanson, MD – Steve Honeywell Jr, BS – Joseph Martinez, BS – Kevin Volpp, MD, PhD

• Funding

– University of Pennsylvania Health System – National Institute on Aging (P30AGO34546)

Change in Generic Medication Prescribing Rates After Health System-Wide Redesign of Default Options within the Electronic Health Record Mitesh Patel, MD, MBA, MS

June 28, 2016

Assistant Professor of Medicine and Health Care Management, Perelman School of Medicine and The Wharton School, University of Pennsylvania; Staff Physician, Crescenz Veterans Affairs Medical Center, Philadelphia

Appendix

State Laws for Generic Substitution

• State laws for generic substitution

– All 50 states have generic substitution laws

• Only 15 states have mandatory laws

– 35 states have permissive laws

• Nearly 80% require patient consent at the pharmacy, which reduces

conversion to generic by 25%

• Overspending on brand name medications

– Medicaid wasted $329 million annually on brand names with existing generic equivalents – Medicare spent $1.4 billion more on brand name medications than the VA for diabetic patients

www.ncsl.org www.pcmanet.org Shrank et al. Health Affairs. 2010;29:1383-90 AEI 2011. http://www.aei.org/paper/10020 Gellad et al. Ann Intern Med. 2013;159:105-14.

Excess Spending from Brand Name Medications

Johansen et al. JAMA Internal Medicine. 2016.

Pilot Study Among General Internists

Medications IM Providers x Post-Intervention P-value All

5.4 <0.001

Beta-Blockers

10.5 <0.001

Statins

4.0 <0.01

PPIs

2.1 0.47