Table 2.

Survival models were generated using Lasso regression in 269 patients with malignant pleural mesothelioma

Predictor variables included in final model	OS	OS <6 months	OS <12 months
Age	0.086070	0.146336	0.176899
White cell count	0.245527	0.436034	0.182477
Albumin	−0.198633	−0.264057	−0.273290
Epithelioid subtype		−0.311515	−0.191842
C-reactive protein			0.110628
Platelet count		0.000774

Cells report coefficients associated with each predictor; these are weighting factors relative to the units of the variable after scaling. Positive coefficients describe a positive association between the predictor variable and mortality risk; negative coefficients describe the opposite. The sum of the weighted coefficients produces an estimate for the outcome of interest.

Lasso, least absolute shrinkage and selection operator; OS, overall survival.