The question is, will all these data aggregated together actually give us any actionable data that we can and will act on? If there's one thing Google Analytics has taught me, it's that it's quite possible to have a lot of detailed data at your fingertips, but still not be able to answer any valuable questions with it.
Many studies show that modern machine learning could revolutionize healthcare. However, many hospitals and governments are extremely protective of patient data. And many hospitals don't even know how to exploit their data through deep learning. Enter Apple and Google and others with big machine learning departments. They are finding new ways of collecting health-relevant data. For example:
Through Apple Heart Study, Stanford Medicine researchers show wearable technology can help detect atrial fibrillation
https://med.stanford.edu/news/all-news/2019/11/through-apple...
https://www.apple.com/healthcare/docs/site/Apple_Watch_Arrhy...
An OS with a focus on health seems to make a lot of sense.
I hate Apple's approach of doing 70% of the work all the time. If they truly cared about health, they'd make it incredibly easy for any company to add health data in to their ecosystem and not just try to sell more Apple Watches. As a Whoop user, it's so much more beneficial than their basic Apple Watch stats (which btw are fine for most people, I just want more/better HRV data)
Apple Health is one of those products that's useless until it suddenly isn't. We're reaching that point for some folks and I only expect that group to expand rapidly in the coming years.
Ever since I heard Mary Lou Jepsen (OLPC/Oculus Rift) talk about her work on a revolutionary holographic medical imaging system, I've been expecting Apple to buy her Openwater startup, add resources and introduce cheap, safe MRI-like body/brain scanners for the masses. If it works, it could be bigger than the iPhone. https://www.openwater.cc/technology
Traditionally, these data sets have lived in silos.
But what could happen if it all was aggregated safely
in one place?
What happened to Palantir handling NHS data in UK? There was some privacy concern there but it happened anyways. Call me a pessimist and this a rant but it seems like the public doesn't really control what their governments decide to do with their data. The best the tech firms do is issue an apology for infringing on privacy or pay a light fine while they get to keep majority of their profits.
Apple has a talent for putting themselves right at the center of important technologies. I wonder if this strategy plays out as integrated medical devices and a full-blown healthcare platform. Their reputation for top-notch hardware, good UX, and their shift towards in-house SoCs gives them a serious edge there. Everything about their brand feels well suited for healthcare, and IMO they still don't have a killer enterprise app. This could be what Watson was promising to be.
The notion that the industry isn’t exposing any data is patently false. FHIR and HL7 have been around for years now and I’ve connected many data sources into both Apple and other services that use those APIs.
I really welcome the Health Kit or Google Fit approach for sharing data among health app, I actually want to use that for health research. The main problem is that as soon as you mention the fact that patients data is going to be shared with a company (besides in the US) doctors shy away because of data protection, regulations etc. We need the data to be on the device and possibly outside of the control of one (or 2) companies. I know Apple says that data is encrypted end to end, and I wish Google did the same, but even so it is still problematic. When you add the fact that every manufacturers of wearables obliges users to share data on their dodgy platforms... I don't see a solution to this. At least not until the mindset changed completely.
Feature request: Heat stroke prevention.
Just heard an interview with author of recent paper about global risk of increased mortality due to climate change. (Radio, no cite, sorry.) Research team has clarified impact of "wet bulb", which is somehow combo of both ambient temperature and humidity, and prevalence of heat related deaths.
Not just the Middle East, but also provided examples from Europe, Toronto, Louisiana.
While wet bulb temp of 35 °C is life threatening, paper shows that with higher humidity even 27 °C can be dangerous for at-risk persons and has led to greater mortality.
While the risk will certainly get worse, people have no intuition for it. Especially for outdoor work and activities.
Any way.
My immediate thought was my watch can alert me to wet bulb risks, a combination of weather and health monitoring.
https://en.wikipedia.org/wiki/Wet-bulb_temperature#Wet-bulb_...
What is all this "health data" people are talking about? When I go to a new doctor that has none of this apparently super complicated, dense health data, they give me a single page form to fill out where just about 90% of people below the age of 50 will tick every box with "no" or leave the field empty.
Apple could be planning to make an EHR system, like Epic and Cerner. Google is supposedly working on one, too. (I don’t think the article really arrives on that possibility)
Most electronic health record systems are really terrible software, and they’re also very profitable pieces of software. Google, Apple, and Microsoft are probably the only companies with enough money and software engineering talent to start from scratch and disrupt that space.
Some interesting further reading: https://medium.com/design-and-tech-co/how-google-ehr-could-f...
If you want to learn more about where research in this area is headed:
Santosh Kumar has been heading a large NSF research center investigating wearable medical sensing and how to turn sensor data into actionable health advice:
Tanzeem Choudhury has been doing some work at the intersection of AI and sensing and health:
Deborah Estrin's Small Data Lab is doing some interesting work too:
I really wish Apple would get on with it - the process for trying to upload medical records was super broken for me (plus as an American, I can't even find most of my medical records since they're in different doctors offices across the 4 states and many cities I've lived in). That said, I am nervous about just how much data they'll own. Does anyone know if there are government backed open systems that someone from another country could use? I know that Estonia for instance has an okay patient health portal where you can see all your info but it's not useful unless you live there.
