Declaration of Innovation FAQs
Frequently Asked Questions
Declaration of Innovation
Example Submission
Idea Title
The National Nutrition Counseling Benchmark: One Shared Measure, Whatever You Teach
Two Sentence Summary
Every college of osteopathic medicine teaches nutrition differently, and not one of them can show whether its students counsel patients any better than anyone else's. The Nutrition Counseling Benchmark gives all of them one set of shared assessments, so each college keeps its own curriculum and still learns how its graduates compare.
Option 1: 90-Second Video Pitch
Create your video and upload it to YouTube. Please include a link to your video on the form.
Option 2: 500 Word Written Pitch
Problem
The HHS Medical Education Nutrition Competency Framework lists seventy-one competencies. The four that decide whether a patient changes anything (counseling, behavior change, food-security screening and referral), are the hardest to assess, so most colleges do not assess them. The field has real, valuable variation in how nutrition is taught in our differing communities, and no way to learn from it. Colleges that invested cannot show a return. Colleges that did not cannot see the gap. And the colleges least able to answer are the ones with no research office to run a study of their own.
Solution
Standardize the measurement, not the curriculum. Every participating college administers the same self-assessment, the same set of AI virtual patient encounters scored with the same rubric, and the same OSCE cases: once before the nutrition curriculum and once after, using two matched forms in counterbalanced order. Curriculum stays local. Outcomes become comparable.
How It Works
EXAMPLE PATIENT
Dale, 52, a long-haul driver from eastern Kentucky with an A1c of 6.1. He feels fine and cannot eat healthy on the road. Three sodas a day, released only if the student asks about drinks. Food insecurity, released only if the student screens for it. Back pain that undoes any activity plan built without it. And one line Dale delivers unprompted: My dad lost his foot to sugar. What the student does in the ten seconds after that line is the assessable, and vital, action.
THE SCORE
Eight domains, seven adapted from instruments that already carry validity evidence. Empathy from the Empathic Communication Coding System, behavior change from MITI 4, history and communication from published OSCE work, guideline concordance from each case's own cited sources. Every AI score cites the transcript passage that justifies it, so any score can be audited and any score can be appealed. Each case also names its harmful-advice flags. Telling Dale to cut all carbohydrates is a critical error, and no flag affects a score without human review.
THE OSTEOPATHIC DOMAIN
Scored on its own, never folded into a total, then tested for discriminant validity against empathy and engagement. If it measures nothing those two already capture, we publish that finding.
The Pilot
The self-assessment, the virtual AI cases with feedback, one summative OSCE case. Three colleges. One academic year. The deliverable is the concordance study: does AI scoring of nutrition counseling agree with trained human raters closely enough to trust? Nobody has published that answer. And in something as difficult to assess as nutrition counseling, whether we can trust AI matters.
Who Receives It
Curriculum committees, who can finally read their own before-and-after scores against a national average instead of against nothing. Faculty who often hold no performance measure for nutrition counseling at all. After the pilot, the rubric and the formative cases publish openly, so adoption never depends on anyone continuing to fund a central program.
Closing
Keep your curriculum. Share the measure. Find out what works.