[Skip to Navigation]
Comment & Response
January 2015

Criteria for Waiver of Informed Consent for Quality Improvement Research

Author Affiliations
  • 1Division of General Internal Medicine and Geriatrics, Department of Medicine, Feinberg School of Medicine, Northwestern University, Chicago, Illinois
  • 2Institute for Public Health and Medicine, Feinberg School of Medicine, Northwestern University, Chicago, Illinois
JAMA Intern Med. 2015;175(1):142-143. doi:10.1001/jamainternmed.2014.6977

To the Editor Pletcher and colleagues1 addressed the important issue of informed consent in quality-improvement research. We have often faced the issue of whether a waiver of consent is justified in our research. For the last several years, we have used 3 guiding criteria. First, the research must be minimal risk. Although this seems self-evident for most quality-improvement projects, it should not be taken for granted. There is always a risk of loss of confidentiality, and for some sensitive topics such as human immunodeficiency virus care or substance use treatment, the risk of loss of confidentiality must be weighed against the advantages of conducting a study with a waiver of consent. Institutional review boards (IRBs) should ensure that proper safeguards are in place to prevent accidental disclosure of sensitive data (eg, extraction of data from the electronic health record without patient identifiers). In addition, IRBs should consider the risk to the control and usual care groups that will not receive a promising intervention, including (1) plans to provide the intervention to the control group after trial completion if it is successful and (2) whether the likelihood of benefit is high enough that it threatens equipoise and it would be unethical to conduct a traditional randomized clinical trial, in which case an alternative study design (eg, stepped wedge, time series) should be recommended. The second criterion is that it would not be possible to obtain informed consent without threatening the validity of the trial. This clearly applies when the process of obtaining consent would make a person aware that she needed a clinical service, eg, telling a patient that she is eligible for a study of outreach to improve colorectal cancer screening because she had not been screened.2 But, it is also appropriate to request a waiver of consent for effectiveness studies, which would be invalid if a minority of eligible patients consented to participate. Third, we require that all data for the study will be collected as part of routine care, including patient demographics, comorbidities, and outcomes. If additional data are needed from patients, we require that informed consent be obtained for the part of the study that requires these data.3 Using these principles, we believe it is possible to conduct large, rigorous, generalizable quality-improvement research while maintaining stringent protection for human participants. Because of the subtleties in judging whether these criteria are met, we believe this decision is best left to an IRB.

Add or change institution