Clinical Research in Vulnerable Populations: Variability and Focus of Institutional Review Boards' Responses

Abstract

Background: Children and patients with cognitive deficits may find it difficult to understand the implication of research. In the European Union (EU), clinical studies outside the EU directives concerning medicinal products or medical devices, i.e., "miscellaneous clinical studies", have no legally mandated timelines for institutional review boards' (IRB) decisions.

Goal: To evaluate the review process of IRBs for two different "miscellaneous" multicenter clinical research protocols involving vulnerable subjects (children and adult stroke patients).

Methods: Descriptive and comparative statistics. Protocol 1 is a prospective, multicenter, cross-sectional screening study of a symptomatic pediatric population at risk for Fabry disease involving genetic testing (NCT02152189). Protocol 2 is a prospective, multicenter, randomized, controlled, open-label, blinded endpoint post-market study to evaluate the effectiveness of stent retrievers (NCT02135926). After having obtained positive initial IRB votes at the main study site, both protocols were subsequently submitted to the remaining IRBs.

Results: Protocol 1 was submitted to 19 IRBs. No IRB objected to the study. Median time-to-final vote was 34 (IQR 10-65; range 0 to 130) days. Two IRBs accepted the coordinating center's IRB votes without re-evaluation. Changes to the informed consent documents were asked by 7/19 IRBs, amendments to the protocol by 2. Protocol 2 was submitted to 16 IRBs. Fifteen decisions were made. No IRB objected to the study. Median time-to final vote was 59 (IQR 10 to 65; range 0 to 128) days, which was not statistically significantly different compared with protocol 1 (Wilcoxon test). Two IRBs accepted a previous IRB decision and did not conduct an independent review. Eight/16 IRBs required changes to the informed consent documents; two IRBs recommended an amendment of the protocol.

Conclusion: Both clinical research protocols involving vulnerable populations were well accepted. IRB workflows and decision times varied substantially. Some IRBs accepted a previous IRB decision without the necessity of another reevaluation process. Requested changes were focused on the informed consent documents. A more standardized approach across jurisdictions is desirable.

Conflict of interest statement

Competing Interests: Bärbel Kästner, Simone Behre, Nadine Lutz, Friederike Bürger, Steffen Luntz, Katrin Hinderhofer, Martin Bendszus, and Georg F. Hoffmann report no competing interests. Markus Ries received consultancy fees or research grants from Alexion, GSK, Shire, and Genzyme. This does not alter the authors’ adherence to PLOS ONE policies on sharing data and materials. There are no restrictions on sharing of data/or materials.

Figures

Fig 1. Time-to-final response [days] of two study submissions to 34 IRBs by IRB type (university or state medical board, p = NS).

The overall median time-to-final response for the two studies was 38.5 (IQR 10 to 83; range 0 to 130) days.

Fig 2. Total number of requested changes in the submission packages for two clinical studies submitted N = 35 IRBs.

Fig 3. Time-to-final response [days] of N = 35 IRBs as function of quality of available web information about the submission process (p = NS, non-parametric ANOVA).