Development and Validation of a Nomogram Prediction Model for Key Symptoms of Post-Intensive Care Syndrome-Family in Family Members of Critically-Ill Patients: Focusing on Sleep Disturbance, Fatigue, Anxiety, and Depression

Haili Dong,1,2 Li Liu,1 Shasha Ma,3 Haixia Han,4 Jiadong Zhang,5 Jordan Tovera Salvador,6 Xiaoxiao Liu2 1Department of Nursing, Binzhou Medical University Hospital, Binzhou, Shandong Province, People’s Republic of China; 2School of Nursing, Binzhou Medical University, Binzhou, Shandong Province, People’s Republic of China; 3Department of Neurosurgery, Binzhou Medical University Hospital, Binzhou, Shandong Province, People’s Republic of China; 4Department of Emergency Intensive Care Unit, Binzhou Medical University Hospital, Binzhou, Shandong Province, People’s Republic of China; 5Department of Intensive Care Unit, Binzhou Medical University Hospital, Binzhou, Shandong Province, People’s Republic of China; 6Department of Nursing Education, College of Nursing, Imam Abdulrahman Bin Faisal University, Dammam, Saudi ArabiaCorrespondence: Li Liu, Email [email protected]: To construct and validate a nomogram model predicting the risk of post-intensive care syndrome-family (PICS-F) in family members of critically ill patients.Methods: This study was conducted on family members (parents, spouses, or children) of critically ill patients in the three intensive care units of Binzhou Medical University Hospital from December 2023 to June 2024, responsible for medical decisions and primary care. The sleep disturbances, fatigue, anxiety, and depression were assessed using the Pittsburgh Sleep Quality Index, the Subscale of Fatigue Assessment Instrument, and the Hospital Anxiety and Depression Scale, respectively. Predictive factors were identified through univariate and multivariate logistic regression, and a nomogram was constructed using R4.2.3. Internal validation used the Bootstrap sampling method, and external validation employed the time-period method.Results: The study involved 567 participants divided into a modeling group (n = 432; median age = 46 years; 209 males, 223 females) and an external validation group (n = 135; median age = 45 years; 70 males, 65 females). The model included five predictive factors: family age, patient age, APACHE II score, average monthly income per family member, and PSSS score. The AUC of the modeling group was 0.894 (0.864 ~ 0.924), with a specificity of 85.4%, a sensitivity of 78.0%, and a maximum Youden index of 0.634. The H–L test revealed a good fit (X2 value = 9.528, P = 0.300). The internal validation results of the Bootstrap sampling method showed that the calibration curve of the model was close to the ideal curve, and the DCA curve results indicated high clinical practicality. Moreover, the external validation results showed that AUC was 0.847 (0.782 ~ 0.912), with sensitivity and specificity of 74.5% and 86.3%, respectively. The H–L test results indicated a good fit (X2 value = 9.625, P = 0.292).Conclusion: The nomogram demonstrated strong predictive performance for PICS-F risk in ICU patients’ families, offering a valuable tool for clinical assessment.Keywords: post-intensive care syndrome-family, nomogram model, prediction model, nursing