Verification and classification bias interactions in diagnostic test accuracy studies for fine‐needle aspiration biopsy

Abstract

BACKGROUND: Reliable estimates of accuracy are important for any diagnostic test. Diagnostic accuracy studies are subject to unique sources of bias. Verification bias and classification bias are 2 sources of bias that commonly occur in diagnostic accuracy studies. Statistical methods are available to estimate the impact of these sources of bias when they occur alone. The impact of interactions when these types of bias occur together has not been investigated. METHODS: We developed mathematical relationships to show the combined effect of verification bias and classification bias. A wide range of case scenarios were generated to assess the impact of bias components and interactions on total bias. RESULTS: Interactions between verification bias and classification bias caused overestimation of sensitivity and underestimation of specificity. Interactions had more effect on sensitivity than specificity. Sensitivity was overestimated by at least 7% in approximately 6% of the tested scenarios. Specificity was underestimated by at least 7% in less than 0.1% of the scenarios. CONCLUSIONS:Interactions between verification bias and classification bias create distortions in accuracy estimates that are greater than would be predicted from each source of bias acting independently. Cancer (Cancer Cytopathol) 2015;123:193

The impact of imperfect screening tools on measuring the prevalence of epilepsy and headaches in Burkina Faso

“…Such discrepancies may be due to spectrum bias where populations in clinical settings have more severe disease and acknowledge their symptoms more, thereby increasing the test sensitivity. The observed estimates from clinical settings may also result in part from verification bias, which typically leads to an overestimation of the sensitivity and underestimation of the specificity [22]. Our sensitivity estimates for epilepsy were similar to the two other community-based studies with sensitivities of 72.4% (95% CI: 52.8–87.3) and 79.3% [20,21].…”

“…The adjusted prevalence estimate was obtained by running a Bayesian latent-class model [22]. In this model, the probabilities that participants were selected for the confirmation test at step two were specified by a set of conditional distributions.…”

“…To our knowledge, only two studies have reported validity estimates of epilepsy screening questionnaires in the general population; however, neither addresses potential verification bias in their validity estimates [20,21]. Verification bias occurs when the participants from step one have different probabilities of being selected for step two [22]. For example, individuals screened positive at step one have a much higher chance of being confirmed by a physician or neurologist, compared to those screened negative, who are rarely, if at all, examined at step two.…”

Use of HPV testing in cervical cancer screening services in Mexico, 2008-2018: a nationwide database study

“…37 The negative test results are differentiated by follow-up in the NCSP. 38,39 It is also possible that the rate of completed colposcopy referrals may differ between screening strategies. We will use the method proposed by Begg and Greenes to correct the partial verification bias.…”

Correcting for partial verification bias in diagnostic accuracy studies: A tutorial using R

“…Biased estimates are misleading, which may result in the premature implementation of the new tests and lead to wrong decision making by the clinicians 4,15 . However, the assessment of index tests in diagnostic accuracy studies often suffer from bias because the verification of the target outcomes by gold standard tests can be clinically infeasible due to cost, ethical, and clinical considerations, as well as time‐consuming and invasive procedures 4,5,9,16 …”