Diagnostic testing is the next big area in evidence-based medicine. In clinical medicine, using diagnostic testing efficiently and effectively is possibly more important than therapeutics. It has enormous implications for health care at both the macro and micro levels. This user friendly book on how to evaluate diagnostic testing is written by statistical experts involved in the Cochrane Diagnostic group and provides a step by step guide to the statistical methodology. It is suitable for health professionals undertaking clinical research at all levels.
Marianne Empson, Department of Clinical Immunology, Auckland Main Hospital, Auckland is a practising clinical immunologist and immunopathologist with an interest in clinical epidemiology and diagnostic test interpretation etc. She regularly teaches and presents material on the topics in the proposed book to a pathology and clinical immunology audience in Australasia. Jonathan Craigholds a personal chair in Clinical Epidemiology at the University of Sydney, and is a practising paediatric nephrologist at the Children s Hospital at Westmead, Sydney, Australia Petra Macaskill is an Associate Professor in Biostatistics at the University of Sydney. Her research is primarily in the area of diagnostic test evaluation and the meta-analysis of diagnostic test performance
1) Introduction. An outline of purpose of the book, target audience and why this is a necessary book for anyone involved in the performance and interpretation of diagnostic tests (pathology and radiology) and/or reading / analysing published studies or designing studies using diagnostic tests. It will include important definitions such as reference standard, index test etc. Part 1: Characteristics and interpretation of test results. 2) Estimates of test accuracy. a. Sensitivity and specificity. b. Receiver Operating Characteristic (ROC) Curve. c. Diagnostic odds ratio. d. Likelihood ratio. 3) Precision of test estimates. 4) Interpretation of test results. a. Reference range. b. Bayes theorem / Post test probability. Part 2: Comparing tests. 5) Test accuracy. 6) Association. a. Categorical data. b. Continuous data. 7) Agreement. a. Categorical data. b. Continuous data. Part 3: Reading the literature and designing studies. 8) How to read a paper on a diagnostic test. 9) Bias and Methodological Standards. a. Selection bias. b. Spectrum bias. c. Verification bias. d. Review bias. e. Gold standard. f. Reproducibility. g. Indeterminate results. h. Misclassification. 10) Sample size. 11) STARD. Part 4: Systematic reviews. 12) How to read a diagnostic test systematic review. 13) Meta-analysis of diagnostic tests