Development and Validation of a Surgical Quality Assurance Tool for the Linx Procedure as Part of the Golf Trial

Background The quality of operative interventions in surgical randomised controlled trials (RCTs) is often unclear. Therefore, these trials are frequently criticised, and the results may not be implemented. Surgical quality assurance (SQA) is the method of assessing the delivery of surgical interventions in trials by measuring adherence to the protocol. The GOLF trial is an international, multi-centre (31 sites), phase III RCT, comparing fundoplication to LINX. In the GOLF trial, a SQA tool will be developed and validated—this is the first study of its kind to do this. The development of a competency assessment tool (CAT) for LINX is described. Methods The LINX CAT operative protocol was developed following a virtual Delphi Consensus. This consisted of questionnaires to ten international experts in hiatal surgery, identified through the literature and peer nomination. To establish inter-observer reliability of the LINX CAT, six surgeons of varying expertise and training were identified by peer-nomination and invited to use the LINX CAT and the previously validated fundoplication CAT developed by the Society of American Gastrointestinal and Endoscopic Surgeons (SAGES). They each reviewed four anonymised, unedited surgical videos (two LINX videos and two fundoplication videos submitted by surgeons participating in the trial), followed by a virtual workshop. Results During the Delphi Consensus, each operative step was classified as mandatory, optional or prohibited, with levels of agreement for each statement. Key anatomy and possible errors for each operative step were listed and free-text boxes were provided to fine-tune this. The results were reviewed qualitatively and amendments to each iteration of the LINX CAT were undertaken using a collaborative approach. A 100% consensus was reached after 3 rounds. The average Cohen’s kappa coefficient for inter-observer reliability when comparing poor/ineffective scores vs sufficient/ideal scores were 0.785 (range: 0.582–0.892) and 0.827 (range: 0.591–0.912) respectively for the two LINX videos. Conclusion This work describes the novel CAT for the LINX procedure. The development and content validation of the LINX CAT will form part of the SQA programme in the GOLF trial. Ongoing work as part of the trial will be to correlate key operative steps with post-operative clinical and patient-reported outcomes, specifically gastro-oesophageal reflux disease health-related quality of life.