For the last week I’ve been largely wiped out with flu. This is particularly problematic for me as an asthmatic. (Yes, I fulfil many geek stereotypes).
Fortunately, being a geek, I can address this problem scientifically. I know from repeated experiments — albeit ones enforced upon me by bacterial infections, how to respond.
If my peak flow (a measure of lung function) drops by more than 30%, I need antibiotics. Because it’s not long then before my blood oxygenation is going to tumble.
And then I end up in hospital. Like I say, I’ve done the experiments.
Now, to get the antibiotics from my normal doctor, I just need to phone her.
Me: “Hey, my peak flow’s down 30%, I’m using my reliever eight times a day, I think I need antibiotics.”
Doc: “Yes you do. I’ll have a script waiting for you at the front desk in half an hour.”
But when my regular doctor’s not available, I have to go to the surgery. That means I have to leave my sick bed, share my germs around and take up the valuable time of a doctor who could be seeing someone else, when it’s clear what I need.
Why? Because while my normal doctor knows that I know how to manage my condition, another doctor doesn’t. My patient records don’t tell her that. She can’t see the readings from my peak flow meter, or pulse ox meter.
This will be different in the future.
Firstly, all of my medical devices will be connected. Secondly, I will be able to choose to share their data with my doctor. Thirdly, my devices, or some personal AI connecting them, may have picked up on the pattern of my infection earlier than I was able to, and ordered antibiotics even quicker — assuming we still have any functioning antibiotics at this point.
This all implies a lot of change. The connected devices, the piece that most people seem to be working on developing and getting excited about, are really the least challenging aspects of making a vision like this, real.
We will need a radical shift in the way we store data, moving it from the healthcare provider to the individual.
We will need to change the way we provision (and buy) the cloud-computing intelligence to handle tracking and diagnosis, from ad-supported centralised services to paid and personal distributed ones.
And we will need to place a lot of faith in the decision-making capabilities of machines — faith that we don’t have today.
In twenty years time when self-driving cars are the norm, we may be a little more willing to let machines prescribe our medicine on a doctor’s behalf. But I don’t think we’re ready to make that leap just yet.
This blog post came out of my preparations for the Tug Life event in London next week. Tickets are still available if you’re interested here: http://www.tugagency.com/life/