Figure 3.

Published results on the relative risk (RR) of false positive recall in relation to use of hormone replacement therapy (HRT). Test for heterogeneity for current use of HRT versus never, χ²₃ = 13.9, p = 0.003. Test for heterogeneity for past use of HRT versus never χ²₃ = 6.8, p = 0.08. Relative risks are shown as black squares with areas inversely proportional to the variance of the log relative risk, indicating the amount of statistical information available for that particular estimate. The corresponding confidence intervals (CI) are shown as horizontal lines. An arrow is used where the upper confidence limit exceeds the scale of the figure. The diamonds represent the summary relative risks and 95% CIs of false positive recall in current and past versus never-uses of HRT.