Apple’s EKG is out, AWS is going to fix EMR, and the medical school business model is dead
|Dec 9||Public post|
Things That Happened
The sound of cardiologists everywhere screaming. The Apple Watch EKG feature went live this week. Consumers seem to love it, while physicians are less enthusiastic. Want to scare a cardiologist? Show her this.
Expanding Circles. Oscar Health is adding a new service tier to its small employer insurance offering. Circle Plus is an upgrade to the standard Circle plan, and includes expanded provider choice. Oscar’s narrow-network plans are one of their key strategies for containing costs, so this could indicate they’re finding it difficult to court employers to their plans. Most healthcare consumers associate the quality of a health plan with the breadth of choice of in-network providers and the size of the deductible, so offering a costlier plan with more choice may be enough to convince employers, and their employees, that Oscar is a viable option for employer-sponsored insurance.
Rideshare Healthcare. Lyft hired Megan Callahan, the former Chief Strategy Officer at Change Healthcare, to run their healthcare business. Given her experience at Change, Callahan is an expert on the billing and revenue side of healthcare. And that’s why Lyft, and other ride-share firms, are so interested in the area of non-emergent medical transport; much of it is billable to insurance companies. And even when it isn’t directly reimbursed, it’s still well worth it for a physician to provide transport when it earns them a visit fee they wouldn’t have seen if the patient didn’t show up for the appointment.
More reason to stare at your phone. Researchers at Stanford designed a machine learning algorithm to detect depression using video recorded from a conversation with a physician using a smartphone. While it’s still very much in the research phase, it’s demonstrative of the reality that our smartphones are, and will be, excellent diagnostic sensing tools.
Share it on Facebook. The initial products of the FastMRI collaboration between NYU medical school and Facebook have emerged in the form of an open sourced set of MRI data and baseline machine learning models. Facebook claims in their blog that it’s the largest MRI data set ever open-sourced.
Psychiatry in Aisle 5. Walmart, continuing their inexorable march towards healthcare dominance, is piloting an in-store behavioral and mental health clinic in a single store in Texas. The clinic is focused on less complex mental health diagnoses, and will refer patients to standard mental health practices should they require more complex care. It’s essentially the same model Walmart is already employing with medical care.
Robots giving you medication? Pillo, the automated pill dispenser slash chatbot, is now available for purchase. The major question Pillo’s executive team should be sweating is whether Pillo offers an experience superior enough compared with PillPack and an Amazon echo dot ($30 total cost) to justify the huge price delta. Pillo starts at $500, and has a $40/month subscription fee starting in year 2.
Things Worth Understanding
Last week Amazon Web Services announced a new HIPAA-compliant service that employs deep learning to interpret and structure physicians’ notes.
In tangible terms, AWS Comprehend Medical will take physician visit notes and output a structured list including, according to the AWS blog post, “patient diagnosis, treatments, dosages, symptoms and signs, and more. “ Should it work as advertised, and for the purposes of this piece we’ll give AWS the benefit of the doubt, this would allow physicians to return to taking notes as they please and not concern themselves as much with the formatting and structure of their EMR.
It almost sounds too good to be true.
EMRs are a huge problem point for physicians. They’re time consuming, difficult to navigate, and pull physician focus away from patients. When there’s 15 minutes to complete a patient visit, it’s a zero-sum game, and the EMR is winning.
Comprehend Medical isn’t the first solution to the problem. Atul Gawande explored a number of fixes in his recent piece in the New Yorker: medical scribes who are trained to take notes for the doctor live during the visit, and even an elaborate system in which the visit is recorded and transmitted to a physician in India who creates the medical notes, then sends them back to the physician. Those are fine as band-aids, but they don’t scale.
Comprehend Medical has the potential to make the interface layer of the EMR user friendly by allowing the user to dictate the terms on which information is inputted rather than the other way around. It’s similar to the promise of the voice interface - you should be able to ask Alexa your weird question, and the burden is on Alexa to figure out what you want. The machine must learn how to respond to your input, rather than the user learning how the machine likes input.
What’s more interesting here is that it seems AWS has found a way around one of the major strategic problems facing big tech companies as they enter healthcare.
Most tech giants - Facebook, Amazon, Apple, Microsoft, Google - have been built using user experience as their key strategic focus. When shopping at a competitor’s website is as easy as clicking to the next tab, Amazon has always had to focused from top-to-bottom on the customer experience. In healthcare, where consumers have diminished agency to make purchasing decisions (because of information asymmetry, and because they often don’t directly foot the bill), user experience isn’t nearly as much of a wedge - if consumers aren’t the ultimate deciders, then you don’t have be as concerned about keeping them happy. Thus tech companies’ greatest strategic advantage, and the central thesis around which their operations are organized, doesn’t hold true in healthcare.
One would think that EMRs, which are primarily used by physicians, would need to win physician approval to gain customers. The worst kept secret in health IT is that EMRs are billing systems first, and clinical record keepers second. As the multi-million dollar contracts for hospital EMRs are being negotiated it’s quite obvious from the outcomes that user experience is not a major factor in the decision.
With Comprehend Medical, however, Amazon has found a way to so dramatically improve user experience for physicians that it can, in fact, act as a wedge. It’s not just that Comprehend will make the physicians’ life easier, but that it will make them happier, more efficient, and could even help health systems avoid expensive intermediary solutions like scribes.
If Comprehend ends up being the blockbuster it could be, it will put Amazon in a position to dictate the market. They could use it was the killer-feature exclusively available on an AWS electronic medical records, or they could simply sell it as a service to existing EMR players as a way to lure them further into using AWS as their complete platform. Given EMRs are headed in the direction of more customization for different specialties, AWS might be avoiding this product headache and positioning themselves as the common foundational infrastructure on which all EMRs are built.
Things To Read
“The key to a successful recovery after illness may be a less stressful, more supportive, more humane experience during the hospitalization.” Austin Frakt wrote this week in the NY Times how hospital standard operating procedures like regular vitals checks and blood draws get in the way of patients’ sleep. Consider this another side-effect of the fact that patients’ aren’t the hospitals customers; health plans are.