Research andQuality Improvement: Drawing Lines in the Grey Zone

The ethical conduct of research is a core requirement that must be followed by all researchers. To that end, the National Association of EMS Physicians andPrehospital Emergency Care require that all “manuscripts reporting results of investigations of human subjects must indicate approval by an institutional review board (human subjects committee).”¹ The organization places the same requirement on research abstracts submitted for presentation at the annual meeting.

Frequently, authors of manu-scripts or abstracts inquire about the applicability of this requirement to studies using quality improvement data. The typical question is: “These data were collected as a part of our ongoing quality improvement processes, andnot as part of any preplanned research project. Thus, institutional review board (IRB) approval was not obtained. Is this a problem?” The difficulty with differentiating quality improvement activities from research is neither new nor unique to emergency medical services. Several authors have addressed this issue in other disciplines,², ³, ⁴ anda 2006 Hastings Center report on the ethics of quality improvement methods also explored this problem.⁵ While there is a natural interplay between research andquality improvement, separating the two is important. Different authors suggest different guidelines for distinguishing between the two activities,², ³, ⁴, ⁵ but the core differences are clear: Quality improvement is part of the clinical enterprise; it is making sure the best possible care is delivered using optimal processes. Research is part of learning andknowledge development; it is determining what interventions andprocesses are best. Quality improvement generates data specific to the organization involved, for the distinct purpose of improving the performance of that specific organization. Research generates data that are (or at least should be) generalizable to other organizations, for the purpose of improving the care of all similar patients.⁶

This is not to say that quality improvement data have no place in research. Most quality improvement data derive from either clinical data, such as medical records, or operational data. These are both legitimate sources of research data, particularly for retrospective studies. Thus, existing quality improvement data can form the basis for research. For example, quality improvement data on cardiac arrest,⁷, ⁸ intubation,⁹ andnon-transport¹⁰ have been used in research.

The key is distinguishing the collection anduse of data for clinical or operational purposes (including quality improvement) from searching, retrieving, andanalyzing those data for research purposes. Quality improvement gives one license to track patient data andoutcomes; it does not give license to share what is found with the rest of the world. IRB approval must be obtained before using quality improvement data for research, just as it must be obtained before using medical records for research.

A project using data collected as part of the patient care process to evaluate existing internal programs against an accepted best practice would be quality improvement. An example would be an EMS agency abstracting data from EMS call reports to evaluate compliance with a refusal of transport protocol andensure delivery of quality care. In contrast, a project using those same data in an effort to determine a best practice or best process would be research. For example, evaluating two training programs on refusal of transport in an effort to determine which produces better protocol compliance would be research.

Evaluating how frequently andaccurately providers document pain management would be a quality improvement initiative. Implementing a new pain management protocol to see if it results in better analgesia administration than the old protocol would be a research effort. Yes, they are fine lines. That is why, if there is any doubt at all, IRB approval should be sought. There is no shame in having the IRB reply, “This activity, as described, does not represent human subjects research.” (Such a letter is very important andshould be saved in case any questions come up later!) This can be the difference between being able to publish your findings andhaving a paper that is unpublishable.

Although it is not a perfect indicator, another thing to consider in determining whether a project is quality improvement or research is how the results of that project will be used. If the findings are going to be shared outside of the organization, through presentation as an abstract at a professional meeting or through publication in a journal, the presumption should be that the project is attempting to produce generalizable data andthus qualifies as research.

Research based on quality improvement data will usually be research based on existing data. Because studies of existing data can be structured to ensure minimum risk andprovide for the de-identification of data, the required IRB process is typically quite easy. Depending on the institution, a simple letter of explanation or a “short” form application may be all that is required to obtain approval. This process is sometimes referred to as “exempt” review. Exempt review means that the proposal is exempt from having to go through full IRB review andcan be reviewed in a more streamlined fashion by the IRB chair and/or designees. It does not mean that the proposal is exempt from having to be reviewed.

One concept that many seem to misunderstand is whether IRB approval can be obtained after a project is completed, the data collected andanalyzed, andthe results submitted for presentation or publication. There is no such thing as retrospective or retroactive IRB approval; that would potentially undermine the protection of human subjects. IRB approval must be obtained before the findings are shared outside the routine internal QI processes.

Particular care must be taken to avoid implementing research under the mistaken umbrella of quality improvement. While quality improvement is an important andlegitimate activity, implementing a quality improvement program to intentionally circumvent consent requirements or IRB requirements is unethical andimproper.

Certainly, the differentiation between research andquality improvement is not always as clear as the examples used above. Making the distinction is still important—indeed, being clear about the distinction might be most important in those situations that are most gray. Quality improvement andresearch are both important to improving andensuring optimal care andsystem performance. The ethical framework for research, however, does differ from that for clinical activities.¹¹ Institutional review board approval is critical for ensuring that research activities—including those based on quality improvement data—are conducted within that framework.