Changing Practice Patterns Among Canadian Primary Care Physicians: Cross-Provincial Analysis of Linked Health Data

Abstract

Introduction
Canadians report persistent problems accessing primary care despite an increasing per-capita supply of primary care physicians (PCPs). There is speculation that PCPs, especially those early in their careers, may be working less and/or choosing to practice in focused clinical areas rather than comprehensive family medicine.

Objectives and Approach
Our objectives were to compare practice patterns among early-career (<10 years in practice) and established (10+ years) PCPs and to determine if any changes over time reflect cohort effects (attributes unique to recent cohorts of PCPs), or period effects (changes over time across all PCPs). We accessed linked population and physician registry data, physician billing data, hospital discharge records, and vital statistics in each of three Canadian provinces from 1996/7 to 2017/8. Physician fee schedules are negotiated separately by province which means that while fields capturing service dates, dollars billed, and diagnosis codes are directly comparable, other fields capture similar, but not identical information across provinces (e.g. specific fee codes and service location codes). We developed measures of services volume, continuity, and comprehensiveness of primary care that could be applied across provinces and used these to describe practice patterns over two decades and explore cohort and period effects.

Results
Measures of service volume (total patient contacts, number of unique patients seen) and continuity are lower among early-career physicians but the gap between early-career and established physicians has not changed. Measures of comprehensiveness of care, including provision of services in home, long-term care, hospital, and emergency department, as well as indicators of focused practice, hospitalist and locum care show more varied patterns, with declines in comprehensiveness across all physicians, but more evidence of focused practice among early-career physicians.

Conclusion / Implications
Findings have implications both for health workforce planning and organizational reforms to support access to comprehensive primary care.