Wrong rosacea diagnosis often starts with an incomplete log, not a bad doctor.

The appointment that ends with the wrong prescription

Picture a 15-minute dermatology appointment. The patient has been flaring for eight months. She brings 30 photos saved in a camera roll, plus a notes-app entry that reads something like: 'Face gets red after wine, hot showers, sometimes for no reason. Had bumps on chin last month.' The dermatologist, working from that material, does their best. They might land on the right answer. But they also might not, and the cost of a miss here is specific: six weeks on the wrong drug before anyone knows.

Erythematotelangiectatic rosacea (ETR) and papulopustular rosacea (PPR) look related on the surface. Both involve facial redness. Both flare. But their first-line treatments move in opposite directions. ETR's persistent background flush and episodic vasodilation respond to vasoconstrictors, brimonidine tartrate being the most established topical option. PPR's inflammatory papules and pustules respond to topical antimicrobials: metronidazole, azelaic acid, ivermectin 1% cream (Soolantra). Give a PPR patient brimonidine and you are treating vascular tone in tissue that doesn't need it. Give an ETR patient metronidazole and you are targeting inflammation that isn't the primary driver. Topical ivermectin was recommended as a first-line therapy for rosacea inflammatory lesions and is FDA-approved for this indication. ([source](https://emedicine.medscape.com/article/1071429-treatment))

The differential isn't hard when the clinical picture is clean. It becomes hard when the data is thin.

What actually separates ETR from PPR in a clinical setting

The 2017 global ROSacea COnsensus (ROSCO) panel moved dermatology away from the older four-subtype model toward a phenotype-based approach precisely because the subtypes were overlapping in ways that muddied treatment decisions. ETR and PPR survive that shift as clinically meaningful distinctions, not because the names are tidy, but because the underlying biology and therefore the treatment logic diverges. The ROSCO panel recommended an approach for diagnosis and classification of rosacea based on disease phenotype, published in BJD doi: 10.1111/bjd.15122. ([source](https://pubmed.ncbi.nlm.nih.gov/27718519/))

ETR centers on persistent central facial erythema: background redness that doesn't fully clear between flares, flushing episodes that are episodic and often trigger-correlated (heat, alcohol, UV, exercise, temperature change), and visible telangiectasia (broken capillaries) in established cases. The primary dysfunction is vascular, neurovascular dysregulation driving cutaneous vasodilation.

PPR centers on inflammatory papules and pustules, without necessarily the same persistent erythema baseline. The papules can occur on a relatively normal-between-flares background. They can be mistaken for acne vulgaris, particularly in younger patients. The primary dysfunction is inflammatory, not purely vascular.

A patient can have features of both. But the first-line treatment question is which phenotype is dominant, and that question requires temporal data, not just cross-sectional photos.

Why most 'prepare for your derm visit' advice stops too short

Search 'how to prepare for a rosacea dermatology appointment' and you'll find a consistent answer: bring photos, keep a diary, note your triggers. This advice isn't wrong. It's incomplete in a specific way that matters clinically.

Photos are cross-sectional. A photo taken during a flare shows the dermatologist what a flare looks like. A photo taken on a calm day shows the between-flare state. But a folder of 30 unlabeled images, no timestamps in the metadata that mean anything, no context about where in the day or the trigger sequence the photo sits, doesn't answer the questions that separate ETR from PPR. The dermatologist can't tell, looking at a flare photo, whether the redness in that image is the baseline that never fully clears or a peak sitting on a clear foundation.

Written diaries have a different failure mode. A notes-app entry captures what the patient remembers, which is usually the dramatic moments: the worst flare, the most obvious trigger. It doesn't capture the between-flare state with any reliability, because on a calm day there's nothing that prompts writing. And it almost never captures time-of-day or duration in enough granularity to show whether the pattern is episodic or persistent.

The gap isn't effort. Most patients who prepare for dermatology appointments are trying. The gap is that the standard advice wasn't designed around the clinical differential that actually drives the treatment decision.

The three elements that make the differential possible

A dermatologist working toward the ETR-PPR differential needs three specific things from a patient's documentation. They are achievable before a first appointment. None of them require clinical training to produce.

One: A documented baseline showing the between-flare state.

This is the single most underrepresented data point in patient-reported documentation. The between-flare skin matters enormously for ETR because one of its defining features is that the background erythema doesn't fully clear. If a patient comes in with a flare photo but no between-flare comparison, the dermatologist is estimating whether the redness they see is persistent or episodic. A photo taken on a low-symptom morning, same lighting, same angle, within 30 seconds of waking and before anything has had a chance to trigger flushing, answers this question directly. One or two of these per week for three to four weeks gives a reliable baseline.

Two: Trigger-correlated events with timestamps showing episodic versus persistent pattern.

