to the federal regulations (45 CFR 46 102.d),
means a systematic investigation (including research development, testing and
evaluation) designed to develop or contribute to generalizable knowledge.
are two significant points to consider:
is a systematic investigation?
Systematic investigations can usually be recognized by the fact that they
have a question they are attempting to answer, or a hypothesis they are attempting
is developing or contributing to generalizable knowledge?
You are developing/contributing to generalizable knowledge if you intend on
sharing the information you produce with others, be it within a classroom,
at a poster presentation, at a conference, or in a publication.
examples of what is not research:
reports - are not considered research in that although they contribute
to generalizable knowledge, they are not systematic investigations. The clinician
is simply sharing information about interesting cases for educational purposes.
people go wrong: if you attempt to answer a question, or prove/disprove
a hypothesis, you are no longer dealing with a case report and instead
are now dealing with research. In addition, case reports are usually limited
to no more than a few patients. Once you start to exceed three subjects
it appears that you are doing something systematic (attempting to prove
a point or answer a question) and IRB approval is required.
Assurance - is not considered research in that although it is a systematic
investigation, there is no intent to share the information with others (contribute
to generalizable knowledge). For example, the pharmacy can conduct quality
assurance and compare the use of a medication PO vs IV administration and
not require IRB approval provided they do not publish the results. The results
of QA can only be published if IRB approval is obtained prior to publication.
people go wrong: if you know you are likely to publish your work provided
the results are worthwhile, then the project does not meet QA guidelines
and should be submitted as research prior to initiation. To meet quality
assurance criteria there must be no intent to publish at the onset of