Are the steroid creams for eczema actually safe or not? Understanding the JAK black box warning, public dermatology triage, and how to apply for the subsidies

TL;DR Eczema (atopic dermatitis, AD) is a chronic, relapsing inflammatory skin disease that can be controlled but not cured. The Hong Kong figures have to begin by asking which question was put: among the same group of 6–7-year-old primary pupils (2015–16, n=3,698), "eczema ever" was 30.9%, while "a chronic rash in the typical sites" was only 8.1% — the same children, a difference of about fourfold. There is no incidence figure at all for adult eczema in Hong Kong, a point stated in two first-hand documents: the Hong Kong College of Paediatricians' review and a HKU/PolyU study protocol. The greatest misconception is that the risks and the benefits of steroid creams come from opposing camps. They do not — the people who say that used properly the benefits outweigh the harms, and the people who say skin thinning, adrenal suppression and withdrawal reactions, are the same regulators and the same guidelines; and the UK Medicines and Healthcare products Regulatory Agency (MHRA) writes as well that using too little equally prolongs the course of treatment and raises the risk of certain side effects. The international research on steroid phobia (16 studies) found prevalence ranging from 21.0% to 83.7%, and the conclusion that fear affects adherence rests on only 2 of those studies, the review stating for itself that they could not be pooled. As for oral JAK inhibitors: the FDA in the United States, the EMA in Europe and the MHRA in the United Kingdom give different answers to whether that warning concerns eczema at all, but all three impose a population-level restriction, differing only in the axis they draw the line on — the United States puts the restriction in the indication itself (refractory disease, inadequately controlled by or unsuitable for other systemic drugs including biologics, aged 12 or above), with two further directions on smoking and thrombotic risk in the warnings section; while Europe and the United Kingdom additionally write age and cardiovascular, malignancy and smoking risk factors into an access condition of "only where no suitable alternative is available". Both drugs' labels also set out the tests required before and during prescribing (latent tuberculosis, full blood count, lipids, vaccination), which belong with the black box warning as one set. And Hong Kong's only professional body guideline was written before the entire era of oral JAK inhibitors — a Hong Kong reader asking how Hong Kong views JAK inhibitors has no Hong Kong answer. On cost: public dermatology is a Department of Health service, not a Hospital Authority one, and the Department's own document shows that as at 30 June 2026, new case appointment dates ran from December 2027 to February 2029 — but that is the figure for stable cases. The Department triages: a new case of severe skin disease is pledged to be seen within 8 weeks, and the medians for 2022 and 2023 were 2.7 and 2.9 weeks. Dupilumab, abrocitinib, upadacitinib and lebrikizumab are all Self-financed Items with Safety Net on the Hospital Authority Drug Formulary (the formulary exists in English only), and the annual fee cap of ten thousand dollars introduced from January 2026 states in terms that it excludes self-financed drugs. This article will not tell you which drug to use, at what strength, for how long, or which doctor to see — potency, quantity and duration are matters of prescription, and this article only reports what each guideline says for itself, leaving the decision to a doctor.

#What kind of disease is eczema? And why does the Hong Kong government's own eczema page say nothing at all about the mechanism?

Eczema is not simply "sensitive skin"; it is the product of a skin barrier defect plus an immune response. But if you read only the Hong Kong government's eczema page you would never encounter that layer at all.

The only government eczema page for the public in Hong Kong is the Department of Health Student Health Service's "Eczema" (the page's own revision date: 「二零二二年六月修訂」). It lists possible causes and lists treatments, but nowhere mentions the skin barrier, filaggrin, the epidermis, cytokines, IL-4 or IL-13, Th2 or immune dysregulation, and gives no potency grading, quantity or duration guidance. So not one sentence of the mechanism below can be cited to a Hong Kong government publication — which is itself something a reader should know.

The original of that page's account of causes (note its own 「可能」 and 「如」 — possible, and such as, which is to say examples rather than a complete list) is at [Note 1]: the possible causes of eczema include heredity; allergens such as dust, wool, pollen, and animal fur and dander; and food proteins such as egg, seafood, dairy, beef and peanut.

The mechanism has to come from the international literature. The Lancet review of atopic dermatitis of 2020 (Langan et al., Lancet 2020;396:345-360) states [Note 2]: the pathophysiology is complex, involving a strong genetic predisposition, epidermal dysfunction and T-cell driven inflammation; and although type 2 mechanisms predominate, there is increasing evidence that multiple immune pathways are involved.

In one sentence: it is not one immune pathway that has gone wrong, so no one drug will hit the mark for everyone. Popular articles frequently write eczema up as a purely Th2 disease; that review itself does not.

#Filaggrin — and a limitation that matters a great deal to Chinese readers

Palmer et al.'s study in Nature Genetics in 2006 is the one that established the skin barrier defect as primary rather than secondary [Note 3]: two independent loss-of-function variants of the filaggrin (FLG) gene (R510X and 2282del4) are very strong predisposing factors for atopic dermatitis, and about 9% of people of European descent carry those two variants.

Note that last sentence: the 9% is a figure for people of European descent, and both variants are European variants. Chen et al.'s study in the British Journal of Dermatology in 2011 (Singaporean Chinese, discovery set n=92; a screening set of 425 patients against 440 controls) puts it plainly: the FLG mutations common in Europeans are rare or absent in Chinese patients with ichthyosis and atopic dermatitis [Note 3].

That study found a further 22 FLG null mutations in Chinese subjects (14 of them novel), the combined genotype being strongly associated with AD (odds ratio 3.3, P=5.3×10⁻⁹). These are Singaporean Chinese, not Hong Kong Chinese — there is no FLG study of a Hong Kong population among the material cited here.

#Why it "clears up and comes back"

The Hong Kong College of Paediatricians' review of 2020 (HK J Paediatr 2021;26:42-57) makes the point about mechanism that is of most everyday use [Note 4]: atopic dermatitis is a chronic inflammation, and subclinical inflammation persists even in skin that appears free of lesions; and so the same document writes that it must be stressed that, because subclinical inflammation persists, topical steroids should not be tapered too early.

In other words: "it looks better" and "the inflammation has stopped" are two different things — and every argument about how long to keep going has its root there.

#Food and prevention: Hong Kong has a guideline, but it is about prevention, not treatment

That Department of Health Student Health Service page lists food proteins such as egg, seafood, dairy, beef and peanut among the possible causes (the page's own 「可能」 and 「如」 make these examples rather than a complete list).

As for whether avoiding foods in pregnancy prevents eczema in the child, Hong Kong does have a local guideline dedicated to prevention — the Hong Kong Institute of Allergy's "Guidelines for allergy prevention in Hong Kong" (Hong Kong Med J 2016;22:279-85, published June 2016) states [Note 5]: a healthy and unrestricted maternal diet during pregnancy and lactation is recommended, since preventive dietary restriction is unlikely to reduce the development of allergic disease; breastfeeding is recommended for the first 6 months; and in high-risk infants, feeding a hydrolysed formula rather than a standard cow's milk formula has been shown to reduce atopic eczema significantly.

Note the scope of that guideline: it deals with how to prevent the disease before it appears, not with how to eat once you have eczema. Among the material cited here, no Hong Kong guideline addresses diet in those who already have the disease.

⚠️ If you take eczema to be sensitive skin that will clear once the allergen is avoided: the barrier defect and the immune layer do not disappear because one thing is avoided, and the Hong Kong government's patient page never mentions that layer at all.

#How many people in Hong Kong actually have eczema? Why can the same children answer 30.9% and also 8.1%?

The best eczema figure for Hong Kong is not a number but a set of them; depending which questionnaire item is used the answer differs fourfold — so no Hong Kong eczema percentage that fails to say which item it used should be taken at face value.

The clearest side-by-side comparison comes from the Hong Kong ISAAC series (Lee SL et al., Int J Environ Res Public Health 2022;19:16503). The third round surveyed 6–7-year-olds at 33 primary schools, n=3,698 (51.4% boys). The same children, the same questionnaire:

A contrast: the same 3,698 pupils aged 6–7 — asked "has your child ever had eczema?", 30.9% said yes; asked "has your child ever had a chronic rash in the typical sites, the elbow creases and behind the knees?", only 8.1%. The first measures how many people accept a label, the second how many meet a case definition. Neither is wrong, but they are not interchangeable.

#The Hong Kong survey of 2021 said two different things in two languages

The survey of allergic disease in schoolchildren released by the University of Hong Kong on 13 January 2022 (February to July 2021, 19 primary schools, 1,165 parent–child pairs, mean age 7.03, 62% boys) says in its Chinese release: the results show prevalence rates for eczema, asthma, rhinitis and allergic rhinitis of 41.6%, 5.5%, 59.4% and 46.1% respectively; and that within the past year, 16.3% of the pupils were troubled by an itchy rash (eczema) [Note 6].

The English release writes the same sentence as "the prevalence rates for eczema … are 41.6%", without saying whether that is ever or current; and the sentence immediately following in the Chinese is 「相比2004年香港的ISAAC數據，曾患濕疹，鼻炎以及鼻敏感的學童有上升的趨勢」 — that against the earlier Hong Kong ISAAC data, there is a rising trend in pupils who have ever had eczema, rhinitis and allergic rhinitis.

What settles it is the study appendix (in English only), where "Eczema ever" (boys 41.9 / girls 41.1) and "Chronic rash in past year" (boys 17.2 / girls 15.0) are two separate rows. That is: 41.6% is "ever in a lifetime", and 16.3% is the past year. Those two figures are frequently cited as though they were the same thing.

The same release also states its own sampling difficulty [Note 6]: because of a low participation rate, a third round of recruitment was conducted; nearly 900 invitation letters were sent to schools, and schools were also recruited through personal networks. No peer-reviewed paper for this set of 2021 figures could be found among the material cited here; it is a press release plus an appendix.

#The Chinese-language questionnaire itself misses people

This is the piece of methodological research of most use to a Chinese-language reader. Chan HH et al. (Clin Exp Allergy 2001;31:903-7) validated the Chinese version of the ISAAC questionnaire against clinical examination by dermatologists as the gold standard, in 1,920 children aged 3–5 at 13 kindergartens, and their conclusion states [Note 7]: the low scores are believed to stem from the reduced sensitivity of the Chinese questionnaire, ranging from 23.5% to 70.6%; and that the low prevalence of atopic eczema reported in Chinese populations in previous studies is at least in part a product of the translated questionnaire itself.

Read together: put that validation study beside the table above, and the historical Hong Kong figures of 3.6%–4.2% for current eczema are quite likely underestimates, with the cause lying not in the patients but in the measuring instrument. That inference rests on Chan et al.'s own concluding sentence, and is not this article's own calculation.

#And adults? Hong Kong has no figure, and that absence can be verified

The Hong Kong College of Paediatricians' review of 2020 writes of itself: local prevalence data have not been updated for more than a decade [Note 8]. A HKU/PolyU study protocol registered on ClinicalTrials.gov (NCT04154839, record last updated 2022-05-05) likewise gives as its rationale that Hong Kong lacks complete epidemiological data on atopic dermatitis in children and adults [Note 8].

The international range cited in that same record is "about 15–30% in children and about 0.3–14% in adults" — an adult range running from 0.3% to 14% is not by itself a number that can stand for Hong Kong.

As for the Centre for Health Protection: in the whole text of the April 2024 issue of its Non-Communicable Diseases Watch, "eczema" and "dermatitis" appear zero times. The reason lies in that centre's own definition of non-communicable disease — cancer, cardiovascular disease, diabetes, chronic respiratory disease — within which eczema does not fall. The Centre for Health Protection has published no eczema prevalence figure at all.

