# Predicting Perioperative Respiratory Adverse Events in Children Undergoing Elective Surgeries Under General Anesthesia Using COLDS Score: A Prospective Observational Study

## TL;DR

This study shows that the COLDS score can help predict breathing problems in children with colds undergoing surgery, but it should be used with other clinical assessments.

## Contribution

The study evaluates the effectiveness of the COLDS score in predicting perioperative respiratory adverse events in children with URTI.

## Key findings

- The COLDS score had a high specificity (97.96%) in identifying children without postoperative complications.
- The score showed strong discriminatory ability with an AUC of 0.92.
- However, its sensitivity was low at 40%, indicating it missed many children who experienced complications.

## Abstract

Background and aims

Pediatric patients undergoing surgeries under general anesthesia (GA) who have a recent upper respiratory tract infection (URTI) pose a unique challenge for the anesthesiologists. We aimed to utilize the COLDS score as a pre-anesthetic risk assessment tool to predict the likelihood of perioperative respiratory adverse events (PRAEs) in children with URTI.

Methods and materials

After ethical approval, we prospectively collected data over six months from children undergoing various ear, nose, and throat (ENT) surgeries under GA. Children above one year undergoing elective ENT surgeries were included. Children less than one year, non-ENT surgical pediatric patients, and those undergoing emergency surgeries were excluded.

Results

A total of 270 patients were included in the analysis, among whom 25 (9.3%) experienced postoperative events. The COLDS score was able to effectively distinguish between patients who did and did not experience postoperative complications. The receiver operating characteristic (ROC) curve (AUC) of 0.92 (95% CI: 0.86-0.99) suggests a strong discriminatory ability of the score. The specificity of the model was high at 97.96% (95% CI: 95.9-99.6), suggesting a strong ability to correctly identify patients without events. The sensitivity was lower at 40.0% (95% CI: 20.7-60.0), i.e., only 40% of patients who experienced complications were detected by the score.

Conclusions

COLDS score can help in distinguishing patients who will and will not have PRAEs. But as the sensitivity is moderate, we recommend using the COLDS score along with a detailed clinical assessment based on the type of surgery planned.