Eric Nauenberg1, Audrey Laporte, Leilei Shen. 1. Department of Health Policy, Management and Evaluation, University of Toronto, Ontario, Canada. eric.nauenberg@utoronto.ca
Abstract
OBJECTIVES: We examine the relationship between social capital, community size and GP visits, and conceptualize social capital as a stock variable measured at a prior point in time. METHODS: Data from the 2002 Canadian Community Health Survey and the 2001 Canadian Census are merged with GP visit data from the Ontario Health Ministry. Negative binomial regression is used to measure the impact of community-level (CSC) and individual-level social capital (ISC) on GP visits. CSC is measured with the Petris Index using employment levels in religious and community-based organizations, and ISC is measured along multiple dimensions. RESULTS: The effect of social capital varies by community size. A one standard deviation increase in the Petris Index in larger communities (population>100,000) leads to a 2.6% decrease in GP visits with an annual offset in public spending of $66.4M. Tangible social support-a measure of ISC-also exhibited large effects on GP visits. In smaller communities (population 10,000-100,000), only increased ISC exhibited an impact on GP visits. Age had no effect on the association between social capital and GP visits. CONCLUSIONS: Each form of social capital likely operates through different mechanisms and impact differs by community size. Stronger CSC likely obviates some physician visits in larger communities that involve counseling/caring services while some forms of ISC may act similarly in smaller communities.
OBJECTIVES: We examine the relationship between social capital, community size and GP visits, and conceptualize social capital as a stock variable measured at a prior point in time. METHODS: Data from the 2002 Canadian Community Health Survey and the 2001 Canadian Census are merged with GP visit data from the Ontario Health Ministry. Negative binomial regression is used to measure the impact of community-level (CSC) and individual-level social capital (ISC) on GP visits. CSC is measured with the Petris Index using employment levels in religious and community-based organizations, and ISC is measured along multiple dimensions. RESULTS: The effect of social capital varies by community size. A one standard deviation increase in the Petris Index in larger communities (population>100,000) leads to a 2.6% decrease in GP visits with an annual offset in public spending of $66.4M. Tangible social support-a measure of ISC-also exhibited large effects on GP visits. In smaller communities (population 10,000-100,000), only increased ISC exhibited an impact on GP visits. Age had no effect on the association between social capital and GP visits. CONCLUSIONS: Each form of social capital likely operates through different mechanisms and impact differs by community size. Stronger CSC likely obviates some physician visits in larger communities that involve counseling/caring services while some forms of ISC may act similarly in smaller communities.