Systematic Reviews

Oregon Health and Science University (OHSU) and the University of York completed independent reviews of the entire body of scientific evidence regarding the safety and effectiveness of recombinant human bone morphogenetic protein-2 (rhBMP-2). The teams had full access to all Medtronic’s clinical trial, post-marketing and safety data about rhBMP-2. 

OHSU and University of York independently developed final reports based on their analyses of the rhBMP-2 data. These reports were formally reviewed by expert members of the YODA Project Steering Committees, including orthopedic surgeons, statisticians, and experts in clinical trial design and analysis, as well as by representatives from the manufacturer of rhBMP-2, Medtronic, Inc. The centers refined their reports based on reviewers’ feedback and peer-review comments made on their manuscripts. Their aims included the following:

  1. Identify all relevant studies, across all users and sponsors (i.e., Medtronic sponsored and others).
  2. Determine the questions that were addressed by these studies. 
  3. Evaluate the quality of the studies. Assess the risk of bias associated with the design, conduct, and reporting of each clinical study, including those identified via the systematic review and those provided by Medtronic, and, if present, how bias may have affected assessment of effectiveness and harms. 
    1. Assessment of study design and conduct should include evaluation of internal validity, methods used to ascertain outcomes and other policies and procedures for data collection, as well as the integrity of case report form (CRF) adjudication.
    2. Assessment of study reporting should include selective publication and selective reporting.
    3. Summary of these findings should include: 
      1. What conclusions can be drawn by assessing the full body of data and what gaps in knowledge remain, taking into account results from the evaluation of quality and risk of bias
      2. An assessment of applicability of these studies. 
  4. Conduct meta-analyses from studies identified via the systematic review, if appropriate and using patient-level data, if possible. If not appropriate, there should be another approach to summarizing the data. The analysis should consider the following: 
    1. For effectiveness, meta-analysis should consider patient-centered outcomes (i.e., quality of life and functional status), as well as surrogate outcomes (i.e., fusion as determined by radiography).
    2. For safety, meta-analysis should include all harms described in the Request for Proposal.