Measurement error in epidemiology: The design of validation studies - II: Bivariate situation

The bias in relative risk estimates caused by errors in measurement of the relevant exposure is being increasingly recognized in epidemiology. Estimat ion of the necessary correction factor to remove this bias for univariate e xposure has been considered in an earlier paper. We consider here the multi variate situation in which non-differential errors in measurement can lead to incorrect identification of the variable most closely associated with di sease. Estimation of the necessary correction factor when the true exposure is unobservable necessarily requires assumptions. We explore the robustnes s of the estimation to departures from a range of assumptions. The value of good biomarkers is demonstrated. We present a bivariate example in which f ailure to take account of measurement error leads to the incorrect exposure being identified as the important determinant of disease risk. Copyright ( C) 1999 John Wiley & Sons, Ltd.