I read with interest an excellent article by Dr Newhouse and her colleagues discussing the difference between quality improvement (QI) and research activities.1 The authors warned that using nursing QI activities as evidence of a research program for satisfying Magnet Hospital criteria may have unintended and deleterious effects upon the host institution. They list "[horizontal ellipsis](1) poorly designed and interpreted studies, (2) lack of consideration of subject rights, and (3) Institutional Review Board (IRB) or other regulatory sanctions[horizontal ellipsis]" as potential negative consequences. All 3 of these untoward consequences are legitimate concerns, and I applaud the authors for raising the issue.

However, I was initially shocked at the sweeping set of criteria proffered by the authors distinguishing QI activities as research and subject to IRB review. I say "initially" because when I reviewed a number of documents and similar articles, I noted that these authors were well within the mainstream of published opinion. Ambiguous criteria are frequently used to define QI studies as research, and thus, subject to IRB approval. They include "systematic investigation," creation of "generalizable knowledge," large sample sizes, and studies of long duration. If the intent to publish findings exists, the QI study is usually assumed to require IRB approval.

The problem with this interpretation of the federal human rights law is that in our community hospital (and indeed many others), most of our QI projects would require IRB review prior to their initiation. We customarily apply inferential statistical analysis at all phases of our projects, study multiple process factors, conduct pilot tests of our process changes, and measure the results for at least one year following closure of the project. Furthermore, we assume an obligation to share our findings with other hospitals in our corporate system. Just as any deviation from these standards would constitute "bad research," their absence in a quality improvement program would constitute "bad QI." Newhouse et al accurately describe the wording of federal law, but we differ on its practical interpretation.

To submit all of our projects to the IRB would have 2 potentially devastating effects on our organization. First, it would overwhelm our IRB staff and (physician) volunteer panel members. But worse than this, it would threaten the viability of many QI projects that consistently work against the forces of cultural inertia that restrain efforts to promote change. The delay in project implementation alone could kill the motivation of the team members. Getting people excited about change and maintaining that level for an extended period without team action would be difficult and exhausting for team leadership.

Our hospital has instead adopted the following 5 criteria, any one of which will trigger the need to submit for IRB review: (1) subject recruitment, (2) subject randomization to treatments, (3) treatments or tests exceeding "usual care," (4) any risk to subjects beyond usual care, and (5) any use of personal health information in the writing of journal articles.

Lowell C. Wise, DNSc, RN

Six Sigma Black Belt

Saint Alphonsus Regional Medical Center

Boise, Idaho

([email protected])

Reference