Hermly Early access

Doctor reports

Fifteen minutes worth
fifteen months of context.

A 30-day migraine PDF your neurologist or headache specialist can read at a glance. Plain-English summary, attack calendar, cycle and pressure overlays, trigger correlations with odds ratios, medication-overuse headache (MOH) exposure, and topics worth discussing — never prescribing. Generated on your iPhone.

Reading time · 6 min Last updated · May 2026

In one paragraph

The Hermly Doctor Report is a one-page clinical PDF summarising the patient's migraine activity over the last 30 days. Sections include a plain-English narrative summary, an attack calendar with menstrual-cycle and barometric-pressure overlays, key metrics with prior-month comparison, trigger correlations expressed as odds ratios across a 90-day window, medication-overuse headache exposure flagged against IHS ICHD-3 criteria (≥10 triptan-days/month, ≥15 OTC-analgesic-days/month), and a short list of topics worth discussing with the clinician. Every page is labelled self-reported, not medically diagnosed. Hermly generates the PDF entirely on the iPhone and hands it to the iOS share sheet — the user decides where it goes. Hermly does not diagnose, treat, cure, or prevent any condition.

What this exists for

A 15-minute visit can't survive on memory.

Migraine visits are short. A neurology follow-up runs 15 to 20 minutes; a primary-care appointment, less. The patient walks in in some level of brain fog — sometimes literally mid-prodrome — and is asked: how many attacks this month, how long, how severe, what triggered them, did the prevention work? Recall is reconstructed verbally, in real time, under time pressure.

That reconstruction is unreliable. Episodic-migraine patients live in episodes, not calendars; a paper diary is the standard recommendation but most patients abandon paper logging within weeks. The clinician then adjusts therapy on partial data — doses titrated, preventives swapped, triptan limits set — using the best evidence available in the moment, which is often a month-old verbal recollection.

The Doctor Report exists so the patient and the clinician start the visit looking at the same numbers. The summary is three sentences the clinician can read in fifteen seconds. The trigger correlations are an empirical observation, not a theory. The medication-overuse panel is a fact the patient already saw before the visit — so the conversation about it isn't a confrontation.

What the PDF looks like

A sample report, layout-true.

The example below uses illustrative numbers. The actual PDF uses only the attacks, pain logs, and medications the user has recorded in Hermly — and the calendar / pressure / cycle overlays come from the same HealthKit and WeatherKit signals Hermly's forecast uses on Today.

Self-reported, not medically diagnosed.

April · Migraine summary

For Dr. Patel · Generated May 1, 2026
Plain English summary

April had 6 attacks, 2 fewer than March. 4 of 6 occurred between cycle days 25–28. Average duration was 5h 12min, down from 7h 04min last month. Sumatriptan was logged within the first hour in 5 of 6 attacks.

Attack calendar
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
Attack Cycle (menses or PMS window)
Key metrics
Attacks
6 −2 vs March
Mean duration
5h 12m −1h 52m
Topics worth discussing
  • 4 of 6 attacks fell on cycle days 25–28. Pattern is consistent across the last 3 months.
  • Mean barometric drop in the 24h before each attack was −9.2 hPa.
  • Sumatriptan timing improved (median 27 min into attack, was 64 min in March).

The PDF is one to two pages. The plain-English summary and the attack calendar are the only two visual anchors a busy clinician needs; everything else is supporting detail in named, scannable sections.

What's inside

Eight sections, all derivable.

Every section in the PDF is derived from data the user logged. When a section needs more data than is available, Hermly says so honestly — it does not fabricate numbers.

  1. Plain-English summary

    Three to four sentences: attack count this period, comparison with the prior period (direction + delta), mean duration, mean peak pain, mean time-to-triptan. The same narrative the in-app surface shows — single source of truth, no drift.

  2. Attack calendar

    A 30-day grid with one cell per day. Attacks are highlighted; menstrual-cycle phases (menses days 1–3 and luteal-late days 22–28) are overlaid so cycle-phase patterns surface visually. Visible even when zero attacks occurred, so the clinician can see that tracking was active.

  3. Cycle pattern heatmap

    For users with menstrual-cycle tracking on, a phase × cycle-day matrix across multiple cycles. A pattern that drifts across calendar months (e.g. attacks always falling on cycle days 25–28 regardless of the month's dates) shows up as a vertical stripe a calendar grid would hide.

  4. Temporal patterns

    Most-frequent onset window (4-hour buckets), morning-onset count, day-of-week skew, weekend frequency. Surfaces only when there are enough attacks to be statistically meaningful.

  5. Acute medication

    Triptan doses by family (sumatriptan, rizatriptan, etc.), naproxen doses, other doses. Mean time-to-dose after onset. Triptan-free days. Sourced from the medication logs the user recorded during each attack.

  6. Trigger correlations (associated factors)

    Up to five ranked findings (cycle phase, barometric drop, sleep deficit, perceived stress, HRV pattern) with attack shares and odds ratios. Computed over a 90-day window so there's enough data; surfaces only when at least four attacks are logged in that window.

