Canadian family physicians' decision to collaborate: age, period and cohort effects

Abstract

One of the core primary care reform initiatives seen across provinces in Canada is the introduction of inter-professional primary healthcare teams in which family physicians are encouraged to collaborate with other health professionals. Although a higher proportion of physicians are collaborating with various health professionals now compared to the previous decade, a substantial number of physicians still do not work in a collaborative setting. The objective of this paper is to examine the age, period and cohort effects of Canadian family physicians' decisions to collaborate with seven types of health professionals: specialists, nurse practitioners, nurses, dieticians, physiotherapists, psychologists and occupational therapists. To this end, this paper employs a multivariate probit model consisting of seven equations and a cross-classified fixed-effects strategy to explain the collaborative decisions of family physicians. Utilizing three cross-sectional physician surveys from Canada over the 2001-2007 period, cohorts are defined over five-year intervals according to their year of graduation from medical school. We find that newer cohorts of physicians are more likely to collaborate with dieticians, physiotherapists, psychologists and occupational therapists; newer female cohorts are more likely to collaborate with nurses while newer male cohorts are less likely to collaborate with nurses but more likely to collaborate with specialists. Older physicians are more likely to collaborate with specialists, physiotherapists, psychologists, and occupational therapists; the age effect for nurses is U-shaped for male physicians while it is inverse U-shaped for females. Family physicians are collaborating more with all seven health professionals in 2004 and 2007 compared to 2001. Belonging to a group practice has a largely positive influence on collaborations; and being paid by a fee-for-service remuneration scheme exerts a negative influence on collaboration, ceteris paribus. The findings suggest that combining a non-fee-for-service remuneration arrangement with a group practice structure would facilitate effective collaboration.