Nudges built into the electronic health record (EHR) can guide health care workers and patients toward evidence-based choices for better health outcomes and more ease on practitioners. A nudge alters how information is presented; in the EHR examples include default selections, automated alerts, language choice, and workflow changes. When health systems share successes and challenges with EHR nudging, those lessons can inform where new nudges could be used or uncover ways that enable clinicians and researchers to test, implement, and scale nudge interventions more easily and effectively.

That premise led to a two-day on-site exchange between Penn Medicine and its EHR partner, Epic, this February. At the gathering, Penn researchers and operational leaders shared examples of EHR nudges they had implemented and discussed “nuts and bolts” that determine whether a great idea becomes a usable workflow. Epic R&D leads and software developers showcased current and planned capabilities – such as automations, AI-generated insights, and A/B testing for alerts – to enhance care delivery with a focus on nudge opportunities.

Epic developers heard directly from Penn clinicians and researchers about what they see as the most high-impact opportunities for consideration:

• streamlined bulk workflows and “pended order” automation to make it easier for care teams to act on high-priority opportunities at scale;
• stronger built-in experimentation capabilities (e.g., A/B testing and randomization) so teams can learn quickly and rigorously which approaches work best, for whom, and in what contexts;
• clear pathways to share learnings – including negative findings – so effective interventions can be adapted and scaled across organizations; and
• more accessible prototyping resources to facilitate user-centered design and accelerate translation of ideas to testable, workflow-ready builds.

The February visit, hosted by Penn Medicine’s M. Kit Delgado, MD, MS, and Srinath Adusumalli, MD, MSHP, MBMI, grew out of conversations begun at Penn’s Nudges in Health Care Symposium, where leaders from health systems and industry gather to share what’s working (and what’s not) in behaviorally informed care delivery. 

At its core, the exchange reinforced a shared ambition: to translate behavioral science into everyday clinical impact, while keeping workflows usable and evaluation rigorous. A crucial part of that work is moving from bespoke, one-off interventions to reusable patterns that can be applied across care sites and tasks. 

“There are not just tens, hundreds, or thousands of these evidence-based practice measures that we could implement or that we could test and ideate and develop new research around,” said Srinivas Sridhara, PhD, chief data and analytics officer at Penn Medicine. “A fundamental ask is: Help us go from 10 to 10,000.”