Therapy guide · medical education
Perioral dermatitis: the full therapy pathway
The one facial rash where the cream that clears it is usually the reason it keeps coming back — what withdrawal actually involves, and what prescriptions are for once the cause is gone.
Before anything else
Not everyone should start at the bottom of this pathway. These presentations are reasons to
see a clinician without waiting out an over-the-counter trial first.
- Eye pain, light sensitivity, or blurred vision Seek care quickly
A red eyelid has several possible explanations and only some of them are skin. Pain in the eye, sensitivity to light or any change in sight needs assessing the same day by a doctor, an optometrist or an eye service — not folded into a dermatology appointment several weeks away.
- A facial rash arriving with fever, spreading swelling, or feeling unwell Seek care quickly
Skin that becomes hot, tender and swollen over hours, particularly alongside a temperature, is far more likely to be an infection than a flare of this condition. That needs urgent assessment and usually antibiotics rather than the removal strategy described below.
- Painful blisters grouped on one side of the face or near an eye Seek care quickly
A one-sided, painful, blistering eruption in a nerve distribution — especially involving the nose or the eyelid — is treated as shingles until proven otherwise, and shingles near the eye is sight-threatening. Antiviral treatment works far better started early, which makes the delay the main risk.
- Any temptation to stop a prescribed inhaler, nasal spray or steroid tablet See a dermatologist soon
An inhaled or oral corticosteroid may well be contributing, and it is still holding down asthma, a lung condition or something else that is more dangerous than a facial rash. Changing it is a conversation with the prescriber — about technique, rinsing, a spacer, or a different preparation — and never a unilateral decision made because of the skin.
- Firm red-brown lumps in a child, or lesions leaving marks See a dermatologist soon
The granulomatous variant runs longer than the ordinary form and scarring has been reported in the more severe childhood cases. Timing genuinely matters here: assessment while it is mild is a substantially easier conversation than assessment after it has been mistaken for eczema for six months.
- A single crusted, bleeding or non-healing spot that behaves differently from the rest See a dermatologist soon
The central face is the commonest site on the body for skin cancer, and a single lesion can hide comfortably inside a background of many similar-looking ones. Anything that persists, scabs repeatedly or bleeds while everything around it comes and goes needs a clinician's eye.
Which route does this look like?
Four routes, and the questions that separate them are mostly about what has been applied or inhaled rather than about how bad it looks. Answers stay in this browser and are never sent anywhere.
Eight questions, none about identity and none required. The result names a route and a position on it, then filters the page down to that column; every other route stays one button away.
One thing to be clear about. This places a presentation on a map rather than naming what it is. Several eruptions look alike in this part of the face and at least two of them are treated in opposite directions, so a result here is something to take to a conversation rather than a substitute for one.
What this looks like
There is no formal staging system in general use. Trials score the eruption with a severity index built from three things counted separately — how many papules, how red the background skin is, and how much it is scaling — and that turns out to be a more useful way to describe a presentation than a label, because the three do not move together. What follows describes what is present rather than predicting what follows it.
A scatter of papules
A handful of small uniform bumps in the usual distribution, on skin that is pink rather than red, without much scaling.
What it looks like: Ten or twenty pinhead papules around the mouth or beside the nostrils, easy to miss at conversational distance and impossible to ignore in a mirror. This is the presentation that most often settles on removal alone, and the one most often treated hardest.
Confluent papules and pustules
Numerous bumps, many pus-topped, sitting on a background that has become uniformly red and often finely scaly.
What it looks like: Crops arriving faster than they resolve, spreading from the mouth toward the chin and the nasolabial folds, with a persistent burning or tight sensation between them. Makeup stops sitting on the skin properly, which is often what finally prompts an appointment.
The withdrawal flare
A dense eruption that appears within days to a fortnight of a facial corticosteroid being stopped, worse than anything the steroid was originally applied to.
What it looks like: Sheets of papules and pustules on hot, angry-looking skin, frequently with weeping or crusting. This is not a new disease and not evidence the wrong thing was stopped — it is the predictable withdrawal, and it is the moment at which most people restart the cream and reset the clock.
Firm red-brown lumps
A variant, commonest before puberty, in which the lesions are solid dome-shaped papules the colour of brick rather than pustules.
What it looks like: Discrete firm bumps around the mouth, nostrils and eyelids that do not come to a head and do not fluctuate day to day. Lesions may appear beyond the face. This form runs longer than the ordinary kind, is more likely to leave a mark, and is the one presentation here where waiting it out is a poor plan.
Three things that change the plan regardless of severity
A corticosteroid used on the face
This is the modifier that reorganises the whole page. A facial steroid suppresses the eruption convincingly for a few days and then requires more of itself to hold the same ground, so the tube gets stronger and the intervals get shorter. Mite densities are also markedly higher on faces that have had prolonged topical steroids, and unremarkable on faces that have not — which reframes the mite as a consequence of the treatment rather than a cause of the rash.
An inhaler, a nasal spray, or oral steroids
Corticosteroid reaching the face indirectly does the same job as one rubbed into it, and it is the version nobody thinks to mention at a skin appointment. In a large matched database study of children with asthma or allergic rhinitis, those on corticosteroid inhalers or nasal sprays were markedly more likely to be given a new diagnosis of perioral dermatitis within six months than matched children on non-steroid preparations. What changes practically is that this steroid is not the reader's to stop — the route is a conversation with whoever prescribed it about technique, rinsing, spacers or a swap.
Eyelids and the skin around the eyes
The same eruption reaching the periorbital skin is why the broader term periorificial dermatitis exists. It matters for three reasons: the skin there is thinner and less forgiving of anything applied to it, steroid eye drops and eyelid creams enter the picture as possible drivers, and several conditions that are genuinely time-critical also present as a red eyelid. Involvement of the eye itself belongs with a clinician rather than a skincare plan.
Children, especially under eight
Two things differ. The granulomatous variant is disproportionately a childhood presentation, and it carries a scarring risk that the adult form does not. And the oral drug with the best evidence in adults — a tetracycline — is off the table below roughly eight years of age because of its effect on developing teeth and bone, so systemic treatment in this age group is built around macrolides instead.
Eczema or seborrhoeic dermatitis underneath
A history of atopic dermatitis turns up in a substantial minority of series, and it creates the hardest version of this problem: the steroid that is driving the papules is also the thing holding the underlying condition down. Stopping it releases both at once. That combination is the clearest indication on this page for prescription cover during withdrawal rather than withdrawal alone.
Pregnant, breastfeeding, or planning to be
The oral tier narrows sharply: tetracyclines are avoided and oral isotretinoin is out entirely. Removal loses nothing, macrolides are the systemic option that survives, and the topical choices are a discussion with whoever is providing antenatal care rather than a shelf decision.
What else it could be
- Acne vulgaris — Blackheads and whiteheads. This condition produces neither, so their presence points elsewhere — and the drying acne routine that follows an acne assumption is one of the reliable ways to make these papules worse.
- Rosacea — Colour and bumps centred on the cheeks, nose and forehead, with flushing and often visible vessels. The distribution is close to the inverse of this one, which classically spares the cheeks and leaves an untouched ribbon at the lip margin.
- Seborrhoeic dermatitis — Greasy yellowish scale in the brows, beside the nostrils and at the hairline. Worth identifying carefully, because it is one of the commonest reasons a facial steroid was prescribed in the first place.
- Irritant or allergic contact dermatitis — A reaction that maps to what touched the skin — a lip balm, a toothpaste, a nail product transferred by hand — and that settles when the source goes. In children, a ring of dryness produced by lip-licking is a frequent look-alike.
- Demodex folliculitis — Follicle-centred papules with fine scale, sometimes with a gritty feel, that can sit in exactly this distribution. It is diagnosed by looking rather than by guessing, and it responds to a different drug.
- Sarcoidosis and lupus miliaris disseminatus faciei — Considered when firm red-brown papules persist, particularly around the eyelids. Both are diagnosed on biopsy rather than on appearance, which is why the granulomatous presentation earns an appointment rather than a longer wait.
One physical sign carries more weight here than any other: the narrow band of clear skin immediately bordering the lips. Papules that crowd right up to the vermilion edge argue for something else; papules that stop a millimetre or two short of it argue for this. It is worth photographing close up before an appointment, because the gap is easier to see in a still image than in a mirror.
Why it happens
Perioral dermatitis is a papular facial eruption that behaves like acne, is treated like rosacea, and is neither. Small uniform pink bumps, some pus-topped, appear around the mouth — classically leaving a narrow ribbon of untouched skin right at the lip edge — and the skin between them often burns, tightens or flakes rather than itches. It affects mostly women between roughly twenty and forty-five, and a smaller number of children. The mechanism is not settled, but a damaged, over-worked epidermal barrier sits underneath most descriptions of it, and the single factor that appears in case series more than any other is a corticosteroid: put on the face, breathed in through an inhaler, sprayed up the nose, or swallowed.
Anyone with small repetitive bumps around the mouth, nose or eyes that a steroid cream flattens within days and that returns harder every time the tube is put down. It is also for parents of a child whose facial rash has been treated as eczema for months with a cream that keeps almost working, and for anyone who was told this was adult acne and has been scrubbing at it ever since.
Orientation, not diagnosis. This page is a map rather than a diagnosis. Several eruptions occupy the same patch of face: rosacea, which favours the cheeks and forehead this condition tends to spare; seborrhoeic dermatitis, which is frequently what the offending steroid was prescribed for in the first place; irritant reactions to a lip balm or a toothpaste; and, in a child with firm red-brown lumps, a granulomatous picture that needs a clinician's eye rather than a shelf.
No prescription needed for any of this The over-the-counter pathway
The non-prescription tier here is unusual in a way worth stating plainly: its most effective component is not a purchase. Across four decades of published reports, the finding that has held most consistently is that stopping topical corticosteroids and cosmetics resolves a large share of mild cases without any drug at all — the literature calls this zero therapy, and it is treated as the treatment of choice in mild disease rather than as a preparatory step. The corroborating detail is buried in a trial of something else: in the vehicle-controlled study of pimecrolimus cream, the dummy arm — a plain cream base applied to people who had also stopped their steroids — cleared or nearly cleared a majority of participants. Beyond removal, the shelf offers plain barrier care with two small open trials behind it, and two acne actives whose evidence in this specific condition is weak. That order is the honest one.
Appropriate for: Mild disease of any origin, and the foundation underneath every prescription on this page. Removal is not made redundant by a prescription — every trial that produced a good result did so on people who had also stopped the driver, and a topical asked to out-treat a steroid still going on nightly is being set up to fail.
