Behavioural disorders in children with pectus excavatum in China: a retrospective cohort study with propensity score matching and risk prediction model

OBJECTIVES Surgical repair of pectus excavatum is typically carried out immediately prior to puberty. However, at the time of surgery, some psychosocial issues, such as behavioural disorders may already have developed and the likelihood of these psychosocial disorders resolving after surgery is unclear. For this reason, some surgeons choose to perform surgical repair at an earlier age in some children. The study retrospectively compares the rate of behavioural disorders in children undergoing the Nuss procedure at 4 vs. 10 years of age. We also attempted to develop a model to predict the risk of behavioural disorders in 10 year-old patients. METHODS The current study included children receiving Nuss procedure for pectus excavatum at either 4 or 10 years of age. The presence/absence of behavioural disorder was assessed preoperatively, and in the third year, after removal of the bar. A propensity score matching (PSM) analysis was conducted to reduce the potential for confounding by baseline factors. Multivariable logistic regression was conducted to establish a model to predict the risk of behavioural disorders in the third year after the removal of the bar. The model was accessed by discrimination and calibration. A formula and a nomogram were developed based on the results. RESULTS The number of patients in each group was 45 after PSM. The rate of behavioural disorders at the baseline was significantly higher in the children undergoing Nuss procedure at 10 years of age [36% vs 20%, odds ratio (OR) 2.21, 95% confidence interval (CI) 0.85–5.72; P = 0.157]. The rate of behavioural disorders in the third year after the removal of the bar was 36% and 18% in children undergoing surgery at 10 and 4 years of age, respectively (OR 2.55, 95% CI 0.96–6.79; P = 0.094). The rate of persistent behavioural disorders, defined as continuing to have behavioural disorders in the third year after the removal of the bar in those with behavioural disorders at the baseline, was 88% vs 56% (OR 3.47, 95% CI 0.56–21.36; P = 0.204). Two patients (4%) relapsed in each group. A risk prediction model by variables of gender, Haller index, pulmonary function and score of Child Behaviour Checklist at the baseline was provided. CONCLUSIONS The rate of behavioural disorders was considerably lower in children who underwent the Nuss procedure at 4 years of age than at 10 years of age. Behavioural disorders may not readily resolve after surgery. Performing surgery at an early age rather than just before puberty may be better for psychosocial development. Psychosocial aid is necessary in addition to surgery to address behavioural disorders.