Lung Organ Failure Score for Predicting In-Hospital Mortality after Blunt Thoracic Trauma in Indonesia: A Retrospective Validation Study

Abstract

Background: The Lung Organ Failure Score (LOFS) predicts pulmonary organ failure; its performance for in-hospital death after blunt thoracic trauma in Indonesia is uncertain.

Objective: To evaluate LOFS discrimination and classification of in-hospital death.

Methods: This retrospective cohort included adults with blunt thoracic trauma admitted to Dr. Mohammad Hoesin General Hospital, Palembang, Indonesia, from January 2023 to January 2025. LOFS used six-hour fluid data and operative records. Discrimination and classification accuracy were estimated. Because patient-level cross-classifications were unavailable, sensitivity analyses fitted Firth-penalized models across every joint table consistent with the observed margins.

Results: Among 32 patients, 7 died (21.9%; 95% confidence interval [CI] 11.0–38.8). Mean LOFS was 21.00 ± 5.29 in non-survivors and 14.16 ± 3.92 in survivors (difference 6.84; 95% CI 3.15–10.53; p < 0.001). The area under the receiver operating characteristic curve was 0.840 (95% CI 0.553–0.957). LOFS ≥18 had sensitivity 85.7% (95% CI 48.7–97.4), specificity 80.0% (60.9–91.1), and negative predictive value 95.2% (77.3–99.2), matching severe thoracic injury (Abbreviated Injury Scale thorax ≥4) alone. Across admissible datasets, adjusted odds ratios remained above 1 after adjustment for severe head injury (12.3–64.4) or sex (14.4–143); incremental value beyond severe thoracic injury was unidentifiable.

Conclusion: LOFS discriminated death imprecisely and, at ≥18, performed no better than severe thoracic injury alone. Validation with at least 100 deaths is needed before clinical use.