Written by Akanksha Jayanthi (Twitter | Google+) | June 23, 2016
Health IT often finds itself in the crosshairs of stakeholders’ fire. Vendors and developers are creating solutions for providers, but providers either don’t want them or want different versions of them. All the while, the government mandates adoption of certain tools — for better or for worse. What results is a muddled digital health landscape, one that lends itself to poor-quality solutions, which James Madara, MD, executive vice president and CEO of the American Medical Association, said was akin to snake oil remedies of times past.
Dr. Madara made the comparison in his address at the 2016 AMA Annual Meeting. “From ineffective EHRs to an explosion of direct-to-consumer digital health products to apps, some of which are poor quality — this is the digital snake oil of the 21st century,” he said.
But not all stakeholders share Dr. Madara’s criticism of the digital health landscape. John Halamka, MD, CIO of Boston-based BethIsraelDeaconessMedicalCenter, said federal incentives and regulations, though challenging, have largely achieved what they set out to do.
“Rather than suggest that vendors are selling the electronic equivalent of snake oil, the AMA should recognize that the regulatory efforts of the past several years have achieved exactly the result that was intended and that stakeholder organizations such as the AMA should suggest a small number of desirable outcomes as our next goals,” Dr. Halamka wrote on his blog “Life as a Healthcare CIO.”
Here, other healthcare stakeholders submitted comments to Becker’s Hospital Review, weighing in on “digital snake oil,” and whether that’s a fair description of the digital health landscape.
Note: Responses have been lightly edited for length and clarity. They are presented here in alphabetical order.
Ashish Atreja, MD. Chief Technology Innovation and Engagement Officer and Director of the Sinai AppLab at Mount Sinai Health System (New York City): I see Dr. Madara’s statement as a positive call to action for all startups in healthcare that there needs to be proven evidence before digital medicine is widely adopted. While most of the 165,000+ apps in the market today are not validated or “peer reviewed,” there are examples of apps that have gone through rigorous clinical trials, such as Bluestar Diabetes and AliveCor. Instead of labeling all innovation as “snake oil,” I see the challenge and opportunity for all of us scientists and digital medicine evangelists is to combine innovation with the rigor of evidence-based medicine to create a new science of evidence-based digital medicine.
Paul Brient. CEO of PatientKeeper: The physician frustration with healthcare IT that Dr. Madara articulated at the 2016 AMA Annual Meeting is spot on, and the improvements he called for are completely reasonable and appropriate. No other industry on Earth has accepted lower productivity and efficiency as a result of computerization; healthcare shouldn’t either — especially when it comes to physicians, who are healthcare’s most valuable and expensive resource.
The fact is we’re still in version one of automating healthcare, which puts us well behind other industries. Take transportation, for example: Uber has completely changed the paradigm for ordering local transportation. There’s no tipping, there’s no dispatcher; instead there’s an incredibly new experience propelled by the technology that enables it. Today’s EHRs have gotten rid of paper, but that’s just table stakes; EHRs currently are doing the moral equivalent of electronic dispatching of taxi cabs. We have yet to Uber-ify healthcare. That’s where dramatic gains in both physician productivity and quality of patient care will be realized.
Jonathan Chen. Executive Vice President of Medullan: As with any new disruptive technology, there is a slew of innovations in the digital health space today, including thousands of apps that claim efficacy in helping patients manage their health. The number of solutions is overwhelming, with little oversight over their efficacy. Providers, patients and caregivers alike struggle with determining what solutions to use or adopt. What large provider systems and payers want most out of digital health is not a solution, but a process and framework for determining their efficacy. The industry will naturally consolidate, but we need to set the bar higher — we aren’t offering solutions to share breakfast pictures; we’re aiming to improve the health of those with chronic disease! Our solutions need to go through as many regulatory approvals as possible, to build credibility and to separate the snake oil from those that truly advance health outcomes.
Travis Good, MD. CEO of Catalyze, Inc.: Much of the impact of the HITECH has been felt on the front lines of healthcare, between the doctor and the patient. Somewhat ironically, practicing physicians have not had a major voice in digital health or EHR development. Many of the health IT solutions being sold and implemented are driven from outside the care experience. I think this outside disruption has driven the recent pushback on health IT and digital health, and the subsequent comments by Dr. Madara.
