The Herxheimer Reaction Tracker: A Patient's Guide

You're three weeks into a new protocol. Last Tuesday felt like the worst day of your year. By Saturday you were upright again, almost human. This morning the knees are loud, the head is a swamp, and you're back at the same question you asked last time, and the time before that: is this a Herx, or am I dying?

That cycle is its own kind of damage. Every flare gets re-litigated from scratch because there's nothing on file. You can't compare this Tuesday to last Tuesday. You can't tell your LLMD whether the new med is harder than the old one or just harder this week. You're running a long, painful experiment and not writing down what happens.

This guide is about fixing that part. Not the suffering. The record. Specifically, how to build a Herx log that earns its keep over months of treatment, surfaces patterns across rotations, and gives you and your doctor something concrete to decide on.

Why a dedicated Herx tracker, not just your daily log

If you already do a daily check-in, you might wonder why Herxes don't just live in there. They can. They just won't tell you much later.

Daily logs answer the question "how was today." That's the wrong scale for a Herx. A Herx is an event with a shape: an onset, a peak, a tail, a trigger, and a list of symptoms that flared above your baseline. Flatten that into seven daily rows and you lose the shape. Two months later you see "bad week in late June" and have no idea whether it was the new antibiotic, a heat wave, or your period.

A dedicated Herx tracker treats each reaction as one record with all of its parts in one place:

• When it started and what changed in your treatment around then
• How bad the peak got and which day it hit
• How long the whole thing lasted
• Which of your baseline symptoms got louder, and by how much
• What you tried to manage it
• What, if anything, helped

The daily log keeps showing your baseline drift. The Herx log captures the spikes. They sit on parallel tracks and can be lined up later. That separation is the whole point. When you average a Herx into your daily data, your baseline trend looks worse than it is and your Herxes lose their fingerprints. When you keep them apart, you can answer questions like "is my baseline actually improving between flares," which is often the question that matters.

The data fields that actually matter

Every Herx log gets bloated if you let it. You don't need a 30-field form. You need the ones that future-you and your LLMD will actually read.

Start date. The day things shifted off baseline. Approximate is fine; you don't need a timestamp. If you can't tell whether yesterday's tiredness was the start or just a bad sleep night, leave it as "around the 14th." Future-you would rather have approximate than nothing.

Treatment trigger. This is half the value of the entire entry. Be specific. Not "antibiotic," but "doxycycline 100mg twice daily, started Monday." Not "added a herb," but "cat's claw 500mg, doubled from once to twice daily on Tuesday." Dose changes count. Brand changes count. Adding a binder counts. Coming off a med counts. Anything that changed in the chemistry goes here, even if you don't think it caused this Herx. "Nothing changed" is also a useful data point.

Peak severity, 1 to 5. One number for the worst day. Same scale you use everywhere else. A 5 means you couldn't do basic self-care. A 3 means you got through the day badly. Don't average the whole episode into one score. The peak is what tells you how hard this reaction hit.

Duration. Days from "things shifted" to "back to baseline." Days, not hours. If it tapered ambiguously, pick the day you stopped having to plan around it.

Which baseline symptoms flared. Not every symptom you've ever had. The ones that got loud during this Herx. Joints, brain fog, headache, sweats, sleep, mood, GI, skin, neuropathy. For each one, a quick comparison to your normal range. Knees were a 4 instead of the usual 2. Brain fog was a 5 instead of the usual 3. Sleep dropped from 7 hours to 4. The pattern of which symptoms light up during a Herx becomes one of the most useful signals you'll have. Most patients have a Herx fingerprint, a small set of symptoms that always get loudest. Yours will emerge after two or three logged reactions.

Interventions tried. Rest days, extra hydration, electrolytes, Epsom baths, heat, lymphatic drainage, binders, anti-inflammatories, an approved dose adjustment, a temporary pause. List what you actually did, not what you meant to do.

What helped. One word per intervention is enough. Helped. Neutral. Didn't help. Made it worse. Over time some interventions will reliably move the needle for you and others won't. Knowing your shortlist saves real suffering during the next reaction. The Illinois Lyme Association's overview is a starting menu, but the only menu that matters is the one you've tested on yourself.

That's it. Seven fields. Five minutes when you're feeling decent, three when you're not. Skip days while it's happening if you have to and back-fill at the end of the episode. An imperfect log written from memory three days later is still worth more than no log.

The patterns that emerge over months

A single Herx is a story with no comparison. The third Herx you log against the same treatment is the start of a pattern. The sixth is a chart your LLMD can actually read. These are the patterns that show up in the data, none of which you can see in any single bad week:

A treatment that produces one strong Herx and then tapers. New antibiotic, hit hard around day three, ride it out for a week, then the response gets quieter on subsequent doses. Common, often a good sign. The data shows a tall first spike followed by smaller bumps that fade.

A treatment that keeps flaring. Same dose, same med, every week or every cycle, the reaction comes back at roughly the same intensity. The data looks like a sawtooth. Some patients see a roughly four-week cycle that some clinicians link to the Borrelia life cycle. Worth flagging either way; it's easy to dismiss in conversation and obvious on a timeline.

A treatment you adjust to. Reaction one is a 4. Reaction two is a 3. Reaction three is a 1. By month two you barely notice the dose. The data shows a clean shrinking trend. This is one of the easier wins to miss without a record, because by the time you've adapted, your memory of the early reactions has softened.

A treatment that's getting worse. Reaction one was a 3 and lasted five days. Reaction four is a 4 and lasted ten. Same dose. Hard to argue from memory, clear on paper. The kind of thing that often prompts a real plan change.

