Estimation of the intervention effect in a non-randomized study with pre- and post-mismeasured binary responses

Measurement error and outcomes defined by exceeding a threshold: biased findings in comparative effectiveness trials

In medical studies, the long-term level of a risk factor exceeding a threshold is often an outcome of interest. In practice, such a risk factor may not be directly measurable. Instead, outcome variables are based on a single or multiple biomedical measurements that have substantial variability. This variability is due to measurement error in a strict sense, true day-to-day variability, or a combination of the two. Estimates of prevalence based on such outcomes are biased; some individuals with long-term levels below the threshold will be diagnosed, and some with long-term levels above the threshold will not be diagnosed. From a public health point of view, this is a relatively minor concern; it is much less important than the fact that many individuals are not tested at all. However, in comparative effectiveness research studies, such as clinical trials evaluating a new treatment as compared with a placebo or a gold standard treatment, the combination of a noisy measurement and a threshold can distort the studies' conclusions in important ways. Using simulations and theoretical formulas, we systematically describe the bias of prevalence difference and prevalence ratio when comparing arms and its effect on trial conclusions.

Bayesian Estimation of Intervention Effect with Pre- and Post-Misclassified Binomial Data

We consider studies in which an enrolled subject tests positive on a fallible test. After an intervention, disease status is re-diagnosed with the same fallible instrument. Potential misclassification in the diagnostic test causes regression to the mean that biases inferences about the true intervention effect. The existing likelihood approach suffers in situations where either sensitivity or specificity is near 1. In such cases, common in many diagnostic tests, confidence interval coverage can often be below nominal for the likelihood approach. Another potential drawback of the maximum likelihood estimator (MLE) method is that it requires validation data to eliminate identification problems. We propose a Bayesian approach that offers improved performance in general, but substantially better performance than the MLE method in the realistic case of a highly accurate diagnostic test. We obtain this superior performance using no more information than that employed in the likelihood method. Our approach is also more flexible, doing without validation data if necessary, but accommodating multiple sources of information, if available, thereby systematically eliminating identification problems. We show via a simulation study that our Bayesian approach outperforms the MLE method, especially when the diagnostic test has high sensitivity, specificity, or both. We also consider a real data example for which the diagnostic test specificity is close to 1 (false positive probability close to 0).