Mariska Tuut

168 Chapter 4 experts who are familiar with the relevant technical details related to their implementation. - For systematic reviews of test accuracy, author teams should include members with expertise in: literature searching, completing systematic reviews, test research methods, and statistics. The information specialists and statistical experts must be aware of the particular methodology for searching and data analysis for systematic reviews of test accuracy. - Tests may directly harm patients if they are invasive. It is important to be aware of what these harms may be and how frequently they are encountered. - It is also important to be wary of the harms associated with false positive and false negative diagnoses. - It is important to assess failure rates and non-diagnostic findings. - Evaluation of the acceptability of a test is important to assess whether patients are willing to undergo a procedure. - Before starting a review, it is essential to understand what kind of primary study would ideally fit the review question. - Before undertaking a test accuracy meta-analysis, it is necessary to understand the distinct types of data, as well as the presentation and meaning of statistical summaries of test accuracy reported in the primary studies. - To inform decision-making, researchers should report the results or outcome for all participants undergoing testing. For many tests, this means that researchers should also report the number of persons tested for whom a conclusive result – a clear positive or a clear negative – could not be obtained, and the reasons why. - The clinical utility of a test will always depend on both sensitivity and specificity and will also be influenced by the proportion with the target condition among those tested. It is therefore crucially important to always report sensitivity and specificity in pairs. - Accuracy measures:  Sensitivity  Specificity  True positives, false positives, true negatives, false negatives  Positive predictive value  Negative predictive value  Positive likelihood ratio  Negative likelihood ratio  Youden’s index  Overall accuracy

RkJQdWJsaXNoZXIy MTk4NDMw