Levels of evidence available for techniques in antireflux surgery.

The objective of this study was to determine the levels of evidence and grades of recommendations available for techniques in antireflux surgery. Areas of technical controversy in antireflux surgery were identified and developed into eight answerable questions. The external evidence was surveyed using the databases Medline and EMBASE. Abstracts and appropriate articles were identified from January 1966 to December 2005. A set of search strategies was systematically employed to determine the levels of evidence available for each clinical question. Primary outcome measures included the determination of levels of evidence and grade of recommendation based on The Oxford Center for Evidence-Based Medicine. Secondary outcome measures included for randomized controlled trials were Jadad scores, noting the presence of a sample size calculation, and the determination of an effect estimate and the reporting of a confidence interval. Higher quality randomized controlled trials (mostly level 2b, occasional level 1b) existed to answer three questions: whether to complete a 360 degrees or partial wrap; whether or not to divide the short gastric vessels; and whether to perform laparoscopic or open surgery. Lower quality randomized controlled trials were available to determine whether the use of mesh was helpful, whether or not to use a bougie catheter for calibration of the wrap, and whether an anterior or posterior wrap results in a superior outcome. This was deemed to be of inferior grade of recommendation due to the lack (< 2) of trials available and the sole presence of level 2b evidence. The final two questions: whether to complete fundoplication using a thoracic or abdominal approach and whether to use intraoperative manometry relied exclusively upon level 4 evidence and thus received a lower grade of recommendation. A higher Jadad score seemed to be associated with studies having a higher level of evidence available to answer the question. Sample size calculations were given to answer three questions. Effect estimate was difficult to interpret given inconsistent findings, composite outcomes and lack of reported confidence intervals. In conclusion, antireflux surgery has many randomized controlled trials available upon which to base clinical practice. Unfortunately, these are generally of poor quality. We recommend that esophageal surgeons determine consistent outcome measures and endeavor to improve the quality of randomized controlled trials they perform.