The clinical question isn't just 'does heat trigger flushing?', it's 'does the flushing resolve completely between triggers, and how long does resolution take?' A log entry that reads 'got hot in the car, face went red' has limited diagnostic value. A log entry that reads '2:40pm, sun exposure for 20 minutes, flush onset at 2:55, peak at 3:15, back to baseline by 4:30' tells a dermatologist something about vascular recovery kinetics. Multiple entries with timestamps build the pattern: episodic with full return to baseline (which can appear in either phenotype but in ETR often leaves persistent erythema even at nominal baseline) versus persistent elevation with superimposed spikes.

Three: A timeline long enough to distinguish rosacea papules from acne vulgaris.

PPR's inflammatory lesions are frequently misread as acne, particularly in patients who first present in their 20s or early 30s. The differential between acne vulgaris and PPR papules includes several clinical cues, but one of the most useful for patient-reported documentation is duration and location pattern over time. Acne lesions tend to occur at different sites, follow a comedone-to-papule-to-resolution arc, and correlate differently with hormonal cycles. PPR papules tend to cluster in the central face, lack comedones (blackheads, whiteheads), and don't follow the same hormonal distribution. A documented log across eight to twelve weeks, with consistent photo angles and location notes, lets the dermatologist see the pattern rather than a single cross-section.

What this looks like in practice: a concrete scenario

A patient in her early 40s has had flushing episodes for about two years. She's been told by her GP that it might be rosacea. She's been prescribed nothing yet. She has a dermatology appointment in six weeks.

In the current default scenario, she takes photos during flares, writes some notes, and arrives at the appointment with a phone camera roll and a general sense of her triggers. The dermatologist, in 15 minutes, sees her skin on that particular day, looks through a handful of photos, and makes their best call. Without a between-flare baseline, they're estimating her resting state. Without timestamped trigger events, the pattern is anecdotal. The ETR-PPR call might be right, but it might not be, and the 6-week trial window for whatever they prescribe will tell them, eventually.

In the informed documentation scenario, she starts six weeks out with a simple system. Every morning, before any triggers, she takes a photo with consistent positioning and lighting. When a flare happens, she logs the time, duration, suspected trigger, and estimated return to baseline. She notes which days her skin looked and felt genuinely calm versus just 'not flaring actively.' Over six weeks, a pattern emerges: her background erythema never fully resolves even on calm days, flushes reliably within 15 minutes of temperature change or alcohol, and returns to that elevated-but-not-peaked baseline within a couple of hours. No papules or pustules during the entire period.

That pattern, presented with the photo evidence, makes the ETR call much cleaner. The dermatologist has a baseline. They have episodic event data with time resolution. They have enough duration to be confident the absent papules aren't just a gap in the log. The brimonidine prescription, or whichever vascular-targeted option they choose, is grounded in actual patient data rather than a 15-minute cross-section.

The 6-week trial window is the real cost of a mis-read

Rosacea treatment trials are long. The typical assessment period for a first topical prescription is six to eight weeks, long enough for the drug to work if it's going to, long enough to declare it a failure if it isn't. NYU Langone doctors typically prescribe topical antibiotics for rosacea for six to eight weeks. ([source](https://nyulangone.org/conditions/rosacea/treatments/topical-treatment-for-rosacea))

A mis-read at the first appointment means the patient spends that window on the wrong drug, presents at the follow-up without improvement, and then starts the actual correct treatment. If the follow-up is eight weeks out, a mis-read can cost a patient three to four months before they're on the right thing. In a condition that affects quality of life significantly, NRS patient survey data consistently flags the psychosocial burden, including impacts on work, relationships, and self-confidence A 2025 NRS survey of 703 rosacea patients found 91% said rosacea affected their mental health, with 20% reporting impact all the time. ([source](https://www.rosacea.org/rosacea-review/2025/spring/new-survey-explores-rosaceas-impact-on-mental-health)), that's not a trivial delay.

The documentation gap isn't the dermatologist's problem to fix. They work from what they're given. Making that material better is something the patient can do before the appointment, and it's well within reach.

What structured tracking is actually designed to capture

The three elements above, between-flare baseline, timestamped trigger events, long-enough duration to pattern-match, aren't naturally produced by a general-purpose notes app or a camera roll. They require a consistent capture format: fixed-angle photos on a regular schedule, event logging with time and duration fields, and a view that shows baseline versus event days side by side.

We built Skinframe specifically around this clinical need. The log is structured for dermatologist-ready documentation: morning baseline photos, trigger event entries with onset/resolution timestamps, and a timeline view that makes the episodic-versus-persistent question answerable from the data. The photo capture is on-device and private, no cloud analysis, no skin-AI, no algorithm making calls about what your photos mean. That's the dermatologist's job. Our job is to make the data clean enough that they can do it well.

Skinframe is coming to iPhone. If you have a first dermatology appointment in the next six to twelve weeks, the waitlist is worth joining now, the baseline-building period starts before the appointment, not the week of.

And regardless of whether you use our app: talk to your dermatologist. Bring the three elements. The appointment is 15 minutes; the documentation is the part you can control.