Predicting Prolonged Hospitalization in Pneumonia: The Role of Lung Injury Score

Abstrak

Introduction: Pneumonia is a lung infection with a high prevalence and is associated with prolonged length of stay (LOS). An excessive LOS increases the risk of complications, while a short LOS may lead to inadequate treatment. The lung injury score (LIS), which assesses the severity of lung injury and respiratory dysfunction, is expected to be an accurate predictor of LOS in patients with pneumonia. This study aimed to determine the correlation between LIS and LOS in hospitalized patients with pneumonia. Methods: This retrospective cohort study included adult patients hospitalized at Dr. Soetomo General Academic Hospital, Surabaya, Indonesia, with a diagnosis of pneumonia between January and December 2023. The LIS component scores (chest X-ray findings, arterial partial pressure of oxygen-to-fractional inspired oxygen ratio, positive end-expiratory pressure (PEEP), and static lung compliance) were obtained from electronic medical records and used to calculate LIS values. The correlation between LIS and LOS was analyzed using the Spearman test, while the cut-off point was determined using receiver operating characteristic (ROC) curve analysis. Results: A total of 152 patients fulfilled the inclusion criteria. The mean LIS was 1.82±0.89, while the mean LOS was 10.23±6.27 days. A statistically significant correlation was found between LIS and LOS (p<0.001) with a correlation coefficient of 0.317. Receiver operating characteristic curve analysis identified the optimal LIS cut-off point for an LOS of 10 days as 1.75. Conclusion: There was a statistically significant correlation between LIS and LOS in patients with pneumonia. The LIS cut-off point of 1.75 for an LOS of 10 days was statistically significant.