This chart is mostly a list of things to stop, and the sequence matters more than it looks. Removal comes first and completely, the flare that follows is planned for rather than reacted to, and the question of whether to add anything at all is left until week four — by which point a good number of people find it has become academic.
The failure mode: the cure and the cause are the same tube
A corticosteroid cream flattens this eruption within days, which is exactly why it keeps producing it. Each stop is followed by a rebound worse than the original rash, each rebound sends the tube back on, and the strength climbs while the intervals shorten — a loop that feels like a worsening disease and is actually a treatment. The escape is the one thing nobody wants to do, because it means accepting a fortnight visibly worse than the starting point in exchange for the thing ending. Nothing further down this page substitutes for that, and the prescriptions above it exist largely to make that fortnight survivable.
Anchored to national primary-care guidance and to the reviews that call removal alone the treatment of choice in mild disease.
Four rules that decide this more than any product does
- A corticosteroid cream on the face is a decision that can be made at the bathroom cabinet. An inhaler, a nasal spray, eye drops or tablets prescribed for something else are not — those are reviewed by whoever prescribed them, and a facial rash is never a reason to interrupt asthma treatment.
- The rash gets worse before it gets better, and the worsening is the withdrawal rather than a wrong turn. It typically builds over the first one to three weeks and then recedes. Restarting the cream during it resets everything.
- Everything comes off at once and nothing goes back on for eight weeks. A partial removal produces a partial answer, and a routine rebuilt while the skin is settling makes the eventual culprit impossible to identify.
- Lighter is better than richer, which reverses the usual instinct for a dry, tight, flaking face. Occlusive creams and balms sit among the reported triggers for this specific eruption, so the useful product is a thin fluid or nothing at all.
Answer as you go and the chart tracks where you are. Nothing is sent anywhere.
- Day 1You are here
Find the steroid
Establish whether a corticosteroid is in the picture, since that single answer reorganises everything below.
What this step involves
Morning
- Read every tube in the bathroom Hydrocortisone, betamethasone, clobetasone, mometasone, fluocinolone — anything whose active ends in -asone, -onide or -ol propionate. Combination antifungal or antibacterial creams are worth reading twice, because several sold outside prescription systems contain a potent steroid that the front of the box does not mention.
- Count the indirect routes A corticosteroid inhaler, a nasal spray for hay fever, steroid eye drops, a recent short course of tablets. In matched cohort data, children using corticosteroid inhalers or nasal sprays were around three times more likely to receive a new diagnosis of this condition within six months than matched children on non-steroid preparations.
Evening
- Write the list down before anything changes Two columns: what goes on the face, and what is prescribed for something else. Only the first column is a decision that can be made without an appointment, and separating them now prevents the mistake that matters.
Why this step: Which of the four routes on this page applies turns on this answer, and so does the entire expectation for the next month. A face with a steroid behind it is heading for a rebound and needs to know that in advance; a face without one is running a different experiment entirely.
Ready for the next step?
Is it clear whether a corticosteroid has been involved, and by which route?
Yes — the list is written: That answer sets the route. Go to step 2 ↓ Go to step 1 ↓
Not sure: A pharmacist can read the labels in a few minutes, and it is worth the trip. Anything prescribed rather than bought is a question for the prescriber, and the rest of this pathway works either way in the meantime.
- Week 0You are here
Stop the facial steroid, and take everything else off with it
Remove the driver and every candidate around it, in one move, so that what follows is readable.
What this step involves
Morning
- The facial corticosteroid stops Where it was prescribed, and particularly where it is potent or has been in use for months, a clinician may choose a stepped reduction instead — that decision belongs with them. What is not an option is continuing at the same strength while everything else on this page is tried.
- The actives stop Acids, retinoids, vitamin C, anything marketed as resurfacing or brightening, and any acne treatment bought for what looked like acne. All of them, not a reduced frequency.
- Foundation and concealer come off the area This is the hardest instruction on the page and the one most often quietly ignored. A mineral powder is a defensible compromise where cover is not optional; a liquid foundation worn over the papules every day is not.
Evening
- Rich creams, balms and heavy lip products stop Occlusive cosmetics appear in the trigger lists for this condition specifically, which is why the advice runs opposite to the advice for ordinary facial dryness. Heavy lip balm reapplied through the day sits exactly where the rash is.
- Switch the toothpaste for a fortnight Tartar-control and high-fluoride formulations have been reported as triggers in this precise distribution more than once. A plain paste for two weeks costs nothing and settles the question, and the foam running over the chin at night is the part that matters.
Why this step: Every published account of this condition puts topical corticosteroids and cosmetics at the head of the trigger list, and a formal appraisal of thirty reports found their discontinuation to be one of the few points the literature agrees on. Removing them one at a time over months is the version that fails: it stretches the withdrawal out and never produces a clean baseline.
What's normal: Nothing here is expected to feel better immediately, and most of it feels worse. That is not a sign the plan is wrong; it is the plan.
Ready for the next step?
Has the facial steroid stopped, and is the routine down to cleansing alone?
Yes — everything is off: The clock starts today, not from when it improves. Go to step 3 ↓ Go to step 2 ↓
Not yet: Nothing below this reads properly until it is. Where a prescribed steroid is involved and stopping it feels unsafe, that is a reason to make an appointment rather than a reason to wait.
- Weeks 1–3You are here
Plan for the flare, then hold
Get through the withdrawal without restarting the thing that caused it.
What this step involves
Morning
- Wash with water or a non-foaming cleanser Lukewarm, fingertips, patted dry. Nothing exfoliating, nothing foaming, no cloth or brush. On days when the skin feels raw, water alone in the morning is the better option rather than the lazier one. Browse all cleansers →
- Photograph it, weekly, in the same light A flare that is receding and a flare that is still building look identical on any given morning. Counting papules in a photograph from a week ago is the only reliable way to tell which is happening, and it is what stops the cream going back on in week two.
Evening
- Cool compresses rather than anything active A cool damp flannel for a few minutes helps the burning and does nothing to the eruption, which is an honest description rather than a criticism. Nothing applied at this stage is going to shorten the withdrawal.
Why this step: Withdrawal from a facial corticosteroid produces a predictable flare of erythema, burning and papules, described consistently in the systematic reviews of topical steroid withdrawal. It is worst in people who used the most potent preparations for the longest, and it is self-limiting. The reason it deserves its own step is that it is the point at which the majority of attempts end, and it ends them by looking exactly like failure.
What's normal: Burning, tightness and a visibly angrier face are expected here. What is not expected is spreading swelling, fever, weeping over most of the face, or an eruption that is still intensifying at week four — those are reasons to be seen.
If only one thing gets done this fortnight
Take the photograph. Everything else on this step is comfort; the photograph is the only thing that will still be useful in a month, both for deciding whether this is working and for showing a clinician what a flare actually looked like on a day that was not the day of the appointment.
Ready for the next step?
Has the flare passed its peak — fewer new papules this week than last?
Yes — it is receding: That is the withdrawal completing. Go to step 4 ↓ Go to step 3 ↓
Not yet, or it is still building: Before week four, that is still within the expected range and holding is the right call. Still intensifying at week four, or severe enough to be unmanageable, is a reason for an appointment rather than more waiting.
- Weeks 2–8You are here
One product, or none
Support a barrier that is failing without reintroducing anything that could be keeping the eruption going.
What this step involves
Morning
- The same plain cleanser, unchanged Deliberately identical every day. The value of a boring routine is that any change in the skin has an obvious cause, and nothing is a boring routine's rival at that job. Browse all cleansers →
- A thin bland fluid, if anything at all Light rather than rich, fragrance-free, and the shortest ingredient list available. The two published trials in this condition used ordinary pharmacy facial fluids and reported severity falling alongside measurable improvement in the barrier over eight weeks. Applying less is usually better than applying more. Browse all moisturizers →
Evening
- The same two things, or one of them Skin that feels genuinely comfortable bare can be left bare. There is no obligation to apply something at night, and on this condition specifically the obligation runs the other way. Browse all moisturizers →
Why this step: Barrier dysfunction runs through nearly every account of how this eruption arises, and water loss through the barrier is measurably raised on affected skin. Both published trials of drug-free facial products reported it falling as the eruption settled. The evidence is open-label and modest, and it points at the mechanism rather than around it — which is more than can be said for the actives in the next step.
What's normal: Nothing at this stage should sting. A moisturiser that stings is the wrong moisturiser, and the answer is to remove it rather than to persist with it.
If even a light fluid seems to keep it going
Dropping to a cleanser alone for a fortnight is a legitimate experiment and occasionally the answer. Rich and occlusive products sit on the reported trigger lists for this condition, and a fluid that suits most faces can still be one product too many for this one.
Ready for the next step?
Has the routine been two products or fewer, unchanged, for at least a fortnight?
Yes — it is stable and boring: That is the baseline established. Go to step 5 ↓ Go to step 4 ↓
No — things are still changing: Hold here until they stop. A routine that changes weekly cannot be interpreted at the checkpoint, and the checkpoint is what decides whether an appointment is worth making.
- Week 4 onwardYou are here
Decide whether to add anything at all
Make an honest decision about the one optional purchase on this pathway, knowing what it is and is not backed by.
What this step involves
Morning
- First, check whether it is still needed A meaningful share of mild cases have substantially settled by week four on removal alone. Adding an active to an eruption that is already resolving guarantees the credit goes to the product, and guarantees it gets bought again next time.
Evening
- If something is added, azelaic acid, once daily, alone A thin layer over the affected area only. Ten children with steroid-driven disease cleared completely on the twenty per cent prescription cream in four to eight weeks; the one systematic review to assess a gel formulation found no change in severity at six weeks. Both of those are the honest picture, and neither is a reason to add a second product alongside it. Browse all azelaic acid products →
Why this step: This step is placed last and framed as optional because that is where the evidence puts it. Removal is the intervention with consistent support across four decades of reports; the non-prescription actives are not. Putting the purchase at the end also solves a practical problem — an active introduced during the withdrawal flare will be blamed for it, and abandoned.
What's normal: A warm prickle for the first fortnight is expected and settles. A brief worsening in the first week was described in a minority of the treated children and passed. Burning that intensifies week on week, or spreads, means stopping rather than persisting.
Sulfur preparations are the other non-prescription option and sit in the same place: available, traditional, and unsupported by any controlled trial in this condition. One or the other, never both.
Ready for the next step?
Has the decision about an active been made and held for at least four weeks either way?
