Research vs. public health practice: when does a study require IRB review?

Although guidelines for the ethical review of researchare continuously evolving, there is one point upon whichthere has long been general consensus: that researchprojects involving human subjects require prior review andapproval by an appropriate institutional review board (IRB).The very definition of research, however, involves someambiguities. The collection or manipulation of data involv-ing human subjects may or may not always be consideredresearch per se. Research, as currently defined, occurs whena study is designed to contribute to generalizable knowledge[1–4,4a]. “Non-research” activities generally take the formof patient treatment, public health practice, program evalu-ation, or population surveillance [2,5]. Public laws providefor oversight of the collection of confidential information bypublic health authorities without consent, and confer specialprotection of the information from public disclosure. This isgenerally because many public health efforts essential toprotect the public health involve the routine collection ofhighly confidential and sensitive personal and medical in-formation (e.g., mandatory reporting of communicable dis-eases). In a similar vein, other public health efforts, such asinvestigation of disease outbreaks, must occur quickly toreduce the spread of the disease and find the source asquickly as possible. Activities undertaken to investigatedisease outbreaks involve application of proven publichealth strategies, not research. Although the activities mayinvolve case-control or cohort study designs, formal statis-tical analysis of data, and publication of findings and controlmeasures, the purpose of the work is to apply public healthpractice, not to contribute to generalizable knowledge, as ina research project [2,5]. While these types of public healthactivities are generally not designed to contribute to gener-alizable knowledge, they may often result in publication offindings in the peer-reviewed literature. Thus the distinctionbetween research and non-research is anythingbut distinct.As a consequence, the responsibility for determiningwhen and whether a manuscript submitted for publicationhas had appropriate ethical review (or an appropriate ex-emption to this requirement) all too often falls on the shoul-ders of journal editors. The editors of Preventive Medicine[6] and other journals [7] frequently receive manuscripts forconsideration that do not document whether or not the studyreceived appropriate ethical review. Frankly, checking forethical review at such a late stage defeats the purpose. It isvery difficult for editors to differentiate research from non-research unless the authors provide explicit information tohelp them. Furthermore, it is simply not appropriate for aresearch study involving human subjects to proceed with nooversight from an IRB with only an after-the-fact opinion ofa journal editor as the sole inducement to comply withhuman subjects protections.Why do we need ethical review of human research stud-ies in the first place? Ethical review accomplishes severalpurposes. It provides expert assessment of the safety of anyprocedures used in a study and it ensures that the autonomyof subjects under study is maintained and that the rights ofindividuals with diminished autonomy are likewise pro-tected (i.e., prisoners, children). It allows an evaluation ofrisk vs. benefit to ensure that benefits to subjects are max-imized while harms are minimized. It helps ensure that anyresearch risks are equitably distributed among populationsmost likely to benefit from the results. Finally, it ensuresthat the research design is sound and that those who conductthe research are competent both to conduct the research andto assure the well being of the research subjects, includingobtaining proper informed consent when appropriate.If we proceed first from the tenet that the common goal