Algorithmic Identification of the Daily Primary Frontline Clinician during Inpatient Medicine Encounters Using EHR...

Area of Science:

• Clinical Informatics
• Health Services Research
• Patient Care Management

Background:

• Identifying frontline clinicians is crucial for patient care but traditionally labor-intensive.
• Electronic Health Record (EHR) audit logs offer a scalable solution, yet their application for this purpose is underdeveloped.

Purpose of the Study:

• To develop and validate an algorithm using EHR audit logs.
• To accurately identify the daily frontline clinician for inpatient medicine encounters on a per patient-day basis.

Main Methods:

• A cross-sectional cohort study was conducted across 12 hospitals within a single health system.
• Four algorithm iterations were designed and compared against manual chart review for accuracy.
• The study included adult inpatient medicine encounters with a minimum 3-day length of stay, excluding ICU admissions.

Main Results:

• The best-performing algorithm achieved 91% accuracy in identifying the daily frontline clinician, significantly outperforming a previous method (78% accuracy).
• Common algorithm errors included misidentified specialties and ambiguity during care transitions.
• The algorithm was applied to over 34,000 patient-days, identifying attending physicians (79%), resident physicians (9%), and advanced practice providers (12%) as frontline clinicians.

Conclusions:

• A scalable, EHR audit log-based algorithm was successfully developed.
• The algorithm demonstrates high accuracy in identifying daily frontline clinicians compared to manual chart review.
• This tool can enhance the understanding of clinical workflows and patient care delivery.