2.6. Main candidate predictor and adjustment variables

Candidate predictors (main candidate predictor and adjustment variables) for the primary outcome should be clinically and biologically plausible, and their relationships with all-cause death should be demonstrated in previous studies [^{[15], [16], [17], [18], [19], [20]}]. We considered all candidate predictors that we included in the model under these principles and finally draw a directed acyclic graph to inform regression models (Supplementary Fig. 1). The main candidate predictor, cTn measurements, was included in the model in three ways:

we divided the measured cTn value by the ULN of each cTn assay and produced the result as “x ULN” (model-1)

we categorized cTn measurements ordinally as <1/2 x of ULN, >1/2 x of ULN, >1–2 x ULN, >2–5 x ULN, >5–10 x ULN, >10–50 x ULN and >50 x ULN (model-2)

we dichotomized each measured cTn value according to ULN as “negative or positive” (model-3).

Since the cTn measurements calculated as x ULN were significantly right-skewed, they were included in modeling as log (x ULN). Adjustment variables were determined as age, sex, neutrophil-lymphocyte ratio (NLR), D-Dimer, lactate dehydrogenase (LDH), C-reactive protein (CRP), hemoglobin, platelet count, coronary artery disease, heart failure (HF), chronic obstructive pulmonary disease (COPD), cerebrovascular disease, hypertension, diabetes mellitus, and chronic kidney disease (CKD).