Bias correction methods for test-negative designs in the presence of misclassification

Abstract

The test-negative design (TND) has become a standard approach for vaccine effectiveness (VE) studies. However, previous studies suggested that it may be more vulnerable than other designs to misclassification of disease outcome caused by imperfect diagnostic tests. This could be a particular limitation in VE studies where simple tests (e.g. rapid influenza diagnostic tests) are used for logistical convenience. To address this issue, we derived a mathematical representation of the TND with imperfect tests, then developed a bias correction framework for possible misclassification. TND studies usually include multiple covariates other than vaccine history to adjust for potential confounders; our methods can also address multivariate analyses and be easily coupled with existing estimation tools. We validated the performance of these methods using simulations of common scenarios for vaccine efficacy and were able to obtain unbiased estimates in a variety of parameter settings. 10 A. Endo et al.

Test negative designs with uncertainty, sensitivity, and specificity

“…2C,D). The uncorrected estimates systematically underestimated vaccine effectiveness aligning with analytical predictions [4].…”

“…2). Thus, while several preceding studies have shown that estimates can be made unbiased by adjusting for sensitivity and specificity [5,3,4], here we further showed that test performance affects vaccine effectiveness uncertainty even after bias adjustment, leading us to conclude that there is independent value in high-performance diagnostic testing, even if signals may be statistically debiased a posteriori.…”

“…where n u = n u+ + n u− is the total number of unvaccinated individuals, and n v = n v+ + n v− is the total number of vaccinated individuals. Previous work by Endo et al showed that Equations ( 1) and ( 4) can be subtracted to quantify the substantial estimation bias that results if one fails to correct for the diagnostic test [4].…”

“…Previous work has shown that vaccine effectiveness estimates are more affected by poor sensitivity in some scenarios and poor specificity in others [3,4], thereby informing test and test cutoff selection.…”

“…Unfortunately, an imperfect diagnostic test may result in false positives (if specificity is imperfect) and false negatives (if sensitivity is imperfect). In turn, these false positives or false negatives, alternatively called outcome misclassifications, will lead to biased estimates of vaccine effectiveness unless statistical corrections are made [3,4,5,6,7]. Thus, appropriate statistical consideration of imperfect diagnostic tests, as well as potential confounders, is important for vaccine effective estimation and public health decision making.…”