Tracking an Antibiotic Rotation for Lyme

It's month three of the protocol, and your LLMD asks how the cefdinir round went. You stare at her. Was that the one with the GI flare, or was that amoxicillin? The doxy round was rough but you can't remember if it was rough at the start or the end. There were two weeks of pulse dosing somewhere in there. Or three. You took notes for the first month, and then life happened.

You're not failing. Tracking a rotation by memory is just impossible. By the time you're four or five drugs in, the rounds blur together and the decisions about what to do next get made on a fog of half-remembered impressions.

Why a rotation is harder to track than a single course

A 21-day course of doxy is one thing to log. A rotation is a different problem.

You're often on more than one agent at a time. Doxy plus a cyst-buster like tinidazole on weekends. A cell-wall antibiotic stacked with an intracellular one. Herbal antimicrobials like cryptolepis or Japanese knotweed running underneath the prescription drugs. A binder taken away from everything else.

The schedules vary by drug. Some are continuous. Some pulse: five days on, two off, or two weeks on, one off. Pulse schedules exist partly to hit Borrelia in different stages of its life cycle, since spirochetes, round bodies, and biofilm communities don't all respond to the same drug at the same time. That's the theory behind rotating in the first place.

The Herx pattern is different per drug. Doxy might give you a clean three-day flare that lifts. Cefdinir might be silent for ten days and then crash you at the end of week two. A pulsed protocol can give you a small Herx every cycle, and the cycles start to blend.

Side effects also change which drug you blame for what. GI upset on amoxicillin can look like a Herx, or it can be the antibiotic itself wrecking your gut. Doxy nausea isn't a Herx. Sun sensitivity isn't a Herx. Telling the two apart matters because one means the drug is working and one means the drug is the problem.

If you want a primer on Herx itself, the Herxheimer reaction guide covers the mechanism in detail.

What's actually worth logging

You don't need to log everything. You need to log the things that, three months from now, will let you and your doctor look back and tell rounds apart.

For each drug or combo:

• The name and dose. "Doxy 100mg twice daily" beats "doxy."
• The schedule. Continuous, or pulsed (and the pulse pattern). Whether you took it with food.
• Start and stop dates. If a round was cut short or extended, why.
• Herx severity, on a consistent scale. The 1–5 severity used in the rest of your tracking is fine. Note when it started and how long it lasted.
• Baseline symptom severity during the round, separate from the Herx. How were you when the Herx wasn't acute? Better than the prior round, worse, the same?
• Notable side effects. GI symptoms, rashes, yeast issues, candida flares, headaches that aren't part of your usual Lyme pattern.
• A one-line gut read at the end. "Felt like this one helped." "Couldn't tolerate after week two." "Nothing changed." Your impression at the time is data, even if it's wrong, because it's what you'll otherwise misremember.

That's it. Five or six fields per round. Not a research study.

The question the data is supposed to answer

Across three to six months of rotation, your doctor is going to want to know a few things to make the next decision:

Which rounds correlated with the biggest drops in baseline symptoms? Not the worst Herx, not the easiest. The ones where, two weeks after stopping, your fatigue or joint pain or fog was measurably better than before.

Which triggered the worst Herx, and did the Herx come with a clinical improvement afterward? A brutal Herx that yields nothing is different from a brutal Herx that breaks a symptom plateau.

Which did you adjust to, and which never settled? Some drugs ramp up Herx, peak, and then your body gets through it. Others stay rough the whole round. That's a different signal.

Which side effects were dose-dependent? Sometimes a drug looks like it's failing when really the dose was too high or the schedule was off.

You can't answer any of that from memory at month four. You can answer it from a log.

Things that make the data lie

A few traps to know about, because chronic Lyme tracking has all of them at once.

Sample size. You're running an N-of-1 experiment with maybe four to eight rounds total. That's not a lot of data points to draw conclusions from. A round that looks great might have looked great because of something else.

Confounders. You started cefdinir the same week your kid got the flu and you didn't sleep for four nights. The "bad cefdinir round" might have been a bad sleep round. Track the obvious ones: sleep, stress, big life events, season, menstrual cycle if relevant. The Herx guide on environmental factors gets into this.

Nocebo and expectation. If you read online that cefdinir is "the hard one," you'll watch for hard. You'll find it. The reverse happens too. Your treating doctor's enthusiasm about a drug shapes what you notice. This isn't a flaw, it's just how human brains work. Logging in real time, before you've decided how a round went, helps a little.

Carry-over. The drug you were on last month might be partly responsible for what's happening this month. Borrelia die-off doesn't end when you stop the antibiotic. A "good first week" of the new drug might be the tail end of the old one.

None of this means tracking is pointless. It means you treat your data with appropriate humility and let your LLMD weigh it against everything else they're seeing.

How LymeTrack handles a rotation

The app was built around the idea that rotations need their pieces tracked separately, then pulled together at the end.

In the 5-step daily check-in (CheckIn1Screen through CheckIn5Screen), each treatment you're currently on gets tagged on the day you take it. If you're running doxy continuously and a tinidazole pulse on weekends, both show up on weekend check-ins, only doxy on weekdays. You can add custom treatments for herbal combos, binders, or anything your LLMD has prescribed that isn't on a default list.

The Herxheimer reaction screen (CheckInHerxScreen) is separate from the daily check-in for a reason. Each Herx logs which drug or combo you were on when it started, how severe it got, which symptoms flared hardest, and how long it lasted. When you rotate, the next Herx logs against the new drug. At month four, you can see a per-drug Herx history rather than one undifferentiated mess.

Factor tracking runs underneath all of it. Sleep, stress, weather, activity, diet. These are the confounders. When the Compass shows you that the cefdinir round looked rough, factor data lets you check whether your sleep tanked that month, or your pollen allergies were spiking, or you were traveling.

The Compass and Insights view (InsightsScreen, HealthDayDetailScreen) is where rotation patterns become visible. Across months, it correlates each drug round with baseline symptom trends, Herx episodes, and factors. Not as a verdict, as a starting point for the conversation with your doctor. The doctor-shareable export pulls all of it into one document so you don't have to summarize from memory in the appointment.

None of this tells you which drug to take. That's not the app's job and not your job either. It's your LLMD's call. The app just makes sure that when she asks how cefdinir went, you have an actual answer.

Further reading

• ILADS Treatment Guidelines. The full guideline document on individualized antibiotic treatment for persistent Lyme, including the rationale for tailoring regimens to the patient.
• Evidence assessments and guideline recommendations in Lyme disease (Cameron et al., Expert Review of Anti-infective Therapy, 2014). The peer-reviewed paper underlying the ILADS guidelines, with a long discussion of treatment for persistent disease.
• Superior efficacy of combination antibiotic therapy versus monotherapy in a mouse model of Lyme disease (Frontiers in Microbiology, 2023). Preclinical evidence on why combination and rotation approaches are being studied. Mouse model, not human, but useful context.

Track the rounds, track the factors, and bring the picture to your doctor. The decisions about what comes next aren't yours to make alone, and you don't have to remember it all.