New prognostic score for mortality in critically ill patients. Development and validation

Area of Science:

• Critical Care Medicine
• Prognostic Modeling
• Health Informatics

Background:

• Predicting mortality in critically ill patients is crucial for effective resource allocation and treatment strategies.
• Existing models like APACHE II have limitations in capturing dynamic risk factors throughout the ICU stay.

Purpose of the Study:

• To develop and validate a novel prognostic model (NMP) for predicting ICU mortality.
• To improve upon the predictive accuracy of the APACHE II score by incorporating dynamic risk factors.

Main Methods:

• A post-hoc analysis of multicenter, prospective data from 167 Spanish hospitals (193 ICUs) over 7 years.
• Multivariable logistic regression was used to develop the model in an estimation group and validated in a separate cohort.
• The NMP incorporated APACHE II and nine additional factors assessed throughout the ICU stay.

Main Results:

• The study analyzed 137,666 patients, with 91,777 in the estimation group and 45,889 in validation.
• The NMP demonstrated superior discriminatory ability (AUROC = 0.872) compared to APACHE II alone (AUROC = 0.826).
• The NMP improved reclassification by 52%, identifying survivors and non-survivors more accurately.

Conclusions:

• A validated novel prognostic model (NMP) was developed, incorporating nine additional risk factors alongside APACHE II.
• The NMP offers enhanced risk stratification for critically ill patients, aiding clinical decision-making.
• This model provides a more comprehensive evaluation of mortality risk throughout the ICU stay.