Dataset Information

Prior Precision, Prior Accuracy, and the Estimation of Disease Prevalence Using Imperfect Diagnostic Tests.

ABSTRACT: Estimates of disease prevalence in any host population are complicated by uncertainty in the outcome of diagnostic tests on individuals. In the absence of gold standard diagnostics (tests that give neither false positives nor false negatives), Bayesian latent class inference can be applied to batteries of diagnostic tests, providing posterior estimates of the sensitivity and specificity of each test, alongside posterior estimates of disease prevalence. Here we explore the influence of precision and accuracy of prior information on the precision and accuracy of posterior estimates of these key parameters. Our simulations use three diagnostic tests, yielding eight possible diagnostic outcomes per individual. Seven degrees of freedom allow the estimation of seven parameters: sensitivity and specificity of each test, and disease prevalence. We show that prior precision begets posterior precision but only when priors are accurate. We also show that analyses without gold standard can use imprecise priors as long as they are initialised with accuracy. Imprecise priors risk the divergence of MCMC chains towards inaccurate posterior estimates, if inaccurate initial values are used. We note that inaccurate priors can yield inaccurate and imprecise inference. Bounded priors should certainly not be used unless their accuracy is well established. Inaccurate estimates of sensitivity or specificity can yield wildly inaccurate estimates of disease prevalence. Our analyses are motivated by studies of bovine tuberculosis in a wild badger population.

SUBMITTER: McDonald JL

PROVIDER: S-EPMC5958675 | biostudies-literature | 2018

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Prior Precision, Prior Accuracy, and the Estimation of Disease Prevalence Using Imperfect Diagnostic Tests.

McDonald Jenni L JL Hodgson Dave James DJ

Frontiers in veterinary science 20180511

Estimates of disease prevalence in any host population are complicated by uncertainty in the outcome of diagnostic tests on individuals. In the absence of gold standard diagnostics (tests that give neither false positives nor false negatives), Bayesian latent class inference can be applied to batteries of diagnostic tests, providing posterior estimates of the sensitivity and specificity of each test, alongside posterior estimates of disease prevalence. Here we explore the influence of precision ...[more]

PMID: 29868615

Similar Datasets

Project description:Background and Aims: Cystic fibrosis-related liver disease (CFLD) is one of the leading causes of morbidity and mortality in cystic fibrosis (CF). Several non-invasive diagnostic methods have been proposed as screening tools for CFLD. Our aim was to rank all available non-invasive modalities for diagnostic performance. Methods: A systematic search was performed in five medical databases to find studies which reported on any single or composite non-invasive diagnostic test (as an index test) compared to the Debray, the EuroCare or the Colombo criteria (as a reference standard). Ranking was carried out with a Bayesian diagnostic test accuracy network meta-analysis based on superiority indices, calculated for pooled sensitivity (Se) and specificity (Sp) with a 95% confidence interval (CI). The study was registered under CRD42020155846 in PROSPERO. Results: Fifteen studies with 15 index tests and a combination of them were included. The New criteria proposed by Koh et al. - which represent a composite diagnostic definition for CFLD including liver biochemistry, ultrasonography, transient elastography and fibrosis markers-had the best performance for detecting CFLD (Se:94%[CI:58-100], Sp:72%[CI:52-84]); while transient elastography (Se:65%[CI:56-74], Sp:88%[CI:84-91]) and a combination of it with a tissue inhibitor of metalloproteinase-4 measurement (Se:78%[CI:30-100], Sp:64%[CI:18-95%]) proved to be the second and third best options, respectively. In the imaging techniques subgroup, transient elastography (Se:66%[CI:57-72], Sp:88%[CI:85-91%]), acoustic radiation force impulse in the right lobe (Se:54%[CI:33-74], Sp:88%[CI:66-96]) and that in the left lobe (Se:55%[CI:23-81], Sp:82%[CI:50-95]) were ranked the highest. Comparing biochemical markers/fibrosis indices, the measurement of the Forns index (Se:72%[CI:25-99], Sp:63%[CI:16-94]), the aspartate aminotransferase-to-platelet ratio (Se:55%[CI:41-68], Sp:83%[CI:66-89]) and alkaline phosphatase (Se:63%[CI:18-93], Sp:64%[CI:19-95]) were ranked the highest. Conclusion: The New criteria show the best diagnostic performance. In clinical practice, transient elastography seems to be a simple, cheap and non-invasive tool, outperforming imaging, biochemical and fibrosis tests for detecting CFLD. Further studies are needed to validate our findings.

