This is Part 2 in a four-part Q&A Series on the future of digital health. You can read Part 1, Mobile Endpoints, here. And if you’d like to contribute to any of these discussions, we’d love to hear from you in the comments section.
Who will lead the charge to bring patients’ data online?
In 2009, Meaningful Use was enacted as part of the American Reinvestment & Recovery Act, effectively requiring all healthcare providers to use electronic health record (EHR) technology.
Since then, more and more health records are now available in digital formats, and regulators are increasingly incentivizing interoperable standards for health data, such as the FHIR protocol (which follows the middling C32 and the CCDA standards). Yet the reality is that “standardization” is difficult to achieve. Just think about how hard it is to get a standard protocol in the consumer technology world — and that market moves a lot faster than healthcare.
There have been some substantial efforts via the CommonWell Alliance among private EHRs to ameliorate the situation. However, the absence of Epic, the largest EHR in the U.S., precludes any notion of consensus. In fact, the spat between CommonWell partners and Epic have surfaced criticisms of Epic’s perceived closed nature and relunctance to get on board with interoperabiity, while others believe the Alliance may be purposely excluding Epic as a means to compete with the dominant vendor.
Politics aside, Epic and other legacy EHRs are profitable selling traditional IT solutions that don’t live in the cloud or adhere to modern, developer-friendly standards. We’d like to see cloud-based models like Practice Fusion’s come to overtake the legacy EHR, IT model, but are also aware of the concerns around privacy and data ownership (see below). Tyler Hayes of Prime has an excellent summary of how fragmented and legacy-riddled this market is today.
Of course, the ability to share health data across networks is one thing; giving consumers a user-friendly solution for storing and managing their data is another altogether. Imagine a sick patient whose health records are spread across 10 different EHRs. How does a doctor know where to look for a patient’s records? And if the patient hasn't explicitly taken control of his or her own data, how does s/he know where to look? By digging through his/her inbox to find post-visit emails from various doctors?
The solution may lie with patients owning their own database of personal health data and then distributing it to doctors as needed. We have already seen some tools to facilitate this, such as consumer-facing EHR-aggregation apps like Prime and Picnic Health. Such apps have proven particularly useful for chronically ill patients who see multiple doctors on a regular basis.
Yet while chronically ill patients have been the early adopters for these apps, healthy people just don’t think about their health too often and thus feel much less urgency to aggregate their medical records.
So if the vast majority of patients are poorly motivated to take control of their own health data, then doctors and/or their IT staff may need to be the driving force in getting EHRs online and encouraging patients to aggregate their health data. The benefits for them are clear: lower costs, greater convenience, and potentially better care.
If we had all that data, what could we do with it?
Modern, efficient data management not only benefits doctors and patients in overseeing individual cases, but also promises a wealth of data that can be anonymized and aggregated for analysis and diagnosis. In a similar vein, Apple’s ResearchKit is a positive step toward generating valuable research data, though its adoption remains uncertain.
Consider how RxRevu aggregates and learns from the history of medical prescriptions and subsequent outcomes, enabling a doctor to choose a medication that is right for a person based on one’s age, gender, physique, and medical history. This is a vast improvement over the current “This is the medication I learned about in medical school” or “This is what’s on the pad of Post It’s that the sales rep gave me” mindset that guides how doctors prescribe medicine today. Pharmaceutical companies are also beginning to leverage 3D printing and personal data to deliver customized doses of medicine, rather than the traditional one-size fits all approach.
There are tradeoffs, of course: while we lament all the personal data we give to, say, Google and Facebook, we also derive a tremendous benefit from their intelligent services. We won’t take a position on the sharing of health data but hope that the debate we’re having now and the lessons we’ve learned regarding data ownership and privacy in general will inform EHRs’ policies and architecture, particularly given the extraordinary security and privacy concerns surrounding personal health data.
The U.S. electronic medical records hacked vs patient accessed ratio is > 5:1 pic.twitter.com/bM8zg5afRC— Eric Topol (@EricTopol) July 4, 2015
As an illustration, you might call Practice Fusion is the “poster child” of free EHRs: physicians can use the product for free as the company’s generates revenue from pharmaceutical and healthcare companies who pay to advertise (to doctors) on their user interface. Another obvious model in this space is to connect end users to clinical trials. These business strategies are not uncommon as we’ve even encountered startups with very modest userbases — a community 10,000 with a particular illness is enough for a clinical trial — selling anonymized user data. We’d encourage every healthcare startup to exercise robust caution when pursuing this model.
Moreover, search for “Practice Fusion” in this post for a taste of how asymmetric the relationship can be in terms of the platform owners’ access to personal data, compared to the patients’ themselves. If consumer technology in healthcare lags behind consumer technology in social media on the order of years, we’d hate to think that policy and data ownership matters will as well. (Please note: We’re not endorsing the comments here and do not intend to single out Practice Fusion)
How many consumers will actively monitor their risks and conditions?
Despite the growing number of consumer fitness devices, mobile health-monitoring technology, and even genomics services, it’s not yet clear to what extent people will start proactively monitoring their health on a regular basis.
Cost is one consideration. Genetic kits like 23andMe serve a niche audience today, but if such services were covered in full or maybe even incentivized on every insurance plan (as they may soon be), how many more people would opt in? Perhaps the cost of sequencing a genome will continue to fall, such that genomic testing becomes a no-brainer for the provider, or an easy choice for the consumer to pay for out of pocket.
A change in health status can be a strong impetus to get people monitoring their health. For instance, say that you use your Apple Watch to monitor your heart rate during exercise. One day you notice you have an irregular heart rate, as this San Diego man did before checking himself into the hospital. Chances are that you’ll start monitoring your heart for the rest of your life, and perhaps upgrade to a premium heart rate monitor. It’s also possible that health monitors will be prescribed by doctors for diagnostic reasons, and perhaps by insurance companies for risk assessment.
In addition to physical devices and trackers, there are also apps that rely purely on users to manually enter their data. Rise and TwoGrand users log their meals and exercise. Women are manually entering their health and behavioral data in fertility and period tracking apps like Glow, Ovuline, and Clue. You can imagine how valuable this data could be to a physician, fertility doctor, or gynecologist.
With all that data in consumers’ hands, however, we're wary of a future where patients push physicians on data that they have collected because of the time and money spent in doing so. There have already been cases where patients have mistakenly relied on false positives or false negatives generated by apps that have no clinical merit. One study was unsurprisingly underwhelmed by the 39 melanoma detection apps on the App Store, some of which date back to 2009, when mobile cameras still had trouble detecting faces let alone cancer. The FDA has since cracked down on unfounded melanoma detection apps.
Or maybe we’re overestimating how much data consumers will want to access. Maybe this category will end up looking a lot like the productivity market today: A small group of people who are really keen on preventative health (as there are a subset people who love their calendar app), while most people will just work with whatever’s handed to them.
Stay tuned for Part III of our series “On Digital Healthcare”. You can get notified via email when Part II is out.