Abbreviation for Standards for Reporting of Diagnostic Accuracy, a set of recommendations on the design and presentation of diagnostic markers for disease.
In essence, any study involving diagnostic tests should aim to have:
- Internal validity (avoidance of bias)
- External validity (generalisability & applicability)
STARD also suggests that published studies should make explicit 25 items relating to several stages in the research (e.g. inclusion/exclusion criteria, recruitment, sampling, data collection, detail of the test methods, reference standards, etc.). The full recommendation is available free.