Last week I covered the Grammys, now I’m giving you my Oscar pick for healthcare-related film of the year. But first, a preface:
One of the most pervasive healthcare business trends I have seen in the last year or two is the formation of companies seeking to help people “age in place,” aka live out their golden years comfortably at home rather than in a facility. Nearly 15 years ago, when I first started in the health investment area, this was also a rich target for entrepreneurs, whose business models mainly focused on service innovations involving actual people visiting seniors at home. Psilos also made one such investment, in Caregiver Services, Inc., which has grown to a business of considerable size.
The challenges in these human-based businesses were and remain how to scale cost-effectively and maintain quality as the business grows. In other words, how do you ensure availability of an appropriate supply of appropriate people to care for seniors at home in a reliable and effective way? Despite a burst of initiatives, start-ups in this area trailed off and the various companies across the country have consolidated significantly.

15 years later, in response to the advent of the iPhone/iPad/iWTF, combined with pervasive home WIFI and broadband availability, we have seen a second cycle of this aging in place business plan phenomenon. However, this time the model centers on how to replace people with technology. If caregivers are worried they might be replaced by machines, they are rightly concerned. I have lately seen a plethora of companies seeking to put technology-based methods of monitoring into seniors’ homes to either passively or actively engage with seniors to ensure they eat properly, take their medicines, and maintain their basic functions of daily living. One of the most unique companies I have seen (at last year’s AARP Health Innovations @50 Live Pitch event), Evermind, puts a small piece of hardware on the senior’s basic appliances, whether it be coffee pot, microwave or light switch, to signal to someone somewhere that grandma has in fact risen from bed and started her day. This eliminates the need to even make the phone call to check in. Love is complicated.
I recently attended a brainstorming discussion at Qualcomm Life’s Advisory Board meeting that focused on what technologies are most suited for home monitoring of seniors and what aspects of a home monitoring system are most important for caregivers. In the end it became very clear that technology can’t replace people and people can’t do all that technology can do. Some combination of the two is probably ideal if such a combination could be balanced to its best effect.
As I pondered this it reminded me of a truly phenomenal movie I saw earlier this year called Robot and Frank. Starring the very great Frank Langella (yes, he played Dracula once upon a time) as a former cat burglar* now living alone and suffering from early stage Alzheimers and other ailments, the movie is an incredible story about the pros and cons of people and technology in the lives of seniors seeking dignity and happiness as they become more dependent on others. If you are interested in businesses that help seniors to “age in place,” including issues around caregivers, home monitoring, health sensors, and all of the related topics, you absolutely must see this movie. It is incredibly moving and funny, but also very poignant and pointed about the issues associated with both people technology in the home.
In the movie, which takes place at an unspecified time in the “near future,” Frank has two adult children, a son and daughter, who struggle with how to care for their father from afar. The son brings Frank a highly intelligent robot caregiver because he doesn’t have time to personally attend to Frank. When Frank sees the robot he is horrified.
Frank (to his son): “You have got to be kidding me. I’m not this pathetic.
Son (to dad): “It’s like a butler” “How can you not be excited about this?”
Robot: “Hi Frank, it’s a pleasure to meet you.”
Frank (to the robot): “How do you know?”
The robot, more advanced than what we might find at Best Buy today but probably not more than 10-15 years off based on the way technology is evolving, engages with Frank like a person but without the hang-ups or emotional baggage of the actual people who care for and about him. The robot is designed to personify the perfect balance of humanity and high technology, eventually becoming both nurse and friend and relying on a combination of facts and guilt to lure Frank to cooperate in his own best interest. The robot puts him on a regular schedule, feeds him nutritious and artfully arranged food, creates a project for mental stimulation (gardening), and gets him to exercise. The robot is also a trained healthcare aide programmed to improve Frank’s physical and mental health, “Frank, I’ve reviewed your medical records; are you finding your episodes of disorientation increasing in frequency?”
Most importantly, the robot gives Frank a form of companionship that allows him to re-engage with his one greatest passion, which I will not reveal here for fear of ruining the plot. Truly one of the best parts of the movie.
