|| Checking for direct PDF access through Ovid
This Open Forum illuminates shortcomings with the basis for determining degree of oversight of health services research and quality improvement activities. Using a federally regulated definition of research rather than a direct appraisal of risk to patients can misallocate effort from activities with higher risk for patients to those with lower risk. The case of the Johns Hopkins multicenter study of central line safety checklists in intensive care units is cited. Definitions of research promulgated by the Office of Human Research Protection are reviewed, and an alternative model based on patient risk is proposed. Suggestions for how quality improvement work fits into the larger paradigm of research are made.