Spatial Clustering of Suicides and Neighborhood Determinants in North Carolina, 2000 to 2017

Few studies in the Southeastern U.S. have examined county-level spatial patterning in suicide clusters, and no studies have examined clustering at the census block group. The objective of this retrospective ecological study is to identify high-risk suicide clusters and characterize the community-level factors associated with suicides inside and outside spatial clusters. We used the discrete Poisson SatScan statistic to identify spatial clusters in suicide for North Carolina, 2000–2017. A suicide cluster was defined as a statistically significant cluster of suicide events. Community-level determinants were obtained from the American Community Survey, and logistic regression models were used to examine the association between community-level determinants and suicide clusters. A total of 12 statistically significant high-risk spatial clusters were identified. Clusters were also identified for specific age-demographics, including adolescents ( ). The risk ratios of suicide varied from 1.27 to 2.05 in high-risk clusters, and spatial clustering was positively associated with being male or residence in a rural area. Multivariable logistic regression found strong associations with income, population change, and educational attainment. Our results highlight the significant geographic heterogeneity in suicide across North Carolina and the need for more research that identifies localized suicide clusters for targeted public health interventions.