A nonparametric method of comparing the partial areas under two correlated receiver operating characteristic curves

A non-parametric method was developed and tested to compare the partial areas under two correlated Receiver Operating Characteristic curves. Based on the theory of generalized U-statistics the mathematical formulas have been derived for computing ROC area, and the variance and covariance between the portions of two ROC curves. A practical SAS application also has been developed to facilitate the calculations. The accuracy of the non-parametric method was evaluated by comparing it to other methods. By applying our method to the data from a published ROC analysis of CT image, our results are very close to theirs. A hypothetical example was used to demonstrate the effects of two crossed ROC curves. The two ROC areas are the same. However each portion of the area between two ROC curves were found to be significantly different by the partial ROC curve analysis. For computation of ROC curves with large scales, such as a logistic regression model, we applied our method to the breast cancer study with Medicare claims data. It yielded the same ROC area computation as the SAS Logistic procedure. Our method also provides an alternative to the global summary of ROC area comparison by directly comparing the true-positive rates for two regression models and by determining the range of false-positive values where the models differ. ^

Catchment area of oral cleft cases for two major hospitals in Harris County

Background. Orofacial clefts are among the most common birth defects and considered to be of complex etiology with both genetic and environmental factors.^ Objectives. The purpose of this study was to describe maternal and infant characteristics, examine the catchment area, and determine if there are any geospatial patterns among infants with an orofacial cleft delivered at two major hospitals in Harris County, The Woman's Hospital of Texas and Memorial Hermann Hospital-Texas Medical Center, from January 1, 2003 through December 31, 2007.^ Methods. Data were obtained from two major hospitals in Harris County and included all babies delivered in the period from 2003 through 2007 with an orofacial cleft. Residential addresses were mapped using MapInfo GIS software and the cluster analysis performed with SaTScan software.^ Results. Ninety-nine cases were identified spanning nine counties. 26% of cases resided within a 5-mile radius of the Texas Medical Center. Birth rates ranged from 1.4 to 16.5 per 10,000 total births. A cluster was identified in southwest Harris County, however, it was not significant (p=0.066).^ Conclusion. This study encourages further focus on linking cleft cases to environmental factors in order to determine potential risks. ^