Primary care continuity and potentially avoidable hospitalization in persons with dementia

Abstract

Background/objective: To measure the association between high primary care continuity and potentially avoidable hospitalization in community-dwelling persons with dementia. Our hypothesis was that high primary care continuity is associated with fewer potentially avoidable hospitalizations.

Design: Population-based retrospective cohort (2012-2016), with inverse probability of treatment weighting using the propensity score.

Setting: Quebec (Canada) health administrative database, recording most primary, secondary and tertiary care services provided via the public universal health insurance system.

Participants: Population-based sample of 22,060 community-dwelling 65 + persons with dementia on March 31st, 2015, with at least two primary care visits in the preceding year (mean age 81 years, 60% female). Participants were followed for 1 year, or until death or long-term care admission.

Exposure: High primary care continuity on March 31st, 2015, i.e., having had every primary care visit with the same primary care physician, during the preceding year.

Main outcome measures: Primary: Potentially avoidable hospitalization in the follow-up period as defined by ambulatory care sensitive conditions (ACSC) hospitalization (general and older population definitions), 30-day hospital readmission; Secondary: Hospitalization and emergency department visit.

Results: Among the 22,060 persons, compared with the persons with low primary care continuity, the 14,515 (65.8%) persons with high primary care continuity had a lower risk of ACSC hospitalization (general population definition) (relative risk reduction 0.82, 95% CI 0.72-0.94), ACSC hospitalization (older population definition) (0.87, 0.79-0.95), 30-day hospital readmission (0.81, 0.72-0.92), hospitalization (0.90, 0.86-0.94), and emergency department visit (0.92, 0.90-0.95). The number needed to treat to prevent one event were, respectively, 118 (69-356), 87 (52-252), 97 (60-247), 23 (17-34), and 29 (21-47).

Conclusion: Increasing continuity with a primary care physician might be an avenue to reduce potentially avoidable hospitalizations in community-dwelling persons with dementia on a population-wide level.

Keywords: dementia; health service research; potentially avoidable hospitalization; primary care continuity; propensity score.