Easy-to-use Scoring System Predicting 30-day Mortality of Patients with Acute Exacerbation of Idiopathic Pulmonary Fibrosis = Easy-to-use Scoring System Predicting 30-day Mortality of Patients with Acute Exacerbation of Idiopathic Pulmonary Fibrosis

다국어 초록 (Multilingual Abstract)

Background
Acute exacerbation (AE) of idiopathic pulmonary fibrosis (IPF) is the leading cause of death in IPF patients. While AE-IPF in-hospital mortality reaches near 50%, what distinguishes non-survivors from survivors is less known. Here, we present an easy-to-use scoring system predicting mortality of AE-IPF patients with six clinical markers that can be quickly acquired at patient’s initial presentation.
Methods
A retrospective derivation cohort of 769 IPF patients between January 2001 to December 2019 was constructed. Following the latest 2016 AE-IPF diagnostic criteria, we focused on 134 patients who experienced AE for the first-time. Medical history, vital signs, and laboratory Results from emergency department visits or hospital admissions were collected via medical record review. We used beta-coefficients of statistically and/or clinically significant factors from multiple logistic binary regression analysis (Primary outcome: 30-day mortality) to design a scoring system. Its performance was assessed with receiver operating characteristic (ROC) curve analysis.
Results
Patients with diabetes [adjusted odds ratio (aOR) 2.728, p=0.042], current cigarette use [aOR 4.085, p=0.023], PaO2/FiO2 (PF) ratio above 100 and less than or equal to 200 mmHg [aOR 3.108, p=0.016], and bilirubinemia over 1.2 mg/dL [aOR 2.767, p=0.040] were associated with higher 30-day mortality. Admission within previous 6 months [aOR 2.147, p=0.082], initial systolic blood pressure over 140 mmHg [aOR 2.432, p=0.061], and PF ratio below 100 mmHg [aOR 2.746, p=0.215] were also included in the modeling due to clinical significance. Area under the ROC curve of the scoring system was 0.770 (95% CI 0.679- 0.862, p<0.0001). Derivation cohort patients were divided into quartiles according to their score. Interquartile mortality differed (Q1: 7.69%, Q2: 17.74%, Q3: 54.35%, Q4: 100%) with statistical significance.
Conclusion
We report a quantitative scoring system predicting 30-day mortality of AE-IPF patients. After validation, this easy-to-use scoring system will serve as a useful aid for clinical decision making.