83 Diagnostic utility of molecular and imaging biomarkers 2 Domain 3: Conduct and interpretation of the reference standard 1. Is the reference standard likely to correctly classify thyroid malignancy or a benign thyroid nodule? If the reference standard is histopathology, answer signalling question with ‘yes’. If the reference standard is repeat FNAC, core needle biopsy or a standardized, well-described follow-up ≥12 months, answer with ‘unclear’. If the reference standard is none of the above, answer with ‘no’. 2. Were the results of the reference standard interpreted without knowledge of the results of the index test? If this type of blinding was reported in the article, answer signalling question with ‘yes’. If it is not reported, answer with ‘unclear’. If it is reported that this type of blinding is not performed, answer with ‘no’. Domain 4: Patient flow and timing 1. Did all patients receive an appropriate reference standard? If histopathology is applied as the reference standard in the entire patient population, answer signalling question with ‘yes’. If: 1. histopathology is applied as the reference standard in more than 50% of the patient population, and 2. repeat FNAC, core needle biopsy or a standardized, well-described follow-up of at least 12 months is the reference standard for the majority of the remaining study population, and 3. did inevitable integration into daily practice already exist for the investigated determinant at the time of investigation (i.e. to subject all patients to diagnostic surgery despite the benign outcome of an readily established testing method, would be unethical), answer question with ‘unclear’. If neither of the above applies, answer question with ‘no’. 2. Did all patients receive the same reference standard? With regard to histopathology, performance of hemithyroidectomy or total thyroidectomy was considered ‘the same reference standard’. 3. Were all patients included in the analysis? Were any patients excluded from the analysis because the result of the index test was nondiagnostic, missing on otherwise not available? Specifically for mutation analysis: were any patients excluded from the analysis because the FNA sample contained to little nucleic acids to perform the index test? Were any patients lost during follow-up, before the reference standard could be conducted? If loss to follow-up of study participants was below 10%, answer signalling question with ‘yes’. If loss to follow-up of study participants was between 10-20%, answer with ‘unclear’. If loss to follow-up of study participants was above 20%, answer with ‘no’. *Modified from Whiting et al.[346]
RkJQdWJsaXNoZXIy MTk4NDMw