Introduction
The end of the last full week in February has been and gone, featuring a double feature 'Pizza and Prosecco', too many meetings with England at 6am my time, and prepping for being back in England in two weeks (6th-20th March). The absolute highlight, however, was seeing the Dance Theatre of Harlem perform at the Schubert Theatre here in New Haven: they were phenomenal. I used to dance pretty seriously until I was in my mid-20s and I've always loved it. In fact, I wrote my undergrad dissertation on how the arrival of social media was facilitating the diversification of dance styles - in particular the development of fusion styles. Still, hands down, one of the best performances I've ever seen.
Anyway, that's not why you're here.
On with the notes.
Things I worked on
Prepping for the UK. I'm about to be back in England for a flying visit filled with a plethora of talks, lectures, panels, etc. So I have been prepping quite extensively for the trip, making slides, preparing talks, pre-meeting, making sure I am up to date with my reading etc.Specifically, I'll be:
At the Nuffield Trust Summit on 7th March speaking on the topic of integrating AI into the NHS. I'll be talking for 15 minutes, and then there will be a panel discussion hosted by Nenna Osuji with myself, Prof Neil & Dr. Simon Roberts. You can register to watch the event virtually here.
At Trinity Hall in Cambridge, for the AI ethics society, panel discussion on the Ethics of AI in Healthcare. You can sign up to attend in person or watch on zoom here.
At Digital Health Rewired on 13th March talking on the main digital transformations stage with Jon Hoeksma, CEO, Digital Health; Dr Vin Diwakar, interim national director of transformation, National Transformation Directorate, NHS England; Matthew Taylor, CEO, NHS Confederation; and Daniel Johnston MRes, RN, senior clinical workflow specialist & UK NHS clinical safety officer, Imprivata. Again I'll speak for approximately 15 minutes on the ethical implications of digital transformation and the importance of getting the governance and framing right. The event is in Birmingham, but you can register to attend here (it's free for public sector and NHS staff).
hh
At the Department of Continuing Education in Oxford on the 15th March for the Mst in practical ethics on The Ethics of AI in Healthcare.
At the Big Data Institute in Oxford on 20th March on the topic of ensuring successful implementation of algorithmic clinical decision support software to transform the NHS into a National Learning Healthcare System (NLHS)
Finishing the literature review for the apps paper: If you've read weeknotes before you will know I've been working for some time on a comprehensive paper auditing the quality of evidence supporting claims made by apps on the iOS app store. The paper is nearing its final stages of drafting and I have just been finishing the last bit of reading. I plan to make the reading list public once I'm finished.
Protocol for meta analysis: There has been a lot of discussion recently about polarisation on social media - with there being a particular focus on political polarisation. I want to see if there are identifiable trends across different topics rather than just within topics. So, with this in mind, I have been scoping a protocol for an adapted version of a meta-analysis to test exactly this. Will see how it comes out.
21st century healthcare paper: I have, a couple of times now, given my talk regarding the problems with the current framing of digital health - how I believe it is leading to personalised unwellness via the creation of the inverse data quality law - rather than personalised medicine. I am now working on turning the theoretical part of that talk into a paper that will hopefully be finished relatively soon.
Things I did
Published a BMJ Editorial on why NHS data should not be sold. This is in response to the implication that it should be to fund a tech revolution from the Tony Blair Institute. Read it here.
Presented to the Yale Collaboration for Regulatory Rigor, Integrity, and Transparency (CRRIT) on the challenges of governing and regulating digital health transformation - particularly with the rapid advancements of LLMs, and what I think we should be doing about it.
Wrote an X thread on the great set of recommendations regarding "People, first" Government Digital 2.0 from Rachel Caldicott, Sarah Gold, and Dr. Natalie Byrom to which I contributed. The full set of recommendations can be read here.
Began Recording all lectures. As not everybody can afford to attend conferences (either for financial or time reasons) or has the option/feels comfortable attending lectures at universities, I have begun self-recording all my talks on digital health framing, ethics of AI and implementing AI into the NHS so that they can be available on YouTube and free for anybody who wants to watch them. I'm using a combination of software to make it as accessible as possible - for example, making sure there are always transcripts and captions. I'll share them as and when they are ready.