Yes — added, or deliberately not: That completes what is available without a prescription. Go to the checkpoint below ↓ Go to the checkpoint ↓
Still undecided: Doing nothing is a legitimate answer and the one the evidence supports best. The checkpoint below works the same either way, and an eruption that settles without a purchase has still settled.
The eight-week checkpoint
Eight weeks counted from the day the corticosteroid stopped and the routine came off — not from when the skin started improving, and not from when an active was added.
How to judge it: Count papules in photographs rather than judging colour in a mirror. Trials score this condition by counting the bumps, the background redness and the scaling separately, because the three improve on different schedules and an overall impression is dominated by whichever is worst that morning.
Eight weeks after the driver came off, with the routine held short throughout — what is the papule count doing?
Nothing to add. This is the expected result.
Complete or near-complete resolution on removal alone is the outcome this approach is built to produce, and it is common in mild disease. The useful move now is not a maintenance product but a decision: what happens the next time a facial rash appears. A corticosteroid applied to it without a clinician looking first is how this starts again, and it starts again faster the second time.
The approach is right and there is more available.
Partial resolution at eight weeks is a good position to bring to an appointment rather than a disappointing one. It demonstrates that the driver has gone and the foundation is in place, which is precisely what a prescription topical is meant to be added to — and it distinguishes an eruption that needs more treatment from one that was never going to respond because the cause was still running.
This is the handover point.
Eight weeks of complete removal with nothing to show for it is exactly the result this checkpoint exists to produce, and it is not evidence that anything was done badly. Two things become worth asking at that appointment: whether a prescription topical or a short oral course is appropriate, and whether this is the right diagnosis at all — several eruptions occupy this patch of face and one of them responds to a different drug entirely.
This one does not wait for the next appointment slot.
A withdrawal flare that is still intensifying at week four, an eruption that has spread to the eyelids, or skin that is weeping or crusting across the face are all reasons to be seen sooner rather than at eight weeks. Worse is not a stronger version of no-change and does not call for a stronger product; it usually calls for someone to look at it, both to cover the withdrawal properly and to confirm nothing else has been added to the picture.
What does not wait eight weeks
Anything involving the eye itself — pain, light sensitivity, blurring — is a same-day question rather than a checkpoint one. A child with firm red-brown lumps does not wait either: that variant runs long, responds slowly, and has left scars in the more severe reported cases, so it earns an appointment while it is still mild. Nor does an eruption arriving with fever or spreading swelling, which is more likely to be infection than a flare. And a corticosteroid inhaler or nasal spray that looks implicated is a conversation to start now rather than at week eight, because that conversation is with a different prescriber and takes longer to arrange.
One photograph a week, same window, same time of day, close enough that individual papules can be counted. Colour is what artificial light distorts most and what memory reports least reliably, and the difference between a flare that has peaked and one that has not is a number rather than an impression. Those photographs are also the single most useful thing to arrive at an appointment with, since this condition looks its calmest on the days it is convenient to be seen.
Where the over-the-counter pathway ends
It ends at the withdrawal, and that is a narrower ceiling than most pages on this site describe. Removal is genuinely the treatment of choice in mild disease and genuinely resolves a large share of it, but nothing available without a prescription reliably covers the fortnight in between, and nothing available without a prescription has convincing controlled evidence in this condition at all. What a clinician adds is specific: a topical that suppresses the inflammation without being another corticosteroid, and — for anything past mild — a finite oral course with the only randomised evidence this condition has behind it.
See the prescription pathway below →Perioral dermatitis: the full therapy pathway · dermapproved.com/treatments/perioral-dermatitis/ · not a diagnosis — confirm anything you act on with a clinician
What to expect
The honest shape of this is a curve that goes down before it goes up. Where a corticosteroid is involved, the first one to three weeks after stopping are worse than the starting point, and the recovery over the following four to six weeks is steady rather than dramatic. Mild disease frequently resolves on removal alone within that window. Moderate disease usually needs a prescription and then resolves, with the randomised evidence showing an oral tetracycline separating from placebo from around day twenty. The granulomatous variant is the slow one: published cohorts describe treatment running a median of ten weeks and clearance measured at six months. What almost nobody needs is indefinite treatment — this is one of the few conditions on this site that genuinely ends, provided the thing that started it does not come back.
- The week-two flare is the single commonest reason a good plan is abandoned, and it is the point at which the plan is working
- An oral course clears the eruption and does not prevent it returning if the driver is still in use, which is why the sequence matters
- Marks left behind after the papules go are post-inflammatory and fade over months; treating them as active disease leads to unnecessary treatment
- Reintroducing several products at once after clearance is how the eventual trigger stays unidentified
- Photographs beat memory here more than in most conditions, because papule counts are countable and impressions are not
Worth going back to a clinician if the eruption returns after a completed course, if it reaches the eyelids at any point, if firm lumps appear or anything is leaving a mark, if a prescribed treatment is not tolerable, or if a corticosteroid has been restarted for any reason. Recurrence is usually traceable to something that came back into use, and identifying it is a more useful outcome than another course of the same drug.
What would bring you back sooner
Whatever step you are on, these do not wait for the end of a trial period. They are the same
presentations listed at the top of this page — repeated here because that is where they
matter, rather than where they were easiest to put.
- Eye pain, light sensitivity, or blurred vision Seek care quickly
- A facial rash arriving with fever, spreading swelling, or feeling unwell Seek care quickly
- Painful blisters grouped on one side of the face or near an eye Seek care quickly
- Any temptation to stop a prescribed inhaler, nasal spray or steroid tablet See a dermatologist soon
- Firm red-brown lumps in a child, or lesions leaving marks See a dermatologist soon
- A single crusted, bleeding or non-healing spot that behaves differently from the rest See a dermatologist soon
Prescription only — from here on, this is a clinician's decision The prescription pathway
Why this tier exists. Two limits, and neither is usually stated. The first is that no product available without a prescription has convincing controlled evidence in this condition — the shelf's contribution is almost entirely subtraction, and after that it thins out fast. The second is more practical: nothing sold over the counter reliably covers the withdrawal flare. That flare is the hardest fortnight of the whole process and the reason most attempts fail, and covering it is the clearest thing a prescription actually buys.
Four routes, and which one applies turns almost entirely on two questions: whether a corticosteroid was involved, and whether the eruption has reached the eyes or turned granular. Read across the top for the presentation, down the side for how far past the aisle it goes.
The same pathway in full below — each step names what it involves, why it sits where it does, and what tells you it's time to move up. Tap any therapy to jump to its detail.
With a corticosteroid behind it
A corticosteroid — a facial cream, an inhaler, a nasal spray or a course of tablets — was in use in the months before this started, or is still in use now. This is the commonest version of the condition and the one where the first move is the largest.
You reach this column when: The flare has not begun to recede by week four, the eruption is severe enough to interfere with work or sleep, or the corticosteroid cannot realistically be stopped without something to cover the withdrawal.
- Prescription topical
A prescription topical during withdrawal
Why this step sits here: This rung exists for the specific problem the withdrawal creates: something is needed and it must not be another corticosteroid. The trial evidence in steroid-driven disease points most directly at the calcineurin inhibitors, which is one of the few places in this condition where a subgroup result is clear enough to guide a choice.
Three topical approaches sit on this rung, and a prescriber chooses between them on what drove the eruption rather than on a ranking:
- Pimecrolimus or tacrolimus backed by smaller studies
The eruption followed a facial corticosteroid. This is the subgroup in which the vehicle-controlled trial found the clearest separation, and the class carries none of the atrophy or rebound that made the original problem.
- Metronidazole, erythromycin or clindamycin backed by smaller studies
A familiar, inexpensive and widely stocked starting point, and what national primary-care guidance names for milder disease. Often the practical choice when the appointment is with a family doctor rather than a dermatologist.
- Ivermectin 1% cream limited evidence
There has been a long run of potent topical steroid on the face, which is the situation in which mite densities are documented to rise. Reached for later than the other two, on case-series evidence only.
When to move up
- A fair trial
- Four to eight weeks of consistent use, applied to skin from which the driver has genuinely gone.
- Improvement looks like
- Fewer papules week on week and a background that is settling from angry red toward ordinary pink.
- Move up if
- New crops are still arriving after eight weeks, the eruption is widespread, or the topical is not tolerable enough to be used consistently.
→ The next step up this pathway.
-
- Prescription oral
A finite oral course
Why this step sits here:
The class is settled by who is taking it rather than by comparative performance, and the two have never been tested against each other here:
- A tetracycline backed by smaller studies
An adult who is not pregnant or breastfeeding. This is where the randomised evidence in this condition actually sits, and national primary-care guidance names a tetracycline for anything past the milder cases.
- A macrolide limited evidence
A young child, a pregnancy, or a tetracycline that cannot be tolerated or interacts with something already prescribed. Almost the entire paediatric literature for this condition rests on this class.
Two questions decide this rung and neither is about which drug is stronger. The first is age and pregnancy, which determines the class outright. The second is what happens at the end of the course — an oral antibiotic reliably clears this eruption and reliably does not prevent it returning if the driver is still in the bathroom cabinet, so a course started before the withdrawal is complete tends to buy a good three months and then a repeat.
-
Staying off the steroid, which is the whole job
Once this settles there is no ongoing treatment, and that is genuinely the position rather than a euphemism for being abandoned. What replaces treatment is a decision made in advance: the next time a facial rash appears, whatever it turns out to be, it does not get a corticosteroid without a clinician looking at it first. That is the entire maintenance plan and it is the part people get wrong, because the cream that produced this problem is also the cream that is easiest to find. The routine that carried the withdrawal — one plain cleanser, one light fluid, nothing else — is worth keeping for a couple of months past clearance before anything is reintroduced, one item at a time.
With no corticosteroid involved
The same eruption in someone who can account for the preceding months and finds no steroid in them — no facial cream, no inhaler, no nasal spray, no tablets. Roughly two in five of the children in the largest published series had no steroid history either, so this is a real presentation rather than a failure of recall.
You reach this column when: The eruption is unchanged at eight weeks, is spreading toward the eyes, or has never been comfortable enough to leave alone.
- Prescription topical
Prescription topicals
Why this step sits here: In the absence of a steroid history the subgroup argument for the calcineurin inhibitors loses its force, and the older antimicrobial topicals — cheaper, familiar and named in primary-care guidance — are the usual starting point. The choice between them here is genuinely closer than the marketing for either would suggest.
When to move up
- A fair trial
- Eight weeks of twice-daily use, which is the duration the comparative trial ran to.
