Expert Consensus and Equipoise: Planning a Randomized Controlled Trial of Magnetically Controlled Growing Rods

Study design: Expert consensus building using combined Delphi method and Nominal group technique.

Objectives: To identify the current state of equipoise surrounding the use of magnetically controlled growing rods (MCGRs) and to determine consensus for planning a randomized controlled trial (RCT) with MCGRs.

Background: The use of MCGRs for the treatment of early-onset scoliosis (EOS) is a new technology. Optimal use has not been thoroughly investigated and much uncertainty exists. Areas of uncertainty include construct architecture, timing of lengthenings, and amount of distraction per lengthening. Expert discussion and consensus is useful at this early juncture and necessary when designing an RCT.

Methods: Two rounds of surveys were administered to a group of experienced pediatric spine surgeons, followed by a 2-hour, face-to-face meeting in November 2015 and a 1-hour, face-to-face meeting in February 2016. The first survey used example cases to establish agreement around the proper use of MCGRs and identified areas of equipoise and disagreement. The second survey again used example cases-this time selected for their equipoise status-to solicit trial arms for a potential RCT of MCGRs and identified important open questions in the use of MCGRs. Lastly, the face-to-face meetings employed iterative voting to preliminarily plan an RCT of MCGRs.

Results: Following the Delphi survey rounds and the two Nominal face-to-face meetings, the group of experts decided on an MCGR RCT design that standardized all patients to bidirectional constructs, and randomized to a lengthening interval of 6 versus 16 weeks with a standardized equation for calculating the total yearly lengthening that approximates normal spine growth.

Conclusion: This endeavor indicates expert support for the use of MCGR in children older than 6 years, with curves greater than 60°. The uncertainty surrounding frequency of lengthening justifies an RCT of MCGRs.

Level of evidence: Level V.