#The global background, and the burden in Hong Kong clinical samples

The Global Burden of Disease study (GBD 2021, Lancet Respir Med 2025;13:425-446): in 2021 an estimated 129 million people worldwide (95% UI 1.24 hundred million to 134 million) had atopic dermatitis, up from 107 million in 1990; but the age-standardised prevalence fell over the same period by 8.3% (from 1,885.4 to 1,728.5 per 100,000). Another GBD analysis (Laughter et al., Br J Dermatol 2021;184:304-309) puts atopic dermatitis 15th among all non-fatal diseases, and, measured in disability-adjusted life years, the heaviest burden of any skin disease.

Hong Kong's local burden data are not prevalence figures, but they are worth knowing: at one Hong Kong paediatric dermatology clinic, 432 children and 380 parents recruited between November 2018 and October 2020 (Lam PH et al., Hong Kong Med J 2024;30:362-70) showed eczema severity and quality of life scores significantly positively correlated with levels of depression, anxiety and stress in both the parents and the children (children r=0.28–0.72; parents r=0.20–0.52). This is a clinic sample, not a population sample, so what it speaks to is that the worse the disease the heavier the psychological burden on the family, not what proportion of Hong Kong eczema families have emotional difficulties.

⚠️ If you come across a sentence of the form "X per cent of Hong Kong people have eczema": ask three questions first — which age group? which questionnaire item? ever, or in the past year? If any of the three cannot be answered, the number cannot be used.

#Are steroid creams safe or not? — both answers come from the same bodies

This section is not "the benefits first, then the harms". It is here to point out something frequently overlooked: the people who say that used properly the benefits outweigh the harms, and the people who say skin thinning, adrenal suppression and withdrawal reactions, are the same guidelines and the same regulator; and that same regulator also says that using too little carries a risk of its own.

The timeline first, because every piece of this argument has a date: the original fingertip unit study was published in 1991; NICE issued CG57 in 2007 (last updated 22 September 2025); the Hong Kong College of Paediatricians issued its local guideline in 2013 (endorsed 14 December 2012); the systematic review of steroid phobia was published in 2017; the Hong Kong College of Paediatricians' review of 2020 appeared in 2021; and the MHRA issued two drug safety updates, in September 2021 and May 2024. Every figure, every body site and every number of days below is written by one of those documents; this article only reports them, and does not convert them into instructions for you.

#The first side: what the guidelines say for themselves

The whole topical steroid passage of NICE CG57 (issued 2007, last updated 22 September 2025; its scope is children under 12) opens with a sentence about communication [Note 9]: discuss the benefits and harms of treatment with topical corticosteroids with children with atopic eczema and their parents or carers, emphasising that the benefits outweigh possible harms when they are applied correctly (1.5.1.12).

Only then does it come to potency following severity and site. This has to be read from its own opening sentence, because that opening sentence is the rule governing the whole recommendation — potency follows severity, and site is a modifying condition, not the other way about. The provision in full is at [Note 9]: the potency of topical corticosteroids should be tailored to the severity of the child's atopic eczema, which may vary according to body site — mild potency for mild severity, moderate for moderate, potent for severe; use mild potency for the face and neck, except for short-term (3 to 5 days) use of moderate potency for severe flares; use moderate or potent preparations for short periods only (7 to 14 days) for flares in vulnerable sites such as the axillae and groin; and do not use very potent preparations in children without specialist dermatological advice (1.5.1.13).

Why the opening sentence has to be quoted with it: quoting only the three bullets that follow reads as though NICE assigns potency by body site; but the provision's own structure sets potency by severity first, and then narrows it by site. Without the opening sentence, a reader whose severity is different will carry the site rules across wholesale.

The corresponding part of the Hong Kong College of Paediatricians' guideline of 2013 is the passage in the material cited here that speaks most directly to Hong Kong [Note 10]: face and neck — moderate or potent for 3 to 5 days; axillae and groin — moderate or potent for no more than 1 to 2 weeks; trunk and limbs — potent for no more than 2 weeks.

Other provisions written very plainly: apply only to areas currently flaring, or that have flared within the past 48 hours (NICE 1.5.1.15, and the Hong Kong College of Paediatricians to the same effect); apply once daily, at most twice (NICE 1.5.1.14, and the Hong Kong guideline of 2013); and do not use potent preparations in infants under 12 months without specialist dermatological supervision (NICE 1.5.1.17).

As for what to do when one or two weeks have not brought it under control, NICE 1.5.1.16 is a three-step recommendation, and it is the second and third steps a reader can use [Note 9]: where a mild or moderate topical corticosteroid has not controlled the eczema within 7 to 14 days — first exclude secondary bacterial or viral infection; in a child aged 12 months or over, a potent topical corticosteroid may be used for as short a time as possible (no more than 14 days, and not on the face or neck); and if it remains uncontrolled, review the diagnosis and refer the child for specialist dermatological advice.

The third step is the one most easily left out of that recommendation and the most practical: where it remains uncontrolled, the provision says review the diagnosis and refer, not increase the potency by yourself. As for proactive maintenance once control is achieved, NICE writes [Note 9]: after control, in children with frequent flares (2 to 3 per month), consider a topical corticosteroid applied for 2 consecutive days per week to the sites prone to flaring to prevent them, and review this strategy within 3 to 6 months (1.5.1.19).

#Quantity: the only Hong Kong source with figures, and its own two versions

The Hong Kong College of Paediatricians' review of 2020 sets out a set of figures that is the only local one with an upper limit on quantity in the material cited here, but it has to be read with the situation it applies to [Note 11]: as maintenance treatment applied two or three times a week, a monthly quantity of 15 g in infants, 30 g in children and 90 g in adolescents is generally safe.

Note the scope of that sentence: it is about maintenance treatment two or three times a week, not a general monthly ceiling for any situation. Quantities during an acute flare are not covered by it.

As for the fingertip unit (FTU) — nine popular accounts in ten write it up as about 0.5 g covering two palms. That formulation has to be taken apart:

• The original study has no figure of 0.5 g in it. Long CC and Finlay AY (Clin Exp Dermatol 1991;16:444-7, 30 adult patients) measured area: one fingertip unit covered 286 square centimetres (standard deviation ±80, n=30); 312 square centimetres in men (±90, n=16) and 257 in women (±55, n=14) [Note 11]. Weight is nowhere in the paper. What the original measured was an ointment, in adults.
• The "0.5 g" and "two palms" formulations come from downstream guidelines — in Hong Kong, from the Hong Kong College of Paediatricians, whose own two documents word it differently. The 2013 one writes 「weighs approximately half a gram; enough to cover a surface area of two adult palms including fingers」; the 2021 one writes 「approximates to 0.5 gram … adequate to cover skin area equivalent to two adult hands with fingers together」. Both are the same college.
• Those same two documents also give two different intervals between applications: the 2013 one writes 「wait for at least 30 minutes between application of emollients and TCS」; the 2021 one writes 「At least 15 minutes should lapse between application of different topical medications and emollients」.

These are not quibbles: a patient following "30 minutes" rather than "15 minutes", and measuring by "palms" rather than "hands", arrives at a different quantity and a different schedule — and the divergence comes from one professional body. This article sets the two versions side by side and does not choose between them for anyone.

The review of 2020 is itself candid about the state of the evidence: the evidence supporting recommendations of this kind on how to apply topical corticosteroids is limited [Note 11].

#Emollients: the only provision with a quantity figure, and the safety warning beside it

The argument about steroid quantities is a lively one, but the provision in NICE CG57 that actually writes down how much to prescribe is the emollient one, not the steroid one.

CG57 1.5.1.4 writes [Note 12]: offer children with atopic eczema a choice of unperfumed emollients to use every day for moisturising; this may be a combination of products or a single product for all purposes; leave-on emollients should be prescribed in large quantities (250 to 500 g weekly), and made easily available to use at nursery, pre-school or school.

⚠️ And above that recommendation, in the opening sentence of the same section, NICE hangs a safety warning that bears directly on quantity; the two have to be read together [Note 12]: emollients are important in the management of dry skin conditions, but the MHRA has warned that a build-up of emollient residue on clothing and bedding is a fire hazard.

In one sentence: the provision tells you to prescribe large quantities with one breath and warns you that residue on clothes and bedding is a fire hazard with the next — the two sentences sit in the same section of the original, and the first should not be quoted alone. The material cited here goes no further than that sentence of CG57; the text of the MHRA fire warning itself was not obtained, so this article makes no statement about which products it applies to or what precautions to take.

#Two formulations corrected by a Hong Kong professional body itself

The Hong Kong College of Paediatricians' review of 2020 writes [Note 13]: diluting a topical corticosteroid with an emollient does not reduce its potency or its side effects; and tachyphylaxis to topical corticosteroids is unproven, while there is good evidence supporting the safety of the appropriate use of topical corticosteroids in children with AD.

There is a divergence between guidelines here worth pointing out: NICE CG57 provision 1.5.1.20 writes that where tachyphylaxis to a topical corticosteroid is suspected, an alternative preparation of the same potency may be considered rather than a stronger one [Note 13] — that is, NICE writes a course of action for the situation of suspected tachyphylaxis, while the Hong Kong College of Paediatricians says it is unproven. The difference in tone between the two documents can be seen in the originals at [Note 13].

#The second side: the harms the regulator itself has recorded

The MHRA's drug safety update of 29 May 2024 both affirms and warns within one passage [Note 14]: topical corticosteroids are considered safe and effective, but rarely can cause serious side effects through systemic absorption, such as skin thinning, adrenal suppression and, very rarely, Cushing's syndrome; and the incidence of those more serious side effects is related to the quantity, the potency and the duration of use.

The same document's description of two specific harms (reproduced in full at [Note 14]): thinned skin is translucent with small visible blood vessels, is more fragile and more prone to stretch-mark-like striae, and is difficult to detect on brown or black skin, so careful monitoring is needed; and adrenal suppression arises from excessive use of topical corticosteroids, with symptoms including low blood pressure, dizziness and fainting — a serious, potentially fatal condition needing emergency treatment, and where adrenal suppression is already present, stopping topical corticosteroids abruptly is dangerous, the person likely needing oral steroid replacement.

The corresponding description in the Hong Kong College of Paediatricians' guideline of 2013 [Note 14]: cutaneous complications include striae, atrophy, telangiectasia and perioral dermatitis; while systemic complications such as Cushing's syndrome, hypothalamic-pituitary-adrenal axis suppression, growth retardation, osteoporosis and hypertension are rare.

⚠️ That list is not complete, and no document treats it as complete. The two passages above are what those two documents each wrote down; not being on a list does not mean something does not matter. Any new change should be assessed by a doctor.

#The third side — the one most often left out: using too little carries a risk too

The same MHRA, in the public version of its update of 15 September 2021, wrote this [Note 15]: besides the known side effects of using too much or for too long, remember also that using too little can prolong treatment and increase the risk of certain adverse reactions. Its advice to healthcare professionals in the same update includes telling patients how much to apply, since applying too little can prolong the course of treatment.

In one sentence: those two sentences and the "skin thinning, adrenal suppression" above come from the same regulator — the difference is that they are in two MHRA drug safety updates, of 15 September 2021 and 29 May 2024. So "the risks of steroids" and "the risks of too little steroid" are not the claims of two camps; they are positions the same regulator set down in two documents. Either quoted alone is not that regulator's whole position.

#Steroid phobia: how solid are the figures?

The systematic review by Li AW, Yin ES and Antaya RJ (JAMA Dermatol 2017;153:1036-1042) — the same one the Hong Kong College of Paediatricians itself cites — retrieved 490 papers and included 16 (all cross-sectional), with these results [Note 16]: the prevalence of steroid phobia ranged from 21.0% (95% CI 15.8%–26.2%) to 83.7% (81.9%–85.5%); the definitions of "phobia" varied greatly between studies, from worry to irrational fear; and the questionnaires used ranged from 1 to 69 items. In the 2 studies comparing adherence between phobic and non-phobic groups, non-adherence was significantly higher in the phobic group (49.4% against 14.1%; 29.3% against 9.8%).

