Bridging the 'Digital Divide' in Diabetes Technology
In this video, David Kerr, MD, MBChB, DM, FRCP, FRCPE, addresses the challenges of using consumer health technologies, such as smartphone apps and smartwatches, to manage diabetes for underserved populations. He explores the role of AI and continuous glucose monitoring in transforming care and stresses the importance of improving digital health literacy and making FDA-approved technologies more accessible. Dr. Kerr also advocates for a greater focus on lifestyle interventions and collaboration with underrepresented communities to create more equitable diabetes care solutions.
Additional Resources:
- Ebekozien O, Fantastia K, Farrokhi F, et al. Technology and health inequities in diabetes care: How do we widen access to underserved populations and utilize technology to improve outcomes for all? Diabetes Obes Metab. 2024;26 Suppl 1(Suppl 1):3-13. doi:10.1111/dom.15470.
- Kerr D, Glantz N. Diabetes, like COVID-19, is a wicked problem. Lancet Diabetes Endocrinol. 2020; 8(11): 873–874. doi:10.1016/S2213-8587(20)30312-0.
TRANSCRIPTION:
Consultant360: You mention that many digital tools are available, but some are more accessible than others. How can clinicians optimize the use of low-cost or widely available consumer health technologies, such as smartphone apps or smartwatches, to benefit diabetes management for patients who cannot afford advanced, FDA-approved systems?
David Kerr, MD, MBChB, DM, FRCP, FRCPE: Thank you. There are already huge number and exponentially growing number of devices available which claim to help in the management of diabetes. Some of these claims are dubious, lacking evidence. Some of these devices do the opposite and actually increase the burden of diabetes. The fundamental problem is that diabetes is inequitable recently we published an article that called it a “wicked problem” in that there are no purely scientific solutions to diabetes and if one expects that technology will simply be able to overcome all the existing barriers including those impacting underserved and underrepresented communities, then I think that's a pipe dream.
That's a long-winded way of saying we have a long way to go and there's lots to be done. Ideally, of course, if FDA -approved technologies were more accessible, meaning they were cheaper, that would go a long way to improving outcomes and making health outcomes more equitable. The best example of that of course is continuous glucose monitoring which has really revolutionized diabetes care and is now hopefully, because I'm a believer, going to expand into the other populations of people who are at risk of diabetes or of prediabetes or simply want to maintain metabolic wellness.
So it's actually not just up to the clinician. We need help from the creators, the developers, the funders, the regulators, the clinical trialists. But we also need help from people with diabetes. And they need to be involved really at the very early stage in the development of new technologies across the whole spectrum of digital health. Because without representation by those currently underrepresented groups, we're just going to widen the already existing digital divide.
C360: Are there any emerging trends or innovations in digital health, such as AI or machine learning, that you believe hold promise for overcoming these barriers in underserved populations? How can clinicians actively participate in leveraging these innovations to improve patient care?
Dr Kerr: There are existing technologies which are being expanded into other populations. I already mentioned that continuous glucose monitoring, which has the ability, if we apply high levels of computer science, digital science to this, to begin to tease out metrics which allow stratification of people into different risk groups. Once you just get people in different risk groups you can then stratify them into different treatment groups and then you get into where we really want to be which is the more personalization of medical care and for diabetes in particular.
The other thing that the technologies are going to do is to create a renaissance for lifestyle interventions. Currently, the focus has always been and continues to be drug A versus drug B, combination of drug A and drug B, et cetera, et cetera, without paying enough attention to lifestyle interventions. The promise of wearable devices, step counters, sleep counters, stress counters, glucose counters, ketone counters, lactate counters, is that you can start to understand human behavior and when there are better opportunities for lifestyle interventions and for these interventions to be more successful and for the success to be sustained. The jargon is “just-in-time adaptive intervention,” but what this really means is that if we improve access to wearable devices that allow people to monitor their physiology and we provide platforms that facilitate understanding and thereby creating actionable information, we could completely revolutionize healthcare in the United States and globally and reduce the reliance and the dependence on pharmaceuticals. I'm not saying pharmaceuticals are bad, I'm just saying that they are the only game in time at the moment. We've really missed an opportunity thus far to offer more personal lifestyle interventions that are going to make a difference for these populations. Risk reduction through lifestyle will be facilitated by wearable technologies going forward.
