Fig. 3
Forest plots showing the results of Cox proportional-hazards regression and subdistribution hazard regression analyses of candidate predictors for a nintedanib-related hepatic injury, b on-treatment acute exacerbation of IPF, and c on-treatment mortality. In addition to the candidate predictors shown, all the multi-variable regression models were also adjusted for gender-age-physiology (GAP) stage, Charlson comorbidity index, and treatment duration. Abbreviation and Notes AE-IPF, acute exacerbation of idiopathic pulmonary fibrosis; BMI, body mass index; BSA, body surface area; DLCO, diffusion capacity for carbon monoxide (in % predicted); HR, hazard ratio; KL-6, Krebs von den Lungen-6; PH, echocardiographic pulmonary hypertension; SPA, surfactant protein A; SPO2, pulse oximetry (while breathing ambient air). *This cut-off value was proposed by Ikeda et al. [26]. **This cut-off value was proposed by Ikeda et al. [27]