This project was nominated by Ranjana Banerjea, HSR&D & HSR&D Research Best Practices Workgroup (RGP). The goal was to describe the evidence regarding areas of interest in the ethical conduct of research, with a particular focus on research and quality improvement initiatives. Further discussions resulted in the following key questions:

Key Question #1.

What has been published regarding the IRB and each of the following issues:

  • Quality improvement initiatives conducted as research studies and therefore requiring IRB approval
  • Managing conflict of interest
  • Studies requiring approval of multiple IRBs
  • Genetic issues
Key Question #2.

What is the actual evidence regarding the issue with the largest literature which may inform VA policy?


Search Terms and Algorithms

“institutional review board” Or “institutional review boards” OR advisory committees OR ethics committee OR ethics committees OR “national commission for the protection of human subjects”


“informed consent”[ti] OR informed consent[mh] OR selection bias OR selection biases OR privacy OR confidentiality OR “using data” OR “data use” OR “data usage” OR “human experimentation” OR “patient safety” OR registry OR registries OR medical records OR ethics[ti] OR ethics[mh] OR ethical* OR “conflict of interest”


qi OR “quality improvement” OR “quality-improvement” OR mortality OR quality assurance, health care OR total quality management OR “quality management” OR quality of health care OR “health care quality” OR quality control

OR quality[ti]

Databases Searched

We searched PubMed databases from 1/01/2000-2/11/2011 for English-language articles related to our key questions. Search terms were chosen that pertained to institutional review boards and informed consent, data use, conflict of interest, ethics, and quality improvement initiatives.

Related Articles Search

We also completed related article searches in Pub Med on four key articles.1215

Internet Search

We searched the websites of relevant organizations, including Association of American Medical Colleges (AAMC), Public Responsibility in Medicine and Research (PRIM&R), the Association for the Accreditation of Human Research Protection Programs, Inc (AAHRPP), and the National Institutes of Health (NIH), for empiric studies of COI.


We completed a title screen, excluding any article that, from its title alone, indicated it was unrelated to Key Question #1. We then performed another round of screening on the abstracts of articles that passed the first round, again excluding articles not related to Key Question #1. Full articles included from this stage were then categorized according to their topic of interest: the challenges of IRB approval in multi-site studies; when quality improvement initiatives become research; conflicts of interest; payments to patients, providers, or study participants (which we split off from conflicts of interest as it seemed to be a distinctive topic in and of itself); genetics research (this category mainly pertained to data repositories); miscellaneous articles pertaining to Key Question #1; background articles; and articles rejected at that stage as not being relevant. Articles were rejected if the scope was not relevant to our population (e.g. articles about children) or setting (e.g. if the article was not pertinent to IRBs), or if they were commentaries, opinions, or did not present original data, i.e. documents presenting an organizational COI policy or a position paper were rejected.


We coded each article according to the issue it addressed (per Key Question #1) or as “miscellaneous.” To the four issues addressed in Key Question #1, we added a 5th category, “payments to patients or providers.” For articles about conflict of interest, the topic selected for detailed review, we abstracted information about the study design, participants, and findings, and constructed an evidence table (Appendix A).


After assessing the output of our literature searches, we then selected for detailed review the area that had the greatest number of relevant published studies that did not already have VA policies in place, namely conflict of interest. Because of the nature of these studies, no quantitative synthesis was possible; therefore our synthesis is narrative.


A draft version of this report was reviewed by seven technical experts as well as VA Central Office leadership. Their comments and our responses are presented in Appendix B.