The same review's own limitation statement [Note 16]: the lack of standardisation in the naming and assessment of steroid phobia across studies made quantitative comparison and extrapolation of the data impossible.

So: there is no pooled estimate, and this article will not pool one. A range as wide as 21% to 83.7% reflects, in itself, 16 questionnaires asking different things. And the conclusion that fear makes people not take the drug rests on 2 of those studies.

When the Hong Kong College of Paediatricians' review of 2020 cites that same review, it gives the upper bound as 83.9% (the original: 「ranged from 21% to 83.9%」), differing from the 83.7% of the source paper; this article takes the original paper as authoritative and records both. That review also writes [Note 16]: steroid phobia is an important cause of treatment failure; and in the local population, misconceptions and fallacies about AD are associated with distrust of and unrealistic expectations of western medicine, fear of anti-inflammatory drugs, and the use of alternative and complementary therapies.

⚠️ But: Hong Kong has no local prevalence figure for steroid phobia. What Hong Kong's professional bodies cite is an international range. Any statement that "X% of Hong Kong parents fear steroids" has no source in the material cited here.

⚠️ If you are holding a tube and thinking you had better not use so much: your worry has a name and 16 studies behind it, and you are not imagining it; but note that "using too little prolongs treatment" and "skin thinning, adrenal suppression" both come from the same regulator, the MHRA; and "used correctly the benefits outweigh the harms" is NICE CG57's formulation. How quantity, potency and duration are to be set is a question of prescription.

#When stopping the drug makes the whole face flare red, is that "steroid withdrawal"?

There is such a thing, and the regulator has set down its four phases; but the same regulator also states that there is still no agreed clinical definition, and — the key sentence — that it is difficult to distinguish a flare of the eczema itself from a withdrawal reaction.

The MHRA's description in its update of 15 September 2021 [Note 17]: topical steroid withdrawal reactions are thought to occur after prolonged, frequent or inappropriate use of moderate to high potency topical corticosteroids; they can occur after more than a year of daily use, and in children can occur within as little as 2 months of daily use; and people with atopic dermatitis are thought to be at highest risk.

Before quoting those four phases, the sentence immediately above them has to be quoted — because it defines what the term "TSW" is [Note 17]: patients have also reported a less well understood set of side effects known as topical steroid withdrawal (TSW) reactions; TSW is the widely accepted patient-led term for these reactions — this set of conditions may not meet the medical pharmacological definition of "withdrawal"; and although these reactions remain poorly understood, the available evidence suggests they generally occur in four phases.

The most important words there are "patient-led term": the name "TSW" came from patient communities and was adopted by the regulator afterwards, and the MHRA says plainly that this set of conditions may not meet the pharmacological definition of withdrawal. So the four phases below are a description of an observed course, not an established diagnosis.

The four phases as listed in the update of May 2024 (all reproduced at [Note 17]; the document itself hangs two footnote markers on the "four phases" sentence, whose contents this article has not obtained): an acute flare usually within days of stopping, with the skin burning, red and weeping, which can spread to sites never treated; then the skin becomes dry, itchy and scaly; then recovery begins, though the skin is more sensitive and can flare intermittently; and finally a return to the state before the topical corticosteroid was stopped, a recovery that may be quite prolonged.

The sentence immediately after the four phases is the update's own direction on management [Note 17]: where TSW is suspected, reassess whether to continue the topical corticosteroid and explore other treatment options.

#But how thick is the evidence? The regulator has said

The update of 2021 [Note 18]: over the period reviewed, 55 reports suggestive of topical steroid withdrawal reactions were identified from the Yellow Card scheme, most of them reported by patients themselves; the incidence of such reactions cannot be estimated; but given the number of patients using topical corticosteroids, reporting of severe withdrawal reactions is understood to be very rare.

The update of 2024 updated that figure and added a limitation that cannot be left out [Note 18]: Yellow Card reports up to August 2023 identified 267 reports either named as topical steroid withdrawal or carrying features often thought to be associated with it; since there is still no agreed clinical definition, not every case reported as TSW is a true withdrawal reaction, some possibly being a worsening of the underlying disease or due to other unknown causes; but the review did identify cases of adverse reactions associated with prolonged use of moderate to high potency steroids, and often severe ones.

The two updates do two things at once: they acknowledge that these reactions are real and can be severe, and at the same time refuse to give an incidence for them. The 267 is a number of reports, not an incidence, because the denominator — how many people have used topical corticosteroids — is not published, and because the regulator itself says some of them may not be true withdrawal reactions. So neither "withdrawal reactions are common" nor "withdrawal reactions do not exist" is what those two documents say.

#One thing this article will not do for you

The MHRA's update of 2021 writes for itself [Note 18]: it is difficult to distinguish a flare of the underlying skin condition — in which case further topical corticosteroid would help — from a topical steroid withdrawal reaction.

So this article will not supply any list of features for telling them apart. The management of the two situations runs in opposite directions (one calls for continuing the drug, the other for changing the plan), and telling them apart is not something a reader can do from symptoms. The same update's advice to healthcare professionals includes a corresponding line: tell patients to seek medical advice if their skin condition worsens while using a topical corticosteroid. The update of 2024 writes that where stopping in the past has produced a reaction suspected to be TSW, alternative treatment should be considered — that sentence is written for prescribers, not an instruction to patients to change their own drugs.

⚠️ If you have seen photographs of "steroid withdrawal" online and are wondering whether to stop everything: the same document quoted above also states that where adrenal suppression is already present, stopping topical corticosteroids abruptly is dangerous. Whether and how to stop is something to discuss with a doctor.

#Dupilumab and the JAK inhibitors — what are the actual figures?

The widely circulated "60 to 70% effective" is the figure for the arm that adds a topical corticosteroid; the registration trial figures for the drug alone are 51% and 44%, and the control arm has to be read alongside every time.

Two terms first: EASI-75 means a 75% or greater improvement from baseline in the Eczema Area and Severity Index; IGA 0 or 1 is the lowest two grades on the Investigator's Global Assessment scale used in the trial (the full definitions of that scale's grades are outside the material cited here). Both are proportions at week 16.

#Dupilumab (subcutaneous injection, anti-IL-4/IL-13)

Before the efficacy figures, note how the same label writes its own target population — dupilumab's United States indication is itself a restriction, only drawn on a different line from the JAK inhibitors' [Note 19]: for adult and paediatric patients aged 6 months and older with moderate-to-severe atopic dermatitis whose disease is not adequately controlled with topical prescription therapies, or when those therapies are not advisable; and DUPIXENT can be used with or without topical corticosteroids (1.1).

That is: even on its United States label, dupilumab is not first line; it comes after topical prescription therapy — but the line it draws is "topical prescription therapy has not controlled it", while the JAK inhibitors' line is "other systemic drugs, biologics expressly included, have not controlled it". The two are not at the same level.

The three randomised, double-blind, placebo-controlled trials in the FDA-approved prescribing information (SOLO 1, SOLO 2, CHRONOS) covered 2,119 adults with moderate-to-severe disease (the label states the entry criteria as inadequate control on topical medication, IGA ≥3, EASI ≥16 and body surface area involvement ≥10%), with a mean baseline age of 38, a mean EASI of 33, and 24% Asian:

So the 69% does exist, but it is the CHRONOS arm with concomitant topical corticosteroids (n=106), and that arm's control group (placebo plus topical corticosteroids, n=315) reached 23%. The monotherapy figures of 51% and 44% have control groups of 15% and 12%.

On safety, the FDA label's table of adverse reactions before week 16 (monotherapy N=529, placebo N=517): injection site reactions 51 (10%) against 28 (5%); conjunctivitis 51 (10%) against 12 (2%); oral herpes 20 (4%) against 8 (2%); other herpes simplex virus infections 10 (2%) against 6 (1%); and keratitis 1 (<1%) against 0. In the concomitant topical corticosteroid group (N=110 against N=315), conjunctivitis was 10 (9%) against 15 (5%), blepharitis 5 (5%) against 2 (1%), and keratitis 4 (4%) against 0. Discontinuation for adverse events was 1.9% in both the monotherapy and the placebo groups.

That 10% has to be read with two things: first, the label itself states that "conjunctivitis" is a pooled term (covering allergic, bacterial and viral conjunctivitis, eye irritation and eye inflammation); and second, 27% of the subjects had allergic conjunctivitis at baseline.

#Oral JAK inhibitors

⚠️ Two things must be known before reading the percentages below, and both are written on the same United States labels: first, abrocitinib (CIBINQO) and upadacitinib (RINVOQ) both carry a black box warning — serious infections, mortality, malignancy, major adverse cardiovascular events and thrombosis; and second, the United States indication for both is limited to refractory moderate-to-severe atopic dermatitis, to patients whose disease is not controlled by, or who are unsuitable for, other systemic drugs (biologics expressly included), and to those aged 12 or above. Which is to say that the population the efficacy figures below speak about is, under the framework of the United States label, people who have already tried other systemic treatment.

The head-to-head trial of abrocitinib against dupilumab (JADE COMPARE, Bieber et al., N Engl J Med 2021;384:1101-1112): 838 people randomised — abrocitinib 200 mg 226, 100 mg 238, dupilumab 243, placebo 131, all four groups using topical treatment at the same time. EASI-75 at week 12 was 70.3%, 58.7%, 58.1% and 27.1% respectively. But that trial's own conclusion is narrowly drawn [Note 20]: abrocitinib 200 mg — but not 100 mg — was superior to dupilumab in the itch response at week 2; and in most of the other key secondary endpoint comparisons at week 16, neither dose differed significantly from dupilumab.

Which is to say: the claim that JAK beats dupilumab holds, in that trial, for one dose, one measure and one time point — 200 mg on the itch response at week 2. In most of the secondary endpoint comparisons at week 16, neither dose differed significantly from dupilumab; but the word the original uses is "most", not "all", and it should not be read as "the 100 mg dose was not superior to dupilumab on any measure in that trial" — what happened on the minority of measures not counted in "most" is something that trial's conclusion does not say.

Upadacitinib's two controlled trials (Measure Up 1 and 2, Guttman-Yassky et al., Lancet 2021;397:2151-2168, ages 12–75): in Measure Up 1, EASI-75 at week 16 was 196/281 (70%) on 15 mg, 227/285 (80%) on 30 mg, and 46/281 (16%) on placebo; in Measure Up 2, 166/276 (60%), 206/282 (73%) and 37/278 (13%) on placebo.

⚠️ Those percentages cannot be compared directly with the dupilumab percentages above. Measure Up is a monotherapy trial, all four groups of JADE COMPARE added topical treatment, and CHRONOS is dupilumab plus a topical corticosteroid — the control groups of the three designs run from 13% to 23% by themselves. Cross-trial comparison does not hold, and this article has not attempted one.

⚠️ If you are taking an efficacy figure to a doctor to discuss: ask three things — what was the control group? what was the denominator? was a topical drug used at the same time? If any of the three cannot be answered, it is not clear what that percentage is about.

#Does the JAK inhibitors' black box warning concern eczema patients at all? Three regulators, three answers

The question has an answer, but not one answer: all three regulators impose a population-level restriction, differing only in the axis they draw the line on — the FDA in the United States draws it in the indication itself (refractory disease, not controlled by other systemic drugs including biologics, aged 12 or above), with two risk-factor directions on smoking and thrombosis in the warnings section; while the EMA in Europe and the MHRA in the United Kingdom additionally write the risk factors into an access condition (over 65, cardiovascular or malignancy risk, long-term smoking, to be used only where no suitable alternative is available). And Hong Kong has no answer.

#The warning itself

The United States labels of the two oral JAK inhibitors approved for atopic dermatitis — CIBINQO (abrocitinib) and RINVOQ (upadacitinib) — carry the same black box warning heading: 「WARNING: SERIOUS INFECTIONS, MORTALITY, MALIGNANCY, MAJOR ADVERSE CARDIOVASCULAR EVENTS (MACE), and THROMBOSIS」.

