- A new JAMA article (“Is There an App to Solve App Overload?”) summarizes the current state of medical apps (a “wild, wild West”), shares ongoing challenges, and highlights WellDoc’s BlueStar as a notable exception.
- About 2,000 patients are using the BlueStar app in five states (MD, VA, WV, PA, DC, and NJ) as part of WellDoc’s regional rollout. The product has seen over 50% sequential growth in prescriptions since launching in 3Q14, with 60% of called-on doctors prescribing.
Last week, JAMA published an excellent perspective article entitled, “Is There an App to Solve App Overload?” Author Bridget Kuehn summarizes the current state of medical apps, which is pretty much captured in this remark from U Michigan’s Dr. Joyce Lee: “For right now, it is still kind of a wild, wild West.” The well-written piece points out numerous challenges in the exploding medical app space (40,000+ available for HCPs and patients), including:
- Lack of clinical studies and published data on apps’ effectiveness;
- Little guidance for HCPs or consumers in terms of what apps are useful;
- Low confidence in the security and privacy of patient data; and
- Little clarity on how apps will work in the context of clinical care (especially notoriously difficult-to-use EMRs).
Notably, the article begins and ends with WellDoc’s BlueStar for type 2 diabetes, which is specifically called “an exception” for a few key reasons: prescription-only; published RCTs; and FDA clearance. [To that we would add (i) a real business model, whereby BlueStar is reimbursed like other prescribed diabetes products, like test strips and insulin pens; (ii) a focus on making providers’ lives easier; and (iii) real-time, contextual, personalized feedback and education that is useful for patients.] The JAMA article says that “about 2,000 patients” use BlueStar as part of WellDoc’s initial regional rollout (MD, VA, WV, PA, DC, and NJ). This is the first user number we can recall hearing. Back in June 2013, WellDoc announced that BlueStar would launch later that year, so the number is perhaps lower than we expected. However, in speaking to WellDoc, we learned that the regional launch was significantly delayed by about a year for three main reasons: (i) the timing of the Series A closing; (ii) the hiring of the initial sales team; and (iii) setting up operations, such as Customer Care and Pharmacy. Since the regional launch began in 3Q14, over 60% of called on doctors have prescribed BlueStar (we’re not sure of the base or type of doctors, but assume it is an early adopter population at this stage), and prescriptions have grown greater than 50% every quarter.
Ms. Kuehn notes that 60-69 year-old users have been among BlueStar’s most prolific users in the regional rollout. Certainly, that’s a testament to the effort WellDoc has put into making BlueStar user-friendly and device-agnostic (smartphone, tablet, computer). Overall, we see these as encouraging metrics in the early days of the launch, and based on the initial results, WellDoc is expanding commercial efforts and has initiated a Series B financing round to grow the sales team. Senior management has also changed to reflect the commercial effort with a new CEO (Kevin McRaith) and additions to the leadership team with extensive healthcare backgrounds. Mr. McRaith previously worked in Sales/Marketing at Human Genome Sciences, Genentech, Abbott, and Baxter.
- We are glad to see JAMA digging into digital health with this piece, and it’s excellent to see diabetes and WellDoc featured as a keystone example. We continue to be optimistic on the potential of this area of medicine (something CES and SXSW made loud and clear this year), though the field is still in its infancy, and certainly, is not a silver bullet to tackle all the challenges we are facing in diabetes.
- The JAMA piece quotes FDA’s Associate Director for Digital Health, Bakul Patel, who largely reiterates his comments from a terrific Bloomberg interview a few weeks ago – the FDA has adopted a risk-based approach to regulating medical apps. Apps that behave as medical devices (electrocardiogram machines, like AliveCor) and/or mitigate a disease (e.g., BlueStar) will continue to require premarketing approval. Apps that interact with medical devices, such as glucose meters, fall into a “moderate-risk category” requiring less premarket oversight than high-risk medical devices. For the vast majority of apps that primarily track information or provide access to health information, the FDA will not require premarket review (simply voluntary registration and adverse event reporting). As we noted in our coverage of the Bloomberg interview, we believe the FDA is trending in the right direction on fostering innovation and encouraging more open data, though there are still some areas of regulatory uncertainty (e.g., when does a device cross from a motivational wellness application to a regulated medical device?). Our full coverage of that interview shares some of the key regulatory questions in digital health.
- It was encouraging to see Mr. Patel’s two positive quotes in the JAMA piece: “We want patient engagement and tech to prosper” and “We are focusing on the higher end of the risk spectrum.” Of course, there’s also a practical element here, as the Agency’s limited resources preclude tight regulation of the multi-billion dollar wellness and digital health device industries. Indeed, the JAMA article notes that the FDA may have to rely on patients, physicians, and whistleblowers to call out false marketing claims or apps that don’t function as intended.
- Interestingly, the 21st Century Cures Act in its current form would strip the FDA of its ability to provide oversight of medical app software. Apparently, some industry advocates have argued that “even limited oversight of apps will stifle innovation.” There is a challenging balance between safety and innovation, and we believe the most high-risk devices should need to go through the FDA; the Agency’s device side has impressed us in recent months with its more forward thinking approach, and we believe the regulatory barriers in digital health are lower than they once were.
- We last wrote about this intriguing legislation in February, when the US House of Representatives Energy and Commerce Committee released an initial 393-page discussion document to spur dialogue following nearly a year of Congressional input-gathering. JDRF’s response to the discussion document is on page 72 here; it is very supportive overall and suggests more specifics in a few areas. As we noted in February, the schedule aims to introduce 21st Century Cures legislation and ultimately send a bill to President Obama’s desk for signature by the end of this year.
- “Medical app development is outpacing the vetting process...The literature is not keeping up with the products on the market.” This is unquestionably true, but also reveals a paradox. Digital health is moving quickly as technology develops at a rapid rate, but carrying out studies take a long time. How do we marry the speed of innovation with the need to gather data? Even WellDoc’s widely praised one-year RCT began in 2008 and wasn’t published until 2011.
- According to an IMS Institute for Healthcare Informatics report, more than 40,000 health-related apps were available for patients and physicians in 2013. An estimated 660 million downloads of health-related apps occurred in 2013 alone. The report also performed an assessment of consumer healthcare app functionality: Inform (10,840 apps); instruct (5,823); record (5,095); display (2,302); guide (1,434); remind/alert (1,357); communicate (395). Notably, there was significant skew: more than 50% of available apps achieved fewer than 500 downloads, while five apps accounted for 15% of all downloads in the healthcare category. We wonder how this report has changed since 2013.
- “Apps that haven’t been properly validated may undermine care in more subtle ways.” The JAMA piece notes a 2014 analysis (Kumar et al., J Am Soc Hypertens 2014) of the top 107 apps for hypertension, 14% of which function as a blood pressure-measuring device. None provided documentation that the tool had been validated against a gold standard, such as a mercury sphygmomanometer. Yet, such apps were among the most popular and received good patient reviews. In another study (Wolf et al., JAMA Dermatol 2013), the diagnostic accuracy of four apps for distinguishing benign from malignant skin lesions found that three of four apps misidentified 30% or more of melanoma lesions as being benign. This is perhaps less scary in diabetes, since even smartphone-enabled glucose meters would need FDA clearance; still, it is food for thought, as not all apps are low-risk.
-- by Adam Brown and Kelly Close