Introduction
I am not sure how it has suddenly become the middle of February, especially since January was about a decade long, but here we are - another week has gone. It's been an eventful week during which I pretended to be Italian on TV and the University closed for a day because of a ferocious (from my inexperienced 'old England' perspective) snow storm, that left New Haven and much of the surrounding area under more than 10 inches of snow.
Now on with the notes.
Things I worked on
App Store Audit. If you've read previous weeknotes you will know that I have been working on a project auditing the quality of the evidence available to support health apps' claims of efficacy. (Spoiler alert: the evidence is poor). This week, after some deliberation, we realised that the quality of evidence supporting apps claiming to have digitised cognitive behavioral therapy is more nuanced than we originally thought. There are some apps which have only interpreted CBT in the lightest possible sense, and still claim to be supported by the evidence that CBT (can sometimes) work in face-to-face interventions. These apps need to be downgraded from low evidence, to very low evidence. On the other hand there are others that have conducted a form of trial testing one specific aspect of CBT operationalised in the relevant app. These apps need to be upgraded from low evidence to medium evidence. Thus, I have been working on this downgrading vs. upgrading process.
AI in Healthcare ROI. Another paper I've mentioned previously that is currently in the works attempts to answer the question "is the current level of investment in AI for healthcare justified based on its ability to meet healthcare system needs?" The paper is in its final stages of drawing so this week I have been working on the narrative flow, in particular removing text that is 'nice to have' but unnecessary to the narrative flow and strengthening the recommendations.
Digital recordings of presentations. Recently, I've had a few people ask me if the talks I give and mention in weeknotes are ever recorded. The truth is this is very hit and miss, sometimes yes but only for internal purposes, more often no. Very occasionally things are recorded but if it's a recording of a live talk the quality can sometimes not be great. SO, I have decided to rectify this by embarking on a mission to self-record as many of my talks as possible (in their long and most generalisable form) and make them publicly available. Because I've talked about it a fair few times on Weeknotes and elsewhere, I thought I'd start with my presentation "From personalised unwellness to algorithmically enhanced public health" which provides the outline of my doctoral programme of research. I will make a separate blog post about this, but for now you can watch the video below and the slides are available here. https://youtu.be/40EfNbiBPNk?feature=shared
Editorial. I said in last week's notes I had been thinking about the recommendation from Tony Blair and William Hague to sell NHS data and that more would be coming soon. Well, this week - with some help from colleagues - I wrote an editorial on the topic which should publish soon.
BMJ Future of the NHS Commission. The Future of the NHS Commission that I have been involved in since last June, and for which I co-authored the first paper on the NHS's values which published at the end of Jan, is nearing its final stages so I have been involved in revising the other paper for which I am a co-author. Mostly this involves me trying to say vaguely sensible things about digital/data/tech whilst other far wiser colleagues with infinitely more clinical experience talk about the things that actually matter like workforce and money.
Things I did
Released a new podcast episode with Digital Health Unplugged ahead of me giving a keynote at the Digital Health Rewired next month.
Appeared on SkyItalia24 (scroll for the "informazione" episode at the bottom of the page) with Luciano (who is the host of the series) talking about misinformation. Watching myself dubbed in Italian is quite possibly one of the most entertaining experiences of all time.
Presented my postdoctoral research programme (as in the recording above) to some colleagues at the Med School. Obviously it was a little more formal than the version I've made available online, but it was still very fun and I'm hopeful that this will lead to collaborations between the two groups.
Pre-printed this paper on Regulation by Design with some colleagues from the Digital Ethics Center.
Gave two (virtual) presentations to the Human Dimension of technology workshop which is part organised by Health Education England. The first talk was on the ethics of AI in theory and the second on the practical operationalization of AI.
Things I thought about
The continuing importance (and failure) of diverse representation in technology policy. This week the Alan Turing Institute, the UK's National Institute for AI, announced that it had appointed four new directors of Science and Innovation to its senior leadership team, to help ensure the ATI uses data science and AI for social good. All four are men. All four are based primarily in London. There is no way that four men from the richest city in the country, all working within elite academic institutions, can represent society and therefore ensure ATI uses data science and AI for social good. This was an opportunity for the ATI to overcome the failings of the Government in, for example, its organisation of the AI Safety Summit which had limited representation from civil society and non-diverse representation at that, and it spectacularly failed. The most disappointing thing is that it is not hard to find diverse talent in the field: one simply has to make the effort to look.
The changing meaning of transhumanism. Transhumanism is defined as the belief or theory that the human race can evolve beyond its current physical and mental limitations, especially by means of science and technology. Traditionally, investigations into this phenomena have focused on additions and enhancements to the human body - in a 'bionic' sense. However, I think that this is now changing. Luciano writes about the 'cleaving power of the digital' i.e., the power of digital technologies to uncouple things that have traditionally been stuck together - space, and time, for example. Increasingly I think AI is decoupling technologies of the self from the physical body and this is moving transhumanism away from a focus on additions and enhancements and towards subtractions and outsourcing. Knowledge about the body is being outsourced to apps, wearables, and self-tracking algorithms. Chatbots exist to allow us to 'live on' after our mortal death. If you can't find a romantic partner IRL, then you can get an AI version. On and on we are not just transcending the current physical and mental limitations, but transcending their boundaries entirely, and I am not entirely sure what the implications of this are. I have personally always wanted to exist only as disembodied thought - but I get that this is probably not everybody's dream reality.
