Dataset Information

Measuring rater bias in diagnostic tests with ordinal ratings.

ABSTRACT: Diagnostic tests are frequently reliant upon the interpretation of images by skilled raters. In many clinical settings, however, the variability observed between experts' ratings plays a detrimental role in the degree of confidence in these interpretations, leading to uncertainty in the diagnostic process. For example, in breast cancer testing, radiologists interpret mammographic images, while breast biopsy results are examined by pathologists. Each of these procedures involves elements of subjectivity. We propose here a flexible two-stage Bayesian latent variable model to investigate how the skills of individual raters impact the diagnostic accuracy of image-related testing in large-scale medical testing studies. A strength of the proposed model is that the true disease status of a patient within a reasonable time frame may or may not be known. In these studies, many raters each contribute classifications on a large sample of patients using a defined ordinal grading scale, leading to a complex correlation structure between ratings. Our modeling approach considers the different sources of variability contributed by experts and patients while accounting for correlations present between ratings and patients, in contrast to currently available methods. We propose a novel measure of a rater's ability (magnifier) that, in contrast to conventional measures of sensitivity and specificity, is robust to the underlying prevalence of disease in the population, providing an alternative measure of diagnostic accuracy across patient populations. Extensive simulation studies demonstrate lower bias in estimation of parameters and measures of accuracy, and illustrate outperformance of the proposed model when compared with existing models. Receiver operator characteristic curves are derived to assess the diagnostic accuracy of individual experts and their overall performance. Our proposed modeling approach is applied to a large breast imaging study for known disease status and a uterine cancer dataset for unknown disease status.

SUBMITTER: Kim C

PROVIDER: S-EPMC8277718 | biostudies-literature |

REPOSITORIES: biostudies-literature

ACCESS DATA

Similar Datasets

Project description:BACKGROUND: Available evidence suggest that perceptions or ratings of the neighborhood, e.g. as being green, walkable or noisy, are important for effects on health and wellbeing, also after controlling for objective measures of identical or similar features. When evaluating effects of the perceived environment, it is important that measurement properties and the reliability of the environmental ratings are evaluated before decisions about how these ratings should be handled in the statistical analyses are made. In this paper we broaden the usage of two association measures, the well-known kappa statistic and the novel colocation quotient (CLQ), to studies of inter-rater reliability and of associations between different categorical ratings in spatial contexts. METHODS: We conducted reliability analysis of a survey instrument for assessing perceived greenness at geographical point locations, here the close outdoor environment within 5-10 minutes walking distance from home. Data were obtained from a public health survey conducted in 2008 in Scania, southern Sweden (n =27 967 participants). RESULTS: The results demonstrate the usefulness of kappa and CLQ as tools for assessing reliability and measurement properties of environmental rating scales when used at geographical point locations. We further show that the two measures are interchangeable, i.e. kappa can be accurately approximated from CLQ and vice versa, but can be used for somewhat different purposes in reliability analyses. Inter-rater reliability between the nearest neighbors was demonstrated for all five items of the evaluated instrument for assessing perceived greenness, albeit with clear differences across the items. CONCLUSION: Reliability analysis employing kappa and CLQ can be used as a basis for informed decisions about, for instance, how dichotomizations of the ratings should be defined and how missing or indefinite ratings should be handled. Such reliability analyses can thus serve as guidance for subsequent epidemiological studies of associations between environmental ratings, health and wellbeing.

Dataset Information

Measuring rater bias in diagnostic tests with ordinal ratings.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets