The availability of community ties predicts likelihood of peer referral for mammography: geographic constraints on viral marketing.

Engaging social networks to encourage preventive health behavior offers a supplement to conventional mass media campaigns and yet we do not fully understand the conditions that facilitate or hamper such interpersonal diffusion. One set of factors that should affect the diffusion of health campaign information involves a person's community. Variables describing geographic communities should predict the likelihood of residents accepting campaign invitations to pass along information to friends, family, and others. We investigate two aspects of a community--the availability of community ties and residential stability--as potential influences on diffusion of publicly-funded breast cancer screening in the United States in 2008-2009. In a survey study of 1515 participants living in 91 zip codes across the State of Minnesota, USA, we focus on the extent to which women refer others when given the opportunity to nominate family, friends, and peers to receive free mammograms. We predicted nomination tendency for a particular zip code would be a function of available community ties, measured as religious congregation density in that zip code, and also expected the predictive power of available ties would be greatest in communities with relatively high residential stability (meaning lower turnover in home residence). Results support our hypotheses. Congregation density positively predicted nomination tendency both in bivariate analysis and in Tobit regression models, and was most predictive in zip codes above the median in residential stability. We conclude that having a local infrastructure of social ties available in a community predicts the diffusion of available health care services in that community.

Duke Scholars

Author Brian Glen Southwell Medicine, General Internal Medicine