I read about that last week and it was really one of the most evil-on-a-personal-level things I've seen in a…
Curbside Consult with Dr. Jayne 11/6/23
Everyone is talking about the recent shutdown of Olive. Looking back over the last several months, HIStalk has been full of mentions of companies that abruptly shut down, declared bankruptcy, or are otherwise in bad shape. Plenty of other healthcare and health-adjacent are companies in those situations, but they don’t always come across our radar or aren’t noteworthy in the healthcare IT realm, so it’s difficult to know what the true number of companies in this situation might be.
In talking with friends who know the industry well, most are in agreement that it’s time for a lot of companies to pay the proverbial piper since they can’t deliver on the promises they made in exchange for startup funding. They forecast that many more companies will be trying to reinvent themselves over the coming months. Those that are successful may live to fight another day, but others may become the stuff of fire sales or ultimately closures.
Taking a look at Olive, some of the problems that they experienced are readily visible at other companies. They promised the use of artificial intelligence and advanced technology, but it was more sizzle than steak as it became apparent that they were using screen scraping tools and less-than-intelligent technologies. In visiting their booth at HIMSS, the team was more excited to talk about their big purple recreational vehicle than it was to talk about their actual solutions. When we did get them to discuss their business, there were plenty of vague ideas, punctuated by delusions of grandiosity. I wasn’t surprised when the company had layoffs in the summer of 2022. CEO Sean Lane explained in an interview, “Our fast-paced growth and lack of focus strained our product and engineering resources and prevented us from executing quickly on key initiatives.”
How many other companies out there are guilty of this? Plenty of startups go after what they perceive is the brightest, shiniest object, throwing their muscle behind initiatives without doing the level of due diligence or project planning that is needed to set themselves up for success. Product and engineering teams often work in a certain degree of chaos, but good people in those disciplines aren’t going to do that indefinitely, especially if other things are going on in the company that make things feel unstable.
I have no knowledge of how those particular teams felt at that particular company, but I’ve consulted for enough startups to have seen some wildly inappropriate leadership behavior that would make anyone vote with their feet. People don’t generally enjoy having projects shelved on a whim, or see funding diverted to initiatives that they know deep down are non-starters. They don’t like CEOS who yell at their teams, micromanage, or have Jekyll and Hyde personality swings. People also tend to be uncomfortable with the idea that their company might be selling vaporware, and are more likely to move on when that starts. One can only pull a rabbit out of the hat so many times before it becomes tiresome.
We know from the postmortem of Theranos that some of these high-flying executives may be lying to their board members and to their investors, as well as to their employees and clients. I haven’t seen any salacious tell-all stories yet about Olive, but will be interested to understand whether there was actual fraud involved or just flagrant mismanagement and a lot of excuses. In this situation, I think there is likely to also have been an element of failure to estimate exactly how much it would cost to try to do artificial intelligence work properly. It’s often much more expensive than people think, especially if you want to do the right kind of training and validation with your models. I’ve seen several companies who claim to have “AI-driven” this or that, when what they really have are sophisticated decision trees and a lot of manual intervention and hard work behind the scenes.
It’s particularly grating when I see a company that is operating in what seems to be a somewhat shady fashion, and I look at their list of investors and see only hospitals and health systems represented. These are their possibly spun-off “innovation” arms, or their venture capital funds rather than the care delivery organizations themselves, but the ultimate source of the money being used for investment is the same – payments from patients, insurance carriers, and the US government. These all boil down to being funded by you and me, in the form of our insurance premiums and tax dollars.
It also makes me angry when I see these care delivery organizations throwing major chunks of cash after technology solutions, when they haven’t yet cleaned up messes of inefficiency that could be handled by solutions they have already purchased. They’re not willing to spend the money to hire analysts and trainers to fully implement the EHRs that they’ve spent tens (if not hundreds) of millions of dollars for, or to optimize those systems to actually improve patient care for the physicians and providers or to improve the patient experience. However, they’re willing to buy other solutions that may just make terrible processes run faster and make their patients and caregivers more frustrated than they already are.
Frankly, you can buy a lot of vaccines, deliver a lot of charity care, and discount a lot of procedures for the hundreds of millions of dollars that are spent annually on solutions that fail to deliver value. Looking at some data on the largest organizations in the most recent year I could find (2021), venture funds run by large hospitals laid out more than $2.7 billion in funding rounds. How much prenatal care could that provide, especially since the US now has the highest infant mortality rate among high-income countries?
They can make the argument that the companies in which they invested are going to bring value that lowers the cost of healthcare, improves outcomes, and more, but the proof is ultimately in the pudding as far as what they are actually able to achieve. I can’t think of anyone I’ve talked to in the last several months that thinks the US healthcare system is stronger or better positioned to handle the challenges it is facing despite all this money being spent on technology. All of the practicing clinicians I speak to are on the frazzled edge, constantly being asked to do more with less, being forced to cut staff, or finding that services have been reduced. Maybe it’s time we start spending healthcare dollars on actual bread-and-butter healthcare delivery and ensure that patients are receiving a minimum level of care before we start writing big checks for pie in the sky ideas.
