Louis M. Mayaud
Alumni of Saint-Hilda's College
CDT Postgraduate Studentship in Healthcare Innovation
Role of dynamic information in prediction of acute events in septic patients in Intensive Care Unit
Traditional severity score for predicting mortality in the Intensive Care Unit (ICU) uses patients’ physiological variables at a point in time (admission and first 24 hours). These scores usually underperform on homogenous subpopulation of patients such as severe sepsis patients. The main hypothesis of this work is that patient’s dynamic information in response to acute events (such as hypotension) and its treatment could provide additional predictive information.
We developed a prediction algorithm for hospital mortality in patients with sepsis and hypotension requiring medical intervention using data from the Multiparameter Intelligent Monitoring in Intensive Care II (MIMIC-II database). We extracted 189 candidate variables, including treatments, physiologic variables and laboratory values collected before, during and after a hypotensive episode.
Thirty predictors were identified using a genetic algorithm on a training set (n=1500), and validated with a logistic regression model on an independent validation set (n=613). The final prediction algorithm used included dynamic information and had good discrimination (AUC = 82.0%) and calibration (Hosmer-Lemeshow C statistic = 10.43, p=0.06). This model was compared to APACHE IV using reclassification indices and was found to be superior with a NRI of 0.19 (p<0.001) and an IDI of 0.09 (p<0.001).