A protocol where Herxes hit a different system each time. Joints in March, brain in April, gut in May. Some patients see this when treatment is reaching different bacterial reservoirs in sequence. Without a "which symptoms flared" field, it looks random. With it, it has a shape.

None of these patterns can be diagnosed from data alone. All of them are easier to recognize and discuss with a record. The goal isn't a research paper. It's enough resolution that the next treatment decision is informed instead of guessed.

Common confusions, and how the log clears them up

A lot of suffering early in Lyme treatment comes from not being sure what you're looking at. The log won't make every case obvious, but it narrows the question fast.

Herx versus drug allergy. An allergy usually shows up as hives, swelling, trouble breathing, or a fast-spreading rash, often within hours of the first dose. A Herx looks like a flu flare layered on the symptoms you already had. If your log notes hives or facial swelling on day one of a new med, that's a flag for urgent care, not a Herx entry. The Cleveland Clinic's overview and StatPearls reference both lay out the typical symptom set.

Herx versus drug side effect. Doxycycline causes nausea and sun sensitivity because of how the drug behaves in your body. A Herx happens because of what the drug is doing to the bacteria. If the symptom is "stomach upset every time I take the pill" and goes away if you take it with food, that's a side effect. If it's "joint pain at 4, brain fog at 5, slept four hours, exact pattern as last time the dose went up," that's a Herx. The "which symptoms flared" field separates these for you over time.

Herx versus disease progression. The most exhausting question. Progression usually means new symptoms or a steady worsening over weeks without recovery between them. A Herx, even a long one, has an arc. It peaks. It tails. You come back to baseline, even if baseline isn't great. If your log shows three reactions in a row that never returned to baseline between them, that's worth raising. If it shows three reactions that each ended cleanly, you're probably looking at a Herx pattern, not progression.

Herx versus an unrelated flare. Weather, stress, infection, hormones, a bad sleep streak, overdoing it on a good day. All of these can produce something that looks Herx-shaped in the moment. The treatment-trigger field is what saves you. If nothing changed in your protocol and you logged a heat wave plus three nights of bad sleep, the likely explanation is a non-Herx flare.

You won't always be sure. That's fine. Log it as a Herx with a note saying you're uncertain. Two months from now the pattern of your uncertain entries against your treatment timeline will tell you something either way.

How LymeTrack handles it

LymeTrack splits the two channels by design. Your normal day-to-day data lives in the 5-step daily check-in (CheckIn1Screen through CheckIn5Screen). Herxes live on a dedicated Herxheimer reaction screen (CheckInHerxScreen) that you open when something has clearly shifted off baseline.

The Herx screen captures the seven fields above. Severity is on the same 1 to 5 scale as the rest of the app, so a Herx peak of 4 means the same thing as a daily symptom of 4. Symptom flares pull from your existing tracked symptoms, so "knees flared from 2 to 4" is a real comparison rather than a guess. Treatments come from your own logged protocol, so you don't retype dose details every time.

Then Compass / Insights (InsightsScreen, HealthDayDetailScreen) does the part you can't do in your head. It lines up your Herx entries against your treatment timeline and your daily baseline. You can see whether the second round of a med was milder than the first. You can see your Herx fingerprint. You can see whether your baseline is rising between Herxes even while the Herxes themselves stay rough. That last one is often the question worth asking.

Factor tracking sits on top. Sleep, weather, stress, activity, and diet are tracked alongside symptoms, so when a flare doesn't line up with a treatment change, the confounders are visible instead of invisible. A flare that hits the day after three nights of bad sleep and a barometric pressure drop reads differently than a flare on a calm, well-rested week. Both can be logged as Herxes; the data tells you which is which over time.

You can also define your own symptoms, treatments, and factors through userPreferences. Nobody's Lyme presents like the textbook. If your Herx fingerprint includes burning feet or a specific kind of jaw clench, those belong on your list, not buried in an "other" field.

The doctor handoff

The point of all this isn't a personal archive. It's the fifteen-minute LLMD appointment where you have to decide whether to continue, adjust, or change a protocol.

LymeTrack's doctor-shareable report (gated by users.dataSharing consent) packages your Herx log alongside your daily baseline, treatment timeline, and factor data. Most LLMDs don't need a 40-page export. They need a clean view that answers a few questions. Did this rotation produce reactions that are getting better, worse, or staying flat? Are reactions clustering around a specific med or dose change? Is the baseline between reactions improving, holding, or eroding? What has and hasn't helped?

With that on screen, the conversation moves from "things have been rough" to "round three of this med looks milder than round one and my baseline is up half a point, so I'd like one more cycle and then re-evaluate." Different appointment. The one your doctor wants to be having.

A Herx is awful. It's also the most informative event in a Lyme treatment, because it's the moment when the disease and the medication are both visibly doing something. Throwing that information away because you felt too bad to write it down is the most common mistake patients make and the easiest one to fix.

Further reading

• LymeDisease.org: Understanding the Jarisch-Herxheimer Reaction. Dr. Chris Green's clinician-friendly overview, including why reactions can recur over a course of treatment.
• StatPearls (NIH): Jarisch-Herxheimer Reaction. Peer-reviewed clinical reference covering mechanism, presentation, and management across spirochete infections.
• Daniel Cameron, MD: Herxheimer Reaction Timeline. On how variable Herx timing can be in Lyme and why patient-tracked data often surfaces patterns a single appointment can't.