- Improvement looks like
- A falling papule count rather than a change in overall colour, since the two improve on different schedules.
- Move up if
- Papules are still arriving at eight weeks, or the eruption is dense enough that a topical is not a realistic answer to it.
→ The next step up this pathway.
- Prescription oral
An oral tetracycline course
Why this step sits here:
This is the top of the ordinary adult ladder, and almost nobody goes past it. The trial that compared an oral tetracycline with a topical over eight weeks left the oral arm with essentially no papules and the topical arm with a small residue, which is the clearest single result this condition has produced. What it does not do is explain why the eruption started, so the questions worth raising at the appointment are diagnostic as much as therapeutic — a papular facial eruption that has resisted removal and a topical deserves a second look at the label on it.
What replaces the routine that came off
Clearance here hands over to a rebuilding exercise rather than to a maintenance treatment. Items go back one at a time with a fortnight between them, heaviest last, and anything that coincides with a return of papules stays out — which is how the individual trigger gets identified, since no test does it. In practice most people find the routine they end up with is permanently shorter than the one they started with, and that the specific casualties are the rich night cream and the daily heavy foundation over the same patch of face. Recurrence is not rare, and it is usually traceable.
Around the eyes and nose (periorificial dermatitis)
The same eruption on the eyelids and around the nostrils, alone or alongside the perioral pattern. This is the presentation with the shortest fuse: the skin is thinner, several of the possible drivers are things dripped or sprayed rather than rubbed on, and a red eyelid has explanations that are not skin at all.
You reach this column when: Any pain in the eye, sensitivity to light or change in vision at any point, or eyelid involvement that has not begun to settle by week four.
- Prescription topical
A prescription topical for eyelid skin
Why this step sits here: Eyelid skin is where the case against another corticosteroid is strongest, because it is thinnest and because pressure inside the eye is affected by steroid applied nearby. That narrows the field to the non-steroidal options, and it is one of the settings in which the calcineurin inhibitors are most often chosen despite the licence describing something else.
When to move up
- A fair trial
- Four to eight weeks, with an early review rather than a distant one given the site.
- Improvement looks like
- Fewer lesions along the lid margin and less swelling of the lid itself.
- Move up if
- The eyelid remains inflamed at eight weeks, the eye becomes symptomatic at any point, or the eruption spreads across the rest of the face.
→ The next step up this pathway.
- Prescription oral
An oral course
Why this step sits here:
Systemic treatment is reached sooner in this column than in the others, for a simple reason: the site rules out most of what would otherwise be applied to it, so the choice is between a very short list of topicals and a tablet. The other thing this rung is for is a second opinion on the diagnosis — persistent eyelid papules are one of the presentations where a clinician will consider a biopsy, and a course of tablets that has failed is part of what makes that reasonable rather than premature.
Eyelid skin afterwards
Eyelid skin stays the most easily provoked part of the face long after the papules have gone, which changes what goes back on it and in what order. Eye creams, if they return at all, return last and singly. Anything containing a corticosteroid — including a drop prescribed for an unrelated eye problem — is worth flagging to the prescriber as a known trigger of this specific rash, because the connection is missed routinely and the same sequence repeats. Where a nasal spray was implicated, the durable fix is usually a technique change or a swap agreed with whoever prescribed it, not a permanent stop.
Granulomatous periorificial dermatitis
Firm red-brown papules rather than pustules, usually in a child before puberty, often around all three openings of the face and occasionally beyond it. It behaves differently enough from the ordinary form to be worth its own column: it runs for longer, it responds more slowly, and severe cases have left scars.
You reach this column when: The lumps are firm and persistent, new ones keep appearing, any lesion leaves a mark, or the distribution reaches the eyelids.
- Prescription topical
Prescription topicals
Why this step sits here: The largest paediatric series treated children with a calcineurin inhibitor alone or combined with topical metronidazole, and reported complete response in around seven in ten either way. That is a retrospective figure rather than a trial result, and it is the best available in this age group for topical treatment.
When to move up
- A fair trial
- Six to eight weeks, with photographs, since firm lesions change too slowly to judge by memory.
- Improvement looks like
- Existing lumps softening and flattening, and the arrival rate of new ones falling.
- Move up if
- Lesions are unchanged, numerous, spreading beyond the face, or leaving marks as they resolve.
→ The next step up this pathway.
- Prescription oral
An oral antibiotic course
Why this step sits here: This is where the recent evidence in this variant sits. A cohort of children treated with oral clarithromycin over a median of ten weeks reported more than nine in ten clear at six months, with treatment length tracking how severe and how long-standing the disease had been at the start — which is the argument for reaching this rung sooner rather than after a long topical trial.
Age decides the class here more decisively than anywhere else on this page:
- Clarithromycin, azithromycin or erythromycin limited evidence
A child under about eight, which is most of this variant. The published cohorts in granulomatous disease are macrolide cohorts, so this is where the experience actually is.
- A tetracycline backed by smaller studies
An older child or an adult, where the class is permitted. It brings the randomised evidence from the ordinary form with it, though that evidence was not gathered in the granulomatous variant.
When to move up
- A fair trial
- Weeks to months rather than a fixed course, judged by a clinician against how the lesions are behaving.
- Improvement looks like
- Steady flattening across the whole eruption, with no new firm lesions arriving.
- Move up if
- Disease persists through an adequate course, or scarring is developing while treatment runs.
→ The next step up this pathway.
-
- Prescription oral
Low-dose oral isotretinoin
Why this step sits here:
Genuinely rare territory, and the evidence for it is case reports. It is described where removal, topicals and a full antibiotic course have all been completed without result, and where continuing to wait risks a scar. Two things are worth knowing before the conversation: the doses used in the reports are far below acne doses and run for months, and the pregnancy-prevention requirements apply in full regardless of how low the dose is.
After a variant that took months
Recurrence after a completed course appears to be uncommon in the published cohorts, which is the reassuring half. The other half is what the eruption leaves behind: flat brown marks that fade over months without treatment, and, in the more severe cases, small depressed scars that do not. Sun protection over the affected area while the marks are still fading is the one thing that measurably changes how they look at a year. Where the child uses an inhaler or a nasal spray, reviewing the technique with the prescriber is the step that prevents the same sequence starting again.
Perioral dermatitis: the full therapy pathway · dermapproved.com/treatments/perioral-dermatitis/ · not a diagnosis — confirm anything you act on with a clinician
When the rules change
The four routes above assume an adult face with a routine that can simply be stopped. Three situations break that assumption, and in each of them the removal step gets more complicated rather than less important.
Pregnant, breastfeeding, or planning to be
The ladder narrows at the top and is untouched at the bottom, which is a better position than it sounds — the intervention with the most consistent evidence in this condition costs nothing and carries no pregnancy considerations at all.
What can be used
- Removal in full: the facial corticosteroid, the cosmetics, the actives, the occlusive creams
- Plain cleansing and a light bland fluid, which are considered compatible throughout
- Azelaic acid, generally regarded as acceptable and frequently the topical people are steered toward when others are withdrawn
- A macrolide such as erythromycin where a systemic drug is genuinely needed, decided alongside antenatal care
What comes off the table
- Oral tetracyclines, because of their effect on the developing skeleton and on tooth formation
- Oral isotretinoin, which causes severe birth defects and is governed by a formal prevention programme
- Assuming a topical is fine because it is topical — the calcineurin inhibitors in particular are a prescriber's judgement rather than a class rule
Each of these is a conversation with whoever is providing antenatal care rather than a decision to make from a page. It is also worth revisiting: this eruption behaves unpredictably across a pregnancy, and a plan made early may not be the plan later.
When the steroid was treating something else
The hardest version of this problem is not a cosmetic steroid habit — it is a face where the corticosteroid has been holding down eczema or seborrhoeic dermatitis that will return the moment it stops. Two conditions are then flaring at once, and the reader is being asked to accept both.
What can be used
- A clinician-directed plan that treats the underlying condition with something that is not a corticosteroid while the withdrawal runs
- The non-steroidal topicals, which exist precisely for this situation and have the best trial evidence in steroid-driven disease
- A stepped reduction rather than an abrupt stop, where the prescriber judges the potency and duration warrant it
- Plain barrier care throughout, which serves both conditions and interferes with neither
What comes off the table
- Attempting this alone with a potent facial steroid and no cover arranged
- Concluding the underlying condition was misdiagnosed because it returns — that is what withdrawal looks like
- Substituting a slightly weaker steroid, which keeps the loop running at a lower amplitude
This is the single clearest indication on the page for having a clinician involved before the steroid stops rather than after the flare arrives. It is also the situation in which a stepped reduction is most often the right call, and that call is not one a page can make.
When the corticosteroid is not on the skin
Inhalers, nasal sprays and eye drops deliver corticosteroid to the face by a route nobody thinks to mention at a skin appointment, and matched cohort data now put the association well beyond the case reports it used to rest on. The complication is that these are treating something that matters more than a rash.
What can be used
- A review of technique with the prescriber, which is often where the answer is
- Rinsing the mouth and wiping the face after inhaler use
- A spacer, which changes how much of the dose lands where it was not aimed
- A change of preparation or device, agreed rather than improvised
- The whole removal pathway for everything else, which proceeds normally in the meantime
What comes off the table
- Stopping or reducing an inhaler, nasal spray or steroid tablet because of the skin
- Assuming the connection will be spotted — it is worth naming explicitly at the respiratory or allergy appointment
- Treating the rash harder while the delivery problem goes unaddressed, which is how it recurs after each successful course
Recurrence in children who remain dependent on inhaled steroids is described specifically in the published series, which is the clearest argument for treating the delivery question as part of the treatment rather than as background.
How an appointment actually happens, what to take to it, and how to keep your skin liveable while a prescription or procedure does its work.
What seeing a dermatologist actually looks like
This is diagnosed by examination and history rather than by a test, which puts unusual weight on what can be described and shown. Two things carry a disproportionate share of that: a close-up photograph taken on a bad day, and a complete list of what has been applied to or inhaled near the face in the preceding months. In most systems the first appointment is with a family doctor, who can prescribe essentially everything on this page. Dermatology referral tends to follow diagnostic uncertainty, disease that has resisted a first-line course, granulomatous or periocular presentations, and any question of a biopsy.