The five items of CIBINQO's black box warning are reproduced in full at [Note 21]; in summary: an increased risk of serious bacterial, fungal, viral and opportunistic infections (tuberculosis included) leading to hospitalisation or death, with the drug to be discontinued if a serious or opportunistic infection develops; in patients with rheumatoid arthritis, a higher rate of all-cause mortality (including sudden cardiovascular death) with another JAK inhibitor than with a TNF blocker, and CIBINQO is not approved for rheumatoid arthritis; and malignancies, MACE and thrombosis having occurred with CIBINQO, while another JAK inhibitor had higher rates of lymphoma and lung cancer in patients with rheumatoid arthritis.

Note that the first bullet is more than a statement of risk — it carries three requirements: test for latent tuberculosis before and during treatment, treat latent tuberculosis first, and continue monitoring during treatment even where the initial test was negative. Those are written inside the black box warning itself, not as detail in some other section.

#The passage beside the black box warning: the indication is itself a population restriction

This is the passage most easily skipped when reading a black box warning and the most decisive on whether it applies to you: the United States indication for the two drugs is not "any moderate-to-severe eczema" but states refractory disease, patients not controlled by or unsuitable for other systemic drugs (biologics expressly included), and aged 12 or above.

The originals of the CIBINQO and RINVOQ indications, with the limitation of use that follows, are at [Note 22]: approved for adult and paediatric patients 12 years of age and older with refractory, moderate-to-severe atopic dermatitis whose disease is not adequately controlled with other systemic drug products, including biologics, or when use of those therapies is inadvisable; limitations of use: not recommended for use in combination with other JAK inhibitors, biologic immunomodulators or other immunosuppressants.

In one sentence: "refractory", "not adequately controlled with other systemic drug products, including biologics" and "12 years of age and older" are three conditions written into the indication itself — which is to say that under the framework of the United States label, an oral JAK inhibitor is from the outset neither first line nor a parallel option to dupilumab, but comes after other systemic treatment. That is a population-level restriction, and the line the EMA and MHRA draw by age and risk factor is a different axis, not a matter of having a restriction against having none.

CIBINQO carries a further contraindication in section 4 of the label, immediately after the black box warning [Note 22]: CIBINQO is contraindicated during the first 3 months of treatment in patients taking antiplatelet therapy, except for low-dose aspirin (≤81 mg daily). RINVOQ's section 4 contraindication is a different matter: known hypersensitivity to upadacitinib or any of its excipients. The two drugs' contraindications differ, and neither can be inferred from the other.

#The United States label itself has two risk-factor sentences

Above, the United States restriction was said to be written into the indication itself. To be accurate, one more thing has to be added: sections 5.4 and 5.5 of CIBINQO's warnings write down two risk-factor directions of their own, only not in the form of "use only where no alternative is available" [Note 23]: patients who are current or past smokers are at additional increased risk (5.4); and CIBINQO should be avoided in patients who may be at increased risk of thrombosis (5.5).

So the difference between the three regulators, put accurately, is this: Europe and the United Kingdom write the risk factors into an access condition (where risk factors are present, use only if no suitable alternative is available); the United States writes refractoriness, prior line of treatment and age into the indication, and additionally names thrombotic risk and smoking in the warnings section by way of "avoid" and "additional increased risk". It is not a matter of having against not having, and it is not that the United States never mentions risk factors.

⚠️ All of the above is the United States label's own text, and is not Hong Kong's access rule. The Hong Kong registered drug database has no "indication" field (see below), so this article has no way of saying how those conditions apply in Hong Kong.

There is a structural oddity here worth pointing out. The sentence 「CIBINQO is not approved for use in RA」 appears six times in the label — once in the black box warning, once in the warnings and precautions of the Highlights of Prescribing Information at the head of the label, and once each in sections 5.2, 5.3, 5.4 and 5.5. That is: the comparative data in the black box warning come from a patient population the drug itself may not be prescribed to. The label also records that malignancies (including non-melanoma skin cancer), MACE, deep vein thrombosis and pulmonary embolism have also occurred in CIBINQO's own atopic dermatitis trials — but the warning text gives no comparative incidence for those AD trial events.

RINVOQ's label carries the same black box warning but not the "not approved for use in RA" sentence, because upadacitinib is itself approved for rheumatoid arthritis.

#Both labels set out tests before and during prescribing — a layer outside the black box warning, but belonging with it

The black box warning states the risks; it is the warnings section immediately following that says what to test for, when, and when to stop. Read the black box warning alone and these drugs look like something you simply take. The originals of what each label sets down are at [Note 24]; in summary.

CIBINQO: use is associated with an increased incidence of thrombocytopenia and lymphopenia — a full blood count should be performed before starting, and is recommended again 4 weeks after starting and 4 weeks after a dose increase; users have had dose-dependent increases in lipid parameters, lipids should be assessed approximately 4 weeks after starting treatment, and the effect of those lipid elevations on cardiovascular morbidity and mortality has not been determined; and before starting CIBINQO, all age-appropriate vaccinations should be completed according to current immunisation guidelines, including prophylactic herpes zoster vaccination.

RINVOQ (the laboratory thresholds of section 5.9, with three directions from sections 5.3, 5.4 and 5.7): use is associated with an increased incidence of neutropenia (absolute neutrophil count below 1000 per cubic millimetre), and treatment should not be started, and should be interrupted, in patients with a low count; non-melanoma skin cancer has occurred in users, and periodic skin examination is recommended for patients at increased risk of skin cancer, with reduced sun and ultraviolet exposure through protective clothing and a broad-spectrum sunscreen (5.3); patients who have had a myocardial infarction or stroke should discontinue (5.4); and gastrointestinal perforations have been reported in clinical trials, with monitoring for patients at risk (a history of diverticulitis, or concomitant non-steroidal anti-inflammatory drugs or corticosteroids, for instance) (5.7). The same section also records lymphopenia (ALC below 500 per cubic millimetre), anaemia (haemoglobin below 8 g/dL) and lipid elevations.

⚠️ The above is what each of the two labels sets down, not a complete monitoring list and not a testing schedule to follow. What to test, when, how to read a result, and when to stop are all matters of prescription and clinical judgment. The two drugs' provisions also differ, and neither can be inferred from the other.

#Where the data come from: ORAL Surveillance

The source of the warning is Ytterberg et al.'s ORAL Surveillance trial (N Engl J Med 2022;386:316-326). It is not an eczema trial [Note 25]: a randomised, open-label, non-inferiority, post-marketing safety endpoint trial in patients with active rheumatoid arthritis despite methotrexate, aged 50 or over and with at least one additional cardiovascular risk factor.

The results (both arms with denominators): tofacitinib 5 mg twice daily 1,455 patients, 10 mg twice daily 1,456, and a TNF inhibitor 1,451; median follow-up 4.0 years. MACE in the combined tofacitinib groups was 3.4% (98 patients) against 2.5% (37) on TNF; cancer 4.2% (122) against 2.9% (42). Hazard ratios: 1.33 for MACE (95% CI 0.91 to 1.94) and 1.48 for cancer (95% CI 1.04 to 2.09); the non-inferiority of tofacitinib was not established [Note 25].

This passage has to be read precisely, because it is easily misread in two ways. First, the 95% confidence interval for MACE, 0.91 to 1.94, crosses 1 — on MACE, what the trial shows is a failure to demonstrate that tofacitinib is not worse than a TNF inhibitor, not a demonstration that MACE is increased. The cancer interval of 1.04 to 2.09 does not cross 1. Second, "non-inferiority not established" and "harm demonstrated" are two different things; and at the same time "harm not demonstrated" is not safety. The trial describes its own population as one enriched for cardiovascular risk.

#Three regulators, three responses

The European Medicines Agency (EMA) — an Article 20 review, endorsed by the Committee for Medicinal Products for Human Use (CHMP) on 23 January 2023. It answers directly whether dermatology is concerned [Note 26]: the EMA concluded that these safety findings apply to all the approved uses of JAK inhibitors in chronic inflammatory disorders, including rheumatoid arthritis, psoriatic arthritis, juvenile idiopathic arthritis, axial spondyloarthritis, ulcerative colitis, atopic dermatitis and alopecia areata.

The restriction itself (reproduced in full at [Note 26]): these medicines should be used in the following patients only if no suitable treatment alternative is available — those aged 65 or above, those at increased risk of major cardiovascular problems, those who are current or long-term past smokers, and those at increased risk of cancer.

The United Kingdom's MHRA — its update of 26 April 2023, whose own therapeutic area label reads 「Therapeutic area: Dermatology, GI, hepatology and pancreatic disorders and Rheumatology」, putting dermatology first. Its explanation to patients names eczema in terms — but that sentence is the first bullet of a list, and the second, third and fourth bullets that follow are its qualifications, and the four have to be read together. All five bullets are at [Note 27]; the direction is: JAK inhibitors are effective medicines for chronic inflammatory disease, including atopic eczema; but compared with TNF-α inhibitors, these medicines can increase the risk of major cardiovascular problems, cancer, blood clots, serious infection and death; where risk factors are present — being 65 or above, already at increased risk of cardiovascular problems or cancer, being a current or long-term past smoker — the risk is higher still; where risk factors are present, the doctor will consider other treatment options and will prescribe a JAK inhibitor only if no other suitable option is available; and the doctor may also recommend a lower dose.

Note the fourth bullet. In language a patient can follow, it says the same thing as the EMA's restriction.

But there is a sentence in the same update that is almost never quoted [Note 27]: the review concluded that these effects could be considered a class effect, while acknowledging that the extent to which the findings of study A3921133 apply to all JAK inhibitors depends on how similar the treated populations are in their risk factors.

Taken together with the above: put that sentence beside the ORAL Surveillance population described above (aged 50 or over, rheumatoid arthritis, at least one cardiovascular risk factor) and a very concrete conclusion follows: an eczema patient of 60 with a cardiovascular history and an eczema patient of 25 with no cardiovascular risk factors are not the same case under this class effect — and that difference is exactly what the EMA and MHRA restrictions use to draw their line. This is not this article's inference: the EMA's restriction and the MHRA's qualification are both written by risk factor.

⚠️ If seeing "the JAK inhibitors carry a black box warning" makes you want to rule the option out at once, or if "Europe still approves them" and "the United States imposes no restriction" make you dismiss it: none of those three formulations is what the documents say. What the documents say is that the warning is indeed attached to eczema drugs; that all three regulators impose a population-level restriction, differing only in the axis on which they draw the line; that the comparative data come from a different patient population; and that how far it can be carried across depends, officialdom itself says, on how similar the risk factors are. Note also that both labels set out testing requirements before and during prescribing — latent tuberculosis, full blood count, lipids, vaccination — a layer that belongs with the black box warning and cannot be left unread.

#Does Hong Kong have an eczema guideline of its own? — one, but written before the whole era of oral JAK inhibitors

Hong Kong's only professional body clinical guideline on atopic dermatitis comes from the Hong Kong College of Paediatricians (the guideline of 2013 and the review of 2020). There is no Hong Kong dermatological body guideline, and no Hong Kong professional body has published a position on dupilumab or on JAK inhibitors.

Of the four bodies searched here, the results were: the Hong Kong Society of Dermatology and Venereology's website carries only a directory and a secretariat address; the Hong Kong College of Physicians' position statements are on cardio-renal-metabolic disease, nephrology and palliative care; the Hong Kong Academy of Medicine's site search for "atopic dermatitis" returns no results row; and the Hong Kong Journal of Dermatology and Venereology's site search returns 101 results for "atopic dermatitis", 26 for "dupilumab" and 21 for "JAK", but the leading ones are all editorials or conference reports, not guidelines.

The sentence in the Hong Kong College of Paediatricians' review of 2020 that mentions JAK inhibitors has to be quoted with its context, because it sits in a passage about new topical drugs [Note 28]: other novel topical anti-inflammatory agents still under investigation, including lipoxins and JAK inhibitors (tofacitinib, for instance), require further study to demonstrate efficacy and safety in the paediatric population.

