Development and validation of a score for prediction of postoperative respiratory complications in infants and children (SPORC-C).

ABSTRACT

BACKGROUND: In infants and children, postoperative respiratory complications are leading causes of perioperative morbidity, mortality, and increased healthcare utilisation. We aimed to develop a novel score for prediction of postoperative respiratory complications in paediatric patients (SPORC for children). METHODS: We analysed data from paediatric patients (≤12 yr) undergoing surgery in New York and Boston, USA for score development and external validation. The primary outcome was postoperative respiratory complications within 30 days after surgery, defined as respiratory infection, respiratory failure, aspiration pneumonitis, pneumothorax, pleural effusion, bronchospasm, laryngospasm, and reintubation. Data from Children's Hospital at Montefiore were used to create the score by stepwise backwards elimination using multivariate logistic regression. External validation was conducted using a separate cohort of children who underwent surgery at Massachusetts General Hospital for Children. RESULTS: The study included data from children undergoing 32,187‬ surgical procedures, where 768 (2.4%) children experienced postoperative respiratory complications. The final score consisted of 11 predictors, and showed discriminatory ability in development, internal, and external validation cohorts with areas under the receiver operating characteristic curve of 0.85 (95% confidence interval 0.83-0.87), 0.84 (0.80-0.87), and 0.83 (0.80-0.86), respectively. CONCLUSION: SPORC is a novel validated score for predicting the likelihood of postoperative respiratory complications in children that can be used to predict postoperative respiratory complications in infants and children.