Going by the Book: Rule-Following in Primary Care Teams

Coordinated rule following is often important for ensuring consistent performance by execution-oriented teams in high-reliability industries. In health care, clinical practice guidelines codify best practices as rules for providers to follow. We examine primary care teams that manage new-onset type II diabetes, where diagnosis is based on an abnormally elevated hemoglobin A1c lab value. Using nationwide electronic health record data that include more than 12 million visits by over 1 million patients from 2013 to 2018, we apply a regression discontinuity design—with hemoglobin A1c as the running variable—and a difference-in-discontinuity strategy to compare providers’ rule-following responses to the diabetes diagnosis threshold across different team characteristics, including size, concentration, predominant provider type, and familiarity. We measure team performance on rule following and clinical outcomes, including diabetes diagnosis, prescribing of medications, diabetes monitoring, and guideline-concordant disease control. We find that teams that are larger and those with effort more diffusely distributed across members are more responsive to the diabetes diagnosis threshold than are solo providers. We find no meaningful differences in team performance based on whether the predominant provider is a physician or a nurse practitioner, nor do we see an impact of team familiarity. Health care managers can consider organizing large or diffuse teams—where each member contributes meaningfully but not exhaustively—to promote rule following and manage chronic diseases more effectively.