And the severity ladder in the same document (the SCORAD >50 tier) reads [Note 28]: hospitalisation; potent to very potent topical corticosteroids; systemic immunosuppressants (ciclosporin A, azathioprine, methotrexate, for instance); and dupilumab where the age is appropriate. Oral JAK inhibitors are not on that ladder at all.

Put those two things together and the chronology matters: the review behind Hong Kong's only guideline was published in 2021; the EMA's and the MHRA's JAK restrictions came in January and April 2023. So a Hong Kong reader asking how Hong Kong recommends using JAK inhibitors for eczema has no Hong Kong answer — not because this article could not find one, but because after going through each Hong Kong professional body's website, no document addresses the question. That is a bounded conclusion: it is about the absence of a published body position, not about Hong Kong doctors not using these drugs (they are in fact on the Hospital Authority Drug Formulary; see the next section).

By way of contrast, England's access rules are written with great precision. NICE TA534 (issued 1 August 2018) states [Note 29]: dupilumab is recommended for moderate-to-severe atopic dermatitis in adults only where the disease has not responded to at least one other systemic therapy (ciclosporin, methotrexate, azathioprine or mycophenolate mofetil, for instance), or where those are contraindicated or not tolerated; and it is stopped at 16 weeks if the response is inadequate, an adequate response being defined as at least a 50% improvement in EASI and at least a 4-point improvement in the Dermatology Life Quality Index.

NICE TA814 (3 August 2022) sets rules of the same structure for abrocitinib and upadacitinib (at least one systemic immunosuppressant having failed or being unsuitable, with the 16-week stopping rule), and states its own evidence limitation: abrocitinib and upadacitinib were compared with ciclosporin only indirectly, and the results are highly uncertain [Note 29].

Hong Kong has published no corresponding access rule. The eligibility page of the Hospital Authority's Samaritan Fund says only 「藥物項目：根據醫管局現行的臨床指引，有關病人的治療計劃必須經由一名指定的醫管局醫生簽發」 — that for drug items, the patient's treatment plan must be issued by a designated Hospital Authority doctor in accordance with the Hospital Authority's prevailing clinical guidelines — and those prevailing clinical guidelines are internal documents that have not been published. The same page also provides 「獲基金資助的病人須為醫院管理局（醫管局）病人」 — that a patient assisted by the fund must be a Hospital Authority patient. What that means for Department of Health patients has a section of its own below.

⚠️ If you want to know whether you qualify for a subsidised drug: England has a published threshold (one systemic immunosuppressant tried first, and stop at 16 weeks if EASI 50 is not reached); Hong Kong has a fund, a drug list and an indication, but no published clinical criteria, so in Hong Kong that question can only be answered in the consulting room.

#How long do you wait to see public dermatology in Hong Kong? — the answer is not in the department you would expect

This is the section most easily got wrong: public dermatology outpatient clinics in Hong Kong are a Department of Health service, not a Hospital Authority one. So the Hospital Authority specialist outpatient waiting time table that everybody cites has no dermatology in it at all.

The Department of Health's own page (the page's own revision date: 「二零一九年十月一日」) says: the Social Hygiene Service comprises social hygiene clinics and dermatology clinics, responsible for the treatment, care, prevention and control of sexually transmitted infections and skin diseases.

And the complete list of specialties covered by the Hospital Authority's open data on specialist outpatient waiting times (covering 1 July 2025 to 30 June 2026, 288 rows) is: ear, nose and throat; ophthalmology; gynaecology; medicine; orthopaedics and traumatology; paediatrics; psychiatry; and surgery — eight, with no dermatology.

The real figures are in a PDF of the Department of Health's own: "Department of Health Social Hygiene Service — New Skin Case Appointment Status", the document's own update date being 「更新日期截至 2026 年 6 月 30 日」:

⚠️ The most important thing about that table is not in it: the Department of Health does triage, and the table above is about stable cases. The Chinese original of the government's reply at the Legislative Council Finance Committee special meeting of 2024 (reference HHB094) reads [Note 30]: according to the records kept, in 2022 and 2023 all new cases of severe skin disease were seen within 8 weeks in accordance with the service pledge, with median waiting times of 2.7 and 2.9 weeks respectively; while for other stable new cases the median waiting times were 90 and 93 weeks, and the overall longest waiting times 203 and 183 weeks.

So: if your case is severe, do not give up on the public route because you have seen "running to 2029" — that figure is not about you. Which cases count as severe is triaged by the Department of Health from the referral letter and a clinical assessment; it is not for a patient to decide, and not something this article can judge for you.

How the table is to be read: what the Department of Health publishes is the appointment date currently being scheduled, not how many weeks you will wait. As at the document's own update date of 30 June 2026, new case appointment dates fell between December 2027 and February 2029. This article will not convert that into weeks, because the source publishes dates rather than durations, and converting would introduce an assumption the document never made.

The entry conditions are worth noting too. The Department of Health's clinic page (the page's own revision date: 「二零二五年六月二日」) reads: all new cases require a doctor's letter of introduction; all new and old cases require an appointment; for a new case appointment, bring identification and a referral letter issued by a Hong Kong registered western medicine practitioner and attend the clinic in person, or send them by post to the designated clinic, with appointment and rescheduling applications also accepted through GovHK [Note 31].

The words 「香港註册西醫」 — a Hong Kong registered western medicine practitioner — are fixed on that page; the page does not say that a Chinese medicine referral is accepted.

⚠️ If you are planning to queue for public dermatology: first, the table above gives appointment months for stable cases; a new case of severe skin disease has an 8-week service pledge, with medians of 2.7 to 2.9 weeks — seeing 2029 and not joining the queue may mean misjudging which category you are in. Second, do not go looking at the Hospital Authority's waiting time table; there is no dermatology in it. Third, a registered western medicine practitioner's referral letter is essential. Fourth, the appointment months of the nine clinics differ by more than a year (Wan Chai Male at December 2027 against Cheung Sha Wan at February 2029), and the Department of Health lets you choose which clinic to make the appointment at.

#What do these drugs cost in Hong Kong? — the same patient, two completely different worlds of charging

Under the Hospital Authority's fee structure, a tube of a formulary steroid cream is 20 dollars per item and a dose of dupilumab is entirely self-financed — and the annual fee cap of ten thousand dollars introduced from January 2026 states in terms that it excludes self-financed drugs. ⚠️ But note: eczema patients in the public system are usually followed up at Department of Health dermatology clinics, while the drug formulary, the Samaritan Fund and the ten-thousand-dollar cap are all Hospital Authority mechanisms. How the two join up — the Samaritan Fund takes Hospital Authority patients only in terms, and a Department of Health patient has to be referred into the Hospital Authority to reach it — has a section of its own below.

#The classification first: which cell do these drugs sit in on the Hospital Authority Drug Formulary?

Version 22.1 of the Hospital Authority Drug Formulary (the page's own version statement: 「Version 22.1 – with effect from 25 July 2026」) has four cells: general drugs, special drugs, self-financed items with safety net, and self-financed items without safety net. The formulary's own definition of the third cell [Note 32]: self-financed items with safety net — very expensive drugs with proven significant benefits but beyond the affordability of the Hospital Authority's general subsidised services; they are not items provided at the standard charges of public hospitals and clinics, and patients who need them and can afford them purchase them at their own cost; while the relevant funds provide a safety net for patients in financial difficulty who need them.

⚠️ The formulary's "special drugs" is a Hospital Authority funding classification, not a clinical positioning. The Hospital Authority has not published the specified clinical conditions for tacrolimus and pimecrolimus (internal documents, like the Samaritan Fund's clinical criteria). By way of contrast, NICE CG57 is quite plain about where it positions those two drugs [Note 33]: in accordance with NICE technology appraisal guidance, topical tacrolimus and pimecrolimus are not recommended for the treatment of mild atopic eczema or as first-line treatments for atopic eczema of any severity (1.5.1.21).

And the two provisions that immediately follow in the same guideline say where it does recommend them and who must start them — those two and the one above have to be read together [Note 33]: topical tacrolimus is recommended as a second-line option for moderate-to-severe atopic eczema in children aged 2 and above where it has not been controlled by topical corticosteroids and there is a serious risk of important adverse effects from further corticosteroid use, particularly irreversible skin atrophy (1.5.1.22); and treatment with tacrolimus or pimecrolimus should be started only on the advice of a dermatology specialist, after a careful discussion with the child and their parents or carers about the potential risks and benefits of all appropriate second-line treatment options (1.5.1.24).

So under NICE's framework these two drugs are not "switch to this if you are afraid of steroids" — they are second line, with an age floor of 2 and above, and expressly begun on specialist dermatological advice. That is England's positioning, not Hong Kong's funding condition — neither can be inferred from the other.

Three things read straight off that table: dupilumab, abrocitinib, upadacitinib and lebrikizumab are all self-financed items with safety net; the topical calcineurin inhibitors tacrolimus and pimecrolimus are special drugs (that is, subsidised, but limited to specified clinical conditions); and baricitinib is not in this section at all (it appears under another, musculoskeletal, category of the formulary).

One practical point for Chinese readers in passing: open the Chinese path to the Hospital Authority Drug Formulary and the page title itself reads 「醫院管理局藥物名冊 [只備英文版]」 — English version only. Which is to say that the formulary determining what you pay is one a Hong Kong patient who reads only Chinese cannot read.

#Which fund is the safety net? — the Samaritan Fund, not the Community Care Fund

The Samaritan Fund's list of subsidised drugs (the document's own effective date: 「(於2026年7月25日起生效)」) is a complete, itemised, numbered list, and the Chinese originals of the four atopic dermatitis rows are at [Note 34]: 5 阿布昔替尼 (Abrocitinib), dermatology, atopic dermatitis; 42a 度匹蘆人單抗 (Dupilumab), dermatology/paediatrics, atopic dermatitis; 111d 烏帕替尼 (Upadacitinib), dermatology/paediatrics, atopic dermatitis; and 65 來瑞奇珠單抗 (Lebrikizumab), dermatology, atopic dermatitis (effective from 25 July 2026).

Baricitinib is on the same list too, but its row reads 「21 Baricitinib Rheumatology Rheumatoid arthritis」 — rheumatology and rheumatoid arthritis, not eczema.

As for the Community Care Fund: its first-phase programme is, on the fund's own page, 「(特定自費癌症藥物)」 — specified self-financed cancer drugs; and the Samaritan Fund / Community Care Fund comparison table inside the fund's programme document ticks the Samaritan Fund column on the "Dermatology – Atopic dermatitis" row, leaving the Community Care Fund column blank. So the safety net for the new eczema drugs is the Samaritan Fund.

The Samaritan Fund is subject to a financial assessment. The Chinese version of the fund's financial assessment page states [Note 35]: under the public healthcare fee reform, the eligibility criteria of the Samaritan Fund were relaxed from 1 January 2026; and the annual disposable financial resources are calculated as 80% of the household's annual disposable income plus 50% of the household's net disposable assets (that is, disposable assets less deductible disregards).

And the sentence immediately following is the one that matters most to a patient with a chronic condition, and the one most often left out [Note 35]: where the patient is a continued applicant (that is, has applied for and been granted drug subsidy under the Samaritan Fund or the Community Care Fund medical assistance programme within the past eighteen months), only 60% of the household's annual net disposable income is counted.

In one sentence: a first application counts 80% of income, but if you have applied for and been granted Samaritan Fund or Community Care Fund drug subsidy within the past eighteen months, only 60% is counted. Eczema is chronic and relapsing, and a patient needing the same self-financed drug over the long term is precisely a "continued applicant" — so reading only the 80% sentence overstates the income to be counted for a renewal applicant. The whole assessment is carried out by a medical social worker on a household basis.

