A mathematical model for predicting length of postoperative intensive care requirement following cardiac surgery in an Indian hospital

Abstract

Intensive care unit (ICU) is a critical resource in a hospital, especially in developing countries such as India. The length of ICU stay after a cardiac surgery is an important variable for effective use of this critical resource. In this context, a predictive model can help a hospital to make optimum use of its ICU occupancy. A study was thus conducted on ICU patients and data gather over a 1-year period in a hospital in India. The critical factors for prolonged ICU stay (more than 72 h) were identified using univariate and multivariate logistic regression and a predictive index was built based on model development set. The predictive index was tested on a validation set and the mean length of ICU stay appeared to increase with an increase in the risk score. In addition, the risk score was tested in case of mortality. Efficient use of the ICU facility is possible with the help of this predictive index.