You’re reading the web edition of STAT Health Tech, our guide to how technology is transforming the life sciences. Sign up to receive this newsletter in your inbox every Tuesday and Thursday.
Ransomware plunges rural hospitals into crisis
Ransomware is no longer a threat restricted to large healthcare institutions. Small rural providers are also being hit by a wave of attacks, in some cases forcing them to resort to paper and pen record keeping to continue serving patients. “We were woefully unprepared,” said John Gaede, director of information services at Sky Lakes Medical Center in rural Oregon. The healthcare system was hit by an attack in October 2020 as it responded to its first local wave of Covid cases, making a difficult situation almost unmanageable.
Such attacks not only create logistical challenges, but also cut off access to the electronic medical history needed to safely care for patients. Read the full Marion Renault story.
Are voice biomarkers ready for prime time?
Using AI to find clues of disease in voice samples remains a new area of research, but the effort appears to be progressing rapidly. A growing number of apps promise to detect mental health issues such as anxiety and depression by analyzing voice samples. And researchers from Mayo Clinic and the Tel Aviv University has just published a study showing a strong association between a specific vocal biomarker and health events related to coronary artery disease. The research is extremely promising, but also preliminary. Asking whether AI can do something in a lab and whether it can do it safely and reliably in the real world are very different questions.
It’s not just about your AUC
In early testing, an AI algorithm for detecting a specific type of hip fracture outperformed radiologists and recorded near-perfect CSA. But closer examination of the model by researchers from the Australian Institute of Machine Learning revealed unexpected problems. In validation on an external dataset of Stanford University, the algorithm’s AUC remained very high, but it also experienced a large increase in false positives during its preset calibration. Of course, this can be solved with careful local training.
But a closer audit of its performance also revealed inhuman errors, such as misdiagnoses of serious fractures in some cases. These kinds of problems can lead to unacceptable errors in clinical settings, and they’re easy to miss if algorithm developers and regulators don’t watch carefully.
Virtual care in numbers…
Year-end operating statistics shared with us by UnitedHealth Group offer some ideas about virtual care worth pondering:
- 28 million virtual care visits were recorded in 2021 – a sustained increase of 2,500% over pre-pandemic levels and an indication of the staying power of remote care.
- 14 million of those virtual visits, half were for behavioral health needs, which tells you why behavioral health startups continue to make money.
- 64% of virtual care users were women, up 2% from 2020. The year-over-year growth shows that the momentum can remain sustainable over the long term.
Dollars, data and drones
- Healthcare data brokers Komodo Health and Dating before are expanding their partnership to enable life science companies to connect proprietary, third-party clinical trial data to Komodo’s software platform. The software allows users to link anonymized data from electronic medical records and other sources to get a more complete picture of patient care and outcomes.
- Alphabet-belonging Fitbit has received FDA authorization for an atrial fibrillation detection algorithm. Many other wearables detect A-fib, but Fitbit emphasizes its product’s ability to report the condition while you sleep, potentially allowing for more comprehensive monitoring.
- Another Alphabet-owned company, Winglaunched a project using drones to deliver over-the-counter drugs and other products ordered by Walgreens customers in Little Elm, TX, near Fort Worth. Items are tied to lines dropped by drones and then rolled up for delivery.
- diligent robotics, the maker of a robot that carries supplies for frontline caregivers, raised $30 million in a Series B round led by world tiger. The company pitches the product as a way to address the national nursing shortage.
- Viz.ai, a company known for using AI to speed up stroke detection, raised $100 million in a Series D round led by world tiger and Insight Partners. The company said its technology is now used in more than 1,000 hospitals.
What we read