Journal of Medical Biochemistry 2007, vol. 26, iss. 4, pp. 309-313

Diagnostic performance of a laboratory test is one of the key elements in decision making on diagnosis, screening, monitoring, risk assessment and prognosis of diseases. Sensitivity, specificity, likelihood ratios, diagnostic odds ratios and receiver operating characteristic curves are the measures of diagnostic accuracy of a test. The pretest probability of a disease or a target condition can be enhanced by the use of these measures, and hence the decision is made with the post test probability. These measures are also used for analysis and critical appraisal of literature for finding the best evidence in the five-step model of evidence-based medicine approach, as well as for integrating the research results into clinical usage. In this context, the specialists in laboratory medicine should assess the diagnostic performance of a laboratory test as well as its analytical performance in order to take part in the management of health care services and health care resources. The aim of this review is to summarize the simple Statistics in diagnostic tests.

Clinical Chemistry November 2012 vol. 58 no. 11 1534-1545

Systematic reviews of diagnostic accuracy studies allow calculation of pooled estimates of accuracy with increased precision and examination of differences in accuracy between tests or subgroups of studies. Recently, several advances have been made in the methods used in performing systematic reviews of diagnostic test accuracy studies, most notably in how to assess the methodological quality of primary diagnostic test accuracy studies by use of QUADAS-2 (Quality Assessment of Diagnostic Accuracy Studies 2) instrument and how to develop sound statistical models for metaanalysis of the paired measures of test accuracy (bivariate metaregression model of sensitivity and specificity). This article provides an overview of the different steps within a diagnostic systematic review and highlights these advances, illustrated with empirical data. The potential benefits of some recent developments in the areas of network metaanalysis and individual patient data metaanalysis for diagnostic tests are also discussed.

Clinical Chemistry December 2012 vol. 58 no. 12 1636-1643

Like any other medical technology or intervention, diagnostic tests should be thoroughly evaluated before their introduction into daily practice. Increasingly, decision makers, physicians, and other users of diagnostic tests request more than simple measures of a test’s analytical or technical performance and diagnostic accuracy; they would also like to see testing lead to health benefits. In this last article of our series, we introduce the notion of clinical utility, which expresses—preferably in a quantitative form—to what extent diagnostic testing improves health outcomes relative to the current best alternative, which could be some other form of testing or no testing at all. In most cases, diagnostic tests improve patient outcomes by providing information that can be used to identify patients who will benefit from helpful downstream management actions, such as effective treatment in individuals with positive test results and no treatment for those with negative results. We describe how comparative randomized clinical trials can be used to estimate clinical utility. We contrast the definition of clinical utility with that of the personal utility of tests and markers. We show how diagnostic accuracy can be linked to clinical utility through an appropriate definition of the target condition in diagnostic-accuracy studies.

J Gen Intern Med. 2012 Jun;27 Suppl 1:47-55.

Abstract

INTRODUCTION: Grading the strength of a body of diagnostic test evidence involves challenges over and above those related to grading the evidence from health care intervention studies. This chapter identifies challenges and outlines principles for grading the body of evidence related to diagnostic test performance. CHALLENGES: Diagnostic test evidence is challenging to grade because standard tools for grading evidence were designed for questions about treatment rather than diagnostic testing; and the clinical usefulness of a diagnostic test depends on multiple links in a chain of evidence connecting the performance of a test to changes in clinical outcomes. PRINCIPLES: Reviewers grading the strength of a body of evidence on diagnostic tests should consider the principle domains of risk of bias, directness, consistency, and precision, as well as publication bias, dose response association, plausible unmeasured confounders that would decrease an effect, and strength of association, similar to what is done to grade evidence on treatment interventions. Given that most evidence regarding the clinical value of diagnostic tests is indirect, an analytic framework must be developed to clarify the key questions, and strength of evidence for each link in that framework should be graded separately. However if reviewers choose to combine domains into a single grade of evidence, they should explain their rationale for a particular summary grade and the relevant domains that were weighed in assigning the summary grade.