What to bring
- Close-up photographs from a bad day, taken in daylight — the lip-margin gap is far easier to see in a still image
- Every product used on the face in the last six months, including the ones that have already been stopped
- Every corticosteroid by route: creams, inhalers, nasal sprays, eye drops, tablets
- The date the corticosteroid stopped, which is what the eight-week clock is counted from
- A note of anything already tried, how long it ran, and whether it helped or made things worse
- For a child, the inhaler and spacer themselves, so technique can be watched rather than described
Questions worth asking
- Is this perioral dermatitis, and what else was on the list?
- Which of the things being used is most likely to be driving it?
- Is there anything that can cover the withdrawal without being another steroid?
- How long before this is judged, and what does progress look like at the halfway point?
- If an oral course is used, how long is it and what happens when it ends?
- Does the inhaler or nasal spray need reviewing, and by whom?
- Is any of this affected by pregnancy, now or later?
Cost and access. Almost everything on this page is inexpensive. The topical antimicrobials and the older oral tetracyclines are long off patent and cost little in most systems; the calcineurin inhibitors are more expensive and, being used outside their licensed indication here, are occasionally queried by insurers on that basis. The genuinely free part is the part that does most of the work. Where cost becomes a factor it is usually the appointment rather than the drug, which is worth knowing when weighing whether to wait out a flare that is not settling.
A family doctor can diagnose this and prescribe the topical and oral treatments described here, and national primary-care guidance covers it explicitly. Asking for a dermatology referral is reasonable where the diagnosis is uncertain, where a first-line course has not worked, where the lesions are firm and granulomatous, or where the eyelids are involved. Where an inhaler or nasal spray looks implicated, the more useful referral may be back to whoever manages that condition.
Finding someone to see
Routes to a clinician differ by country, and in most systems the first appointment for this is deliberately with a family doctor rather than a specialist.
- US Find a Dermatologist American Academy of Dermatology Board certification is verified in the directory, and results can be filtered by location.
- Canada Find a Dermatologist Canadian Dermatology Association Listings by city or postal code; in most provinces a family doctor's referral comes first.
- UK Perioral dermatitis — patient information Primary Care Dermatology Society From the society that publishes the primary-care guidance this page is anchored to.
Regions appear here only where a directory or national source could be checked directly this session.
Skincare alongside therapy
Supporting care on this page is defined by what is absent from it. Every other guide on this site builds a routine around a treatment; here the routine is the treatment for the first eight weeks, and its virtue is the number of things it does not contain.
Two products is the target and one is acceptable. Light beats rich, fragrance-free beats soothing-branded, and unchanged beats improved — because a routine that keeps being refined cannot be read at the checkpoint, and reading it at the checkpoint is the entire point of holding it steady.
Through the removal and the flare that follows
The first three weeks after a facial corticosteroid stops are the hardest part of this whole process, and almost nothing applied to the skin shortens them. What supporting care is for here is making them tolerable enough to be completed.
Do
- Wash with lukewarm water and fingertips, using a non-foaming cleanser or nothing at all
- Apply a cool damp compress for a few minutes when the burning is at its worst
- Keep whatever goes on afterwards thin, fragrance-free and identical every day
- Photograph the area weekly in the same light, close enough to count individual papules
- Expect the appearance to worsen for one to three weeks and plan the calendar around it
Avoid
- Any corticosteroid on the face, including a low-strength one bought without a prescription
- Occlusive creams, balms and rich night products, which sit among the reported triggers here
- Exfoliation of any kind, mechanical or chemical
- Introducing an active during the flare, which guarantees the flare gets blamed on it
What that looks like on a shelf: a gentle cleanser, a barrier moisturizer.
While a topical is running
Most topicals used here sting for the first few days, and most of the abandonment happens in that window. Scheduling and buffering fix more of that than switching products does.
Do
- Apply to the affected area only rather than across the whole face
- Let the skin dry fully after washing before anything goes on it
- Buffer with the bland fluid underneath on days the skin is reactive
- Judge the stinging by whether it is decreasing, over days rather than hours
Avoid
- Running two actives at once, which makes any reaction unattributable
- Stopping and restarting repeatedly, which produces eight weeks with no interpretable result
- Adding a rich cream on top to counteract dryness, which reintroduces the occlusion that was removed
What that looks like on a shelf: a gentle cleanser, a barrier moisturizer.
Where the eyelids are involved
Eyelid skin is the thinnest on the body and the least forgiving of anything applied to it, which makes the supporting routine there shorter still — and makes the audit of what already goes near it more important than anything added.
Do
- Include eye drops, eyelid creams and nasal sprays in the audit of possible drivers
- Clean the lids with water alone rather than with a facial cleanser
- Mention any corticosteroid eye drop to the prescriber who issued it, since the link to this rash is missed routinely
- Keep eye makeup off the area entirely until the papules have gone
Avoid
- Any corticosteroid preparation near the eye that has not been specifically directed there
- Eye creams of any kind during treatment, and for a while after it
- Rubbing, which inflames lid skin that is already inflamed
What that looks like on a shelf: a gentle cleanser.
Where the patient is a child
The practical problems in children are different: the driver is more often an inhaler than a cream, the routine to be stripped is usually short already, and the family has frequently been told this is eczema and given exactly the wrong product for it.
Do
- Review inhaler and nasal spray technique with the prescriber, including rinsing and spacer use
- Wipe the face after inhaler use, which is a small change with a plausible mechanism behind it
- Take photographs, since firm lesions change too slowly for anyone to judge from memory
- Protect the area from sun while post-inflammatory marks are fading
Avoid
- Continuing a facial steroid prescribed for a presumed eczema that is not settling
- Waiting out firm red-brown lumps, which is the pattern associated with scarring
- Stopping asthma or rhinitis treatment because of the skin
What that looks like on a shelf: a gentle cleanser, a barrier moisturizer.
The supporting shelf
A gentle cleanser
A barrier moisturizer
The product roles above link to categories rather than to specific bottles, and the category is genuinely the part that matters here — the useful cleanser for this condition is defined almost entirely by what it lacks. A shorter ingredient list beats a better-marketed one, and the cheapest product that meets the description is not a compromise.
Every therapy named above, in full. Look things up here; it is not meant to be read end to end.
Every therapy, explained
Reference, not reading — every therapy named anywhere above, described the way a clinician would describe it. Nobody reads this section end to end; open the one you were sent here for. Which one is right for you is the appointment's job, not this page's.
Over the counter
Available without a prescription. Each one opens with what to look for on a label, which is the part that stays true regardless of which brand is on the shelf that month.
Over-the-counter Stopping the corticosteroid
Removal of the driver
Generally compatible
Stopping the corticosteroid
Removal of the driver
The first and most effective intervention available, and the only one on this page that costs nothing. Where a corticosteroid is involved, it is simultaneously what suppresses the eruption and what keeps producing it.
At the shelf
A cupboard audit rather than a purchase — every tube, spray and inhaler in use.
Facial creams containing hydrocortisone, betamethasone, clobetasone, mometasone or anything ending in -asone or -onide are the first category. The second is easier to miss: corticosteroid inhalers for asthma, nasal sprays for hay fever, eye drops, and short oral courses. Those belong to whoever prescribed them and are raised at an appointment, not discontinued privately. A third category catches people out entirely — combination antifungal or antibacterial creams bought abroad or online, several of which contain a potent steroid the label does not advertise in English.
Watch out for: Products marketed as steroid-detox or withdrawal-support creams. There is nothing in that category with published evidence, and several of them are simply emollients at a premium.
How it works
Chronic topical corticosteroid on facial skin thins it, alters its barrier function and suppresses local inflammatory signalling. When the drug is withdrawn that suppression lifts abruptly, which is what produces the rebound flare, and the flare is what drives people back to the tube. Removing the drug permanently is the only way out of that loop; the mite overgrowth documented on steroid-treated faces recedes with it.
What the evidence says strong
A formal appraisal of thirty published reports found discontinuation of topical corticosteroids and cosmetics to be among the few points on which the whole literature agrees, and the trial evidence for every active treatment sits on top of that step rather than instead of it.
Typical course
Stopped once, permanently. The flare that follows typically builds over the first one to two weeks and settles over the following four to six.
Commonly discussed side effects
- A predictable worsening of the eruption for one to three weeks after stopping
- Burning, tightness and redness through the withdrawal window
- Where the steroid was also treating eczema or seborrhoeic dermatitis, that condition returns at the same time
Worth knowing
- Inhaled, nasal and oral corticosteroids are changed only by the prescriber — a facial rash is not a reason to interrupt asthma treatment
- Tapering rather than stopping outright is sometimes chosen by a clinician for a potent steroid on a face, and that decision belongs with them
- Restarting the steroid during the flare resets the process and generally leaves the next attempt harder
Pregnancy: Stopping a topical corticosteroid raises no pregnancy concerns of its own. It matters more than usual during pregnancy, because the oral tier that would otherwise cover a difficult withdrawal is narrowed, so removal is carrying more of the load.
Why it sits where it does: It sits at the bottom of every column because it changes what everything above it is working against. The trials that report good outcomes for topicals and orals were run on people who had also stopped, which means no result on this page has ever been shown to hold while the driver is still in use.
Over-the-counter Taking the rest of the routine off
Cosmetic and irritant reduction
Generally compatible
Taking the rest of the routine off
Cosmetic and irritant reduction
The second half of removal, and the part most people do incompletely because it involves giving up things that are not obviously medicinal.
At the shelf
Nothing new. The action is a shelf being emptied rather than filled.
Out go the heavy occlusive creams and balms, foundation and concealer worn over the area, the acids and retinoids, anything fragranced, and the thick lip products applied several times a day. Two culprits are peculiar to this condition and get missed: tartar-control and high-fluoride toothpastes, which have been reported repeatedly as triggers in the exact perioral distribution, and sunscreens applied heavily to a face that is already struggling. Chewing gum appears in the reported-trigger lists often enough to be worth a fortnight's abstinence.
Watch out for: A replacement product for every item removed. The point of the exercise is a shorter list, and a routine rebuilt at the same length as the one that came off tests nothing at all.
How it works
Barrier dysfunction runs through nearly every account of how this eruption arises. Occlusive cosmetics, surfactants, actives and repeated low-grade irritation all add to that dysfunction, and each one is another candidate that has to be excluded before anything applied afterwards can be interpreted. Removal also settles the question of contact reaction, which sits high in the differential and is answered by absence rather than by testing.
What the evidence says promising
Cosmetics and topical preparations appear alongside corticosteroids in the trigger lists of every major review of this condition, and their discontinuation is the other half of the zero-therapy approach that reviews describe as the treatment of choice in mild disease.
Typical course
Everything off at once, held for at least eight weeks before anything is reintroduced, and then one item at a time.
