Community-level predictors of pneumococcal carriage and resistance in young children

Pneumococcal carriage and resistance vary markedly among communities and are not fully explained by individual predictors. Community risk factors may explain this variation. The authors geocoded addresses from a multicommunity sample of 710 Massachusetts children previously swabbed for pneumococcal carriage in 2001. Using regression models, the authors evaluated associations between census tract measures and pneumococcal carriage or resistance. Living in census tracts with an average household size of more than 2.9 predicted carriage (odds ratio = 3.0, 95% confidence interval: 1.7, 5.5), and living in socioeconomically disadvantaged census tracts conferred an additional two- to threefold odds of carriage equal to attending child care. Living in a census tract with a median household income of less than $35,000 predicted carriage among nonattendees (odds ratio = 2.7, 95% confidence interval: 1.7, 4.3). The predictive value of a low-income census tract was interchangeable with any of several socioeconomic measures, including poverty, unemployment, low educational attainment, and low owner occupancy, in addition to high density of children and limited household plumbing facilities. Furthermore, living in census tracts with low educational attainment significantly predicted resistance (odds ratio = 4.0, 95% confidence interval: 1.3, 12.7) and was interchangeable with a high density of children (odds ratio = 3.5, 95% confidence interval: 1.0, 11.7). The two- to threefold odds of pneumococcal carriage conferred by certain community characteristics suggest that these measures may target communities for interventions to decrease transmission.