If you have not seen the answers here, please look at this related question first:
What Are Examples of the Different Approaches to Home Eye Pressure Monitoring? | Ask FitEyes
A FitEyes member provided the following answer via our email discussion list:
My IOP follows a certain daily pattern. I am getting pretty much similar numbers every day at certain data points like 7am, 7pm etc. After I change my medications I continue to measure my IOP, but I recognize that it might take a while for the washout of the old medication to complete.
That's a good answer. It might be the complete answer for many of us.
However, I will expand on it for anyone who desires to be more detailed or rigorous. (The word "methodology in the title suggests to me that the asker wishes to be rigorous.) You do not have to be this rigorous to derive a lot of value from your own at-home IOP measurements, but some of us may want a research-quality data set of our own IOP. Here are some ideas that may help you achieve that.
I suggest Dr. Baumgarten's approach is suitable for engineers. But my own approach is also suitable.
For any rigorous approach, you want to use the scientific method, use the appropriate statistics, and use precision.
Before trying to find the difference between two different medications (or any other factor you wish to ask your tonometer about) I suggest you have these things in place:
1. establish your basic methodology. I have some training in science and I generally understand how to apply the scientific methodology. If you don't, read up on it.
Make sure you have defined these specific parts of your methodology:
How do you record your IOP data? For example, I use the FItEyes software. That software eliminates IOP measurements with low scores from reports, but it doesn't delete that data. If you delete or ignore any data, you could be introducing a source of bias into your data. As both the experimenter and the subject, you should never make the decision to ignore certain data.
The FitEyes software doesn't include "bad" measurements in the IOP exam summary. However, if you later decide that a good measurement really wasn't good (such as by looking at its waveform) or that a bad measurement really was acceptable, you can re-examine your data because all the original data is available. If you are simply recording your IOP on paper or in a spreadsheet, you will be missing valuable data for multiple reasons. The FitEyes software for a Reichert 7CR tonometer records and saves over 800 data points for every single IOP measurement. It also saves every measurements, even "bad" ones, while only computing statistics and summaries based on valid measurements.
How many puffs per measurement? How many measurements per lateral exam? How many IOP exams per day?
Do you use the before-after activity approach, the clock-based (time-of-day-based) approach or both? Dr. Baumgarten used a clock-based approach. I use an activity-based approach.
How long will you test a new treatment (e.g., a glaucoma medication)? How many times will you repeat the test before you reach a conclusion?
Your method needs to be repeatable and consistent.
Don't introduce new variables (confounders). For example, if I am looking for the effects of a lifestyle activity (or a food or supplement) I do not alter my glaucoma medication at all. I take the same medication in the same amount at the exact same time of day, eliminating every variable that I can.
2. For things that are not part of your methodology or treatment, randomize as much as you can. For example, I don't start an IOP exam with the same eye every time.
With these factors in place, establish a baseline. You do this by recording your IOP over time.
In regard to your specific question, here's the way I approach it. I do not lump all IOP measurements from a day into one mean (average). I focus on specific measurements that are somewhat repeatable.
1. my upon awakening IOP exam. I take my IOP upon awakening exactly the same way each day. My methodology for this first IOP measurement of the day is very carefully considered. I watch for confounding factors, and that includes thoughts. In IOP research the subject's mental activity level and mind content are confounders that many researchers miss. If I wake up and start thinking about work before checking my IOP, for example, that is a confounder (so, as a rule, I do not do it).
2. my IOP exam after key activities. For example, if you take an evening walk most days, your IOP after that walk could be a good reference point. As another example, I look at my IOP after my morning shower. For many people, it could be important to include a reference point related to work -- maybe that could be your first IOP measurement after work each day.
3. my peak IOP and my lowest IOP each day (regardless of when they occur).
4. if, unlike me, you always take IOP measurements based on clock-time, and your daily routine is consistent, pick times of the day that you know (from your baseline data) to give consistent IOP values. You can compared your 7:00am IOP value over time, and this will very likely reflect changes in your IOP due to the differing efficacy of different glaucoma medications.
