The Haiti Breast Cancer Initiative: Initial Findings and Analysis of Barriers-to-Care Delaying Patient Presentation

Figure 2

Logistic regression analysis. Note: univariate logistic regression for covariates not meeting the inclusion criteria into the multivariate model is shown for reference. Only covariates that exhibit significance in the multivariate analysis are considered as independent risk factors for delayed patient presentation (barriers-to-care). Multivariate logistic regression model: 𝑐 statistic = 0.807, concordant pair percentage = 80.7%, discordant pair percentage = 19.3%, Hosmer-Lemeshow Goodness-of-Fit test: 𝜒²=3.95, 𝑃=0.86 (df = 8).