From an AI perspective, the sky's the limit here. With AI at the moment, most of the emphasis is on prediction, predicting who is going to do well, who is not going to do well, who is doing well but may do less well and so forth. And that's great. The other advantage of AI is a consideration of the use of language as a therapy. I'm very impressed with what's already happening with the ability of AI to reduce the burden for clinicians by capturing conversation, dialogue, words and their meaning between conversations with a clinician and a patient -the human-to-human interaction. We're starting to see AI systems that can capture that dialogue and populate the electronic health record and perhaps do away with the electronic health record in its current form. That will free up time for clinicians to do what they've they've always been good at - being a clinician.
And the other advantage of AI is that it's already showing that it can be empathetic. And empathy is very therapeutic in healthcare setting. I believe there's lots of opportunities. for AI There are lots of warnings related to artificial intelligence about what are the sources of data, what's the representation, what's the risk of hallucinations, will there be an impact on the doctor -patient relationship which could be negative, will we be dependent on machines and therefore free thinking will disappear. But I think all of those pale into the background when you consider the opportunities that AI machine learning can provide in diabetes care.
C360: Based on your research, what specific insights have you gained about the barriers that clinicians face when trying to implement digital health technologies in underserved populations? How can clinicians better understand and address these barriers during patient interactions to enhance technology adoption?
Dr Kerr: For the clinician, there needs to be better education and training for clinicians, meaning new clinicians, as well as existing clinicians about the interaction between a human machine to provide more insight into the social and digital determinants that influence behavior decision making by people with chronic disease such as diabetes. That is a skill set which is not taught very well at the moment. As I've alluded to earlier, you know, trying to understand where the person sitting in front of you is at a moment in their life where the application of a digital health technology will have benefit or whether that application will be at some time in the future. The other question is does the application have to be continuous or should we be thinking more about digital health solutions, solving particular problems rather than the expectation that they're there forever. In general, people do not like to have to use them forever. There's a lot of new learnings that need to take place and there's an opportunity for people who are interested in clinician education, medical education, to actually have a rethink about the curricula and to have much more in teaching around this human-machine interaction.
C360: How should clinicians measure the success of interventions aimed at increasing access and usage of these technologies? How can clinicians balance the need for immediate clinical outcomes with the longer-term goal of improving technology adoption among their patients?
Dr Kerr: Well, it's actually very simple. It's also very difficult. The clinician needs to ask three fundamental questions. Who is the target for the prescription of a digital health intervention? And that's down to the granular, down to the individual level.
The second question is, and clinicians sometimes are not so good at this, what does success look like? What are the metrics of success? Are we going to remain hung up about surrogate measures? Or are we going to look at multiple measures at the same time from the perspective of the person with a chronic disease? So what are the metrics of success?
And the third question is, who's going to pay for this? Now, I don't just mean dollar payment and that's very important particularly in the United States which has a very different system to where I come from— the United Kingdom. But also payment in the form of burden. Burden is a currency in diabetes care. So how much burden will it cost the clinician, the person with diabetes, the system, if they prescribe this particular digital health technology?
So if you can answer those three questions— who's the target, what are the metrics of success, who's gonna pay for it— then you're well on the way to really understanding the potential value of digital health for diabetes care.
C360: Challenges related to digital literacy can limit the effectiveness of diabetes technologies like CGMs and insulin pumps. Are there specific interventions you would recommend to help clinicians bridge digital literacy gaps in these patients, and encourage long-term use of diabetes technologies?
Dr Kerr: Yes, the first, most important thing is don't blame the patient, okay? We are creating UI, user interface, user experiences that have not been thought through as to the target audience. A really useful tool is a readability engine. I you're going to give people information and you write that content or you create that content, test the level of understanding that's required to gain the maximum return from that content, that device, that technology. We're spending too much time and too much effort creating technologies that at the human-machine interaction are just not understandable or are not convertible into actionable information. So be careful about the language, the words, the content, the meaning. And if we start thinking about that, we will reduce a major barrier to digital health equity.