Frank’s daughter, who travels the world and is a fan of the humanistic and holistic (she looks like a Whole Foods fan club member who buys her clothes on Berkeley’s Telegraph Avenue), is astounded that her brother has abandoned dad to a robot. She temporarily suspends her career to come personally attend to Frank, and when she arrives she turns off the robot, feeling her personal touch will be a more humane way to care for Frank. The result is disastrous, as Frank misses the comforting predictability of the robot and has very mixed feelings about living with his daughter. As he becomes more and more frustrated with her and she aims to please, the situation devolves into the same schedule-free, nutrition-poor, lifestyle that prompted his earlier deterioration. In other words, while her intentions are good, the daughter’s hands-on human help is terrible for Frank and both recognize the limitations of life without the robot, despite the advantages of having someone who truly loves you physically present.

The movie has a great subplot about a guy who looks like the poster child for every Silicon Valley start-up yoga retreat. This high-strung hipster dude has acquired rights to turn the local library into what basically looks like an Apple store without the joy. With his overly twee eyeglasses and hipster vibe, he calls Frank an “old timer” and says he’d like to chat with Frank to learn more about his history and his connection to the written word in a time when the library was the only place to get information, as if it were a time when dinosaurs roamed the earth. His plan is to turn the library into an “augmented reality” center dedicated to delivering the “library experience.” The new library will, in essence, replace the dusty bound classics with cold, sterile technologies that take the heart and soul out of learning. It is a very interesting contrasting story to the main one, showing how too much technology can dehumanize in the same way that too much of the wrong human can also bring a person down.
I see a lot of movies and this one was definitely in my top 3 for 2012. And if you have an interest in this field, it is a must see. Here’s the trailer:
httpv://www.youtube.com/watch?v=nJRnMnz7fNo
*a cat burglar breaks into houses but does not steal cats, in case you were wondering
As usual, Lisa, you make great points – and although I was reading this on a plane, I laughed out loud (or at least chuckled audibly to those around me) at “iWTF.” Sardonic genius!
But here’s a serious takeaway, and a parallel to a phrase often used by EMS professionals: “Treat the patient, not the monitor.” This axiom tends to draw knowing nods from emergency service personnel and even (sometimes) the doctors working at the hospitals they service.
There are risks to relying on machines too much: among the most plain, false positives are quite common, and they can have all sorts of deleterious effects. Or none, save obfuscation…but that can be bad enough.
I recall a ride-along in which the monitor-defibrillator warned medics that their patient may be having a heart attack based on the shape of his ECG. But he wasn’t. He had have an irregular heartbeat…and he had had one for nigh on 50 years. This passenger happened to be a chronic care “frequent flier,” so the medics knew to discount the bleeping of the machine. They just ignored it. (N.B. “Frequent fliers” – which includes chronic care transports as well as abuse of the EMS system by people who view ambulances as “free” taxi services to and from hospitals – represent some 25% of EMS traffic annually).
Guess what? If the man with the irregular heartbeat HAD been having a cardiac event, these medics might not have known it. Why? Because they had been “trained” by a preponderance of false positives to essentially ignore this particular patient’s ECG. Protocols must be followed, so these men weren’t in a position to opt out of the assessment; but in practical terms they ignored the bleeping, making it ever-harder to parse a true emergency from a false alarm. Such a cycle is how one gets eaten by a wolf.
If the proverbial wolf in this case is astonishingly fast-rising cost of healthcare that is eating America’s economic vitality, then it is having a smörgåsbord right now. Redundant testing on both sides of the EMS-ED handoff remains the medical order of the day: we test even though we don’t trust. Then we test more to make us trust more. As if MORE technology is the same as BETTER technology, and the tests were not statistically independent (i.e., equally fallible).
The false-positive rate is, in my opinion, the starkest cautionary counterpoint to Vinod Khosla’s controversial assertion that computers will replace doctors. Maybe they will for some things—like the paperwork associated with billing, or finding drug interactions; both would be helpful and cost-effective. But there is still an art to medicine, and machines are pretty inartistic (clearly, since any “artistic sense” they have has been programmed!)!