Commonly discussed side effects
- The skin frequently looks worse in the first fortnight without occlusive cover over it
- Going without foundation over a visible facial rash is the part people find hardest, and it is a genuine cost rather than a trivial one
Worth knowing
- Reintroducing several products together at the end makes the eventual culprit unidentifiable
- A mineral powder is a reasonable compromise where cover is non-negotiable, though it is still a variable in the experiment
- Trigger lists describe what has been reported in populations rather than what applies to any individual
Pregnancy: Simplifying a routine has no pregnancy implications. It becomes proportionally more important in pregnancy, since several of the prescription options above are withdrawn and this rung is doing more of the work.
Why it sits where it does: It sits beside steroid withdrawal rather than after it because doing one without the other produces an ambiguous result. A face that is still being occluded, exfoliated and made up every morning has not had the experiment performed on it.
Over-the-counter Plain cleansing and a bland fluid
Barrier support
Generally compatible
Plain cleansing and a bland fluid
Barrier support
What is left once the removal is complete: one or two products whose job is to be unremarkable while the skin recovers.
At the shelf
A non-foaming fragrance-free cleanser and, if anything at all is applied afterwards, the lightest bland fluid available.
Light is the operative word, and it is the opposite of the advice given for most dry-feeling facial skin. Rich occlusive creams sit among the reported triggers for this condition, so the useful product is a thin lotion or fluid rather than a balm. Short ingredient lists earn their reputation here: every additional botanical extract, essential oil or preservative is another candidate for the thing that keeps it going. Water alone in the morning is a legitimate option rather than a shortcut.
Watch out for: Anything sold as a barrier-repair treatment at treatment prices. What the two published trials tested were ordinary pharmacy facial fluids, and their virtue was that they contained almost nothing.
Browse all glycerin products in our catalog →
How it works
Water loss through the barrier is measurably elevated on affected skin, and both published trials of bland facial products reported it falling as severity fell. The proposition is not that a moisturiser treats the eruption but that supporting a barrier which is failing removes one of the conditions the eruption depends on.
What the evidence says limited
Two small open-label eight-week trials of drug-free facial products reported significant falls in severity scores alongside improved barrier measurements. Neither was blinded or randomised, and one compared its result with historical controls rather than a concurrent arm.
Typical course
Twice daily for the eight weeks the removal is being judged over, and indefinitely afterwards if it suits the skin.
Commonly discussed side effects
- A fluid that is too rich can itself keep the papules going, which is the trap in treating this like ordinary dryness
Worth knowing
- Applying less is often better than applying more, which reverses the instinct that works for most facial dryness
- Fragrance-free and unscented are different claims — the second permits a masking fragrance
- If the skin is genuinely comfortable bare, nothing has to go on it
Pregnancy: Plain cleansers and simple emollient fluids are considered compatible with pregnancy and breastfeeding, and this is the rung that carries most of the weight when the systemic options narrow.
Why it sits where it does: It sits alongside removal rather than above it because it is what makes removal survivable. Skin that has lost its steroid and its cosmetics still has to be washed and still feels tight, and the alternative to a deliberately plain product is usually an improvised rich one.
Over-the-counter Azelaic acid
Non-prescription anti-inflammatory acid
Generally compatible
Azelaic acid
Non-prescription anti-inflammatory acid
The best-known non-prescription option for this eruption, with a more complicated evidence record than its reputation suggests.
At the shelf
A stated percentage on the packaging, in a light cream or gel rather than a wash.
Non-prescription strengths run from around ten to fifteen per cent in most markets; the prescription cream sits at twenty. The paediatric series that reported complete clearance used the twenty per cent cream, so a lower-strength product is a lower dose of the same molecule rather than an equivalent. Suspension-style bases are thick and can sit oddly on skin that is already flaking, which matters more here than it would elsewhere.
Watch out for: Blends that list azelaic acid without a percentage, and anything pairing it with an exfoliating acid. The second active is the problem.
Browse all azelaic acid products in our catalog →
How it works
It dampens inflammatory signalling in the follicle and reduces the reactive oxygen species released by neutrophils, which is the same rationale that supports its use in other papular facial conditions. It is not directed at any mechanism specific to this one.
What the evidence says limited
Ten children with steroid-driven disease cleared completely on twenty per cent cream within four to eight weeks in an uncontrolled series, with a transient worsening in three of them in the first week. The 2022 systematic review, assessing a gel formulation, concluded it may result in no change in severity by either physician or patient assessment at six weeks.
Typical course
Once daily, judged over four to eight weeks. Any benefit in the reported series appeared inside that window rather than after it.
Commonly discussed side effects
- Stinging and warmth for the first fortnight
- A brief worsening in the first week of use, described in a minority
- Dryness and flaking on skin that is already scaling
Worth knowing
- Adding it before the removal is complete confounds the result and is the commonest reason it appears not to work
- A worsening that continues past the second week is not tolerance building, and stopping is the reasonable response
- It is one active, not an addition to a routine that is being rebuilt at the same time
Pregnancy: Azelaic acid is generally regarded as compatible with pregnancy and is frequently the topical people are steered toward when other options are withdrawn. Confirmation belongs with whoever is providing antenatal care.
Why it sits where it does: It sits above removal and clearly below the prescription tier, which is an unusual placement and a deliberate one. It is the most plausible thing on a shelf and its record in this condition specifically is thin, so it is offered as something reasonable to try rather than as the reason the eruption will settle.
Over-the-counter Sulfur preparations
Traditional topical
Individualized — ask a clinician
Sulfur preparations
Traditional topical
An old preparation that survives in second-line lists on clinical custom rather than on trial data.
At the shelf
Sulfur as a stated active in a wash or a lotion, alone or with sodium sulfacetamide where that combination is sold without a prescription.
Availability varies a great deal by market: in some countries the sulfacetamide-and-sulfur combination is a prescription item and only plain sulfur can be bought freely. A wash gives limited contact and a leave-on lotion gives more, which is the practical difference between the two formats. The smell is not a formulation flaw and does not wash out of the memory quickly.
Watch out for: Clay masks marketed on a sulfur content. Occlusion and abrasion are both counterproductive here, and a mask delivers both.
Browse all sulfur products in our catalog →
How it works
Sulfur has mild antimicrobial and keratolytic activity and some effect on follicular mite populations, which is the usual justification for its place in papular facial conditions generally. No mechanism specific to this eruption has been demonstrated.
What the evidence says limited
It appears in second-line lists for this condition on the strength of case series and clinical experience, and features in retrospective paediatric practice alongside other topicals rather than as a treatment tested on its own. No controlled trial supports it here.
Typical course
Once or twice daily, reviewed at four weeks rather than persisted with indefinitely on hope.
Commonly discussed side effects
- Dryness and a fine scale
- Stinging on skin that is already inflamed
- The smell, which is the usual reason it is abandoned
Worth knowing
- Not used alongside another topical active, which would make either impossible to judge
- Wash formats give little contact time and correspondingly little to expect
- Its presence on a second-line list reflects how long it has been used rather than how well it has been tested
Pregnancy: Topical sulfur has a long history of use and minimal absorption, and is often regarded as acceptable in pregnancy, but the products vary and several combine it with other actives. Worth checking the specific preparation with whoever is providing antenatal care rather than assuming from the class.
Why it sits where it does: It sits at the far edge of the non-prescription tier because it is genuinely available and genuinely unproven here. Naming it honestly is more useful than omitting it, since it is what a pharmacist frequently reaches for.
Browse the whole catalog for this condition:
What changed recently: The genuine change over the last few years has been in what gets counted rather than in what gets sold. A 2022 systematic review applied formal certainty grading to this condition for the first time and found the entire evidence base — including the treatments in routine daily use — resting on low and very low certainty. That is not a reason to abandon any of it, and it is a reason to hold the confident-sounding advice on the internet at arm's length. Alongside it, the corticosteroid story widened: inhalers and nasal sprays moved from case reports into matched cohort data, which makes the question worth asking of every reader who cannot identify a facial cream.
Prescriptions and procedures
About the brand names below. Brand names are what the pharmacy label says; the generic name underneath is what actually matters, and it is the part that stays true when a brand is discontinued or a region carries a different one. Where a region is missing from a card below, that reflects what could be verified rather than proof of absence.
Labels linked from the prescription cards below point at national medicines sources rather than at manufacturers. Every drug on this page is used outside its licensed indication for this condition, so a label will describe eczema, rosacea or acne instead — that is normal and lawful, and it is why the leaflet will not mention this rash at all.
Prescription topical Topical calcineurin inhibitors
Non-steroidal anti-inflammatory topical
Individualized — ask a clinician
Topical calcineurin inhibitors
Non-steroidal anti-inflammatory topical
The prescription topical with the most direct trial evidence in this condition, and the one designed for the specific problem of covering a corticosteroid withdrawal without using a corticosteroid.
What it's called at the pharmacy
Generic name Topical calcineurin inhibitors
Sold as
- Elidel — pimecrolimus cream 1%
- Protopic — tacrolimus ointment 0.03% and 0.1%
- Elidel — pimecrolimus cream 1%
A region missing here reflects the limit of what was verified, not the limit of what is sold.
Read the label
- Pimecrolimus cream — US prescribing information DailyMed (US National Library of Medicine)
- Tacrolimus ointment — US prescribing information DailyMed (US National Library of Medicine)
How it works
These block calcineurin inside T cells, which suppresses the cytokine cascade that drives inflammation in the skin without any of the atrophy, vessel dilation or dependence that follows prolonged facial corticosteroid use. That distinction is the whole reason this class matters here: it interrupts the inflammation without re-entering the loop being escaped.
What the evidence says promising
A multicentre vehicle-controlled trial found pimecrolimus cream reduced severity faster than vehicle over four weeks, with the largest benefit in those with prior topical corticosteroid use. A retrospective series of 132 children reported complete response in around seven in ten treated with a calcineurin inhibitor alone. Certainty was graded low in systematic review.
Typical course
Applied twice daily and reviewed at four weeks, most often through the withdrawal window rather than indefinitely.
Commonly discussed side effects
- Burning and stinging in the first days of use, which is the commonest reason it is abandoned
- Transient redness and warmth on application
- Reactivation of cold sores has been reported
Worth knowing
- Use in this condition is outside the licensed indication in most markets, which is routine and lawful and means the leaflet describes eczema instead
- The class carries a boxed warning in some countries based on theoretical long-term risk, and a prescriber will raise it
- The application-site burning is worst in the first few days and is worth expecting rather than discovering
Pregnancy: Systemic absorption from facial application is very low, but human pregnancy data are limited and this class is generally reserved for situations where the alternatives are worse. It is a decision for a prescriber alongside whoever is providing antenatal care.
