Testing effects of community collaboration on rates of low infant birthweight at the county level.

Interorganizational collaboration has become a popular strategy for addressing population health and well-being. However, evidence for its effectiveness in achieving outcomes at the population level is limited, at least in part due to a variety of methodological challenges such as reduced sample size at the population level, the availability of suitable comparison groups of communities, and study durations that are too short to detect slowly emerging outcomes. The present study addresses these challenges by retrospectively examining the effectiveness of a mature network of community collaboratives, using latent growth modeling of longitudinal change in an archival community-level outcome, low infant birthweight, and propensity score matching of comparison communities. A group of 25 Georgia counties with collaboratives targeting low infant birthweight was compared to a weighted comparison group of counties from other southeastern states, using propensity score matching. We report results of full matching methods and outcome analyses examining differences in change in county rates of low infant birthweight from 1997 to 2004 between intervention and comparison counties. Results indicated significantly smaller increases in low weight birth rates in intervention counties than in comparison counties.