⚠️ The Chinese and English versions word the assets item differently, and the Chinese version governs. The English writes "50% of patients' household net assets"; the Chinese writes 50% of 「家庭的可動用資產淨值（即可動用資產扣除可扣減豁免額）」 — the Chinese states in terms how the net figure is arrived at, and the English does not.

#Doing the arithmetic once (an example): a public patient's drug costs for a year, and which cell they stop in

Using the Department of Health's own fee table (the page's own date: 「二零二六年一月一日」). An eligible person attending a dermatology clinic: 250 dollars per attendance; drugs prescribed by that clinic 20 dollars per item, for up to 4 weeks.

• On 4 follow-ups in a year: 250 dollars × 4 = 1,000 dollars
• A year is 52 weeks, that is 13 units of "4 weeks"; for one topical corticosteroid: 20 dollars × 13 = 260 dollars (note: the fee table says 20 dollars per item prescribed at each attendance, for up to 4 weeks, so to reach 13 units one has to assume a corresponding number of dispensing arrangements over the year — an assumption that is not in the fee table)
• The year's total: 1,000 + 260 = 1,260 dollars

(For a non-eligible person the same fee table says 850 dollars per attendance and 90 dollars per drug item; the same arithmetic gives 850×4 + 90×13 = 4,570 dollars.)

⚠️ That calculation covers only attendance and drugs prescribed by that clinic. The same fee table states for itself that the fee at item (a) does not include the related dispensing, radiological imaging services and laboratory charges, and that the fee at item (e) does not include self-financed drugs — which is to say that imaging and laboratory charges are not inside the 250 dollars, and self-financed drugs are not inside the 20 dollars; both are charged separately, and the fee table publishes no amounts for them.

Comparing that figure with the Hospital Authority's annual fee cap of ten thousand dollars from 1 January 2026 is beside the point in this example — but the real point is in the next sentence. The Hospital Authority's page announcing that cap states for itself [Note 36]: from 1 January 2026, an annual expenditure cap of ten thousand Hong Kong dollars applies to designated public healthcare charges for eligible patients, with no financial assessment … charges for self-financed drugs and medical devices are not included.

⚠️ "No financial assessment" does not mean it takes effect automatically — the same page's own eligibility and application arrangements state that it must be applied for, must be applied for afresh each year, and has a deadline. The Chinese version of that page has four eligibility items, reproduced in full at [Note 36]: the patient must be an eligible person; the patient's cumulative payment of "eligible medical fees and charges" within the year must reach ten thousand Hong Kong dollars; there must be no outstanding amount owed anywhere within the Hospital Authority at the time of application; and the hospital services received must not have been determined by the Hospital Authority to be without clinical need. The application arrangements: the annual application period opens on 1 January each year and closes on 31 March of the following year, late submissions are not accepted, and eligible medical fees cover only bills issued between 1 January and 31 December and paid in full at the time of submission; and a fresh application is needed for each year. The English version additionally sets out the channel: an eligible patient whose cumulative valid annual expenditure has reached ten thousand dollars may apply through HA Go or at the shroff office of any hospital [Note 36] — that is, the patient applies themselves, without a medical social worker, unlike the Samaritan Fund's arrangement of a financial assessment by a medical social worker.

Which is to say: the calculation above uses the Department of Health's attendance and drug fee table. The formulary's classification (general, special, self-financed), the $130 self-financed drug administration fee and the ten-thousand-dollar annual cap are all Hospital Authority arrangements — a patient attending the Hospital Authority who moves to dupilumab or an oral JAK inhibitor pays for it entirely, and the ten-thousand-dollar cap does not help. Where that is genuinely unaffordable there is a route: the Samaritan Fund safety net (see the section above) exists for exactly that situation — but note that it is subject to a financial assessment, that not everyone is eligible, and that it is not reached automatically. ⚠️ Department of Health patients do not follow the Hospital Authority's classification — the Department of Health has a drug formulary of its own (see the next section), so the Hospital Authority's four cells, the $130 administration fee and the ten-thousand-dollar cap apply only after you have been referred into the Hospital Authority.

#The bridge between the two organisations — the answer is: you cannot apply directly, but there is a referral route

The two sections above each hold, but they belong to two different organisations: the drug formulary and the Samaritan Fund are Hospital Authority mechanisms; the dermatology outpatient clinic an eczema patient attends in the public system is a Department of Health service; and the Department of Health's own fee table, after listing 「每項目20元」, immediately adds 「上述(e)項的收費不包括自費藥物」 — that the fee at item (e) does not include self-financed drugs.

Part 1: you cannot apply as a Department of Health patient. The first sentence of the Samaritan Fund's "eligibility" page reads: a patient assisted by the fund must be a Hospital Authority patient and meet the conditions below [Note 37]. And the drug item condition reads: in accordance with the Hospital Authority's prevailing clinical guidelines, the patient's treatment plan must be issued by a designated Hospital Authority doctor.

Which is to say that the eligibility condition is tied at once to the patient's status (a Hospital Authority patient) and to the issuing doctor's status (a designated Hospital Authority doctor). Attending only a Department of Health dermatology clinic satisfies neither.

Part 2: but there is a route from the Department of Health into the Hospital Authority, and the government has spoken about it in the Legislative Council. The key fact is that the Department of Health has a drug formulary of its own, a different document from the Hospital Authority's. At the Legislative Council meeting of 28 November 2018, the Secretary for Food and Health's reply is quoted at [Note 38]; its direction is: under the existing mechanism, where a patient needs a drug that is not in the Department of Health's formulary, the patient can be referred to the Hospital Authority, the essential thing being that the doctor considers on clinical assessment that the patient has such a need; and where a doctor considers on clinical assessment that a drug outside the Department of Health's formulary is needed, it has to be obtained from the Hospital Authority, and a mechanism already exists to refer patients to Hospital Authority hospitals for treatment or for those drugs. And in reply to another member, the Secretary spoke specifically about eczema (the only biologic then registered in Hong Kong for severe atopic eczema being dupilumab, registered on 30 October 2018): the Social Hygiene Service under the Department of Health would keep abreast of the latest clinical and research evidence on that biologic, and would in due course collaborate with the Hospital Authority through the existing mechanism to refer such patients to the Hospital Authority.

Part 3: what that route rests on has to be stated plainly. The government does have a formal document setting out the referral mechanism between the Department of Health and the Hospital Authority — "The Department of Health's Collaboration and Referral Mechanism with the Hospital Authority on Dermatological Services", submitted to the Legislative Council Panel on Health Services on 17 December 2018. But that document states for itself that it covers psoriasis only: the collaboration and referral mechanism currently applies to psoriasis patients [Note 38].

So to state it accurately: referral for eczema rests on (i) the general rule that a drug outside the Department of Health's formulary can lead to a referral to the Hospital Authority, and (ii) the Secretary's oral undertaking in the Legislative Council in 2018 about atopic eczema; and not on a published referral guideline naming atopic dermatitis. Whether there is a waiting time, whether there are quotas, and what the clinical threshold is — no published document says. Nor has the government been seen to restate that eczema referral route since 2021.

What that means for a reader in practice: do not assume that attending a Department of Health clinic connects you automatically to the Samaritan Fund; institutionally it does not. The route is to raise it with your Department of Health attending doctor, have them determine on clinical assessment that you need a drug outside the Department of Health's formulary, and have them refer you into the Hospital Authority through the existing mechanism; only then does a designated Hospital Authority doctor issue the treatment plan and a medical social worker carry out the financial assessment. The route exists, but it does not walk itself, and it has no published timetable.

#Registration ≠ having the indication ≠ being subsidised

The Department of Health Drug Office's Hong Kong registered drug database (the page's own marking: 「Last Updated: 31-Jul-2026」) shows that dupilumab (4 products, the earliest HK-65961 registered 30 October 2018), abrocitinib (3, 15 December 2022), upadacitinib (3, 30 October 2020), baricitinib (2, 26 March 2018) and lebrikizumab (1, 24 March 2025) are all registered in Hong Kong; and that tralokinumab (Adtralza) is not registered in Hong Kong — searched once by ingredient and once by trade name, both returning 0 items.

But that database has no "indication" field, and publishes no product information. So this article can say only that these products are registered in Hong Kong, and cannot say that they are registered in Hong Kong for eczema. The only Hong Kong official document that puts these drugs and "atopic dermatitis" in the same place is the Samaritan Fund's subsidy list — and that is a subsidy indication, not a registration indication. The two cannot be run together.

#Private prices: only news reports, no official or institutional price list

Among the material cited here there is no price list published by any Hong Kong official body, hospital or provider. The only Hong Kong figure found is a news report. A report in the Hong Kong Economic Journal's health column of 30 July 2025 states: one injection of an eczema biologic costs about $7,500 to $10,000; and at one injection every two weeks, the monthly drug cost is about $15,000 to $20,000 [Note 39].

That needs three qualifications: first, it is a news report, not a price list; second, it says "eczema biologic" without naming a drug; and third, for the Hong Kong prices of abrocitinib and upadacitinib, the material cited here does not even have a news report.

⚠️ If you are thinking that a public subsidy makes it all right: the divide is not only public against private but which organisation and which formulary cell — within the Hospital Authority, a formulary cream is 20 dollars and a self-financed biologic injection is paid for in full plus an administration fee, and the annual fee cap does not cover the latter (and that cap itself has to be applied for through HA Go or a hospital shroff office, afresh each year, closing on 31 March of the following year); and a Department of Health dermatology patient is not connected to that arrangement automatically — the Samaritan Fund takes Hospital Authority patients only in terms, and it is reached through a referral into the Hospital Authority by your attending doctor.

#What to do next

This section is about process and documents only, and does not touch decisions about drugs.

1. To get into public dermatology: you need a referral letter from a Hong Kong registered western medicine practitioner; bring identification and the referral letter and make an appointment in person or by post at one of the nine Department of Health clinics, or submit an appointment or rescheduling application through GovHK. The appointment months currently being scheduled at the nine clinics differ by more than a year, and the Department of Health's document lets you choose the clinic yourself. And if your case is severe, remember that the table is about stable cases — a new case of severe skin disease has an 8-week service pledge.

2. To find out which cell your drug is in: version 22.1 of the Hospital Authority Drug Formulary (effective 25 July 2026) has four cells, and if you attend the Hospital Authority that classification determines whether you pay a standard charge or the whole cost. Note that the formulary exists in English only; and note that if you are followed up at a Department of Health dermatology clinic, what applies to you is the Department of Health's own formulary, the Hospital Authority's four cells applying only once you have been referred in.

3. If a self-financed drug is unaffordable: the Samaritan Fund is a Hospital Authority mechanism and takes Hospital Authority patients only in terms. If you are currently followed up at a Department of Health dermatology clinic, the route is to raise it with your attending doctor and have them refer you into the Hospital Authority through the existing mechanism — the government has confirmed in the Legislative Council that a drug outside the Department of Health's formulary can lead to such a referral. The financial assessment is carried out by a medical social worker on a household basis; and the subsidy list (effective 25 July 2026) names abrocitinib, dupilumab, upadacitinib and lebrikizumab against "atopic dermatitis". The clinical criteria are Hospital Authority internal documents and have not been published, so eligibility can only be confirmed in the course of a consultation. Remember also the 60% rule: a renewal applicant who has applied and been granted within the past eighteen months has only 60% of disposable income counted, not 80%.

4. If your skin worsens while using a topical corticosteroid: the MHRA's update of September 2021 includes advice to prescribers to tell patients to seek medical advice if their skin worsens during use. And the update of May 2024 states that where adrenal suppression is already present, stopping abruptly is dangerous.

5. If you are taking a JAK inhibitor: the MHRA's update of 26 April 2023 gives patients a list of eight bullets; the three most directly about when to seek help at once are at [Note 40]: during treatment, if you have chest pain or tightness (which may spread to the arms, jaw, neck and back), breathlessness, cold sweats, light-headedness, sudden dizziness, weakness in the arms or legs, or slurred speech, contact your doctor or seek emergency medical advice — these are signs of a medical emergency; check your skin regularly and tell your doctor or nurse if you notice any new growth or change in a mole (including itching, shape and weeping, which may be less obvious on darker skin); and always read the leaflet that comes with the medicine.