I try to be very aware of recurring patterns and delayed affects. For example, some stressful events will continue to show up as elevated "upon awakening" IOP for several days after the stressful event has ended. (I have published specific examples on the FitEyes blog in years past.) One stressful event can negatively affect one's IOP for a full week (or more, maybe). If you don't know your own pattern, you can be misled by a naive look at the data.
Many things could have long lasting effects on your IOP. For you, maybe it is a food. Maybe the food you ate on Monday affects your IOP starting on Tuesday and continuing thru Wed. If you don't recognize that pattern (including especially the delayed onset), you could be misled.
You could call these things confounders. We do need to constantly be on the lookout for confounders. However, what I have in mind for this section specifically are patterns that I have recognized. Don't miss delayed reactions. Know how the patterns unfold for you. The following blog post might be helpful:
Detective Work Required - Eye Pressure Going Up While On Relaxing Vacation | FitEyes.com
I have generally not needed any sophisticated statistics when looking at my own IOP data.
4. I do not use a raw daily mean because I don't take the same number of IOP measurements per day. I could skew a mean by, for example, simply taking a few extra IOP measurements in the evening when my IOP is lowest. However, if you use Dr. Baumgarten's clock-time approach and you perform the same number of IOP exams each day you could use an overall daily mean and related statistics.
5. The statistics I use include mean and range (or standard deviation), but I apply those to items 1, 2 and 3 above, rather than to the entire day. For example, I look at statistics and charts of my upon awakening IOP over periods of time.
One of the challenges is "experimental design". Because this experimental design involves n=1 (one subject -- you) and it is not occurring under controlled laboratory conditions, we have to look at experimental design a bit differently. If any of you scientists have suggestions, please jump into this conversation.
I have always sought to take a lot of IOP measurements and to repeat multiple trials of my experiments. For example, if I want to compare Rocklatan vs the Vyzulta+Rhopressa combo, I would probably repeat the trial multiple times and it will take me multiple months to reach a conclusion. Even if I think I see a clear difference very quickly, I will repeat my trial to be sure. The more trials you perform, the more reliable your conclusion. The repeated trials remove confounders like seasonal allergies, current events, and random effects we could never consciously identify.
Many studies address these random effects by enrolling a lot of people. In self-experimentation, we can address that through repeated trials, each of which needs sufficient data. I always try to take as many IOP measurements as I can.
When testing medications, I do not use a true washout period. As a patient, I don't want to do that to myself. Therefore, as a self-experimenter, I have to find other ways to group my data based on different medications. I have not found any magical method except to be patient. If I have switched medications, I just make sure I have waited long enough to be confident the old medication is out of my system.
I have established some baselines through prior IOP monitoring. For example, for most people timolol would cease to provide any IOP reduction after a day (or less). For me, timolol will continue to reduce my IOP for at least a week after using it. So I wait at least two weeks to be sure the timolol is out of my system. Some studies use a 30 day washout period when changing medications. I don't think I need to wait that long. But not waiting long enough is the way most of us are likely to go wrong.
In contrast, prostaglandins will continue to reduce IOP for more than one week for many people. (Side note: for some people, they could switch from daily use of a prostaglandin to weekly use without any loss of IOP control.) In my case, the effect of prostaglandin medications is gone within about 2 days according to my IOP data.
As a community, we could continue to expand on this topic. If you have a specific approach that is not covered here, please add yours as an answer here, or as an example at this topic:
What Are Examples of the Different Approaches to Home Eye Pressure Monitoring? | Ask FitEyes
We could also ask new questions to go into more detail. For example, one question might be, "When rigorously testing the efficacy of different glaucoma medications, how do you transition from one to the other without a washout period?" Maybe many of you have a better answer on that point than I gave above.
In my case, I simply switch directly from one to the other without any washout, and then I continue with the new medication long enough for the old to be out of my system before I tag my IOP data as belonging to the new medication's data set.