Machine learning regression models and associated variables for predicting preoperative anxiety in paediatric patients

Dear Editor, This journal published ‘Prediction of correlation between preoperative parents’ anxiety and their child’s anxiety before elective surgery under anaesthesia’.[1] The study addresses an important topic by highlighting a strong correlation between parental anxiety and its impact on a child’s preoperative anxiety, which can significantly influence surgical outcomes. However, this study yielded insightful findings on parental anxiety. Additionally, the authors could have considered psychological factors, such as the child’s temperament and coping mechanisms, as these can also have a significant impact. Exploring these factors could provide a more holistic understanding of preoperative anxiety in children. In this approach, linear regression models such as polynomial regression or ridge regression can be applied to develop predictive models and correlations between various dependent and independent variables with anxiety levels. The model learns from historical data by identifying patterns and relationships between input variables. We can consider independent variables for training the model, such as parent’s anxiety (State-Trait Anxiety Inventory STAI scores), sociodemographic factors (child’s age, gender, birth order, previous hospitalizations, type of surgery, residence), and parent’s age, gender, and education level.[1] By predicting the various factors responsible for children’s preoperative anxiety along with anxiety levels ahead of time, healthcare providers can implement targeted interventions such as psychological counselling or pharmacological treatments, improving the child’s surgical experience and outcomes. Furthermore, incorporating large language models such as ChatGPT can enhance this approach by providing real-time communication support for parents, further reducing anxiety and refining the model’s predictions. Financial support and sponsorship Nil. Conflicts of interest There are no conflicts of interest.