Nomograms for Predicting High Hospitalization Costs and Prolonged Stay among Hospitalized Patients with pAECOPD

• Published in: Canadian respiratory journal 2024-09, Vol.2024 (1), p.2639080
• Main Authors: Dilixiati, Nafeisa, Lian, Mengyu, Hou, Ziliang, Song, Jie, Yang, Jingjing, Lin, Ruiyan, Wang, Jinxiang
• Format: Article
• Language: English
• Subjects: Aged; Aged, 80 and over; Analysis; Bacterial pneumonia; Blood cell count; Cardiovascular disease; Care and treatment; Chronic obstructive pulmonary disease; Community-Acquired Infections - economics; Comorbidity; Cor pulmonale; Critical care; Diabetes; Economic aspects; Female; Females; Gastroesophageal reflux; Hospital care; Hospital costs; Hospital Costs - statistics & numerical data; Hospital patients; Hospitalization; Hospitalization - economics; Hospitalization - statistics & numerical data; Humans; Intensive care; Length of Stay - economics; Length of Stay - statistics & numerical data; Leukocyte Count; Logistic Models; Lung diseases, Obstructive; Male; Medical care, Cost of; Medical research; Medicine, Experimental; Middle Aged; Mortality; Nomograms; Nomography (Mathematics); Physiological aspects; Pneumonia; Pneumonia - economics; Pneumonia - epidemiology; Pulmonary arteries; Pulmonary Disease, Chronic Obstructive - economics; Pulmonary Disease, Chronic Obstructive - therapy; Regression analysis; Respiratory failure; Respiratory system; Risk Factors; ROC Curve; Statistical analysis; Variables; Ventilators
• ISSN: 1198-2241; 1916-7245; 1916-7245
• DOI: 10.1155/2024/2639080

This study aimed to develop nomograms to predict high hospitalization costs and prolonged stays in hospitalized acute exacerbations of chronic obstructive pulmonary disease (AECOPD) patients with community-acquired pneumonia (CAP), also known as pAECOPD. A total of 635 patients with pAECOPD were included in this observational study and divided into training and testing sets. Variables were initially screened using univariate analysis, and then further selected using a backward stepwise regression. Multivariable logistic regression was performed to establish nomograms. The predictive performance of the model was evaluated using the receiver operating characteristic (ROC) curve, area under the curve (AUC), calibration curve, and decision curve analysis (DCA) in both the training and testing sets. Finally, the logistic regression analysis showed that elevated white blood cell count (WBC>10 × 10 cells/l), hypoalbuminemia, pulmonary encephalopathy, respiratory failure, diabetes, and respiratory intensive care unit (RICU) admissions were risk factors for predicting high hospitalization costs in pAECOPD patients. The AUC value was 0.756 (95% CI: 0.699-0.812) in the training set and 0.792 (95% CI: 0.718-0.867) in the testing set. The calibration plot and DCA curve indicated the model had good predictive performance. Furthermore, decreased total protein, pulmonary encephalopathy, reflux esophagitis, and RICU admissions were risk factors for predicting prolonged stays in pAECOPD patients. The AUC value was 0.629 (95% CI: 0.575-0.682) in the training set and 0.620 (95% CI: 0.539-0.701) in the testing set. The calibration plot and DCA curve indicated the model had good predictive performance. We developed and validated two nomograms for predicting high hospitalization costs and prolonged stay, respectively, among hospitalized patients with pAECOPD. This trial is registered with ChiCTR2000039959.