The use of prognostic indicators in the development of a statistical model predictive for adrenal insufficiency in trauma patients

Abstract

We performed a retrospective chart review of trauma patients admitted to Palmetto Richland Memorial Hospital and identified 63 cases of adrenal insufficiency along with 65 trauma patient controls. Two statistical models, a neural network and a multiple logistic regression, were developed to predict patients with increased risk of developing adrenal insufficiency. Each model had 11 selected independent variables, along with patient demographic data, to make a probabilistic prediction of patient outcome. The neural network model was trained with 102 patients to identify linear and nonlinear relationships that could yield a predictive capability. The neural network achieved an accuracy of 71 per cent. The logistic regression model achieved an accuracy of 82 per cent. With these models, we have shown the feasibility of a method to more accurately screen patients with an increased risk of adrenal insufficiency. This ability should allow earlier identification and treatment of patients with adrenal insufficiency. Further development with a larger database is needed to improve the accuracy of the present models.