Capture-recapture method for estimating misclassification errors: application to the measurement of vaccine efficacy in randomized controlled trials

Background The measure of efficacy is optimally performed by randomized con trolled trials. However, low specificity of the judgement criteria is known to bias toward lower estimation, while low sensitivity increases the requi red sample size. A common technique for ensuring good specificity without a drop in sensitivity is to use several diagnostic tests in parallel, with e ach of them being specific. This approach is similar to the more general si tuation of case-counting from multiple data sources, and this paper explore s the application of the capture-recapture method for the analysis of the e stimates of efficacy. Method An illustration of this application is derived from a study on the e fficacy of pertussis vaccines where the outcome was based on greater than o r equal to 21 days of cough confirmed by at least one of three criteria per formed independently for each subject: bacteriology, serology, or epidemiol ogical link. Log-linear methods were applied to these data considered as th ree sources of information. Results The best model considered the three simple effects and an interacti on term between bacteriology and epidemiological linkage. Among the 801 chi ldren experiencing greater than or equal to 21 days of cough, it was estima ted that 93 cases were missed, leading to a corrected total of 413 confirme d cases. The relative vaccine efficacy estimated from the same model was 1. 50 (95% confidence interval: 1.24-1.82), similar to the crude estimate of 1 .59 and confirming better protection afforded by one of the two vaccines. Conclusion This method allows supporting analysis to interpret primary esti mates of vaccine efficacy.