estimated cumulative incidence difference is not significantly elevated, and these data are consistent with a wide range of possible values of the cumulative incidence difference with 95% confidence. 0.00204 = 490.3 1 = 137.42 + 183.97 + 96.89 + 72.02 1 = D) I (C ar V ˆ ˆ ) 0.174 0.0035, - ( = 0.00204 1.96 0.085 Harvard TH Chan School of Public Health EPI202 – Epidemiologic Methods II Fall 2018 26 Evans County Study Crude Vs. Stratified CID Estimates Recall that the confidence interval for the crude CID, (5.3%, 20.9%), was indicative of a significant excess risk among subjects with elevated CAT levels. Note that although the p ‐ value from the hypothesis test indicated that there was a statistically significant association and the 95% CI for the CIR excludedthe null, the 95% CI for the cumulative incidence difference included the null value of 0. This can occur when the p ‐ value is close to the threshold cut ‐ point. In this instance, there is a significant association, but the estimation procedure for the difference parameter is less efficient than for the ratio parameter.
Harvard TH Chan School of Public Health EPI202 – Epidemiologic Methods II Fall 2018 27 Stratified Count Data Note that all formulas and strategies discussed in this lecture apply without modification for testing the null hypothesis and for the analysis of difference and ratio measures based upon prevalence data in cross ‐ sectional studies. Harvard TH Chan School of Public Health EPI202 – Epidemiologic Methods II Fall 2018 28 Key Concepts Notation Hypothesis tests Point and interval estimates for the cumulative incidence ratio Point and interval estimates for the cumulative incidence difference
- Summer '14