Table 3.

Time-dependent Cox regressions analysisa of all-cause mortalityb

Baselinec VariablesRR95% CIP
Cumulative mean predialysis serum β2M level over time during follow-up (per 10-mg/L increase)1.111.05 to 1.190.001
Age (per 10-yr increase)1.451.35 to 1.55<0.0001
Gender (female)0.850.71 to 1.020.075
Race (black)0.770.64 to 0.910.003
Diabetes1.361.16 to 1.60<0.0001
Duration of dialysis (per year increase)1.031.02 to 1.05<0.0001
ICED score (per unit increase)1.341.22 to 1.47<0.0001
Serum albumin (per g/dl increase)0.430.34 to 0.54<0.0001
Residual kidney urea clearanced (per ml/min)0.960.80 to 1.150.681
Baseline high-flux dialysis1.040.82 to 1.320.733
Ultrafiltration volume/weight (per L/70 kg)1.161.08 to 1.24<0.0001
Modeled body urea distribution volume (per L)0.990.98 to 1.010.236
Randomization to high urea Kt/V arm0.950.83 to 1.100.501
  • a Stratified by clinical center.

  • b Only patients who had at least one β2M clearance measurement during follow-up were included; n = 1813. RR, relative risk; CI, confidence interval.

  • c All variables were baseline except for the cumulative mean predialysis serum β2M level over time (which was a time-dependent variable calculated as the mean over all preceding and the current sessions; presented in the first row of the table) and randomization to high urea Kt/V arm during follow-up (presented in the last row of the table).

  • d Normalized to 35 L of urea distribution volume.\