  7. Medication-overuse headache (MOH) exposure

    Triptan-days and OTC-days against the IHS ICHD-3 thresholds (≥10 days/month for triptans, ≥15 days/month for OTC analgesics). Hermly flags exposure as a number for the clinician to interpret; it does not diagnose MOH.

  8. Topics worth discussing

    Three to five auto-generated discussion points derived from the period's patterns — for example, "4 of 6 attacks fell on cycle days 25–28; pattern is consistent across the last 3 months." Phrased as observations the patient and clinician can discuss together, never as recommendations.

  9. Trend (last 6 months)

    A trailing-6-month rollup with attack count, mean peak pain, and triptan-days per month. Lets the clinician see where the headline month sits inside the longitudinal curve before reading the rest of the report. Visible only when at least 3 of the trailing 6 months recorded attacks; the section is hidden on near-empty data because a "trend" surface on two spikes is misleading. Pure values — no "improving" / "worsening" labels and no color verdicts.

Who generates one

Three composite scenarios.

The scenarios below are composites — built from the situations the Doctor Report is designed to serve, not real named patients. Each names the underlying clinical problem and what the report does about it.

Composite scenario · neurology follow-up

Quarterly visit, medication adjustment

A patient with episodic migraine sees her neurologist every three months. The previous visit, asked "how many migraines last month?", she said "maybe 5? maybe 7?" — and the preventive dose was adjusted on that estimate. This visit, she opens the Doctor Report on her phone two minutes before walking in.

The problem: medication decisions get made on patient recall that is unreliable for episodic events spread across weeks. Treatment quality suffers when the shared baseline is fuzzy.

What the Doctor Report does: shows 6 attacks this month (down from 8), mean duration 5h 12min (down from 7h 04min), 4 of 6 falling on cycle days 25–28, sumatriptan logged within the first hour in 5 of 6 attacks. The neurologist reads it in twenty seconds and starts the visit at the cycle pattern, not the "how have you been" preamble.

Composite scenario · headache-specialist intake

First visit after a long referral

A patient finally gets a referral to a headache specialist after years of inadequate care from primary care. The intake form asks the questions every headache-specialist intake form asks: attacks per month, characteristics, triggers, medications tried and discontinued, response history.

The problem: the patient has months of context the specialist needs but no structured way to convey it. The intake form ends up populated from memory; the specialist's working picture is fragmented.

What the Doctor Report does: emailed to the specialist's intake nurse three days before the appointment. The trigger correlations highlight a barometric-pressure pattern primary care missed. The MOH-exposure section flags that the patient is approaching the triptan threshold. The visit starts already informed.

Composite scenario · MOH conversation

Catching the threshold before the clinician has to

A patient's triptan use has been creeping up — partly because her preventive isn't quite right, partly because the attacks have intensified. Her last two Doctor Reports show triptan-days approaching the IHS threshold for medication-overuse headache (10 days/month for triptans, per ICHD-3).

The problem: the MOH conversation usually happens reactively — once the clinician notices a refill pattern in the prescription record. By then the dynamic is already in place, and the conversation feels like a correction.

What the Doctor Report does: the patient sees the MOH-exposure panel during her own monthly check-in. She brings it up in the next appointment proactively. The clinician adjusts the prevention plan to address the underlying frequency; the MOH risk doesn't have to become a diagnosis.

How it's made

Three taps. No portal.

  1. 1

    Log attacks as they happen

    Start each attack from the Action Button, Lock Screen widget, Apple Watch, or in-app button. Record pain levels and medications during the attack; end the attack when it resolves. Hermly stores everything on the iPhone.

  2. 2

    Open the Doctor Report

    Reports tab → Doctor report. Hermly aggregates the last 30 days automatically. Optionally type a one-line reason for the visit; it appears at the top of the PDF for the clinician.

  3. 3

    Share — or write to Apple Health

    Tap the share icon. Hermly renders the PDF and hands it to the iOS share sheet — email, AirDrop, Files, patient portal. Headache events are also written to Apple Health, so an iPhone-equipped clinician can read the same source data directly.

Privacy

The report stays on your iPhone.

The Doctor Report is generated by code running inside the Hermly iOS app, against the SwiftData store on the same iPhone. The resulting PDF is written to a temporary file in the app sandbox and handed to the iOS share sheet. From there, the user picks the destination — Mail, AirDrop, Files, or a clinician's patient portal — and the operating system handles the transfer.

Hermly's servers never see the PDF, the underlying attack records, the pain logs, or the medication entries. The backend exists for the landing page you are reading, the StoreKit subscription state for Hermly Pro, and anonymous event counters that contain no health values. This is the same privacy posture documented on the privacy promise page.

The Apple Health writeback is reciprocal: each ended attack becomes a Headache-category entry in Apple Health, so the same source data is visible to a clinician through the Health app rather than the Hermly PDF. Apple Health entries never leave the user's device either — that's Apple's constraint, not Hermly's.

What the Doctor Report is not

Important limits.

