Development and validation of a nomogram for predicting prolonged ICU stays after pediatric cardiac surgery

Area of Science:

• Pediatric Cardiac Surgery
• Intensive Care Unit (ICU) Medicine
• Predictive Analytics

Background:

• Prolonged intensive care unit (ICU) stays pose challenges in pediatric cardiac surgery.
• Accurate prediction of prolonged ICU stays is crucial for resource allocation and patient management.

Purpose of the Study:

• To develop and validate a nomogram for estimating the probability of prolonged ICU stays in pediatric cardiac surgery patients.
• To identify key clinical variables associated with prolonged ICU stays.

Main Methods:

• Retrospective analysis of a pediatric intensive care database.
• Development of a nomogram using a training set (75%) and validation in a testing set (25%).
• Inclusion of 795 patients in the training set and 266 in the testing set.

Main Results:

• The nomogram incorporated eight variables: age, blood pressure, respiratory rate, bicarbonate, bilirubin, C-reactive protein, international normalized ratio, and operation time.
• The nomogram achieved an area under the curve of 0.812 (training) and 0.736 (testing), indicating good predictive performance.
• Decision curve analysis supported the clinical utility of the nomogram for patient management.

Conclusions:

• The developed nomogram is a valuable tool for early identification of pediatric cardiac surgery patients at risk of prolonged ICU stays.
• This tool can facilitate personalized postoperative care strategies, potentially improving patient outcomes.
• The nomogram demonstrated excellent discrimination and calibration, supporting its clinical applicability.