Development of Consensus-Based Best Practice Guidelines for Postoperative Care Following Posterior Spinal Fusion for Adolescent Idiopathic Scoliosis

Nicholas D. Fletcher, MD; Michael P. Glotzbecker, MD; Michelle Marks, PT, MA; Peter O. Newton, MD


Spine. 2017;42(9):E547-E554. 

In This Article

Materials and Methods

An initial investigation of current postoperative protocols provided postoperative pathways from nine major spinal deformity centers. The framework of these pathways was used to create an electronic survey exploring the postoperative care of patients after PSF for AIS. This initial survey of 25 questions was distributed to 26 spinal deformity surgeons via Survey Monkey (Palo Alto, CA) in October of 2015 with 21 members responding (80.7%). This group of surgeons represents 18 unique centers and was selected on the basis of the significant clinical and research experience of each member. Participants were asked to vote on their agreement with 17 unique questions using a five-point Likert scale (Strongly agree, agree, neutral, disagree, strongly disagree). In addition, nine more open-ended questions regarding when to initiate certain interventions (i.e., regular diet, oral narcotics, physical therapy, etc.) were included. The initial responses were collected and served as the basis for modifying and creating further questions seeking agreement within the perioperative period.

An initial face-to-face meeting was convened with 21 members present. A Delphi method was then used to begin construction of a consensus-based BPG for the postoperative management of AIS patients following PSF. The Delphi method is a previously validated system of developing consensus through repeated administration of consensus statements that are modified after multiple rounds of discussion.[37] Unique to this method is the fact that respondents may change their responses to consensus recommendations on the basis of the group discussions and/or rewording of the recommendations.

Each individual question from the electronic survey was discussed at the initial face-to-face meeting after the results were revealed. Modifications to the wording of the initial query were performed on the basis of group consensus and a second round of voting was performed on each question using the revised verbiage when necessary. The group discussion resulted in the formulation of a revised list of 23 questions utilizing a four-point Likert scale (the neutral option was removed). Seven additional survey questions were used to determine specific timing of certain interventions (i.e., how many days postoperative labs are obtained on or how many times physical therapy should see a patient daily). An audience response system (ARS) was then used to vote on this second iteration of the survey with the results again presented to the group for a second round of revisions. At the end of this meeting, the group was provided with relevant best available literature for review over a 6-week period. We chose to provide literature to the group after the initial face-to-face meeting due to the relatively limited material related to many of the areas of discussion and our desire to have an initial discussion of these areas.

A second face-to-face meeting was used to further discuss the initial consensus statements and resolve discrepancies wherever possible. A vote on the 30 items was performed and questions were revised as deemed appropriate by the group. These results were used to generate a final tally of 28 questions that were then posed to the group at this final meeting. Consensus was met on 19 statements as defined as >80% agreement and all members approved the subsequent consensus statement.