Table 2.

Effect on renal outcomes of adjusting for baseline proteinuria as a continuous covariatea

Original AnalysisAdjusted Analysis
RR (95% CI)χ2PRR (95% CI)χ2P
Combined (n = 1513)
    primary16.1 (2.5 to 27.8)5.20.02222.2 (9.5 to 33.1)10.60.001
    ESRD or death19.9 (5.3 to 32.3)6.80.00925.7 (12.1 to 37.3)12.00.001
    ESRD alone28.6 (11.5 to 42.4)9.40.00236.7 (21.3 to 49.0)17.0<0.001
Baseline proteinuria <2000 mg/g (n = 1012)
    primary9.2 (−13.5 to 27.3)0.70.3977.9 (−15.0 to 26.3)0.50.466
    ESRD or death21.1 (−1.9 to 38.8)3.30.06920.4 (−2.7 to 38.3)3.10.079
    ESRD alone44.5 (17.7 to 62.6)8.60.00343.1 (15.6 to 61.6)7.80.005
Baseline proteinuria ≥2000 mg/g (n = 501)
    primary20.8 (3.0 to 35.4)5.10.02529.2 (13.0 to 42.4)10.80.001
    ESRD or death18.1 (−2.3 to 34.4)3.10.07925.1 (6.2 to 40.2)6.40.012
    ESRD alone20.0 (−3.7 to 38.2)2.80.09128.7 (7.4 to 45.1)6.40.011
  • a Original analysis: A Cox regression model with treatment and region as covariates, and baseline proteinuria levels as strata (<2000 and ≥2000 mg/g), which was prespecified in the data analysis plan; adjusted analysis: Cox regression model with treatment, region, and baseline continuous proteinuria as covariates. RR, risk reduction = 100 × (1 − hazard ratio) between losartan and placebo.