Why it sits where it does: It sits on the first prescription rung because it answers the specific difficulty this condition creates. The people who most need something during withdrawal are the people who must not be given another steroid, and this is the class that fills that gap.
Prescription topical Topical metronidazole, erythromycin and clindamycin
Prescription topical antimicrobials
Individualized — ask a clinician
Topical metronidazole, erythromycin and clindamycin
Prescription topical antimicrobials
The topicals most likely to be handed over at a first primary-care appointment, and the ones with the longest record of routine use in this condition.
What it's called at the pharmacy
Generic name Topical metronidazole, erythromycin and clindamycin
Sold as
- Metronidazole cream and gel are the preparations named in primary-care guidance, alongside topical clindamycin and erythromycin
A region missing here reflects the limit of what was verified, not the limit of what is sold.
Read the label
- Erythromycin — NHS medicines NHS (UK)
How it works
The antibacterial action is probably not the whole story, and may not be the important part. All three have anti-inflammatory effects on the skin independent of any organism they kill, which is the usual explanation for why they help conditions with no established infectious cause.
What the evidence says promising
Topical metronidazole reduced papule counts substantially over eight weeks in a double-blind comparison, though oral tetracycline did significantly better in the same trial. Topical erythromycin beat placebo in a randomised study. Reviews consistently describe the topical evidence as the least consistent part of this literature.
Typical course
Applied twice daily and judged over roughly eight weeks, which is the duration used in the comparative trial.
Commonly discussed side effects
- Dryness and local irritation
- Stinging on broken or scaling skin
- Contact sensitisation, which is uncommon but worth considering when a topical stops suiting
Worth knowing
- Prescribed off-label for this indication in most markets, so product leaflets describe rosacea or acne instead
- Topical antibiotic use carries resistance considerations, which is part of why courses are finite
- A topical applied while a facial steroid continues has not been shown to work in that situation and is not expected to
Pregnancy: Topical erythromycin is commonly regarded as among the more acceptable options in pregnancy and topical metronidazole is used with caution, particularly beyond the first trimester. The choice belongs with a prescriber in consultation with antenatal care.
Why it sits where it does: It shares the first prescription rung rather than sitting below the calcineurin inhibitors, because which of the two a prescriber reaches for depends on the case rather than on a hierarchy. Where a corticosteroid is the driver the trial evidence favours the other class; where it is not, these are the long-standing choice.
Prescription topical Topical ivermectin
Antiparasitic and anti-inflammatory topical
Individualized — ask a clinician
Topical ivermectin
Antiparasitic and anti-inflammatory topical
A newer topical borrowed from rosacea practice, used here mainly where the mite population is thought to be part of the picture.
What it's called at the pharmacy
Generic name Topical ivermectin
Sold as
- Soolantra — ivermectin cream 1%, licensed for the inflammatory lesions of rosacea
- Soolantra — ivermectin cream 1%, licensed for inflammatory lesions of papulopustular rosacea
A region missing here reflects the limit of what was verified, not the limit of what is sold.
How it works
It kills follicular Demodex mites and is separately anti-inflammatory. The relevance is specific: mite densities are elevated on faces that have had prolonged topical corticosteroids and not on faces that have not, so this is a drug aimed at a consequence of the driver rather than at the driver itself.
What the evidence says limited
A small case series of adults and a paediatric series describing children treated with topical or oral ivermectin both reported clearance or near-clearance over one to three months. No controlled trial has assessed it in this condition.
Typical course
Once daily, with the reported series running one to three months before a verdict.
Commonly discussed side effects
- Burning or stinging on application
- Transient dryness and peeling
- Worsening in the first week has been described
Worth knowing
- Entirely off-label for this condition, on evidence that is case series only
- It is expensive in several markets relative to the older topicals
- The rationale is strongest where there is a long topical steroid history behind the eruption
Pregnancy: Systemic absorption from the cream is low, and human pregnancy data for the topical are limited. Whether it is used in pregnancy is a prescriber's judgement made alongside antenatal care rather than a class rule.
Why it sits where it does: It sits on the same prescription rung as the older topicals rather than above them, because it has not been compared with any of them in this condition. Its place is a reasoned one rather than an evidenced one, and saying which is which matters.
Prescription oral Oral tetracyclines
Systemic tetracycline-class antibiotics
Avoid in pregnancy
Oral tetracyclines
Systemic tetracycline-class antibiotics
The systemic treatment with the strongest and longest evidence in this condition, and what primary-care guidance names for anything past mild disease.
What it's called at the pharmacy
Generic name Oral tetracyclines
Sold as
- Lymecycline is the tetracycline named in national primary-care guidance for this condition, given once daily for around six weeks
A region missing here reflects the limit of what was verified, not the limit of what is sold.
Read the label
- Doxycycline — NHS medicines NHS (UK)
How it works
The benefit is generally attributed to the anti-inflammatory properties of the class rather than to killing anything — these drugs inhibit inflammatory enzymes and neutrophil activity at doses below those needed to suppress bacteria, which is why sub-antimicrobial regimens are used in several facial conditions.
What the evidence says strong
Oral tetracycline outperformed topical metronidazole in a double-blind randomised comparison of 108 patients, and beat placebo in a separate randomised study. Systematic review graded the certainty low while still identifying it as the intervention with the most supporting data, and formal appraisal of the wider literature found it the point of greatest consistency.
Typical course
A finite course rather than an open-ended one — primary-care guidance describes around six weeks, and the trial evidence shows separation from placebo from about day twenty.
Commonly discussed side effects
- Abdominal discomfort and nausea, the commonest reason a course is not completed
- Photosensitivity, which matters more on a face already being managed
- Oesophageal irritation if a capsule is taken lying down or without water
- Thrush and other consequences of altered flora
Worth knowing
- Not used below roughly eight years of age, or in pregnancy, because of effects on developing teeth and bone
- Interactions with retinoids and with anticoagulants are the ones a prescriber will check
- It treats the eruption and not the cause — a course taken while the driver continues buys time rather than a result
Pregnancy: Tetracyclines are avoided in pregnancy and while breastfeeding because of their effect on the developing skeleton and on tooth formation. Where a systemic drug is genuinely needed during pregnancy, a macrolide is the class usually considered instead.
Why it sits where it does: It sits above the topicals because it outperformed one of them directly in the only head-to-head trial this condition has, and below nothing in the ordinary adult presentation. It is the last rung most people ever need.
Prescription oral Oral macrolides
Systemic macrolide antibiotics
Individualized — ask a clinician
Oral macrolides
Systemic macrolide antibiotics
The systemic option for everyone the tetracyclines exclude — young children and pregnancy — and the drug class with the most published paediatric experience in this condition.
What it's called at the pharmacy
Generic name Oral macrolides
Sold as
- Erythromycin and clarithromycin are the macrolides named in national primary-care guidance for this condition
A region missing here reflects the limit of what was verified, not the limit of what is sold.
How it works
Macrolides share the anti-inflammatory behaviour attributed to the tetracyclines — suppression of neutrophil function and of inflammatory cytokine production — without the effect on developing teeth and bone that rules the older class out in children.
What the evidence says promising
A retrospective series of 222 children reported around seven in ten reaching complete resolution on regimens usually combining oral azithromycin with a topical. A cohort of 43 children with the granulomatous variant treated with oral clarithromycin over a median of ten weeks reported clearance in more than nine in ten by six months, with scarring confined to the more severe cases.
Typical course
Courses in the published paediatric work have run from a few weeks to a few months, with duration tracking baseline severity rather than being fixed in advance.
Commonly discussed side effects
- Nausea, abdominal pain and diarrhoea
- Altered taste with clarithromycin
- Effects on heart rhythm in susceptible people, which is what makes the interaction check matter
Worth knowing
- Macrolides interact with a long list of common medicines, and that check belongs to the prescriber
- Evidence here is retrospective throughout — there is no randomised trial of a macrolide in this condition
- A course does not remove the need to deal with the driver, and recurrence has been described specifically in children who remain on inhaled steroids
Pregnancy: Erythromycin has the longest record of use in pregnancy of any drug on this page and is the systemic option usually considered when one is genuinely needed. Choice and timing belong with a prescriber alongside antenatal care.
Why it sits where it does: It occupies the oral rung wherever a tetracycline cannot go, which is not a demotion — it is simply where the paediatric evidence actually is, since almost every published series in children rests on this class.
Prescription oral Low-dose oral isotretinoin
Systemic retinoid
Avoid in pregnancy
Low-dose oral isotretinoin
Systemic retinoid
The last resort, reserved for eruptions that have resisted everything else, and reported almost exclusively in the granulomatous form.
What it's called at the pharmacy
Generic name Low-dose oral isotretinoin
Sold as
- Prescribed by or under the supervision of a consultant dermatologist, under formal pregnancy-prevention requirements
A region missing here reflects the limit of what was verified, not the limit of what is sold.
Read the label
How it works
Isotretinoin shrinks sebaceous glands, alters follicular keratinisation and has broad anti-inflammatory effects. Why that helps a granulomatous facial eruption is not fully explained, which is worth stating: the reports here are of a result rather than of an understood mechanism.
What the evidence says limited
Published experience consists of case reports in recalcitrant disease, describing improvement over roughly three months and resolution by six on alternating low daily doses. Reviews list it as an option after standard therapies have failed rather than as a step in an ordinary course.
Typical course
Months rather than weeks in the reported cases, at doses well below those used for severe acne.
Commonly discussed side effects
- Dryness of lips, eyes and nasal lining, which is near-universal
- Muscle aches and, less often, disturbed night vision
- Blood lipid and liver changes, which is what the monitoring is for
- Mood changes are discussed at every consultation and are part of the standard conversation
Monitoring
Blood tests before starting and periodically during treatment, plus the pregnancy testing and contraception requirements that apply to everyone who can become pregnant.
Worth knowing
- Absolutely contraindicated in pregnancy and governed by a formal prevention programme
- The evidence base here is case reports, which is a thinner footing than for any other drug on this page
- It does not remove the need for the driver to have gone; it is used when removal plus a full course of antibiotics has not been enough
Pregnancy: Isotretinoin causes severe birth defects and is managed under a mandatory pregnancy-prevention programme with testing and contraception requirements before, during and after treatment. It is not an option at any point during pregnancy or while trying to conceive.