Data Interpretability. Whilst there has been significant attention paid to the need to ensure AI is trained on data of sufficient quantity and quality, far less attention has been to ensure that the data generated and produced by AI is interpretable. There is of course the XAI movement which develops technical solutions to ensuring AI is Explainable but this is not the same thing. Interpretability is far broader than explainability, it has to do with how information influences human decision making not whether the AI's decision-making process is explainable. I think in part this epistemic or heuristic perspective on interpretability has received less attention because the XAI community is primarily composed of data scientists not cognitive scientists and human computer interaction designers. I think there is a need to create a research programme bringing together cognitive scientists, data ethicists, and human computer interaction designers, to develop best practice for data interpretability from high-risk AI systems including - but not limited to - medicine.
Ethnography of Electronic Health Records. In prepping for my presentation to the Med Schoo, I revisited the paper my colleague Joe and I (along with others) published on NHS data flows. Whilst we did mention in the paper that the lack of transparency is likely to hinder public trust in NHS data management, we did not talk about how difficult it is to communicate the complexity of the data flows to the public in a way that is meaningful. Many valid and valiant attempts have been made, but there is a tendency to gloss over the number of journeys a patient record, or elements of a patient record e.g., a diagnosis, may make - bouncing from source to source. The whole concept is exceedingly abstract and this abstraction I think exacerbates the trust problem. I also think that policymakers continue to underestimate the importance of sorting out the mess of data flows and giving patients more meaningful control over where their data is going and what it is being used for. I would therefore, really like someone (maybe me, maybe someone else) to do an ethnographic study documenting the process of trying to find out what had happened to their medical, or GP record - the levers that needed to be pulled, the legal mechanisms that had to be relied upon, how long it too, and what the experience felt like. I think that the existence of such a study would not only help provide patients and publics with something more concrete and less abstract, but also demonstrate to policymakers the importance of fixing the issue.
(A selection of) Things I read All about the apps this week
Ahn, David T. “Benefits and Risks of Apps for Patients.” Current Opinion in Endocrinology, Diabetes, and Obesity 29, no. 1 (February 1, 2022): 17–22. https://doi.org/10.1097/MED.0000000000000701.
Bernstein, Emily E., Hilary Weingarden, Emma C. Wolfe, Margaret D. Hall, Ivar Snorrason, and Sabine Wilhelm. “Human Support in App-Based Cognitive Behavioral Therapies for Emotional Disorders: Scoping Review.” Journal of Medical Internet Research 24, no. 4 (April 8, 2022): e33307. https://doi.org/10.2196/33307.
Bhatt, Paras, Jia Liu, Yanmin Gong, Jing Wang, and Yuanxiong Guo. “Emerging Artificial Intelligence-Empowered mHealth: Scoping Review.” JMIR mHealth and uHealth 10, no. 6 (June 9, 2022): e35053. https://doi.org/10.2196/35053.
Cruz-Cobo, Celia, María Ángeles Bernal-Jiménez, Rafael Vázquez-García, and María José Santi-Cano. “Effectiveness of mHealth Interventions in the Control of Lifestyle and Cardiovascular Risk Factors in Patients After a Coronary Event: Systematic Review and Meta-Analysis.” JMIR mHealth and uHealth 10, no. 12 (December 2, 2022): e39593. https://doi.org/10.2196/39593.
Ford, Emmalee A., Alexandra E. Peters, Shaun D. Roman, Eileen A. McLaughlin, Emma L. Beckett, and Jessie M. Sutherland. “A Scoping Review of the Information Provided by Fertility Smartphone Applications.” Human Fertility (Cambridge, England) 25, no. 4 (October 2022): 625–39. https://doi.org/10.1080/14647273.2021.1871784.
Hatem, Sarah, Janet C. Long, Stephanie Best, Zoe Fehlberg, Bróna Nic Giolla Easpaig, and Jeffrey Braithwaite. “Mobile Apps for People With Rare Diseases: Review and Quality Assessment Using Mobile App Rating Scale.” Journal of Medical Internet Research 24, no. 7 (July 26, 2022): e36691. https://doi.org/10.2196/36691.
Indraratna, Praveen, Uzzal Biswas, James McVeigh, Andrew Mamo, Joseph Magdy, Dominic Vickers, Elaine Watkins, et al. “A Smartphone-Based Model of Care to Support Patients With Cardiac Disease Transitioning From Hospital to the Community (TeleClinical Care): Pilot Randomized Controlled Trial.” JMIR mHealth and uHealth 10, no. 2 (February 28, 2022): e32554. https://doi.org/10.2196/32554.