What do you think about the future of tech unicorns in the healthcare space? Leave a comment or email me.
Email Dr. Jayne.
You have put into words what I have seen over and over again. Staying focused on what they have already purchased is not in their DNA. It’s always the next thing. Thanks.
Re:. “they haven’t yet cleaned up messes of inefficiency that could be handled by solutions they have already purchased.”
Amen to this! I used to have a saying that we never, literally never took full advantage of any system we implemented.
Then when systems replacement time came along, part of the justification for the new system, was the fact that we hadn’t fully implemented the old system! It infuriated me endlessly. Why are we so bad at this?
“System Old could not do X. Therefore we need System New.”
“Well, actually System Old could do X. We just didn’t take advantage of that feature.”
“Look, it’s all the same! Whether it couldn’t or we didn’t, the outcome is the same.”
“No, it’s not the same. If we failed to implement X before, why would we succeed this time?”
“You are not being helpful. I need you get on board with this.”
(Sighs). “OK. Fine.”
I remember a solution that was sold that would deal with lists of patients for followup — a population health solution for large hospital systems with multiple hospital EMRs.
The “Solution” would take aggregate clinical and billing data from across the systems and combine them in a couple different engines to spit out lists of patients that needed followup, including scoring their risk and sending it off to a patient management “Solution” for patient contact and scheduling.
That, was the best PowerPoint I had ever seen. Unfortunately, the PowerPoint to C# compiler wasn’t working during those years and the solution never went from thought to production. It did however, get sold to large systems — I wonder how many of those patients could have been helped by simply combing through the data?
Instead, the solution was “next month” or “next release” and when the release came the implementers would spend a couple months trying to get it to work. All to know avail. The implementors would report the new failures, and note that the old failures were still there — it actually got worse rather than better over time. A couple three years later the “product” was finally killed.
Those hospital systems could have spent that money on a couple data analysts and serviced their community much better. Also, what happened to those patients in that period when the promised “Solutions” met reality?
I love the promise of big data, but I also know that there are people out there who do nothing but smell money and whose scruples are less than whole. Theranos is one level of corrupt, but there are others out there selling PowerPoint “Solutions” as if they were real.
Reminds me of an add-on to a product we used. The core product was pretty good, discounting the usual “well I want it to do X” Wish Lists.
They eventually announced a “helper” add-on, and that never worked right.
The vendor apparently admitted, years later, that they had drastically underestimated the difficulties they had encountered.
I knew their problem. There were tons of unwritten rules that the users just automatically followed. It had to do with what you can realistically ask a human being to do. But being unwritten, you could not write code to do the job.
It all smelled like some customers had requested some “wouldn’t it be nice” features. Vendor salespeople had enthusiastically backed them up. Yet far too much data and logic was implied, making the code either make excessive assumptions, or just give up and issue an error message.
“They’re not willing to spend the money to hire analysts and trainers to fully implement the EHRs that they’ve spent tens (if not hundreds) of millions of dollars for, or to optimize those systems to actually improve patient care”
Sure, Olive isn’t a poster child for HIT but let’s not swing the pendulum in the other direction.
Do you honestly think we have a better/faster/cheaper shot @ improving healthcare by throwing analysts at a clunky monolith of software — the EHR — that was built on 1990s tech, to largely to gov-mandated specs, with a below 0 NPS across all major players?
Interop remains a major challenger for developers. Some of the largest EHR players engage in monopolistic practices that also run afoul of the ONC Final Rule… I predicat an FTC investigation in the near future, interop will get easier, GOOD innovation (not of the Olive type) will flourish, and balance will be restored.
The major EHRs which dominate the US landscape (e.g. Epic, Cerner) are very capable of streamlined and efficient workflows. The issue is not the EHR and it’s not “government-mandated specs” which are killing EHR usability, those are tired and antiquated tropes. It’s the complicated nature of the “art” of healthcare, the inability for MDs to agree and align on a “singular” workflow, and the lack of commitment from health systems, physicians and IT/informaticists to dedicate time to refining workflows. Period.
I guarantee the major EHRs are more than technically capable of supporting beautiful and intuitive workflows, it’s just not that simple when you get the humans in the room.
Trying to declare the dominant EHR products as capable of supporting clinical workflows is an “interesting take”. These reimbursement cash registers have been augmented by hundreds of third-party apps to do a barely adequate job for clinicians.
Seen too many hospitals get into the innovation game, which really was just a way to get a look under the hood of Epic’s software to ‘touch wires together’. Worked at a healthcare system in early-mid 2000s who spun off innovation arm and sold ~100 Order Sets for ~$1,000,000 — they were in MS Word, not in Epic!
Seemingly the only big player that seems to act like adults in the room is Epic…as it relates to disciplined software development, data security, interoperability, and overall long game players). They still have a lot of ugly software that needs to be addressed, but it’s the ‘best of the crap’ from a UX perspective but BEST IN CLASS (and KLAS) for all other areas that we should care about.