Identification of US Counties at Elevated Risk for Congenital Syphilis Using Predictive Modeling and a Risk Scoring System.

Autor:	Cuffe KM; From the Division of STD Prevention, Centers for Disease Control and Prevention, Atlanta, GA., Kang JDY; From the Division of STD Prevention, Centers for Disease Control and Prevention, Atlanta, GA., Dorji T; Oak Ridge Institute for Science and Education, Oak Ridge, TN., Bowen VB; From the Division of STD Prevention, Centers for Disease Control and Prevention, Atlanta, GA., Leichliter JS; From the Division of STD Prevention, Centers for Disease Control and Prevention, Atlanta, GA., Torrone E; From the Division of STD Prevention, Centers for Disease Control and Prevention, Atlanta, GA., Bernstein KT; From the Division of STD Prevention, Centers for Disease Control and Prevention, Atlanta, GA.
Jazyk:	angličtina
Zdroj:	Sexually transmitted diseases [Sex Transm Dis] 2020 May; Vol. 47 (5), pp. 290-295.
DOI:	10.1097/OLQ.0000000000001142
Abstrakt:	Background: Although preventable through timely screening and treatment, congenital syphilis (CS) rates are increasing in the United States, occurring in 5% of counties in 2015. Although individual-level factors are important predictors of CS, given the geographic focus of CS, it is also imperative to understand what county-level factors are associated with CS. Methods: This is a secondary analysis of reported county CS cases to the National Notifiable Diseases Surveillance System during the periods 2014-2015 and 2016-2017. We developed a predictive model to identify county-level factors associated with CS and use these to predict counties at elevated risk for future CS. Results: Our final model identified 973 (31.0% of all US counties) counties at elevated risk for CS (sensitivity, 88.1%; specificity, 74.0%). County factors that were predictive of CS included metropolitan area, income inequality, primary and secondary syphilis rates among women and men who have sex with men, and population proportions of those who are non-Hispanic black, Hispanic, living in urban areas, and uninsured. The predictive model using 2014-2015 CS outcome data was predictive of 2016-2017 CS cases (area under the curve value, 89.2%) CONCLUSIONS: Given the dire consequences of CS, increasing prevention efforts remains important. The ability to predict counties at most elevated risk for CS based on county factors may help target CS resources where they are needed most.
Databáze:	MEDLINE
Externí odkaz:	Zobrazit plný text záznamu