You open your symptom journal three weeks into a treatment change to see if anything's shifting. The last twelve entries say some version of "tired again," "rough day," "headache, kinda bad," and "feeling weird." You stare at it. None of those entries can be compared to any of the others. You have no idea if this week is better than last week, or if you've been at this same level the whole time and just forgotten.
That's the moment most people realize free-text journaling alone isn't enough. Words drift. Numbers don't.
The problem with vague entries
"Kinda bad" today and "kinda bad" three weeks ago look identical on the page, but they almost certainly weren't the same. Maybe today is a 3 and that Tuesday in May was a 4. Maybe your fatigue was a 4 then and it's a 2 now, but the headache flipped the other way. With prose alone, you can't tell. The data is there in your body but it's not in any form you can compare.
This matters in two specific places.
The first is your own pattern recognition. If you can't compare yesterday to last month, you can't see whether a treatment is helping, whether a flare cycle is actually monthly, or whether your bad days really do cluster around weather changes. You're stuck reacting to whatever you feel right now, with no real sense of the trend.
The second is your doctor. An LLMD with a 30-minute slot is going to ask "how's the joint pain been since we changed your protocol?" If you say "up and down, I think mostly down," that's not enough to make a decision on. If you can say "fatigue averaged 3.5 the first two weeks, dropped to 2.2 the next two, joint pain unchanged at about 3," that's a different conversation. They can act on that.
Severity scales aren't about pretending to be a clinical study of one. They're about giving the same answer the same number, so future-you can recognize the pattern.
What works
A handful of small choices make scoring something you'll actually keep doing.
Use a 1 to 5 numeric scale with anchored descriptions. Pick what each number means, in plain language, and write it down once. For fatigue, that might be: 1 is "normal energy for me," 2 is "a little dragging but functional," 3 is "noticeably tired, slowing me down," 4 is "couch-level, hard to do normal things," 5 is "can't get up, need help." The anchors are the whole game. Without them, a 3 today and a 3 next month are different numbers in different moods.
Score per symptom, not "overall." Lyme doesn't move as one thing. Your joint pain might be at 4 while your brain fog is at 2, and an "overall: 3" hides both. Score fatigue, brain fog, joint pain, sleep, headache, mood, and any other symptom that matters to you, separately. That's what makes patterns visible later.
Score at roughly the same time of day. Most patients pick early evening, after the day has played out but before sleep blurs the memory. Morning works too if you're consistent. What you want to avoid is scoring breakfast some days and bedtime others, because Lyme symptoms shift across a day and you'll be comparing apples to a different kind of apple.
Pair the numbers with one short note. A single line of context: "didn't sleep well, big work meeting," or "third day on new dose, knees worse." The number is what you compare. The note is what tells you why.
That's it. You don't need more than that.
The pitfalls
A few things will quietly wreck your data if you let them.
Recall bias is the big one. If you skip a day and try to score yesterday from memory the next morning, your number will be wrong. Research on patient-reported outcomes is clear that people remember symptoms more accurately within the same day than even one day later, and accuracy keeps dropping after that. People also tend to remember symptoms as worse than they actually were when they're currently feeling bad, and milder than they were when currently feeling fine. So either score today, today, or leave it blank. A blank day is honest. A guessed day is misleading and you can't tell which is which a month later.
Drift is the second one. The 3 you score in week one and the 3 you score in week eight may not mean the same thing, because your sense of "normal" has shifted. People who feel awful for a long time recalibrate. What used to be a 4 starts feeling like a 2 because the baseline moved. There's no perfect fix, but re-reading your anchored descriptions every few weeks helps. So does re-anchoring after a clear life event: a treatment change, a long flare, a recovery. Just spend five minutes asking "is my 3 still a 3?"
Over-precision is the third. A 1 to 10 scale sounds more accurate than 1 to 5, but in practice it adds noise. Most people can't reliably tell a 6 from a 7 in their own fatigue. They can tell mild from moderate from severe. The extra granularity ends up being random, and random data washes out real patterns. 1 to 5 is the sweet spot for self-tracking. If clinical research wants ten points for statistical reasons, that's their problem. You're trying to see your own trends.
How LymeTrack handles it
The 5-step daily check-in in the app is built around a 1 to 5 severity scale with descriptive levels for each symptom. The descriptions live in the symptom templates, so when you score "joint pain," you see "mild" through "severe" defined for joint pain specifically, not as generic numbers. Same for fatigue, brain fog, headache, sleep quality, mood, and anything custom you've added. That's the anchoring step done for you.
Scoring is per symptom, not overall. You move through your active symptoms one at a time, tap a level, and move on. Most check-ins take under two minutes once you've done a few.
If your symptoms shift through the day, you can do multiple check-ins per day. Some patients run a morning and an evening, especially during a treatment change or a flare. The app keeps each one separate, so you can see, for example, that your fatigue is a 2 in the morning and a 4 by 6 PM, every day this week. That's a pattern you'd never spot from a single daily score.
After two to four weeks of data, the Compass and Insights view starts showing you trends. Which symptoms are moving, which are flat, which days of the week or treatment cycle they cluster around. The insights aren't worth much in the first week, and the app doesn't pretend they are. You need enough scored days for the noise to average out and the real signal to come through. Once you have it, you can hand a clean trend chart to your LLMD and have an actual conversation about what's working.
The point of all of this isn't to turn yourself into a science experiment. It's to make sure that "I think I'm a little better" can be checked against "am I actually a little better."
Further reading
- Numeric Pain Rating Scale on Physiopedia. A clear breakdown of how the NRS works, why it's a segmented version of the VAS, and how it's used clinically.
- Comparing the VAS and NRS in patient-reported outcomes (Journal of Rheumatology, 2021). A research comparison of the two main severity scales, with patient-preference data.
- Validity of end-of-day recall assessments of pain and fatigue (PMC). The recall-bias research underlying why same-day scoring beats catching up later.
LymeTrack is a tracking tool, not medical advice. Talk to your LLMD or treating physician before changing a treatment plan.