deteriorate is a challenge faced by many emergency physicians. Comparing these patients who die or are transferred to an ICU within 48 hours with those who remain at primary wards may form the basis of a prediction model of ICU transfer or death.

Data Breakdown: In a study,  researchers found that respiratory rate, temperature, and number of failing organs were independent predictors of ICU transfer or death. A prediction model using these predictors had a sensitivity of 63%, specificity of 93%, positive predictive value of 72%, and negative predictive value of 90% for detecting ICU transfer or death.

Take Home Pearl: Among ED patients with suspected infection, readily available clinical and laboratory variables at presentation appear to predict ICU transfer or death.