Scan of clinical trials. In preparation for the talk I gave to the Yale CRITT and in preparation for my next auditing project after the apps store auditing paper is completed, I did a quick scan of the availability of clinical trials results involving AI/ML on clinicaltrials.gov in comparison to the number of AI/ML SaMD products approved by the FDA. These are indicative numbers only, they will need to be checked properly and generated computationally before being cited, but to give an indication of the scale of the problem.
As of 23/03/2023, the FDA had approved 692 AI/ML-enabled SaMD devices
In total, on clinicaltrials.gov there are 336 completed interventional trials involving AI/ML in the intervention itself (i.e. not just in the collection of data or in the analysis of results. OF these approximately 11% have published results.
Within the 'AI' category alone:
There are 159 completed interventional trials
17 of these trials have published results (c.11%)
115 of the 142 unreported trials completed more than 12 months ago (81%)
Things I thought about
Infosphere as a Social Determinant of Health. Last week, some researchers from Google Research published a comment in Nature arguing that Information is a social determinant of health. The comment is here and it makes very similar arguments to those made by myself and colleagues back in 2020 in our paper "Recognizing the Infosphere as a Social Determinant of Health" available here. I've been reflecting on this more about what is the key point to draw out to make it clear to policy makers that in 2024, our informational environments have as much impact on our health as our physical environments. And I think it is about stressing the fact that any information produced by individuals, generated about individuals, consumed by individuals, and recorded about individuals varies enormously in quality depending on a very wide range of sociocultural factors in a way that has a significant impact on health outcomes.
Data interpretability. I think in the context of AI ethics/ digital health ethics, an enormous amount of attention has been paid to data quality, data quantity (i.e. representativeness and more), and model explainability but I think insufficient attention has been paid to data interpretability. There is awareness of issues related to alert fatigue and automation bias, but lack of technical exploration of how to optimise the design of e.g., clinical decision support sustems to support clinician cognitve workflow and to direct attention of the clinician to the most important information in a way that supports their autonomous decision making capacity rather than undermines it or detracts from it. I think to 'solve' this problem would require a multidisciplinary team involving cognitive scientists, human computer interaction designers, software engineers, clinicians, and data ethicists. I hope to make this happen in the near future.
(A selection of things I read)
All from literature alerts this week
Barwise, Amelia K, Susan Curtis, Daniel A Diedrich, and Brian W Pickering. “Using Artificial Intelligence to Promote Equitable Care for Inpatients with Language Barriers and Complex Medical Needs: Clinical Stakeholder Perspectives.” Journal of the American Medical Informatics Association 31, no. 3 (February 16, 2024): 611–21. https://doi.org/10.1093/jamia/ocad224.
Baum, Nancy M., Samantha Iovan, and Marianne Udow-Phillips. “Strengthening Public Health Through Primary Care and Public Health Collaboration: Innovative State Approaches.” Journal of Public Health Management and Practice 30, no. 2 (March 2024): E47–53. https://doi.org/10.1097/PHH.0000000000001860.
Blake, Sarah R, and Neelanjan Das. “Deploying Artificial Intelligence Software in an NHS Trust: A How-to Guide for Clinicians.” British Journal of Radiology 97, no. 1153 (January 23, 2024): 68–72. https://doi.org/10.1093/bjr/tqad043.
Cha, Kenny H. “Software as a Medical Device (SaMD) at the FDA: Regulatory Science and Review.” In 2023 IEEE John Vincent Atanasoff International Symposium on Modern Computing (JVA), 71–74. Chicago, IL, USA: IEEE, 2023. https://doi.org/10.1109/JVA60410.2023.00022.
Chen, John J. “Multicenter Observational Studies: Understanding the Basics of Data Sharing and Data User Agreements.” Journal of Neuro-Ophthalmology 44, no. 1 (March 2024): 1–4. https://doi.org/10.1097/WNO.0000000000002086.
Derraz, Bouchra, Gabriele Breda, Christoph Kaempf, Franziska Baenke, Fabienne Cotte, Kristin Reiche, Ulrike Köhl, Jakob Nikolas Kather, Deborah Eskenazy, and Stephen Gilbert. “New Regulatory Thinking Is Needed for AI-Based Personalised Drug and Cell Therapies in Precision Oncology.” Npj Precision Oncology 8, no. 1 (January 30, 2024): 23. https://doi.org/10.1038/s41698-024-00517-w.
