BONUS POST - The Ninth Principle
A small act of attempted restoration before 2025 draws to a close
I was privileged to work on a great many impactful projects during my time at the NIH. One of the highlights for me was co-leading a working group tasked by Dr. Monica Bertagnolli during her tenure as NIH Director to come up with a core set of principles to guide the NIH’s thinking about how data generated for the purposes of clinical care can/should be used for research purposes. The group was comprised of experts from across NIH, and though I won’t name them individually here (for reasons that will shortly become obvious), they all deserve tremendous credit and recognition for the outstanding work they did.
The working group was convened at a particularly pivotal moment, in my opinion. The explosion of artificial intelligence across sectors was fueling all kinds of excitement in biomedical research about how the incredible quantities of clinical data – electronic health records, medical imaging, pathology slides, waveform data, and more – could power revolutionary discoveries. But, many considerations around permissions, privacy, data security, data quality, representativeness, actionability of findings, cost, data governance, and a whole host of additional issues come into play when you start thinking about using data to power research when the data were originally generated for the purposes of delivering healthcare. Indeed, it sometimes can become unclear where the boundaries are between the Health Information Portability and Accountability Act (HIPAA, which governs many aspects of clinical data use) and the 2018 Revised Common Rule (which governs many aspects of human participant research). I can actually remember being on a call once that required both the HIPAA expert *and* the Common Rule expert from NIH’s Office of General Counsel. (Achievement unlocked!)
Through many hours of deliberations, discussions, and respectful disagreements, the working group developed a set of principles that were intended to guide thinking across all of NIH for how clinical care data could be used for research. The principles not only tried to capture the current state of affairs but also tried to anticipate likely rapid evolution of this part of the research enterprise, looking ahead to potential future issues related to technology developments, the spread of AI, evolving discussions in bioethics, future complexities in privacy and reidentification, and more. The group iterated on the text of the principles, considering the potential subtleties of meaning in each word. After weeks of work, we had the privilege of presenting the principles to the NIH Director and the assembled IC Directors, ultimately arriving at a set of principles that met with the approval of the highest echelons of NIH leadership.
So, I can honestly say that I was absolutely delighted to see our principles appear on the NIH website just a few days ago! Please give them a look here: https://osp.od.nih.gov/policies/healthcare-research/
This was part of NIH’s newly announced Roadmap for Engaging the Public As Partners in Clinical Research, which you can read about here: https://www.nih.gov/about-nih/nih-director/statements/roadmap-engaging-public-partners-clinical-research
There’s a LOT of good stuff here! I really am thrilled that the NIH has put the principles and the roadmap forward! But as I read over the principles again, there was something that caught my attention. It was something that wasn’t there. You see, the working group developed nine principles to guide NIH’s thinking about the use of clinical care data for research, but only eight principles appear on the website. After a close read of the eight principles on the website, I am delighted to say that I look at them now and think, “Yeah, these are really good!” fully acknowledging my own biases. I can also say that I don’t think the important features of the missing ninth principle are fully reflected in the other eight (which I suppose means that we did a decent job of being non-redundant).
I think it’s important to honor the work done by the dedicated and brilliant experts on the working group. I think it’s important that people know there was one other principle that was considered important enough to have its own entry. I think it’s important that people know the breadth and depth of public service that was intended by all nine of the principles. I think the ninth principle is still scientifically and socially important, perhaps even more so now, and I think people ought to have a chance to know what it was.
The ninth principle was about the importance of ensuring that clinical data used for research purposes could be used to support research that promotes health equity. The principle was intended to speak to the fact that data collected from clinical care settings is heavily influenced by who in this country is able to seek and to access healthcare. For any group of people or segment of American society that doesn’t have adequate access to healthcare, their data will also not be adequately represented in any research conducted with clinical care data. So, for the data to represent the American population instead of the portion of the American population able to get healthcare, specific attention to health equity must be present across all phases of the data *and* the research lifecycles. Understanding which communities and groups are underrepresented in clinical care data is vital to knowing what the limitations of applicability and generalizability are for any research findings based on those clinical data. These limitations must be communicated with any research findings. This is a basic tenet of rigorous, reproducible science. Further, as the data and research lifecycles make clearer the populations underrepresented in clinical care data and in research, deliberate efforts to augment the presence of those populations in research data and the access of those populations to healthcare would improve the impact of research findings and the health of the nation simultaneously.
Please understand that I have not sought the opinion or assent of any of the working group members regarding sharing this. I am not speaking for the working group in any way. I am not a federal employee and I do not represent or speak for the NIH or any part of it in any way. There may have been a detailed deliberative process that led to the removal of the ninth principle. I don’t know and so I cannot comment on that. I am only speaking for myself here as a scientist, a physician, and a public servant.
Happy holidays to you and yours, and let’s keep at it in 2026!
