Table 2.

Standardized incidence ratios for malignancy after cyclophosphamide exposure by potentially confounding variables

Risk FactorIncidence Ratio95% CI
Unadjusted4.61.5 to 18.8
Univariate-adjusted
 Age3.31.0 to 10.6
 Men3.31.1 to 10.0
 Smoking5.81.6 to 20.8
 Prior therapy4.21.3 to 13.4
 Family history of malignancy7.11.6 to 32.0
 CKD stage5.01.5 to 18.8
 Nephrotic syndrome4.31.3 to 14.1
Age- and sex-adjusted3.21.0 to 9.5
  • Age was categorized as <45, 45–54, 55–64, 65–74, and ≥75 years. The unadjusted row shows the unadjusted incidence ratio. The univariate-adjusted rows are incidence ratios for cyclophosphamide exposure adjusted for the individual confounder. Ever smokers were compared with patients who never smoked. Prior therapy was defined as immunosuppressive drug therapy other than prednisone before cyclophosphamide therapy was initiated. A multiple adjusted incidence ratio of 3.1 (95% confidence interval [95% CI], 0.9 to 9.9) was obtained with a Poisson regression after multiple imputation of missing values for smoking and family history of malignancy. The most parsimonious model was adjusted only for sex and age (the latter as a continuous variable). The model showed a moderate fit (R2=0.15).