Not a medical device

Hermly is not FDA-cleared as a medical device. The Doctor Report contains no diagnostic language and never uses the words "diagnose", "treat", "cure", or "prevent". It is a patient-collected summary the clinician interprets.

Not a substitute for clinical judgment

The trigger correlations are statistical associations over a 90-day window of self-reported data. Causality, mechanism, and what to do about a pattern belong to the clinician — not to the report.

Not a treatment recommendation

The "topics worth discussing" section names observations, never prescriptions. The Doctor Report never tells the patient to start, stop, switch, or escalate a medication. All medication decisions belong to the prescriber.

Not a continuous-care platform

Hermly does not transmit reports between visits, does not monitor for emergencies, and does not alert clinicians to changes. The patient brings the report; the patient owns the share. Continuous care is the clinic's job.

Common questions

FAQ.

What is a Hermly Doctor Report?

A 30-day patient-collected PDF summary of the user's migraine activity, designed for a clinical visit. It includes a plain-English summary, attack calendar with cycle overlay, key metrics with prior-month trend, trigger correlations with odds ratios, medication-overuse headache (MOH) exposure, and topics worth discussing with the clinician.

Will my neurologist accept the Doctor Report?

Most neurologists and headache specialists welcome structured patient logs — they would otherwise be reconstructed verbally during the visit. The Doctor Report is explicitly labelled 'Self-reported, not medically diagnosed' on every page so the clinician knows exactly what they are looking at: patient evidence, not a diagnostic claim.

How is this different from a paper headache diary?

A paper diary has the same content shape but requires the patient to fill it out by hand, which most users abandon within weeks. Hermly populates the report automatically from attacks the user logged in the app, then derives the temporal patterns, trigger correlations, and MOH exposure that a paper diary cannot compute.

Does the report diagnose anything?

No. Hermly is not a medical device. The Doctor Report contains no diagnostic language. It surfaces patterns in patient-reported data and flags concepts like medication-overuse headache risk against IHS ICHD-3 criteria — but the clinical judgment of whether a diagnosis applies belongs to the clinician, not to Hermly.

What is medication-overuse headache (MOH), and why does the report track it?

MOH is a headache disorder defined by the International Headache Society where over-frequent acute medication use (≥10 days/month for triptans/opioids/combination analgesics, ≥15 days/month for OTC analgesics) actually causes more headaches. The Doctor Report surfaces the patient's triptan-days and OTC-days against these thresholds so the topic can come up in the visit before it becomes a problem.

Does the report send anything to Hermly's servers?

No. The Doctor Report is rendered to PDF on the iPhone and handed to the iOS share sheet. The user picks the destination — Mail, Messages, AirDrop, Files, their clinician's patient portal. Hermly's servers never receive the file, the attack records, or the medication logs.

What's the difference between the Doctor Report and the Evidence Report?

Doctor Report is a 30-day clinical snapshot built for a treating clinician — trigger correlations, MOH risk, topics to discuss. Evidence Report is a 6 to 24-month longitudinal dossier built for a VA adjudicator, HR office, or insurance reviewer — distinct missed-work days, prostrating-attack counts, hospitalization flags. Same underlying data, different audience and shape.

Can I share the report through Apple Health?

Yes. Each ended attack is also written to the Apple Health Headache category with the recorded severity. A clinician with iPhone access can open Health → Browse → Headache and read the same source data the Doctor Report aggregates — useful when the PDF format is not accepted by a particular portal.

How often should I generate a Doctor Report?

Generate one before each clinical visit; the window auto-aligns to the trailing 30 days. Many users also receive an optional monthly notification when the previous month's report is ready, which surfaces it as a checkpoint even between visits.

What if a section is blank or marked as a placeholder?

Hermly never fabricates numbers. If a section needs more data than the user has logged (for example, trigger correlations require at least four logged attacks within the trailing 90 days), the section honestly reports the gap to the clinician instead of guessing — that the threshold was not yet crossed and which number is missing.

Compared to

Doctor Report vs Evidence Report.

Same underlying data store, two different audiences and shapes. Pro users get both — most generate a Doctor Report before a neurology visit and an Evidence Report when a claim or HR office asks for one.

Dimension
Doctor Report
Evidence Report
Audience
Neurologist, headache specialist, primary-care clinician
VA adjudicator, VSO, HR office, insurance reviewer
Window
30 days (current month)
6, 12, or 24 months
Shape
Aggregated stats + trigger correlations + topics to discuss
Monthly rollup + per-attack chronological ledger
Distinctive fields
Mean pain, mean duration, trigger odds ratios, MOH exposure, cycle-phase heatmap
Prostrating count, missed-work days, ER visits, hospitalizations
Tone
"Here's what to discuss"
"Here's what happened, dated and timestamped"

Early access

Hermly is in private beta.

Doctor reports are part of Hermly Pro alongside the Apple Watch app, Spouse Mode, evidence reports, and the year-end summary. Leave your email and we'll write when there's something real to show.

Self-reported, not medically diagnosed. Hermly is not a medical device.