Dobson, Rosie, and Robyn Whittaker. “What Do Health Service Users Think About the Use of Their Data for AI Development?” In Studies in Health Technology and Informatics, edited by Jen Bichel-Findlay, Paula Otero, Philip Scott, and Elaine Huesing. IOS Press, 2024. https://doi.org/10.3233/SHTI231146.
Evans, Ruth P., Louise D. Bryant, Gregor Russell, and Kate Absolom. “Trust and Acceptability of Data-Driven Clinical Recommendations in Everyday Practice: A Scoping Review.” International Journal of Medical Informatics 183 (March 2024): 105342. https://doi.org/10.1016/j.ijmedinf.2024.105342.
Fennelly, Orna, Dearbhla Moroney, Michelle Doyle, Jessica Eustace-Cook, and Mary Hughes. “Key Interoperability Factors for Patient Portals and Electronic Health Records: A Scoping Review.” International Journal of Medical Informatics 183 (March 2024): 105335. https://doi.org/10.1016/j.ijmedinf.2023.105335.
Greener, Ian, and Martin Powell. “Crisis in the UK National Health Service: What Does It Mean, and What Are the Consequences?” Social Policy & Administration 58, no. 2 (March 2024): 313–28. https://doi.org/10.1111/spol.13001.
Heaton, Dan, Elena Nichele, Jérémie Clos, and Joel E Fischer. “Perceptions of the Agency and Responsibility of the NHS COVID-19 App on Twitter: Critical Discourse Analysis.” Journal of Medical Internet Research 26 (February 1, 2024): e50388. https://doi.org/10.2196/50388.
Hill, Scott. “Against the Double Standard Argument in AI Ethics.” Philosophy & Technology 37, no. 1 (March 2024): 23. https://doi.org/10.1007/s13347-024-00720-4.
Hogg, H. D. J., M. Al-Zubaidy, P. A. Keane, G. Hughes, F. R. Beyer, and G. Maniatopoulos. “Evaluating the Translation of Implementation Science to Clinical Artificial Intelligence: A Bibliometric Study of Qualitative Research.” Frontiers in Health Services 3 (July 10, 2023): 1161822. https://doi.org/10.3389/frhs.2023.1161822.
Isasa, Imanol, Mikel Hernandez, Gorka Epelde, Francisco Londoño, Andoni Beristain, Xabat Larrea, Ane Alberdi, Panagiotis Bamidis, and Evdokimos Konstantinidis. “Comparative Assessment of Synthetic Time Series Generation Approaches in Healthcare: Leveraging Patient Metadata for Accurate Data Synthesis.” BMC Medical Informatics and Decision Making 24, no. 1 (January 30, 2024): 27. https://doi.org/10.1186/s12911-024-02427-0.
Laka, Mah, Drew Carter, and Tracy Merlin. “EVALUATING CLINICAL DECISION SUPPORT SOFTWARE (CDSS) – CHALLENGES FOR ROBUST EVIDENCE GENERATION.” International Journal of Technology Assessment in Health Care, February 8, 2024, 1–21. https://doi.org/10.1017/S0266462324000059.
Lyell, David, Adriaan Lustig, Kate Denyer, Satya Vedantam, and Farah Magrabi. “Using Clinical Simulation to Evaluate AI-Enabled Decision Support.” In Studies in Health Technology and Informatics, edited by Jen Bichel-Findlay, Paula Otero, Philip Scott, and Elaine Huesing. IOS Press, 2024. https://doi.org/10.3233/SHTI230975.
Maggio, Lauren A., Karina Villalba, Deborah German, Steven L. Kanter, and Harold R. Collard. “Defining the Learning Health Care System: An International Health System Leadership Perspective.” Academic Medicine 99, no. 2 (February 2024): 215–20. https://doi.org/10.1097/ACM.0000000000005540.
Marco-Ruiz, Luis, Miguel Ángel Tejedor Hernández, Phuong Dinh Ngo, Alexandra Makhlysheva, Therese Olsen Svenning, Kari Dyb, Taridzo Chomutare, Carlos Fernández Llatas, Jorge Muñoz-Gama, and Maryam Tayefi. “A Multinational Study on Artificial Intelligence Adoption: Clinical Implementers’ Perspectives.” International Journal of Medical Informatics 184 (April 2024): 105377. https://doi.org/10.1016/j.ijmedinf.2024.105377.
