We recommend that the protocol include a clear description of study design;
objectives, estimation of performance goals (e.g., desired confidence interval
widths) that are directly related to the intended claims for the IVD device, or
hypotheses; and a statistical plan to be applied to the data.
