Sample size determination in reference-controlled diagnostic trials

Purpose: A tutorial illustration of a flexible approach to determine the sample size in reference-controlled diagnostic trials. Materials and Methods: Assuming the usual setting of a new diagnostic method to be compared with a reference method, the emphasis is on the sensitivity of the new method in comparison with the reference method, using a binary outcome (positive versus negative) for both methods. Based on the confidence interval of the sensitivity, a simple but flexible procedure for determining the sample size is described, which incorporates clinically interpretable information. The procedure is illustrated by the fictious planning of a trial to assess the diagnostic value of MRI versus arthroscopy as a reference, in the detection of meniscal ruptures. Results: The principal investigator merely has to propose the range for the sensitivity in which the new method is considered equal to the reference method. Furthermore, it must be decided in advance how accurate the study outcome should determine the sensitivity of the new method, i.e., how wide its maximum confidence interval may become. The minimum sample size necessary for the trial can be directly derived from this outlined strategy, which can easily be extended by simultaneous consideration of sensitivity and specificity of the new method being tested. Conclusion: The flexible approach to planning by means of the confidence interval of the sensitivity controls the desired confidence of the outcome of the diagnostic trial. It allows a priori evaluation of study budget, study duration, number of study centers and, above all, any ethical limitations. It provides arguments for the investigator to proceed with the comparison and a rationale for the decision to conduct the comparison as mono- or multicentric trial.