Skip to main content

Advertisement

Table 2 Observed relative bias for the simple product-based estimator (ESRS) and the revised product-based estimator (ESRR)

From: Evaluation of exposure-specific risks from two independent samples: A simulation study

Low exposure probability (.05)/Low disease probability in unexposed (.02)
  N1 = 1,000, N2 = 1,000 N1 = 5,000, N2 = 5,000
RR/ESR ESRS ESRR ESRS ESRR
1.0/.02 9.5% 3.9% -1.4% -2.3%
2.0/.04 7.8% -2.6% 4.5% -1.4%
3.0/.06 18.0% 1.8% 12.2% 0.8%
4.0/.08 21.1% 0.0% 17.6% 1.1%
5.0/.10 31.4% 3.4% 22.6% 1.0%
Low exposure probability (.05)/Moderate disease probability in unexposed (.09)
  N1 = 1,000, N2 = 1,000 N1 = 5,000, N2 = 5,000
RR/ESR ESRS ESRR ESRS ESRR
1.0/.09 0.1% -0.9% 0.9% 0.6%
2.0/.18 6.1% 0.1% 5.5% 0.3%
3.0/.27 9.2% -1.4% 10.4% 0.2%
4.0/.36 16.6% 0.4% 15.9% 0.5%
5.0/.45 22.0% 0.7% 21.2% 0.8%
High exposure probability (.20)/Low disease probability in unexposed (.02)
  N1 = 1,000, N2 = 1,000 N1 = 5,000, N2 = 5,000
RR/ESR ESRS ESRR ESRS ESRR
1.0/.02 8.7% 1.1% -0.2% -1.3%
2.0/.04 26.3% -1.4% 21.4% -0.1%
3.0/.06 45.0% -1.8% 41.7% 0.1%
4.0/.08 73.3% 1.2% 61.6% 0.1%
5.0/.10 93.9% 0.3% 82.5% 0.0%
High exposure probability (.20)/Moderate disease probability in unexposed (.09)
  N1 = 1,000, N2 = 1,000 N1 = 5,000, N2 = 5,000
RR/ESR ESRS ESRR ESRS ESRR
1.0/.09 -0.8% -0.4% -1.1% -1.0%
2.0/.18 22.6% 0.7% 20.4% 0.2%
3.0/.27 40.9% -0.2% 40.3% 0.0%
4.0/.36 63.6% 0.7% 60.6% 0.1%
5.0/.45 82.6% 0.2% 81.0% 0.2%
  1. N1 is the sample the overall risk is derived from
  2. N2 is the sample the relative risk is derived from
  3. Relative Risk/Exposure-Specific Risk (RR/ESR) values are the hypothesized values