But more urgently, our ability to trust machines is limited by their reliability – to sense, to assess, to parse the “improbable” from the “impossible.” Even IBM’s Watson is pretty bad at this dance. Throwing more machines against the wall won’t help (well, maybe…but only insofar as catharsis). Increasing the quality and accuracy of the machine systems already in-use will go the distance.
Here’s one example: a two-lead ECG may be “cool,” “neat,” and even “rad,” but using it to diagnose a STEMI event would be as irresponsible as failing to make clear the limitations of such technologies. Why? Because the average person – who can easily buy an AliveCor device and app – either won’t know or won’t understand the device’s limitations, and therefore is frighteningly likely to endow it with more trust than it probably deserves.
Want more proof? The National EMS Information System – the data standard to which emergency medical service documentation systems conform – does not even have a space to enter a 2-lead ECG value! Its list starts at 3-lead and goes up to 12. It does have a section for “other,” though, which suggests that 2-lead is still nonstandard among emergency health practitioners.
Here’s another vivid example: At the 2013 FutureMed conference I heard a story about someone on an airplane who used the AliveCor application to assess whether a passenger was having a heart attack, and advise the pilots accordingly regarding the need for an emergency landing. Good or bad idea?
I don’t know the skill level of the assessor, and studies suggest that – for resting ECG interpretation – a 2-lead ECG like AliveCor may suffice relative to a more elaborate setup. But what if the patient had been experiencing a STEMI [ST-segment Elevated Myocardial Infarction]? Many highly trained medics – let alone non-specialist physicians who don’t often read ECGs – will be challenged when it comes to identifying the elevated ST-segment if it isn’t pronounced. As a recently published (Kayani et al., 2012) study concluded:
“The ability of clinicians to distinguish between ischemic and non ischemic STE varies widely and is affected by prevalence of such changes in patient population. More studies need to be done to delineate the criteria to clearly distinguish between ischemic and non ischemic ST elevation.”
Add the fact that the literature is ambivalent about the technical ability to glean a read of sufficient resolution from a 2-lead ECG (vs. a 12-lead, for instance) so as to diagnose complex cardiac events, and the skeptic quickly reaches an essential question: Will someone who relies on the findings of the AliveCor device – or another such technology – be able to determine whether it is definitively showing a negative read….or if it is simply insufficient to prove a positive one?
=======
SOURCES:
“Computerized two-lead resting ECG analysis for the detection of coronary artery stenosis.” http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2034757/
“Conditions mimicking acute ST-segment elevation myocardial infarction in patients referred for primary percutaneous coronary intervention.” http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2570763/
“A comparison of 2-lead, 6-lead, and 12-lead ECGs in patients with changing edematous states: implications for the employment of quantitative electrocardiography in research and clinical applications.” http://www.ncbi.nlm.nih.gov/pubmed/14665478
Kayani et al. (2012) “ST elevation: telling pathology from the benign patterns.” http://www.ncbi.nlm.nih.gov/pubmed/22980232
Lisa,
I was pretty pumped when Redbox carried Robot and Frank a couple of weeks ago. Between my dad having Alzheimers and a technical background, I probably got something different from this movie. (btw you forgot to mention that Susan Sarandon is the librarian.)
In order to make technology or home health workers useful in these care settings, they have to engage with the people directly. Interact and gain their trust. So often we design our systems to be efficient. But if they are not engaging are they effective?
Hi Lisa, I enjoyed the article and the movie reference! You might like to check out this article from Bloomberg that describes GeriJoy as the present realization of Robot & Frank: http://www.bloomberg.com/news/2012-10-08/remote-control-pets-create-real-companionship-for-seniors.html
Also what do you think of the new model? gerijoy.com/features
Victor, great article and your new model is definitely an improvement, although you I know I loved the demonic Guiness pig! Hope you’re well. Keep me posted on your progress. Lisa
Thanks for directing me to this post with its comments. I need to find a copy of this movie. The article on virtual pets is interesting, too. Wow.
Glad you enjoyed it. Thanks for reading. You can get the movie at amazon for sure and maybe Netflix too