Despite this very legitimate sentiment and pushback, healthcare technology, or digital health, is going to be an increasingly large part of healthcare and care delivery today and into the future. We’re now seeing models and evidence of technology that is helping to scale expert clinical opinion and care, as well as delivering on the triple aim of healthcare reform. Health IT is a very broad industry and we need to be careful not to lump all technology as either “good” or “bad.” Holding digital health to more rigor around efficacy and allowing the market to drive direction and adoption, as opposed to more government mandates, will help.
Bryan Haardt. CEO of Decisio Health: The term digital health encompasses hundreds of different types of healthcare technologies and each one is very different from the next. Broadly categorizing them as one and the same does a disservice to promising technologies that are doing things the right way. But Dr. Madera does call out the two qualities that digital health companies should pursue. They should be trusted, and ideally cleared for use by the FDA. And they should be designed hand-in-hand with the clinician to improve workflow and physician-patient interaction. If our technologies are trusted and unapologetically clinically focused, the rosy future for digital health technologies may still come to pass.
Maulik Majmudar. Associate Director of the Healthcare Transformation Lab at Massachusetts General Hospital; Senior Editor of Ranked Health and Hacking Medicine Institute (Boston): I agree that there needs to be reproducible and rigorous scientific validation for today’s vast number of digital health products and services; however, Dr. Madara’s reference to the field of digital health as “snake oil” was naive and imprudent. While more work needs to be done to establish the clinical and economic value of today’s digital health solutions and to drive large scale clinical adoption, there are many examples of mobile apps and digital services (Omada Health, Twine Health, Sleepio, MediSafe, PillPack, and AliveCor; just to name a few) that have already demonstrated significant impact on disease prevention, disease progression, medication adherence and patient engagement.
Chad Meyerhoefer, PhD. Arthur F. Searing Professorship, Associate Professor of Economics at LeHigh University (Bethlehem, Pa.): When one considers EHR technologies, it must be acknowledged that not all components of the EHR (or the systems they interface with) are designed to improve health outcomes. Some components of these systems are designed to increase productivity or lower costs, so it is reasonable to expect that those components will not positively or negatively affect health outcomes.
There is not a significant amount of research of which pieces of clinical data are most important for decision-making. In the past, information about patients was transmitted by paper (e.g. by mailing a patient chart from one location to another). However, not all the information in the chart was useful or necessary to the receiving providers. EHR systems now electronically transmit all of that information and more. However, some of the clinical data highlighted in the EHR might not be the data that is most important to the provider, or which has the greatest effect on healthcare decisions and health outcomes. For example, our research on pregnancy care suggests that the availability of only some types of clinical data through the EHR measurably improves birth outcomes. In order to refine EHR systems to have the greatest clinical impact, more research needs to be conducted on the importance of specific clinical information in different care settings.
Digital technologies may improve outcomes in some care settings but not others. In many cases, how the hospital or provider network is organized, or how it adjusts workforce practices to utilize new technologies, makes a big difference in how effective the technologies are in improving outcomes. Typically, a workforce change must accompany deployment of the technology in order to achieve the desired result.
Adam Powell, PhD. President of Payer+Provider Syndicate: Dr. Madara is correct in suggesting that there is a lot of “digital snake oil” on the market. However, it is important to also recognize that there are numerous people working to address this issue. In 2014, JAMA published a Viewpoint article and a letter on this issue which I co-authored with Drs. Adam Landman and David Bates. We called for the increased review and certification of mobile health apps. We suggested that multiple types of organizations might play a role in the review process, and that it may also be possible to automate some app evaluations.
Fortunately, there are a number of unbiased organizations working to evaluate the quality of digital tools today. PsyberGuide provides reviews of digital tools for mental health. RankedHealth evaluates the clinical relevance, safety and efficacy of mobile health apps. (I am an advisor to both organizations.) Both organizations are nonprofits without ties to any software developers.
Unfortunately, the evaluation process is hard. A study we conducted earlier this year found 21 of the 22 measures frequently used to evaluate the quality of mobile health apps have poor interrater reliability. That is, the reviewers themselves have trouble agreeing on the proper rating. Furthermore, while the review process often focuses on features which can easily be externally observed, reviews are often unable to evaluate how apps impact outcomes, as randomized controlled trials have often not been conducted. PsyberGuide and RankedHealth both examine the academic literature and highlight peer-reviewed findings when they exist. In cases where such findings do not exist, it is difficult to objectively determine the impact of an app. The AMA should encourage the development of review processes for mobile health apps and should emphasize the importance of clinical trials. Apps should be a part of evidence-based medicine.