Why it sits where it does: It sits at the top of the granulomatous column alone, because that is the only presentation for which its use is reported with any consistency, and because the ordinary form of this condition essentially never reaches it.
On the horizon
Pipeline information current as of 2026-09-01
Only treatments in phase 3 trials, submitted to a regulator, or recently approved for this condition. Earlier work is left out on purpose: most of it never arrives, and listing it turns a research programme into a promise.
This changes nothing about what to do now. Nothing in this section changes what is available or advisable today. It is here so that a reader knows what is moving, not so that anyone defers treatment waiting for it.
In late-stage development
This section is empty, and the emptiness is the finding rather than a gap. No treatment has ever been approved by a regulator specifically for this condition — every drug described on this page is prescribed outside its licensed indication, which is routine, lawful, and the reason the trial evidence is as thin as it is. There is no phase 3 programme to report. What movement exists is at the case-report end: individual reports of oral JAK inhibitors clearing disease that resisted everything else, and two small series of a narrow-spectrum tetracycline. Neither is close to the threshold this section uses, and both are described further down among the lower-evidence options instead.
The other things people try
This condition attracts a wide field of half-treatments, partly because the evidence for the standard ones is thinner than most pages admit. What follows is where the popular options actually stand, so that time and money go somewhere with a result attached.
- Adapalene gel limited evidence
Available without a prescription in several markets, which makes it the retinoid most likely to be reached for. The published support here is a single case report of a woman who cleared over four weeks. Retinoids are also reliably irritating on skin whose barrier is already the problem, which is the argument against trying it early.
- Sarecycline backed by smaller studies
A narrow-spectrum tetracycline, of interest because the case against the broad-spectrum ones is mostly about gut flora and resistance rather than efficacy. Two small reports exist — a retrospective handful of adults and a nine-subject open pilot — with improvement in everyone and no drug-related adverse events. Interesting, and a long way from a trial.
- Photodynamic therapy backed by smaller studies
One split-face study set it against topical clindamycin and reported higher clearance on the treated side. A third of the participants withdrew because of irritation from the treatment, and three developed pigment changes afterward. That combination — better result, worse tolerability, on a face — is why it has not travelled.
- Topical praziquantel backed by smaller studies
An antiparasitic ointment tested against vehicle in a single-blind pilot of forty-six adults, with most of the treated group clear or nearly clear at four weeks. The systematic review rated the certainty very low. It is also not available for topical human use in several countries, which limits how much the result matters in practice.
- Oral JAK inhibitors limited evidence
Two published case reports describe disease that had resisted calcineurin inhibitors and oral antibiotics clearing on a twelve-week course of an oral JAK inhibitor. These are individual patients, the drugs carry substantial monitoring requirements, and nobody has run a trial. Worth knowing the reports exist; not worth asking for.
- Barrier and soothing cosmetic fluids backed by smaller studies
The one non-drug category with published trials in this condition, and both were open-label and eight weeks long. Severity fell steadily and barrier measurements improved. Placed here rather than higher because neither was blinded or randomised — but it is the best-evidenced thing on any shelf, which says as much about the shelf as about the fluid.
- Mite-directed treatment limited evidence
Mite densities are genuinely elevated on affected skin, but the study that measured it found the elevation confined to people who had used topical corticosteroids — in those who had not, counts matched controls. That makes the mite a consequence of the treatment rather than a cause of the disease, and it explains why anti-mite therapy is reached for in one subgroup and not as a general answer.
- Switching toothpaste limited evidence
Tartar-control and high-fluoride formulations have been reported as triggers in this exact distribution in several case reports, and the mechanism is plausible: foam runs over the chin and perioral skin twice a day. There is no trial. It costs almost nothing to test for a fortnight, which is a reasonable basis for including it in the removal step and no basis at all for treating it as the cause.
- A weaker steroid as a bridge limited evidence
The most common self-devised plan and the one that most reliably fails. Substituting a milder corticosteroid keeps the same loop running at lower amplitude and postpones the withdrawal rather than avoiding it. Where a stepped reduction genuinely is the right approach — a potent preparation used for a long time — it is directed by a prescriber alongside something non-steroidal to cover the gap.
- Steroid-detox and withdrawal-support creams limited evidence
A retail category built on a real phenomenon. Topical corticosteroid withdrawal is documented and its features are described in systematic review; no product has been shown to shorten it. Most of what is sold here is emollient at a premium, which is not useless and is not what the label claims.
Sources and further reading
The staging, escalation points, and therapy descriptions on this page are anchored to published clinical guidelines rather than editorial judgment.
What this page is built on
- Concise guidance: National Primary Care Treatment and Referral Guidelines for Common Skin Conditions — section 20, perioral dermatitis
- Pharmacological interventions for periorificial (perioral) dermatitis in children and adults: a systematic review
- Perioral dermatitis: a review of the condition with special attention to treatment options
The studies behind this pathway
There is no single guideline behind this page, so it is built on primary-care guidance and on the trials themselves. Every link below was resolved against the journal record before publication; a citation printed without one is a citation that could not be confirmed rather than a link left out.
- Pharmacological interventions for periorificial (perioral) dermatitis in children and adults: a systematic review The certainty grading behind every evidence tier on this page, and the finding on azelaic acid gel.
- Critical appraisal of reports on the treatment of perioral dermatitis That discontinuation of corticosteroids and cosmetics is among the few points the literature agrees on.
- Evidence based review of perioral dermatitis therapy The zero-therapy concept and the relative standing of the treatments described here.
- Perioral dermatitis: a review of the condition with special attention to treatment options The overall ladder order, and the differences between adult and paediatric options.
- Periorificial dermatitis: Pathophysiology, diagnosis, and management The barrier and microbiome account of mechanism, and the absence of any regulator-approved therapy.
- A randomized, double-blind, vehicle-controlled study of 1% pimecrolimus cream in adult patients with perioral dermatitis The calcineurin inhibitor rung, and the observation that its benefit concentrates in steroid-driven disease.
- Topical calcineurin inhibitors for pediatric periorificial dermatitis The response rates quoted for calcineurin inhibitors alone and in combination in children.
- Topical metronidazole in the treatment of perioral dermatitis The head-to-head comparison of topical metronidazole with an oral tetracycline over eight weeks.
- Steroid-induced periorificial dermatitis in children--clinical features and response to azelaic acid The azelaic acid series, including the transient early worsening described in a minority.
- Pediatric periorificial dermatitis: clinical course and treatment outcomes in 222 patients Paediatric outcomes, the proportion with a prior steroid history, and recurrence in inhaled-steroid users.
- Management of childhood granulomatous periorificial dermatitis with clarithromycin: A retrospective cohort study The granulomatous column: clearance rates, treatment duration, and the scarring risk in more severe cases.
- Density of Demodex folliculorum in perioral dermatitis That elevated mite density is confined to those with prior topical corticosteroid use.
- Association of Nasal and Inhalant Corticosteroids With Pediatric Perioral Dermatitis: A TriNetX Retrospective Cohort Study The magnitude of the association between inhaled or nasal corticosteroids and a new diagnosis in children.
- Topical corticosteroid withdrawal ('steroid addiction'): an update of a systematic review The features and course of the withdrawal flare described in step 3.
- Efficacy and Safety Results of a Drug-Free Cosmetic Fluid for Perioral Dermatitis: The Toleriane Fluide Efficacy in Perioral Dermatitis (TOLPOD) Study The bland-fluid evidence, and the historical-control comparison behind it.
- Efficacy of a skin care cream with TRPV1 inhibitor 4-t-butylcyclohexanol in the topical therapy of perioral dermatitis The barrier measurements reported alongside falling severity on a drug-free cream.
- Ivermectin therapy for papulopustular rosacea and periorificial dermatitis in children: A series of 15 cases The ivermectin card, and the case-series basis on which it sits.
- Topical praziquantel as a new treatment for perioral dermatitis: results of a randomized vehicle-controlled pilot study The praziquantel entry among the lower-evidence options.
- Photodynamic therapy for perioral dermatitis The split-face photodynamic therapy result, including the withdrawals for irritation.
- A Retrospective Review of a Cohort of Patients with Periorificial Dermatitis Treated with Sarecycline The sarecycline entry among the lower-evidence options.
- Perioral dermatitis successfully treated with topical adapalene The single case report behind the adapalene entry.
- A Case of Perioral Dermatitis Successfully Treated with Abrocitinib The oral JAK inhibitor case reports mentioned in the pipeline note and the lower-evidence list.
- PERIORAL DERMATITIS: STILL A THERAPEUTIC CHALLENGE The isotretinoin card, and the description of zero therapy as the treatment of choice in mild disease.
- Periorificial dermatitis in infants and preschoolers - a narrative review The paediatric differential, the lip-margin sparing sign, and the tetracycline age restriction.
This condition has no dedicated national or international guideline of the kind acne and eczema have. What exists instead is primary-care guidance, one systematic review with formal certainty grading, and a small number of trials underneath both — which is the honest footing for everything on this page.
Guidance and evidence reviews
The first is the guidance a family doctor is most likely to be working from in the UK; the second is the only formal certainty assessment this condition has had.
- Concise guidance: National Primary Care Treatment and Referral Guidelines for Common Skin Conditions Primary Care Dermatology Society (UK) Section 20 covers this condition: stop the steroid, a topical antibiotic in milder cases, an oral course for everyone else.
- Pharmacological interventions for periorificial (perioral) dermatitis in children and adults: a systematic review Journal of the European Academy of Dermatology and Venereology Eleven studies, 733 participants, GRADE assessment throughout — and the source of the uncomfortable finding that all of it is low or very low certainty.
- Evidence based review of perioral dermatitis therapy Giornale Italiano di Dermatologia e Venereologia Where the term zero therapy is set out plainly, alongside a ranking of what the evidence actually supports.
Patient information
Written to be read rather than decoded, and useful for showing a family member what the withdrawal is going to look like.
- Peri-oral dermatitis British Association of Dermatologists Short, clear, and explicit about the steroid connection.
- Perioral dermatitis — patient information leaflet Primary Care Dermatology Society The patient-facing companion to the guidance above.
- Periorificial dermatitis DermNet Filed under the broader term, with images covering the periorbital and perinasal patterns as well as the perioral one.
Take the charts with you
Print exactly the piece you need. Each chart is one landscape page — no prose, no controls, sized to pin up or hand to someone — and the full document is the whole pathway as it reads on screen. Every printout carries the date, the source, and a line making clear it is not a diagnosis.
Printing happens in your browser. Your answers stay private — saved on this device, and to your free account when you're signed in.