   ⚠️ That list is what the MHRA set out in that update, and is not a complete list. A symptom not listed does not mean it does not matter. And the United States labels set out a layer that the MHRA list does not mention: the testing requirements before and during prescribing (latent tuberculosis, full blood count, lipids, completing age-appropriate vaccination before starting), and RINVOQ's provisions on non-melanoma skin cancer, on discontinuing after a myocardial infarction or stroke, and on gastrointestinal perforation. Those are matters for the prescribing doctor to arrange.

6. If you cannot make sense of a package insert or a dermatology report: upload the document to GoodDoctor.hk document reading for a plain-language explanation.

7. If you need a specialist referral: use the GoodDoctor.hk doctor matching tool. This article does not and cannot compare or recommend any doctor or provider.

#Frequently asked questions

How many people in Hong Kong have eczema?

It depends on the age group and on how the question is asked. Pupils aged 6–7 (2015–16, n=3,698): "eczema ever" 30.9%, "chronic rash in the typical sites" 8.1%. The HKU survey of 2021 (the first two years of primary school, 1,165 parent–child pairs): "ever" 41.6%, "past year" 16.3%. For adults there is no Hong Kong figure at all — both the Hong Kong College of Paediatricians' document and the HKU/PolyU study protocol state that local data are missing, and the Centre for Health Protection has published no eczema statistics.

How long is it safe to use a steroid cream?

The numbers of days each guideline writes for itself are these. NICE CG57 (under 12s) — the general rule for the face and neck is mild potency, with moderate potency only for short periods (3 to 5 days) in severe flares; moderate or potent for 7 to 14 days for flares in the axillae and groin; and where mild or moderate potency has not controlled it in 7 to 14 days, exclude secondary infection. The same guideline also states that children should not use very potent preparations without specialist dermatological advice; the Hong Kong College of Paediatricians likewise states that potent and very potent preparations should be used only on specialist advice. The Hong Kong College of Paediatricians' guideline of 2013 — face and neck 3 to 5 days, axillae and groin no more than 1 to 2 weeks, trunk and limbs potent for no more than 2 weeks. Potency, quantity and duration are matters of prescription, and this article only reports what each guideline says.

Are "one fingertip unit = 0.5 g covering two palms" and "diluting with an emollient means fewer side effects" accurate?

The original fingertip unit study (Long & Finlay 1991, 30 adults) measured area: one fingertip unit covered 286 square centimetres (standard deviation ±80), and weight is nowhere in the paper; the "0.5 g" and "two palms" formulations come from downstream guidelines, and the Hong Kong College of Paediatricians' own two documents word it differently (2013, two adult palms including fingers; 2021, two adult hands with fingers together), while giving intervals between applications of 30 minutes and 15 minutes respectively. As for dilution, the same college's review of 2020 writes: 「Dilution of TCS with emollients will not reduce its potency or side effects.」

My skin flared badly after stopping — is that steroid withdrawal?

The MHRA acknowledges that such reactions exist and lists four phases, and the same update of 2024 also states that there is still no agreed clinical definition, and that not all of the 267 reports up to August 2023 are true withdrawal reactions; while the sentence that it is difficult to distinguish a flare of the eczema itself from a withdrawal reaction is from its update of 2021. This article supplies no method of telling them apart; and the same document states that where adrenal suppression is already present, stopping abruptly is dangerous.

Does the JAK inhibitors' black box warning concern eczema patients?

The EMA's conclusion, endorsed by the CHMP on 23 January 2023, expressly covers all approved uses including atopic dermatitis, and provides that for those aged 65 or above, at increased cardiovascular risk, current or long-term past smokers, or at increased cancer risk, they are to be used only where no suitable treatment alternative is available. The MHRA followed at the same time, but added a sentence: how far the class effect holds depends on how similar the treated populations are in their risk factors.

Does the United States FDA only attach a warning, without imposing any restriction on use?

No. The United States indications for CIBINQO and RINVOQ are themselves a population restriction: limited to refractory moderate-to-severe atopic dermatitis, to patients not controlled by or unsuitable for other systemic drugs (biologics expressly included), and to those aged 12 or above, with a stated limitation against combination with other JAK inhibitors, biologic immunomodulators or immunosuppressants. CIBINQO has a further contraindication: patients taking antiplatelet therapy during the first 3 months of treatment, except for low-dose aspirin (≤81 mg daily). The difference between the FDA and the EMA/MHRA is not a matter of restriction against no restriction, but of the axis on which the line is drawn. Note also that the first bullet of the black box warning itself carries three requirements: test for latent tuberculosis before and during treatment, treat latent tuberculosis first, and continue monitoring for active tuberculosis during treatment. And the source trial, ORAL Surveillance, was in rheumatoid arthritis patients aged over 50 with a cardiovascular risk factor; its hazard ratio for MACE was 1.33 (95% CI 0.91–1.94, crossing 1) and for cancer 1.48 (1.04–2.09), and its own conclusion was that non-inferiority was not established — which is not the same thing as a demonstrated increase in MACE.

Does avoiding foods help?

The Department of Health Student Health Service's eczema page lists food proteins such as egg, seafood, dairy, beef and peanut among the possible causes, and the page's own 「可能」 and 「如」 make these examples rather than a complete list. As for prevention, the Hong Kong Institute of Allergy's local prevention guideline of 2016 recommends a healthy and unrestricted maternal diet during pregnancy and lactation, and states that preventive dietary restriction is unlikely to reduce the development of allergic disease. Note that that is a prevention guideline, not a dietary guideline for those who already have the disease; among the material cited here, no Hong Kong guideline addresses diet in those who already have it.

#Notes: the official originals

[Note 1] Department of Health Student Health Service, "Eczema" (the page's own revision date: 「二零二二年六月修訂」): 「濕 疹 的 可 能 成 因 包 括 ： 遺 傳／致 敏 源 如 灰 塵 、 羊 毛 、 花 粉 、 寵 物 的 毛 和 皮 屑／食 物 中 的 蛋 白 質 如 蛋 、 海 產 、 奶 類 、 牛 肉 、 花 生」 (the spaces between the characters are that page's own typesetting, reproduced as found).

[Note 2] Langan SM et al., Lancet 2020;396:345-360: 「The pathophysiology is complex and involves a strong genetic predisposition, epidermal dysfunction, and T-cell driven inflammation. Although type-2 mechanisms are dominant, there is increasing evidence that the disorder involves multiple immune pathways.」

[Note 3] Palmer CNA et al., Nature Genetics 2006: 「two independent loss-of-function genetic variants (R510X and 2282del4) in the gene encoding filaggrin (FLG) are very strong predisposing factors for atopic dermatitis. These variants are carried by approximately 9% of people of European origin.」 Chen H et al., British Journal of Dermatology 2011: 「the prevalent European FLG mutations are rare or absent in Chinese patients with IV and AD.」

[Note 4] Hong Kong College of Paediatricians review, HK J Paediatr 2021;26:42-57: 「AD is a chronic inflammatory condition with subclinical inflammation present in skin without visible lesions.」「It must be stressed that tapering of TCS should not be done too early as subclinical inflammation persists.」

[Note 5] Hong Kong Institute of Allergy, "Guidelines for allergy prevention in Hong Kong", Hong Kong Med J 2016;22:279-85: 「Mothers are advised to enjoy a healthy diet with no restrictions during pregnancy and lactation, as prophylactic dietary restriction is unlikely to reduce the development of atopic disease. Breastfeeding is recommended for the first 6 months of the newborn's life.」「Atopic eczema has been shown to be significantly reduced in high-risk infants fed a hydrolysed formula compared with standard cow's milk formula.」

[Note 6] University of Hong Kong press release of 13 January 2022: 「研究結果顯示，濕疹、哮喘、鼻炎和鼻敏感的發病率分別為41.6%、5.5%、59.4%和46.1%。」「在過去一年內，有16.3%的學童受到痕疹的困擾（濕疹）……」「相比2004年香港的ISAAC數據，曾患濕疹，鼻炎以及鼻敏感的學童有上升的趨勢。」 The English release: 「the prevalence rates for eczema … are 41.6%」; and the sampling note: 「Due to low participation rates, a third round of recruitment was conducted. Nearly 900 invitation letters were sent to schools. Schools were also recruited from personal network.」

[Note 7] Chan HH et al., Clin Exp Allergy 2001;31:903-7: 「The low scores were likely to be due to a reduction in the sensitivity of the Chinese questionnaire, which ranged from 23.5% to 70.6%.」「The indication of a low prevalence of atopic eczema among the Chinese population reported in previous studies was at least partially due to problems with the translated questionnaire.」

[Note 8] Hong Kong College of Paediatricians review of 2020: 「There has been a lack of more recent local prevalence data for more than 10 years.」 ClinicalTrials.gov NCT04154839 (record last updated 2022-05-05): 「the lack of complete epidemiology data on childhood and adult AD in Hong Kong」 and the international range it cites, 「around 15-30% among children and 0.3-14% among adults」.

[Note 9] NICE clinical guideline CG57: 「Discuss the benefits and harms of treatment with topical corticosteroids with children with atopic eczema and their parents or carers, emphasising that the benefits outweigh possible harms when they are applied correctly.」 (1.5.1.12) 「Tailor the potency of topical corticosteroids to the severity of the child's atopic eczema (which may vary according to body site): use mild potency for mild atopic eczema / use moderate potency for moderate atopic eczema / use potent for severe atopic eczema / use mild potency for the face and neck, except for short-term (3 to 5 days) use of moderate potency for severe flares / use moderate or potent preparations for short periods only (7 to 14 days) for flares in vulnerable sites such as axillae and groin / do not use very potent preparations in children without specialist dermatological advice.」 (1.5.1.13) 「If a mild or moderately potent topical corticosteroid has not controlled the atopic eczema within 7 to 14 days: exclude secondary bacterial or viral infection / for children aged 12 months or over, use potent topical corticosteroids for as short a time as possible (no longer than 14 days, and not on the face or neck) / if the atopic eczema is still uncontrolled, review the diagnosis and refer the child for specialist dermatological advice.」 (1.5.1.16) 「Once the atopic eczema has been controlled, consider treating problem areas with topical corticosteroids for 2 consecutive days per week to prevent flares in children with frequent flares (2 or 3 per month). Review this strategy within 3 to 6 months.」 (1.5.1.19)

[Note 10] Hong Kong College of Paediatricians, Clinical Guidelines on Management of Atopic Dermatitis in Children, HK J Paediatr 2013;18:96-104: 「♦Face and neck: moderate or potent TCS for 3-5 days ♦Axillae and groin: moderate or potent TCS no more than 1-2 weeks ♦Trunk and limbs: potent TCS for no more than 2 weeks」

[Note 11] Hong Kong College of Paediatricians review of 2020: 「TCS applied two to three times per week as maintenance treatment with a monthly amount of 15 g in infants, 30 g in children and 90 g in adolescents is generally safe.」「Evidences to support recommendations on how to apply TCS are limited.」 Long CC, Finlay AY, Clin Exp Dermatol 1991;16:444-7: 「One FTU covers 286 cm2 (s.d. +/- 80, n = 30). In males one FTU covers 312 cm2 (s.d. +/- 90, n = 16) and in females 257 cm2 (s.d. +/- 55, n = 14).」 The Hong Kong College of Paediatricians' two versions of the fingertip unit description: 2013, 「weighs approximately half a gram; enough to cover a surface area of two adult palms including fingers」; 2021, 「approximates to 0.5 gram … adequate to cover skin area equivalent to two adult hands with fingers together」. The two versions of the interval between applications: 2013, 「wait for at least 30 minutes between application of emollients and TCS」; 2021, 「At least 15 minutes should lapse between application of different topical medications and emollients」.

