Keeping in line with its risk-based approach to regulation, the agency issued guidance on planning and implementing adaptive designs — as opposed to unchanged designs — for clinical studies in medical device development programs, saying these can reduce resource requirements and/or increase the chance of study success.

An adaptive design for a medical device clinical study allows for “prospectively planned modifications” based on study data that’s being accumulated, without undermining the trial’s integrity and validity. While modifications should be “prospectively planned” and described before a study is initiated, post-trial commencement study changes can be scientifically valid if the trial design decision-makers haven’t had access to outcome results by treatment, the guidance explains. Adaptive study design planning is focused on anticipated modifications that could potentially be desirable based on the data accrued throughout the study. The FDA advises sponsors to expect and plan for changes based on a range of potential scenarios and to go over planning with the agency, noting unplanned modifications may not be approved by the agency.

The guidance lists several advantages of adaptive designs — compared to unchanged designs — noting that ultimately they may help accelerate device development decision-making, and thus enable more efficient resource investment in a study. It also provides guidance on how to determine whether an adaptive design is “feasible and advantageous,” noting studies enrolling subjects quickly or complex studies with multiple endpoints and secondary endpoints for claims may not lend themselves to adaptation. However, studies with shorter endpoints and longer recruitment times, as well as studies where the time to the primary endpoint evaluation is long and the accrual is longer, may benefit from adaptation. The FDA recommends that the choice of an adaptive design be considered during the planning of a pivotal study, using a number of realistic scenarios to make a determination.

The document — applicable to PMA applications, 510(k) submissions,de novo submissions, HDE applications and IDE submissions — also goes over principles for adaptation in the design of clinical studies, including controlling the chance of flawed conclusions and minimizing operation biases. The document also covers the use of unblended data, with group sequential designs, sample size adaptation and group sequential design with sample size reassessment as the most widely used, as well as special considerations. The FDA also lists regulatory considerations, recommending interactions and communication with the agency, and having a risk-based monitoring plan established. The guidance further describes best practices to protect study blinding, and the submission of an adaptive design’s content to the FDA.