How to Drive Department Collaboration and Physician Adoption: Dr Howard Dinh’s Quest at Kaiser Permanente

Oct 6, 2026

21:10

Digital Health & Health Tech Health IT Primary Care

Episode Description

In this episode of MDAware: Upgrading Clinical Judgment, Dr. Graham Walker talks with Dr. Howard Dinh about how to get physicians to better use clinical decision tools, and how a health system decides which tools are most worth building in the first place. A conversation about why low-value testing often has less to do with a doctor's knowledge than with time pressure, anxious patients, and the incentives they work under, and what it takes to put guidance at the point of care in a way clinicians trust.

Dr. Dinh is Chair of Cardiology for Kaiser Permanente Northern California and leads the region's clinical decision support subgroup. He spent ten years in practice before joining Kaiser, and he and Graham are former colleagues at TPMG. He traces the group's start a decade ago, when about 70% of ischemic testing across KP Northern California was going to patients who didn't need it, and explains how talking with primary care doctors, and understanding what they were dealing with, led to a stable chest pain tool that does the pretest probability math for them. From there, he covers the practical side of scaling and driving adoption across 22 medical centers: Dr. Walker and Dr. Dinh also take on whether clinical decision support only makes sense in value-based care, or whether fee-for-service practices face the same pressure. The conversation then turns to AI: why answers that shift with the prompt worry Dr. Dinh, where machine learning already supports KP's risk models, and a new joint venture that will let Kaiser test its tools against ten years of real patient outcomes. He closes with advice for physicians who want to start a decision support group at their own institution.

What You'll Learn

Why much of the overuse of ischemic testing came from system pressures rather than physician judgment
How to get medical centers aligned on one algorithm while leaving room for local differences
Why a tool that takes six clicks can still save time if it documents the visit for you
How value-based and fee-for-service incentives shape whether decision support gets used
How KP plans to check its tools against a decade of patient outcomes so it can help set guidelines instead of only following them