Reporting Interim Results Can Show the Feasibility of Practicing Outcome-Based Neurovascular Care Within Randomized Trials: An Opinion.

Abstract

BACKGROUND\nRandomized trials of commonly performed surgical interventions are notoriously difficult to conduct. The trial methodology may nevertheless be the best way to offer outcome-based neurovascular care in the presence of uncertainty. One obstacle to promoting such trials is the conventional prohibition of publication and dissemination of interim results as the trial progresses.\n\n\nMETHODS\nWe review the scientific and statistical reasons against the publication of interim analyses as well as exceptions that can occur when 1 treatment is unexpectedly shown to be harmful or when the results of other trials have convincingly shown the comparative benefits of a new intervention. We also discuss the promotion of difficult surgical trials.\n\n\nRESULTS\nReasons to support the conventional ban on publication of interim results include control of statistical errors, prevention of invalid conclusions, and dissemination of false claims of equivalence of rival interventions. In the early phases of a trial, usually 1 treatment cannot be shown superior to the other. We believe, contrary to the received view, that a transparent report of the early progress of certain trials can be justified, even when interim results are inconclusive, to promote the recruitment of participating centers and the practice of a novel way to offer neurovascular care in the presence of uncertainty in the best medical interest of patients.\n\n\nCONCLUSIONS\nIn our opinion, the early publication of inconclusive interim results may increase awareness of the feasibility of surgical care trials.