Just last night I read an essay by Galloway where he pointed out that the really huge companies (Apple, Google, etc) have a hard time maintaining growth so have to move into different sectors where there are huge pools of money. Health is the most obvious one.
The epistemology of the thing...
In the immediate human sense recall that this was the health app that launched with no way to record your menstrual cycle (a very important aspect of health for about half of humanity, the half that makes new humans I should add.)
In the deeper sense, we might encode "humors" into the design of our system. Our knowledge is incomplete. (E.g. the microbiome, or the "interstitium" which has been called "a new organ" https://www.livescience.com/62128-interstitium-organ.html )
It's important that our data systems handle open-world rather than closed-world data, eh? https://en.wikipedia.org/wiki/Open-world_assumption
(As a practitioner of Reiki I can attest that scientific medicine does not yet capture all relevant variables.)
Last but not least, the crux of the issue turns on what Martin Buber called "I and Thou": https://en.wikipedia.org/wiki/I_and_Thou
> Buber's main proposition is that we may address existence in two ways:
> The attitude of the "I" towards an "It", towards an object that is separate in itself, which we either use or experience.
> The attitude of the "I" towards "Thou", in a relationship in which the other is not separated by discrete bounds.
There is a subtle trap in thinking of the body as "it" (and also a trap in thinking of the body as "thou", and another as "I"). It behooves us to be careful what we automate.
Microsoft's almost always either too late, or in this case too early.
Reminds me of The New, New thing!
I would be happy if they would still support older machines with macOS. A macmini of 2010 that is still working good can't be left without security updates!
The USA has been making plans for widespread sharing of Electronic Health Record data for many years:
https://www.healthit.gov/hitac/committees/health-it-policy-c...
This is seen as desirable for more than one purpose, including allowing people to move their medical records to a new doctor or hospital easily, and for doing research on what makes for the best care, as well as giving doctors the ability to look at case histories similar to that of a patient they are treating.
Computerization of medicine has been resisted by doctors for a long time, which has meant that all the data and observations and diagnosis made over the years has essentially been lost when it could have been informing doctors decisions and being used in research. There are definitely privacy and abuse problems with putting everyone's health data in a central database run by the government, and it is very difficult to de-identify health data (how many people have your height, weight, age, and diseases?), but without some way to gather all the knowledge and experience and data from the practice of medicine we have been missing important correlations and creating harm due to each doctor making their own decisions instead of having a set of best practices informed by hard data. Whether the police, FBI, NSA, courts, and politicians should have access to this data are difficult questions and hard to implement so they are not subject to abuse. The police already can make some kinds of queries at hospitals for medical data.
The UK has started a large scale Precision Medicine project that is collecting health data on 500,000 people:
https://en.wikipedia.org/wiki/UK_Biobank
There are other ongoing large scale projects collecting health data also:
and I'm sure there are others, in many countries. A lot of this is part of the move towards Precision Medicine, medicine based on data and personalized to an individual, rather than treating all people as being identical and letting each doctor make their own guess on how to treat a patient (I've been in meetings where doctors say that 80% of the time they are guessing, though guesses get better as they see what works and what doesn't for any one patient):
https://en.wikipedia.org/wiki/Precision_medicine
The space Apple is aiming to dominate is currently one small part of all this. One big question is what do doctors do with all this data? It's too much to look at it all. They have no experience with such detailed info also. What is normal for the way your blood pressure changes as you go about your day? Your doctor doesn't know and hence even if she had continuous blood pressure measurements from you 24/7 wouldn't know what to do with it. What is a reasonable blood pressure while having sex? When climbing stairs? The only measurements like this that we've had historically are from astronauts. Machine Learning may help, but what do you do with an opaque message from an algorithm that there is a 75% chance you have some abnormality in some obscure biological parameter? The chance of 25% of those notifications being wrong means you can't use it as the basis of a diagnosis. Maybe it suggests something you can investigate further, but then how much time will you waste following up on those 25% of notifications that are in error? Was the ML trained for your age, weight, and specific set of health conditions or will it fail to detect anything for people outside the scope of the training data, yet be used to evaluate your health (incorrectly) anyway? And using ML also creates yet another path for bad actors to steal and tamper with your data.
There are a lot of complicated questions surrounding medical data.
I don't see why I'd need Apple to tell my personal trainer I have a herniated disk. I can tell them myself, and I'm looking to be less locked into the Apple ecosystem, not more.
If that's what they're doing (which I bet it's not), I'd consider it as a good sign that Apple has run out of good ideas.
Maybe it's the french in my talking, but I really have a hard time trusting anything else than the state to handle health data.
Anytime you associate health with financial motivation, you end up with something twisted on the long run.
Sure, you can say that Apple, right now, is in the proper state to do that (although given they were part of PRISM, it's debatable).
But what about in 10 years? What if their executives decide that Apple Pay should lead to the creation of a bank? Then from a bank to an insurance company? Then they have a conflict of interest.
Simpler than that, what if they change their stance on user data, and decide to exploit it to understand their customers better? Or sell it? Or just partner it to labs, insurance companies, hospitals?
Like with facebook, it will of course all be "made anonymous", until it's not anymore.
Money and health data are not meant to be mixed IMO.
But even without all that, we are adding yet another tracking system on top of what is already a sensor packed pocket self-spying device.
It's hard to not worry that such concentration of power has a high potential for suffering.