Broad Consent in Healthcare Research: What Is Efficient, What Is Right?

1.

The concept of informed consent in medicine has evolved notably over time. Historically, patients had little control over medical decisions, including participation in clinical experiments, as the physician's authority was largely unquestioned. Fortunately, we have evolved as a society and ensure that patient rights and autonomy are protected. This has led to the concept of informed consent, to verify and document that individuals understand the risks, benefits, and alternatives to any medical procedure or research participation.

Informed consent is a crucial aspect of medical and research ethics. In the medical context, it refers to obtaining a patient's explicit permission for treatment, ensuring they fully understand the potential outcomes and alternatives. In research, informed consent involves providing participants with detailed information about the study's purpose, goals, procedures, potential risks, and their right to withdraw at any time. In both contexts, it is essential that individuals receive sufficient information, presented in a manner that is understandable to them, to make an informed decision and fully volunteer their participation.

Consent generally takes two main forms: implied or written. Implied consent is inferred through actions, such as a patient moving onto an examination table. This action implies their consent to be examined by a physician. However, verbal consent is not sufficient in the majority of cases. Written consent, in contrast, involves a formal, signed agreement and is required for most medical procedures and research. There are exemptions, such as in emergency situations or low‐risk research, where the requirement for written informed consent may be waived or modified. These exceptions, however, raise ethical concerns. Hence, for even low‐risk research, Institutional Review Board (IRB) approval remains necessary. Of note, some JMRI submissions incorrectly conflate clinical medical consent with research consent, underscoring the need for clearer distinctions in publication standards and style guides.

The implementation of informed consent in imaging differs globally, due to varying national laws and local ethics or review board guidelines [1, 2, 3, 4]. In the United States, to date, most institutions obtain a written, study‐specific, or explicit consent for prospective studies [3]. For retrospective studies that use deidentified or anonymized data, a waiver of consent may be granted by the IRB. Investigator integrity is paramount here, as many imaging studies use sequences that may not be part of routine clinical imaging protocol. Editors find substantial variability in the implementation of consent for “sequence”‐focused studies.

In January 2019, the United States revised the Common Rule, introducing the concept of broad consent for the secondary use of identifiable private information and biospecimens in research [5]. This change followed the European Union's General Data Protection Regulation (GDPR), which, effective from May 2018, allows the processing of personal data for scientific research under certain exemptions [6]. Global acceptance of broad consent is inconsistent. For example, Germany's Medical Informatics Initiative (MII) aims to standardize consent across healthcare settings, facilitate data sharing, and promote collaborative research [7]. In contrast, the acceptance and implementation of broad consent in the United States have been more cautious and is highly institution‐dependent.

Broad consent falls somewhere between a waiver and full informed consent, with a general understanding of potential uses and explanation of the process for withholding or withdrawing consent. However, the consent forms typically do not specify how exactly the data will be processed, by whom, or for what exact purpose [2]. While this allows researchers to use the data without reobtaining consent for each new project (e.g., streamline research) it also carries notable ethical risks and may further undermine trust in the medical system and raise investigator probity concerns. Although the Declaration of Helsinki is often viewed as establishing uniform ethical consent standards, practices around consent in imaging research can vary quite a bit. This inconsistency raises important ethical questions that merit further attention. If broad consent forms fail to clearly disclose the possibility of data sharing with external or commercial parties and lack transparency, they may undermine informed consent altogether.

A pertinent example—while not a strict case of broad consent—is the 2015 collaboration between Google's DeepMind and the Royal Free London NHS Foundation Trust [8]. The Trust shared approximately 1.6 million patient records including imaging data and sensitive information, with DeepMind to develop an AI tool (Streams) for detecting acute kidney injury. Patients had consented to medical treatment and the associated use of their data for direct clinical care, but were not explicitly informed of the data transfer; the Trust relied on implied consent under the assumption that the data would be used for “direct care.” Ultimately, the Information Commissioner's Office (ICO) later determined that this was an inappropriate legal basis, as the data were used for testing the app, not direct patient care. This case underscores the need for transparency, allowing patients the opportunity to make truly informed decisions about their data, especially in an age where AI plays an increasingly important role in healthcare and research.

Despite concerns about patient autonomy, privacy, and trust, access to large datasets is paramount for advancing AI‐driven healthcare, particularly in medical imaging. If implemented with care, broad consent can help ensure that data collected from patients as part of clinical studies is used transparently and in alignment with patients' expectations, unlike the situation highlighted above. Broad consent has the potential to reduce unnecessary bureaucracy and accelerate research by enabling responsible data reuse, thereby supporting the development of advanced diagnostic tools with significant potential to improve patient care. Furthermore, broad consent has the potential to expand the inclusion of underrepresented populations and improve research scale—but only if implemented with care.

To balance research efficiency and addressing the ethical challenges of broad consent requires flexible and patient‐oriented solutions. Dynamic consent models—which allow patients to revisit and adjust their preferences over time (enabling real‐time updates) via secure digital platforms—would be well‐suited for addressing transparency and trust in data use. For instance, a patient who underwent an MRI could log into a secure portal, view that their imaging data was used in a specific university‐led AI study on lung cancer, and update their preferences to allow or restrict future use by commercial sites.

Consent documents should be concise, written in accessible language, and clearly distinguish research participation from clinical care. Importantly, patients must be assured that declining participation or changing preferences will not affect their medical treatment. Research protocols should include data minimization, clear sharing agreements, and robust oversight, especially for commercial or AI uses, with rigorous IRB review and transparent reporting to ensure alignment with the original consent.

Given the variation in how broad consent is currently applied across countries, it may be advisable for manuscripts to clearly specify how consent was obtained, in accordance with the legal and ethical standards of the researchers' political location. Ideally, a globally aligned framework would equal rights and protections for data subjects across borders while facilitating ethically sound multinational research collaborations.

Ultimately, in medical imaging, efficiency and ethics should be viewed as complementary rather than conflicting goals in the pursuit of responsible innovation. We hope that with this commentary we spark interest and encourage readers to help shape the future of broad consent in the evolving landscape of imaging research ethics, especially in an era of AI‐driven data use.

Conflicts of Interest

M.E.S. consulting fees from ISMRM and MMI; payment for expert testimony; participation on a Data Safety Monitoring Board or Advisory board for Ankasa, Paradigm, Spine, TOPS, Premia Spine, and Calypso; Carelon Medical board member.

Starekova J. and Schweitzer M. E., “Broad Consent in Healthcare Research: What Is Efficient, What Is Right?,” Journal of Magnetic Resonance Imaging 62, no. 3 (2025): 856–857, 10.1002/jmri.70000.