Jiang, Shaojun, Meina Lv, Tingting Wu, Wenjun Chen, and Jinhua Zhang. “A Smartphone Application for Remote Adjustment of Warfarin Dose: Development and Usability Study.” Applied Nursing Research : ANR 63 (February 2022): 151521. https://doi.org/10.1016/j.apnr.2021.151521.
Kim, Sun Kyung, Mihyun Lee, Hyun Jeong, and Young Mi Jang. “Effectiveness of Mobile Applications for Patients with Severe Mental Illness: A Meta-Analysis of Randomized Controlled Trials.” Japan Journal of Nursing Science : JJNS19, no. 3 (July 2022): e12476. https://doi.org/10.1111/jjns.12476.
Martinengo, Laura, Anne-Claire Stona, Lorainne Tudor Car, Jimmy Lee, Konstadina Griva, and Josip Car. “Education on Depression in Mental Health Apps: Systematic Assessment of Characteristics and Adherence to Evidence-Based Guidelines.” Journal of Medical Internet Research 24, no. 3 (March 9, 2022): e28942. https://doi.org/10.2196/28942.
Mehraeen, Esmaeil, SeyedAhmad SeyedAlinaghi, Zahra Pashaei, Pegah Mirzapour, Alireza Barzegary, Farzin Vahedi, Kowsar Qaderi, et al. “Mobile Applications in HIV Self-Management: A Systematic Review of Scientific Literature.” AIDS Reviews 24, no. 1 (March 1, 2022): 24–31. https://doi.org/10.24875/AIDSRev.21000025.
Nogueira, Francisco M. F., Ricardo P. Martins, Ellen C. H. Pereira Nery, and Anabela G. Silva. “A Systematic Search and Assessment of the Quality and Characterisation of Free Mobile Applications Targeting Knee Pain.” Musculoskeletal Care 21, no. 1 (March 2023): 212–20. https://doi.org/10.1002/msc.1688.
Ouellette, Samantha, and Babar K. Rao. “Usefulness of Smartphones in Dermatology: A US-Based Review.” International Journal of Environmental Research and Public Health 19, no. 6 (March 17, 2022). https://doi.org/10.3390/ijerph19063553.
Ouwerkerk, Maaike, Isaline C. J. M. Eijssen, Meike M. W. van der Linden, Inez M. Wijnands, Frank J. G. Dorssers, Marc B. Rietberg, Heleen Beckerman, and Vincent de Groot. “A Smartphone Application to Assess Real-Time and Individual-Specific Societal Participation: A Development and Usability Study.” Archives of Physical Medicine and Rehabilitation 103, no. 10 (October 2022): 1958–66. https://doi.org/10.1016/j.apmr.2022.01.168.
Philip, Ben Joseph, Mohamed Abdelrazek, Alessio Bonti, Scott Barnett, and John Grundy. “Data Collection Mechanisms in Health and Wellness Apps: Review and Analysis.” JMIR mHealth and uHealth 10, no. 3 (March 9, 2022): e30468. https://doi.org/10.2196/30468.
Porras-Segovia, Alejandro, Isaac Díaz-Oliván, Maria Luisa Barrigón, Manon Moreno, Antonio Artés-Rodríguez, María Mercedes Pérez-Rodríguez, and Enrique Baca-García. “Real-World Feasibility and Acceptability of Real-Time Suicide Risk Monitoring via Smartphones: A 6-Month Follow-up Cohort.” Journal of Psychiatric Research 149 (May 2022): 145–54. https://doi.org/10.1016/j.jpsychires.2022.02.026.
Tarp, Kristine, Trine Theresa Holmberg, Anne Marie Moeller, and Mia Beck Lichtenstein. “Patient and Therapist Experiences of Using a Smartphone Application Monitoring Anxiety Symptoms.” International Journal of Qualitative Studies on Health and Well-Being 17, no. 1 (December 2022): 2044981. https://doi.org/10.1080/17482631.2022.2044981.
Taylor, Greenberry, Carma L. Bylund, Amanda Kastrinos, Jordan M. Alpert, Ana Puig, Joanna M. T. Krajewski, Bhakti Sharma, and Carla L. Fisher. “Practicing Mindfulness through mHealth Applications: Emerging Adults’ Health-Enhancing and Inhibiting Experiences.” International Journal of Environmental Research and Public Health 19, no. 5 (February 24, 2022). https://doi.org/10.3390/ijerph19052619.
Tripathi, Susmit, Ashwin Malhotra, Murtaza Qazi, Jingyuan Chou, Fei Wang, Samantha Barkan, Natalie Hellmers, Claire Henchcliffe, and Harini Sarva. “Clinical Review of Smartphone Applications in Parkinson’s Disease.” The Neurologist 27, no. 4 (July 1, 2022): 183–93. https://doi.org/10.1097/NRL.0000000000000413.
Zhang, Meng, Wei Wang, Mingye Li, Haomin Sheng, and Yifei Zhai. “Efficacy of Mobile Health Applications to Improve Physical Activity and Sedentary Behavior: A Systematic Review and Meta-Analysis for Physically Inactive Individuals.” International Journal of Environmental Research and Public Health 19, no. 8 (April 18, 2022). https://doi.org/10.3390/ijerph19084905.