Quantified Asthma: That time I logged all of my asthma-life for 10 (plus) days
I’ve written before on Quantified Self and the apps that are available to help you quantify your asthma. In that post, I mentioned that the currently available applications—while certainly improved on some past options—don’t hit all of my marks for being a perfect app—an app that would assist me in logging, tracking, journaling, quantifying—whatever you want to call it—my asthma for the longer term to help me find trends to act on (if there are indeed any!).
I’ve had conversations with a few different software developers over the years, who are all trying to fill this void of missing technology in the asthma world—things exist, but as I covered in my previous technology post… they don’t fit what I want. Maybe I’m a nerdy anomaly (okay, definitely), but come on, there have to be more of us out there who want useful, interpretable data like other developers, websites, and tech companies are providing to users.
This usually brings me back to a place where, whenever I do decide to quantify the dozen-plus factors I feel influence my asthma, I start to try to build something from scratch. I wanted to do this with as much automation as possible, but, the asthma world is not as connected as I’d like to see yet. I tried a few systems, but essentially gathered the same data.
Here’s what I logged.
Each aspect was also time-/date-stamped.
- Peak expiratory flow. I intended to log morning and evening values but did this imperfectly, thus the importance of the time stamps.
- FEV1 (Forced expiratory volume in 1-second). Just another lung function measure captured by my digital meter. Since I’m logging PEF, might as well note down the FEV1 as well.
- Oxygen saturation. For the most part, oxygen levels are stable in people with asthma. But, since I have a pulse oximeter at my disposal, I threw it into the mix.
- Humidity (percent). Automatically logged into a spreadsheet each day
- Step count. Tracked via my Fitbit Charge HR, automatically logged into a spreadsheet.
- Active minutes. Also logged by my Fitbit Charge HR.
- Symptoms/severity: Using a Google spreadsheet, I have a column each for coughing, chest tightness, and dyspnea (I don’t wheeze), and a severity ranking of 1 (mild) – 5 (severe) for the time I log that symptom.
I started off to log an overall daily reflection score of how many times I experienced each symptom and the severity, but figured logging as I went would be more accurate and easier to correlate.
What I’d liked to have added.
- Air Quality Health Index. Similar to the Air Quality Index in the United States. Both of these systems use a number to indicate how high or low air quality is—in Canada, ours interprets this data to whether or not people in high risk populations should be active outdoors on a particular day. I unfortunately couldn’t get this to log even pseudo-automatically for the places I spend time (Winnipeg, and out in rural Manitoba).
- Pollen counts. In my area, pollen is ranked as Low, Moderate and High. Since I don’t have significant allergies, the non-specificity worked for me. (Your asthma may vary—I’m not allergic to pollens, at least per my last test, so clumping them together works for me—people with allergies may wish to be more specific if they are nerding out like me.) I’d planned to assign these a 1-3 value for each day, but I couldn’t automate this, as hard as I tried—as it has little value to me, I’ll try this again in the future.
- Location data. I’d like to automatically grab the location tag for where I’m at when I log symptoms and taking my rescue inhaler so I can plot these on a map.
What did I gain?
Well, I got a lot of data, that’s for sure. My logging methods were imperfect, and I switched to a ridiculous system involving QR codes about two weeks in. Here are some visualizations.
A heatmap of sorts I made in Excel… This is a visualization of almost everything I logged over 10 days (or, 9.5 really. This chart concludes around 2 pm local time on the 28th).
Here’s the legend:
The symbols in the PEF, or peak flow, yellow zone box are not quite right but was really just to differentiate from FEV1. I wouldn’t read too much into the FEV1 colours either.
I’m not a stats person. Fortunately, Datasense exists (although it’s in Beta still). Here are some cool correlations I found using Datasense.
X-axis (horizontal) – FEV1 – Y-axis (vertical) – dyspnea