Diagnostic statistics for predictive models of Triatoma dimidiata intradomiciliary prevalence >5%.

Key to department and study abbreviations: Dept, department; BV, Baja Verapaz; JU, Jutiapa; UVG, Universidad del Valle de Guatemala; GNMH; Guatemala National Ministry of Health. Key to model abbreviations: ENV, environmental model; DOM, domicile construction material model; ALL, combination of census and environmental models. Key to accuracy measure abbreviations: AUC, area under receiver-operator curve; Max κ, maximum kappa; PPV, positive predictive value; NPV, negative predictive value.Multivariate logistic regression models were developed to estimate the predictive probability of Triatoma dimidiata intradomiciliary village prevalence >5%. For each department and study, predictive models of environmental and domicile construction risk factors were developed separately and together. Overall model accuracy was compared using the area under the receiver-operator curve (AUC). Sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) were calculated using the probability threshold with maximum value of kappa (κ).