Using Data to track your health

I’ve been having mobility problems for some time now. In 2017 and 2018 I had a nasty case of plantar fasciitis, which is very common among professions who spend a lot of time on their feet – people like teachers, nurses, retail and hospitality workers, etc. If you’ve never had it, be grateful. At its worst it feels like every step you take pushes knives into your heel. Fun stuff.

In November 2018, at the Supercomputing Conference in Dallas, I actually tore my plantar fascia, which was a whole new world of hurt. There’s not much you can do about that except wait it out, so I did a lot of limping for many, many months. That made my left hip grumpy, and it took months of physio to sort that out, but according to my memory, I did sort it out, and got back to normal eventually. Various things then happened over time, making my left hip grumpy, or my back, or my right hip. Each time I worked on it with the help of two exceptionally good physiotherapists.

In 2022 I was nearly back to full mobility (or so I thought), when I got covid, and the combination of being bed ridden for days, plus the inflammation that comes with covid, sent my right hip straight to hell, and from there, despite the best efforts of my physio and the fantastic myotherapist I added to the team, I just didn’t get better. Then my back went out, and a series of scans later, it became clear that the osteoarthritis in my hips had progressed to the point where surgery was the only option.

Marshalling my evidence before my appointment with the surgeon, I took the monthly average step count data from my phone and graphed it, so that I had a relatively objective measure of my progress over time. And here’s the interesting thing: although my personal narrative was “I kept getting almost 100% better, and then something would set me back”, the graph tells a different story.

A graph of average daily step count per month, 2016 to present. Although there are peaks and troughs, the trend line is very clearly heading downwards the entire time.

You can see that there are, indeed, peaks, but the trend has very clearly been downwards the whole time. I never even got close to fully recovered. Also the downwards trend started well before 2018. Bearing in mind that there were lockdowns and things in there as well, the trend has existed for a long time. My surgeon was very excited to see the graph. (None of his patients have arrived with graphs before, can you believe it??) He said it shows very clearly what they say about hips like mine – that there are ups and downs, but that the deterioration is inexorable.

It also highlights something about our memories that we often forget, which is that our memories are not accurate videos of what took place. They are, for the most part, how we interpreted events, rather than what actually happened. So each peak on that graph, in my mind, was a return to almost normal, even though the data shows that “normal” was declining the whole time. It’s so interesting to take my interpreted memory of my progress, and compare it with the objective evidence of the step count from my phone. It’s important to remember that phone step counts aren’t particularly accurate, but they’re pretty good, I suspect, for trends over time.

On May 30th (just 3 weeks and 2 days ago!) I had bilateral hip replacement surgery, meaning both hips were replaced at once. Fortunately my surgeon is at the cutting edge (hah) of medicine, and where possible does anterior hip replacement surgery, meaning he goes in through the front, and does not cut muscles the way the more traditional posterior version does. This means I was able to fully weight bear on day 1, and my recovery has been sensational. It’s too early for monthly average graphs, since it’s only been three weeks since the surgery, but here is my daily step count since the surgery, with a very happy trend line.

A graph of my daily step count since my surgery. Very much up and down, but with a clear trend line going rapidly upwards.

Daily step counts aren’t terribly relevant individually, since we all have off days, and tired patches, and times when we’re out and about more, but the trend line is relevant, even over 3 weeks, as it shows that my mobility is getting better all the time.

Similarly each individual month on the monthly graph isn’t particularly interesting on its own, but the overall trend tells a very clear story.

There’s a lot to be concerned about, with regard to how much data is being collected about us, who owns it, and how it’s being used, but the ability to take the data and use it to observe your own progress is pretty exciting to me.

And it’s important to note that I didn’t need any intense statistical skills or technical skills to make it happen. I literally typed the data into google sheets (because Apple does not make extracting the data from their health app at all easy, which is a crime in its own right, in my view), told google sheets to make it into a chart, and ticked the “add trend line” box.

Data Science is incredibly powerful, and there is a lot of insight to be gained using basic spreadsheeting and graphing skills. It’s within your reach. What might you do with it?

1 thought on “Using Data to track your health”

Leave a Reply