Nagaratnam, Kiruba, Ain Neuhaus, James H. Briggs, Gary A. Ford, Zoe V. J. Woodhead, Dibyaa Maharjan, and George Harston. “Artificial Intelligence-Based Decision Support Software to Improve the Efficacy of Acute Stroke Pathway in the NHS: An Observational Study.” Frontiers in Neurology 14 (January 18, 2024): 1329643. https://doi.org/10.3389/fneur.2023.1329643.
Neff, Michaela Christina, Jannik Schaaf, Richard Noll, Svea Holtz, Dania Schütze, Susanne Maria Köhler, Beate Müller, Najia Ahmadi, Michael Von Wagner, and Holger Storf. “Initial User-Centred Design of an AI-Based Clinical Decision Support System for Primary Care.” In Studies in Health Technology and Informatics, edited by Jen Bichel-Findlay, Paula Otero, Philip Scott, and Elaine Huesing. IOS Press, 2024. https://doi.org/10.3233/SHTI231125.
Parsons, Matthew, James Ratcliff, Bernadette Egan, Hana Hassanin, and Aftab Ala. “The UK Research Ethics Committee: Making the Case for Better Serving the Underserved – Can We Do Better?” Clinical Medicine 24, no. 1 (January 2024): 100012. https://doi.org/10.1016/j.clinme.2023.100012.
Pavlidis, Georgios. “Unlocking the Black Box: Analysing the EU Artificial Intelligence Act’s Framework for Explainability in AI.” Law, Innovation and Technology, February 9, 2024, 1–16. https://doi.org/10.1080/17579961.2024.2313795.
Pruski, M. “AI-Enhanced Healthcare: Not a New Paradigm for Informed Consent.” Journal of Bioethical Inquiry, February 1, 2024. https://doi.org/10.1007/s11673-023-10320-0.
Rakowski, Roman, and Petra Kowaliková. “The Political and Social Contradictions of the Human and Online Environment in the Context of Artificial Intelligence Applications.” Humanities and Social Sciences Communications 11, no. 1 (February 21, 2024): 289. https://doi.org/10.1057/s41599-024-02725-y.
Saba, Charles Shaaba, and Marinda Pretorius. “The Impact of Artificial Intelligence (AI) Investment on Human Well-Being in G-7 Countries: Does the Moderating Role of Governance Matter?” Sustainable Futures 7 (June 2024): 100156. https://doi.org/10.1016/j.sftr.2024.100156.
Saleh, Mansoor, Karishma Sharma, Aisleen Shamshudin, Innocent Obayo, Stacey Gondi, and Noureen Karimi. “Regulatory Approval of Clinical Trials: Is It Time to Reinvent the Wheel?” BMJ Global Health 9, no. 1 (January 2024): e013727. https://doi.org/10.1136/bmjgh-2023-013727.
Shandhi, Md Mobashir Hasan, Karnika Singh, Natasha Janson, Perisa Ashar, Geetika Singh, Baiying Lu, D. Sunshine Hillygus, Jennifer M. Maddocks, and Jessilyn P. Dunn. “Assessment of Ownership of Smart Devices and the Acceptability of Digital Health Data Sharing.” Npj Digital Medicine 7, no. 1 (February 22, 2024): 44. https://doi.org/10.1038/s41746-024-01030-x.
Toader, Alice-Maria, Marion K. Campbell, Jennifer K. Quint, Michael Robling, Matthew R Sydes, Joanna Thorn, Alexandra Wright-Hughes, et al. “Using Healthcare Systems Data for Outcomes in Clinical Trials: Issues to Consider at the Design Stage.” Trials 25, no. 1 (January 29, 2024): 94. https://doi.org/10.1186/s13063-024-07926-z.
Wang, Yueye, Chi Liu, Wenyi Hu, Lixia Luo, Danli Shi, Jian Zhang, Qiuxia Yin, Lei Zhang, Xiaotong Han, and Mingguang He. “Economic Evaluation for Medical Artificial Intelligence: Accuracy vs. Cost-Effectiveness in a Diabetic Retinopathy Screening Case.” Npj Digital Medicine 7, no. 1 (February 21, 2024): 43. https://doi.org/10.1038/s41746-024-01032-9.