This appeared at a New England Journal Of Medicine sub-site.
Why Real-World Results Are So Challenging for Digital Health
Article · February 8, 2017
Joseph C. Kvedar, MD & Alexander L. Fogel, MBA
Partners HealthCare
Stanford University School of Medicine
Stanford University School of Medicine
Companies should look for lessons in other industries, create products that are tailored to population subsets, develop synergistic partnerships, and understand the performance characteristics of products in the real world.
Despite years of hype in the field of digital health products — a term that we use here to describe technologies that are designed to have clinical impact on disease — fewer products than expected are being deployed in real-world clinical settings. Many digital health products that demonstrate impressive results in clinical trials often fail to do so in real-world settings.
Why? Much of the success of digital health products is predicated on patient engagement, and clinical trials are among the most engaging environments in health care. Clinical trials involve the use of a variety of tools (e.g., training, close monitoring, payments) to ensure that patients use the technologies appropriately, but few of these tools are used in the real world. In order to cross the chasm from success in clinical trials to success in practice, digital health companies need to focus on patient engagement.
Digital Health Hinges on Engagement
Engagement is so important because many digital health products are designed to achieve behavioral changes for the purpose of preventing or treating chronic diseases. For patients who are at risk for, or are living with, a chronic disease, successful prevention or management requires minute-to-minute, day-to-day changes in decision-making. Patients need to be highly motivated to make behavioral changes, they need to be praised when they follow through, and they need guidance when they slip up. The process needs to be sticky and self-reinforcing in order to maintain patient interest. In a nutshell, patients need to be engaged.
Patients need to be highly motivated to make behavioral changes, they need to be praised when they follow through, and they need guidance when they slip up. The process needs to be sticky and self-reinforcing in order to maintain patient interest. In a nutshell, patients need to be engaged.”
Engagement is important not only for clinical results, but also for the business models of many digital health companies, which frequently incorporate pay-for-performance contracts. Digital health companies need to track data continuously in order to demonstrate their value to the purchaser or user. Products that do not engage patients drag down performance metrics and reduce the viability of the company.
Contrast this with pharmaceuticals: payors pay for drugs before patients take them, and pay-for-performance contracts are the exception; this is part of the reason why nearly 50% of medications for chronic disease are not taken as prescribed.
The Gap Between Clinical Trials and Real-World Results
Successful clinical trials of both digital health products and pharmaceuticals are engaging by definition. In order for a trial to demonstrate significant results — and in order for the results of the trial to be published — the attrition rate for participants needs to be very low. The entire endeavor is designed around ensuring that patients use the product or service appropriately and follow up regularly.
A clinical trial is designed as follows: an artificial scenario is crafted, willing participants volunteer, participants are carefully selected on the basis of optimal criteria, staffers are thoroughly trained, measurements are taken like clockwork, and participants are induced (by means of monetary payments or a barrage of phone calls, emails, and texts) to use the product or service as directed. As a result, the experience of individuals who are involved in a clinical trial typically is much more engaging than that of patients who use the technology in real-world scenarios.
Tools to Drive Engagement
Extending the engagement-boosting practices from a clinical trial to a much larger population of patients in the real world is not feasible because of cost and logistical complexity. As a result, different approaches are needed to facilitate engagement.
(lots omitted) …...
Making the Leap from Novelty to Necessity
Clinical trials are a critical process in the evolution of digital health products. Evidence is an important means for convincing health care executives to buy these products. However, simply learning that an intervention was successful in a clinical study is not enough. Real-world validation is also important as digital health companies own the validation process from product development to scale. Once clinical validation is achieved, real-world evidence must be gathered to justify success at scale. In order for digital health products to make the leap from novelty to necessity, manufacturers will need to focus on driving patient engagement in real-world settings. Engagement will require taking lessons from other industries, creating products that are tailored to population subsets, developing synergistic partnerships, and understanding the performance characteristics of products in the real world.
The full article is here:
What this article says very clearly to me is that digital health initiatives have both to be proven to work in controlled carefully managed trials and then also shown to work in the field before a full blooded roll out is undertaken.
We really do have to stop assuming that so called ‘good ideas’ are guaranteed to work and be useful. We need evidence both in trials and then in the ‘wild’ before rushing in and then being disappointed – slow may this process be!
This has to be one area that needs extreme rigor in evaluation!
Hospitals forging patient experience of the future with voice AI technologies
Commonwealth Care Alliance and Penn Medicine are already putting Amazon Alexa to work in early-stage projects that promise to reshape the patient experience.
By Tom Sullivan
February 13, 2017 07:40 AM
Commonwealth Care Alliance chief of clinical innovation John Loughnane, MD, said the industry is on the verge of voice technologies that can be used to tailor individualized care regimens.
“Alexa, when will you be in my doctor’s office?”
“Silly human — I already am!”
Okay, so Alexa’s response might not be that sassy precisely, but you get the picture. Artificial intelligence voice technologies are on the verge of changing the dynamics of how we interface with machines and, consequently, how humans interact with each other.
Artificial intelligence voice technologies are on the verge of changing — not to be overzealous here — but just about everything.
We’ve all witnessed how the smartphone has had an impact on society and, similarly, voice technologies are poised to effectively and strategically enchant the human psyche to accept them as humanlike companions.
And healthcare is ripe for artificial intelligence — perhaps even more than other industries when it comes to voice-interaction AI — because of the potential for improving care delivery, optimizing processes and improving the patient experience.
“We’re on the cusp of voice systems and voice learning tools that can tailor individualized care,” said John Loughnane, MD, chief of clinical innovation at Commonwealth Care Alliance.
It’s not just CCA either. Penn Medicine is also starting the voice AI trek.
Lots more here:
Things are really moving along at some considerable speed!
David.
Anyone who has to depend on current voice 'recognition' technology as part of their daily work knows that the first of your excerpts here is particularly applicable to the second. VR has been around for a long time now, perhaps this new version will be an improvement.
ReplyDelete