Skip to main content

Advertisement

Table 3 Simulation results for the estimation of attributable risk A(.) under proportional hazards, increasing baseline hazard (γ=4/3) with regression parameter β= ln(2) and probability of exposure q=0.5

From: Comparison of methods for estimating the attributable risk in the context of survival analysis

Estimation method    n=1, 000 n=10, 000
  Time A(t) Bias SEE SSD CP Bias SEE SSD CP
KM τ/4 0.299 0.000814 0.064311 0.064377 0.947 −0.000024 0.020388 0.020204 0.956
  τ/2 0.250 0.002020 0.043388 0.043169 0.952 0.000210 0.013761 0.013651 0.944
  3 τ/4 0.200 0.001174 0.037152 0.037027 0.955 −0.000469 0.011824 0.011798 0.960
  τ 0.153 0.007382 0.043968 0.054032 0.891 0.000554 0.018140 0.021081 0.939
WKM τ/4 0.299 0.000805 0.064427 0.064296 0.950 −0.000010 0.020380 0.020196 0.954
  τ/2 0.250 0.002055 0.043322 0.042973 0.949 0.000272 0.013722 0.013643 0.947
  3 τ/4 0.200 0.001193 0.036838 0.036463 0.962 −0.000410 0.011739 0.011741 0.958
  τ 0.153 0.005596 0.040652 0.048586 0.898 0.000055 0.017280 0.019095 0.935
COX τ/4 0.299 0.001207 0.041863 0.040891 0.960 −0.000209 0.013250 0.013076 0.962
  τ/2 0.250 0.001321 0.036377 0.035580 0.954 −0.000062 0.011499 0.011341 0.958
  3 τ/4 0.200 0.001300 0.030350 0.029672 0.956 −0.000121 0.009572 0.009502 0.965
  τ 0.153 0.002791 0.027165 0.028199 0.945 −0.000309 0.009206 0.010402 0.945
PCH τ/4 0.299 −0.000084 0.041594 0.040674 0.961 −0.001831 0.013151 0.013022 0.957
  τ/2 0.250 0.000876 0.036176 0.035464 0.956 −0.000759 0.011424 0.011313 0.958
  3 τ/4 0.200 0.001462 0.030163 0.029655 0.959 −0.000051 0.009509 0.009485 0.961
  τ 0.153 0.002572 0.025716 0.024704 0.961 0.000622 0.008058 0.007962 0.945
Simpler 0.333 0.001129 0.044983 0.044481 0.955 −0.000242 0.014226 0.014195 0.957
  1. KM nonparametric approach based on Kaplan-Meier estimation for S(t), WKM nonparametric approach based on weighted Kaplan-Meier estimation for S(t), COX semiparametric approach, PCH parametric approach using a piecewise constant hazards model, Simpler simpler approach based on proportion of exposed subjects, Bias sampling mean of the difference between \(\hat {A}(t)\) and A(t), SEE sampling mean of standard error estimate of A(t), SSD sampling standard deviation of \(\hat {A}(t)\), CP coverage probability of the 95% Wald confidence interval