Receiving this newsletter from a colleague or friend? Subscribe to our Digital Health Digest so you don't miss a beat!
DeciBio Weekly Digital Health Digest Volume 3 — February 4, 2020
Researchers aim to track coronavirus outbreaks via AI & machine-learning powered social media analytics. The FDA clears a wearable that translates vitals data into digital "symptom" biomarkers with initial applications in oncology & pre- and post-operative care. A major EHR vendor busted for rigging clinical decision support in a criminal kickback scheme. Amazon Pharmacy™ plants a flag in the EU and 15 other global markets.
Here's the roundup for the week:
Recent Headlines
DeciBio Original Content
Stay tuned for next week's deep-dive into digital health's opportunity to address limitations of precision medicine
TLDR: BioIntelliSense's BioSticker wearable monitors symptoms like coughing, sneezing, and vomiting, providing physicians with up to 30 days of ongoing patient data. Their partnership with the University of Colorado’s health system aims to enable earlier discharges by monitoring patients for complications while they’re at home instead of in a hospital bed. The company obtained FDA 510(k) clearance for the wearable, and is expecting the full service to be available in the U.S. within a few weeks on a per-patient, per-month basis.
So what? BioIntelliSense’s success echoes an industry trend of scalable remote patient monitoring devices earning regulatory clearance and piloting initial use cases with health systems. BioSticker follows a string of announcements in digital outbreak surveillance (e.g., social media analytics for coronavirus, other cough detection devices).
TLDR: Eko received 510(k) clearance for EkoAI, its AFib and heart murmur detection algorithm. Eko now offers an end-to-end suite of FDA-cleared software and hardware (DUO handheld digital stethoscope & ECG, CORE digital stethoscope and attachment) for use by clinicians or patients for at-home monitoring.
So what? EkoAI’s clearance marks 2020’s first big win in digital diagnostics. The company’s algorithm detects AFib and heart murmurs with 99% and 87% sensitivity, respectively, compared to 70-80% and 50-70% sensitivity by cardiologist and GP exams. While EkoAI augments diagnostic accuracy, DUO enables earlier intervention and improved access through easy-to-use, at-home cardiac testing. DUO + EkoAI joins the ranks of AliveCore KardiaMobile (6L) and Apple Watch Series 4 in the FDA-cleared remote cardiac monitoring space, while differentiating through its offering of AI-assisted stethoscopes for clinician use.
TLDR: Practice Fusion paid a civil settlement of ~$120M and $26M in criminal fines after admitting to a $1M contract with Purdue Pharma (the company behind OxyContin), rigging the EHR with a clinical decision support pop-up designed to increase Purdue’s opioid sales. Although Allscripts claims no knowledge of the shady dealings, reporters have noted suspiciously timed events suggesting the contrary.
So what? The EHR, embedded in nearly 30,000 practices, showed the pop-up an estimated 230M times over the last 3 years, impacting up to 3,000 patients and adding $11.3M to Purdue’s pockets. This case isn’t the first of its kind. eClinicalWorks and Greenway Health recently settled MU and kickback allegations. These cases herald the need for increased technical review and regulation of health tech solutions and partnerships to ensure advances in data-driven care and decision-making do not compromise patient wellbeing and trust.
TLDR: PillPack by Amazon Pharmacy™ is now trademarked in the EU, Canada, Australia, China, Japan, India, and 10 other international markets.
So what? While Amazon’s U.S. PillPack operation has encountered a slew of challenges (e.g., non-compete litigation, dropped beneficiaries, patient data loss), Amazon seems to be doubling down on translating its expertise in distribution, logistics, and AI-ML driven consumer experience to disrupting the global prescription drug supply chain. Amazon’s global reach and unrivaled ability to leverage data to understand us better than we understand ourselves spells trouble for challengers.
TLDR: The collaboration leverages ProofPilot’s clinical trial platform and Google’s cloud-based servers and data products to aid researchers in communicating with participants via text, email, and push notifications for MGH's study, "Piloting Time-Restricted Feeding (TFR) for Bipolar Disorder.”
So what? Researchers are increasingly looking towards digital tools to streamline clinical trial activities, and many startups are popping up in the space (e.g., Science 37, Trials.ai, Medaptive Health, Deep 6 AI). These tools aim to improve trial recruitment efficiency and diversity (e.g., via social media and other online registration), increase patient engagement and adherence (e.g., through nudges and notifications), and collect and utilize "real-world" data (i.e., remote device-enabled data collection outside of clinical supervision).
TLDR:Increased data availability through social media and news posts has opened the door for AI-ML as a surveillance tool for the current coronavirus outbreak. Researchers are using machine learning to help predict where and how fast the disease may spread, predict what types of people are most at risk, evaluate efficacy of treatments, and assess symptoms to distinguish the virus from seasonal flu.
So what? Outbreak surveillance is yet another health application for AI-ML, and activity will be driven by companies that generate novel data. Though it’s likely still early before we see the real impacts of AI in outbreak surveillance, the concepts outlined are still important to note and will evolve over the next year, which many are calling the “year of AI.”
TLDR: The academic medical center joins the ring of innovators looking to develop smartphone-based digital biomarkers and AI-ML solutions for predictive diagnosis of critical mental health events. mindLamp combines active (e.g., psychiatrist-designed surveys, cognitive tests) and passive smartphone use data (e.g, biometric sensors, GPS, text / call activity) to provide treating psychiatrists with a dashboard for data-driven treatment modification
So what? Digital diagnostics have shifted the center of discussion from early to predictive detection. Studies have estimated predictive analytics could save the U.S. healthcare system up to $450B annually, with mental health being a key application. While MindStrong’s “mood monitoring” feature similarly incorporates passive data (more broadly including “gesture” tracking -- e.g., swipes, scrolls, selections), patients may be more keen to trust an academic center with such data.
Reward yourself for powering through all that news. ICYMI — voice tech shines in Super Bowl LIV commercials. Check out Google and Amazon's commercials below:
Google's "Loretta" strikes a sympathetic note
Ellen Degeneres reps Amazon in #BeforeAlexa
Headlines curated weekly by DeciBio's digital health team
Chris Lew
Senior Associate
Fanny Anderson
Associate
Tina Wang
Senior Analyst
Looking for news updates personalized to your needs? Claim your BioTrack invite and get the latest market and research activity customized by news type, source, keywords and more.
