Systematic analysis underlying the quality of the scientific evidence and conflicts of interest in interventional medicine subspecialty guidelines.

OBJECTIVE: To determine the validity of guidelines published by interventional medical societies.

METHODS: We reviewed the interventional medicine subspecialty society websites of the American Association for Bronchology and Interventional Pulmonology (AABIP), American Society of Diagnostic and Interventional Nephrology (ASDIN), American Society for Gastrointestinal Endoscopy (ASGE), and Society for Cardiovascular Angiography and Interventions (SCAI) as of November 15, 2012, for published interventional guidelines. The study was performed between November 15, 2012, and January 1, 2013. The AABIP did not publish guidelines, so American Thoracic Society and American College of Chest Physicians guidelines were reviewed. All the guidelines were reviewed for graded levels of evidence, methods used to grade the evidence, and disclosures of conflicts of interest (COIs).

RESULTS: Of 153 interventional guidelines evaluated, 4 were duplicates. Forty-six percent of guidelines (69 of 149) graded the quality of evidence using 7 different methods. The ASGE graded 71% of guidelines (46 of 65) compared with 29% (23 of 78) by the SCAI and 0 by the ASDIN (n=4) and the pulmonary societies (n=2). Of the 3425 recommendations reviewed, 11% (n=364) were supported by level A, 42% (n=1432) by level B, and 48% (n=1629) by level C. The mean age of the guidelines was 5.2 years. Additionally, 62% of the guidelines (92 of 149) failed to comment on COIs; when disclosed, 91% of guidelines (52 of 57) reported COIs. In total, 1827 COIs were reported by 45% of the authors (317 of 697), averaging 5.8 COIs per author.

CONCLUSION: Most of the interventional guidelines failed to grade the evidence. When present, most guidelines used lower-quality evidence. Furthermore, most guidelines failed to disclose COIs. When commented on, numerous COIs were present. Future guidelines should clearly state the quality of evidence, use a standard grading system, be transparent regarding potential biases, and provide frequent updates.