Project description:Ages of Florida manatees (Trichechus manatus latirostris) can be estimated by counting annual growth layer groups (GLGs) in the periotic dome portion of the tympanoperiotic complex of their earbones. The Florida Fish and Wildlife Conservation Commission manages an archive of more than 8,700 Florida manatee earbones collected from salvaged carcasses from 1989 to 2017. Our goal was to comprehensively evaluate techniques used to estimate age, given this large sample size and changes to processing protocols and earbone readers over time. We developed new standards for estimating ages from earbones, involving two independent readers to obtain measurements of within- and between-reader precision. To quantify accuracy, precision, and error, 111 earbones from manatees with approximately known ages (first known as calves: "KAC") and 69 earbones from manatees with minimum known ages ("MKA," based on photo-identification sighting histories) were processed, and their ages were estimated. There was greater precision within readers (coefficient of variation, CV: 2.4-8.5%) than between readers (CV: 13.1-13.3%). The median of age estimates fell within the true age range for 63.1% of KAC cases and was at least the sighting duration for 75.0% of MKA cases. Age estimates were generally unbiased, as indicated by an average raw error ± SD of -0.05 ± 3.05 years for the KAC group. The absolute error (i.e., absolute value of raw error) of the KAC data set averaged 1.75 ± 2.50 years. Accuracy decreased and error increased with increasing known age, especially for animals over 15 years old, whose ages were mostly underestimated due to increasing levels of resorption (the process of bone turnover that obscures GLGs). Understanding the degree of uncertainty in age estimates will help us assess the utility of age data in manatee population models. We emphasize the importance of standardizing and routinely reviewing age estimation and processing protocols to ensure that age data remain consistent and reliable.

Project description:BackgroundOverall accuracy measures of medical tests are often used with unclear interpretations.ObjectivesTo develop methods of calculating the overall accuracy of medical tests in the patient population.MethodsAlgebraic equations based on Bayes' theorem.ResultsA new approach is proposed for calculating overall accuracy in the patient population. Examples and applications using published data are presented.ConclusionsThe overall accuracy is the proportion of the correct test results. We introduce a clear distinction between the overall accuracy measures of medical tests that are aimed at the detection of a disease in a screening of populations for public health purposes in the general population and the overall accuracy measures of tests aimed at determining a diagnosis in individuals in a clinical setting. We show that the overall detection accuracy measure is obtained in a specific study that explores test accuracy among persons with known diagnoses and may be useful for public health screening tests. It is different from the overall diagnostic accuracy that could be calculated in the clinical setting for the evaluation of medical tests aimed at determining the individual patients' diagnoses. We show that the overall detection accuracy is constant and is not affected by the prevalence of the disease. In contrast, the overall diagnostic accuracy changes and is dependent on the prevalence. Moreover, it ranges according to the ratio between the sensitivity and specificity. Thus, when the sensitivity is greater than the specificity, the overall diagnostic accuracy increases with increasing prevalence, and vice versa, that is, when the sensitivity is lower than the specificity, the overall diagnostic accuracy decreases with increasing prevalence so that another test might be more useful for diagnostic procedures. Our paper suggests a new and more appropriate methodology for estimating the overall diagnostic accuracy of any medical test. This may be important for helping clinicians avoid errors.

Dataset Information

Prior Precision, Prior Accuracy, and the Estimation of Disease Prevalence Using Imperfect Diagnostic Tests.

Publications

Prior Precision, Prior Accuracy, and the Estimation of Disease Prevalence Using Imperfect Diagnostic Tests.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets