Explaining sleep disorders in children after earthquakes using SHAP modelling: Container City modelling
Journal of Pediatric Nursing, cilt.89, ss.414-422, 2026 (SCI-Expanded, SSCI, Scopus)
- Yayın Türü: Makale / Tam Makale
- Cilt numarası: 89
- Basım Tarihi: 2026
- Doi Numarası: 10.1016/j.pedn.2026.04.027
- Dergi Adı: Journal of Pediatric Nursing
- Derginin Tarandığı İndeksler: Science Citation Index Expanded (SCI-EXPANDED), Social Sciences Citation Index (SSCI), Scopus, CINAHL, EMBASE, MEDLINE, Health Research Premium Collection (ProQuest)
- Sayfa Sayıları: ss.414-422
- Anahtar Kelimeler: Boosting machine learning algorithms, Child, Logistic regression, Machine learning, Random forest, Sleep wake disorders
- İnönü Üniversitesi Adresli: Evet
Özet
Background It is well-known that sleep disturbances and sleep-related traumas are the most common symptoms experienced by children following a traumatic event. Our study aims to predict the factors affecting post-traumatic sleep disorders in children living in container cities by using machine learning models. Methods This descriptive, cross-sectional study determined the average age of the children to be 13.72 ± 2.35 years. Research data were obtained using a child information form and a sleep disorder scale for children. A study of 1085 children living in container cities used logistic regression, random forest and XG-Boost models to make predictions and calculate accuracy metrics. Based on the assessment, sleep disorders were present in 600 children (55.3%), while the remaining 485 (44.7%) exhibited no such disorders. Results The comparison revealed that the XGBoost model achieved the best performance, with an accuracy of 0.89, an F1 score of 0.87, a negative predictive value of 0.89, a precision of 0.90 and a recall of 0.85. Conclusions Based on the SHAP importance values derived from the model, ‘sleep problem situation’ was identified as the most influential feature, while ‘having shelter problems in an earthquake’ was found to be the least influential. These findings highlight key associations that can inform the development of rehabilitation service plans and guide further investigation into factors related to sleep disorders in children during traumatic periods. Implications for practice This study suggests that psychiatric and paediatric nursing practice and education should include strategies to intervene with children after natural disasters.