The following table is taken from an article that investigated the effect of preoperative inpatient hospitalization on the risk of perioperative surgical site infection following elective vascular procedures. The investigators were interested in investigating the association between the response variable, whether or not a patient had perioperative surgical site infection (1 if they did, 0 if they did not), and the list of explanatory variables in the table below. From Table V, we see that their final model was a multiple logistic model with 12 predictors.

Table V. Multivariable model for perioperative surgical site infection. Variables listed in order of multivariable model entry using forward selection criterion of p g 0.05 Variable Odds ratio 95% condence limits Rest pain or gangrene 2.67 2.31-3.09 History of revascularizationtamputation for peripheral 2.17 1.91 -2.47 vascular disease Weight (kg)a 1.38 1.31-1.46 Female gender 1.21 1.06-1.39 Current smoker within 1 year 1.49 1.32-1.67 History of transient ischemic attack 0.17 0.07-0.38 Dependent functional status 1.34 1.09-1.64 Steroid use for chronic condition 1.54 1.17-2.03 Transfusion >4 U in 72 hours before operation 4.65 l.48-14.06 Diabetes requiring oral agent or insulin 1.21 1.07-1.38 Preoperative inpatient hospitalization 1.21 1.06-1.39 Pstory of severe chronic obstructive pulmonary disease 1.18 1.01-1.38