[Note 12] NICE CG57 1.5.1.4: 「Offer children with atopic eczema a choice of unperfumed emollients to use every day for moisturising. This may be a combination of products or one product for all purposes. Prescribe large quantities of leave-on emollients (250 g to 500 g weekly) that are easily available to use at nursery, pre-school or school.」 The opening sentence of the same section: 「Emollient creams are vital in helping to manage dry skin conditions, but there are Medicines and Healthcare products Regulatory Agency (MHRA) warnings about fire hazards associated with build-up of emollient residue on clothing and bedding.」

[Note 13] Hong Kong College of Paediatricians review of 2020: 「Dilution of TCS with emollients will not reduce its potency or side effects.」「Tachyphylaxis is not proven in TCS. There is good evidence to support the safety of appropriate TCS use for paediatric AD.」 NICE CG57 1.5.1.20: 「Consider a different topical corticosteroid of the same potency as an alternative to stepping up treatment if tachyphylaxis to a topical corticosteroid is suspected」

[Note 14] MHRA drug safety update of 29 May 2024: 「Whilst considered safe and effective, topical steroids can rarely lead to serious side effects such as thinning of the skin, adrenal suppression or very rarely Cushing's syndrome, due to systemic absorption. The incidence of these more serious side effects is linked to the amount, potency and duration of use of the topical steroid.」「Thinned skin appears translucent with visible tiny blood vessels and may be more fragile and more susceptible to stretch marks. This can be very difficult to see in brown or black skin, therefore careful monitoring is required.」「Adrenal suppression arises from overuse of topical steroids. The symptoms include low blood pressure, dizziness and fainting. This is a serious, life-threatening condition that needs urgent treatment. Stopping the topical steroid suddenly is dangerous if there is adrenal suppression, and it is likely that the individual will require oral steroid therapy replacement.」 Hong Kong College of Paediatricians guideline of 2013: 「Cutaneous complications include striae, atrophy, telangiectasia and perioral dermatitis Systemic complications such as Cushing's syndrome, hypothalamic-pituitary-adrenal (HPA) axis suppression, growth retardation, osteoporosis and hypertension are rare.」

[Note 15] MHRA update of 15 September 2021 (public version): 「as well as the known side effects associated with using too much of a topical corticosteroid or with using it for too long, remember that using too little can prolong treatment time and increase the risk of certain adverse effects」; and the advice to healthcare professionals: 「advise patients on the amount of product to be applied; underuse can prolong treatment duration」

[Note 16] Li AW, Yin ES, Antaya RJ, JAMA Dermatol 2017;153:1036-1042: 「Topical corticosteroid phobia prevalence ranged from 21.0% (95% CI, 15.8%-26.2%) to 83.7% (95% CI, 81.9%-85.5%). There was significant variation in how phobia was defined, ranging from concern to irrational fear. Questionnaires used to assess for TCS phobia included 1 to 69 questions. In the 2 studies that compared nonadherence between a phobia group and a nonphobia group, patients in both phobia groups were found to have a significantly higher rate of nonadherence (49.4% vs 14.1% and 29.3% vs 9.8%).」 The limitation statement: 「The nomenclature and assessment methods for TCS phobia used in studies, however, lack standardization, precluding quantitative comparison and extrapolation of data.」 Hong Kong College of Paediatricians review of 2020: 「ranged from 21% to 83.9%」 and 「Steroid phobia is an important reason for treatment failure. … In our local population, misconceptions and fallacies on AD are related to mistrust and unrealistic expectations about Western Medicine, phobias of anti-inflammatory agents, and use of complementary and alternative therapies.」

[Note 17] MHRA update of 15 September 2021: 「Topical steroid withdrawal reactions are thought to occur after prolonged, frequent, or inappropriate use of moderate to high potency topical corticosteroids. Topical steroid withdrawal reactions can develop after application of a topical corticosteroid at least daily for longer than a year. In children they can occur within as little as 2 months of daily use. People with atopic dermatitis are thought to be most at risk of developing topical steroid withdrawal reactions.」「Patients have also reported experiencing a less well understood group of side effects that has been termed Topical Steroid Withdrawal (TSW) reactions. TSW is the generally accepted patient-led term for these reactions - this group of conditions do not necessarily meet the medical pharmacological definition of withdrawal. Whilst these are still poorly understood groups of reactions, the evidence to date is that these reactions typically occur in four stages:」 The four phases in the update of May 2024: 「A few days (usually) after discontinuation, there is an acute eruption of burning red, exudative skin which may extend to untreated areas／Skin becomes dry and itchy with shedding (desquamation)／Skin starts to recover but is more sensitive and intermittent flares may occur／Skin recovers to the state prior to topical corticosteroid cessation. The recovery process may be prolonged」 and 「Continued use of topical steroids should be re-evaluated and alternative treatment options explored if a suspected TSW has occurred.」

[Note 18] MHRA update of 2021: 「During our review we considered data gathered from Yellow Card reports and identified 55 reports indicative of topical steroid withdrawal reactions, most of which were reported by patients. … We are unable to estimate the frequency of these reactions. However, given the number of patients who use topical corticosteroids, we understand reports of severe withdrawal reactions to be very infrequent.」 and 「It can be difficult to distinguish a flare up of the skin disorder, which would benefit from further topical steroid treatment, and a topical steroid withdrawal reaction.」「inform patients to return for medical advice if their skin condition worsens while using topical corticosteroid」 The update of 2024: 「Yellow Card reports up to August 2023 were reviewed and 267 reports were identified reporting reactions under the term TSW or with features that are often considered to be associated with this term.」「As there is still no accepted clinical definition for these reactions, not all cases of adverse reactions reported via Yellow Cards as TSW will be true withdrawal reactions. Some of the reported cases may be due to the worsening of the underlying condition or other unidentified causes. However, undoubtedly the review did identify cases of adverse reactions, often severe, which were associated with the prolonged use of moderate or high potency steroids.」 and 「if previous discontinuation was associated with reactions that raise suspicion of TSW, alternative treatments should be considered」

[Note 19] DUPIXENT United States FDA-approved prescribing information, 1.1: 「for the treatment of adult and pediatric patients aged 6 months and older with moderate-to-severe AD whose disease is not adequately controlled with topical prescription therapies or when those therapies are not advisable. DUPIXENT can be used with or without topical corticosteroids.」

[Note 20] Bieber T et al. (JADE COMPARE), N Engl J Med 2021;384:1101-1112: 「The 200-mg dose, but not the 100-mg dose, of abrocitinib was superior to dupilumab with respect to itch response at week 2. Neither abrocitinib dose differed significantly from dupilumab with respect to most other key secondary end-point comparisons at week 16.」

[Note 21] CIBINQO United States label, black box warning: 「WARNING: SERIOUS INFECTIONS, MORTALITY, MALIGNANCY, MAJOR ADVERSE CARDIOVASCULAR EVENTS (MACE), and THROMBOSIS」「• Increased risk of serious bacterial, fungal, viral and opportunistic infections leading to hospitalization or death, including tuberculosis (TB). Discontinue treatment with CIBINQO if serious or opportunistic infection occurs. Test for latent TB before and during therapy; treat latent TB prior to use. Monitor all patients for active TB during treatment, even patients with initial negative latent TB test. (5.1) • Higher rate of all-cause mortality, including sudden cardiovascular death, with another JAK inhibitor vs. TNF blockers in rheumatoid arthritis (RA) patients. CIBINQO is not approved for use in RA patients. (5.2) • Malignancies have occurred with CIBINQO. Higher rate of lymphomas and lung cancers with another JAK inhibitor vs. TNF blockers in RA patients. (5.3) • MACE has occurred with CIBINQO. Higher rate of MACE (defined as cardiovascular death, myocardial infarction, and stroke) with another JAK inhibitor vs. TNF blockers in RA patients. (5.4) • Thrombosis has occurred with CIBINQO. Increased incidence of pulmonary embolism, venous and arterial thrombosis with another JAK inhibitor vs. TNF blockers. (5.5)」

[Note 22] CIBINQO indication: 「CIBINQO is indicated for the treatment of adults and pediatric patients 12 years of age and older with refractory, moderate-to-severe atopic dermatitis whose disease is not adequately controlled with other systemic drug products, including biologics, or when use of those therapies is inadvisable. Limitations of Use CIBINQO is not recommended for use in combination with other JAK inhibitors, biologic immunomodulators, or other immunosuppressants.」 RINVOQ §1.3: 「RINVOQ is indicated for the treatment of adults and pediatric patients 12 years of age and older with refractory, moderate to severe atopic dermatitis whose disease is not adequately controlled with other systemic drug products, including biologics, or when use of those therapies are inadvisable. (1.3) Limitations of Use RINVOQ is not recommended for use in combination with other JAK inhibitors, biologic immunomodulators, or with other immunosuppressants. (1.3)」 CIBINQO section 4 contraindication: 「CIBINQO is contraindicated in patients taking antiplatelet therapies, except for low-dose aspirin (≤81 mg daily), during the first 3 months of treatment.」 RINVOQ section 4: 「RINVOQ/RINVOQ LQ is contraindicated in patients with known hypersensitivity to upadacitinib or any of its excipients」

[Note 23] CIBINQO label 5.4: 「Patients who are current or past smokers are at additional increased risk.」 5.5: 「Avoid CIBINQO in patients that may be at increased risk of thrombosis.」

[Note 24] CIBINQO label: 「Treatment with CIBINQO was associated with an increased incidence of thrombocytopenia and lymphopenia … Prior to CIBINQO initiation, perform a CBC … CBC evaluations are recommended at 4 weeks after initiation and 4 weeks after dose increase of CIBINQO.」「Dose-dependent increase in blood lipid parameters were reported in subjects treated with CIBINQO … Lipid parameters should be assessed approximately 4 weeks following initiation of CIBINQO therapy … The effect of these lipid parameter elevations on cardiovascular morbidity and mortality has not been determined.」「Prior to initiating CIBINQO, complete all age-appropriate vaccinations as recommended by current immunization guidelines, including prophylactic herpes zoster vaccinations.」 RINVOQ label: 「Neutropenia Treatment with RINVOQ was associated with an increased incidence of neutropenia (ANC less than 1000 cells/mm3). … Avoid RINVOQ/RINVOQ LQ initiation and interrupt RINVOQ/RINVOQ LQ treatment in patients with a low neutrophil count (i.e., ANC less than 1000 cells/mm3)」「NMSCs have been reported in patients treated with RINVOQ. Periodic skin examination is recommended for patients who are at increased risk for skin cancer. Exposure to sunlight and UV light should be limited by wearing protective clothing and using a broad-spectrum sunscreen.」 (5.3) 「Discontinue RINVOQ/RINVOQ LQ in patients that have experienced a myocardial infarction or stroke.」 (5.4) 「Gastrointestinal perforations have been reported in clinical trials with RINVOQ … Monitor RINVOQ/RINVOQ LQ-treated patients who may be at risk for gastrointestinal perforation (e.g., patients with a history of diverticulitis and those taking concomitant medications including NSAIDs or corticosteroids).」 (5.7)

[Note 25] Ytterberg SR et al. (ORAL Surveillance), N Engl J Med 2022;386:316-326: 「We conducted a randomized, open-label, noninferiority, postauthorization, safety end-point trial involving patients with active rheumatoid arthritis despite methotrexate treatment who were 50 years of age or older and had at least one additional cardiovascular risk factor.」「The hazard ratios were 1.33 (95% confidence interval [CI], 0.91 to 1.94) for MACE and 1.48 (95% CI, 1.04 to 2.09) for cancers; the noninferiority of tofacitinib was not shown.」 and 「a cardiovascular risk-enriched population」

[Note 26] EMA JAKi Article 20 review (CHMP endorsement date 23 January 2023): 「EMA has now concluded that these safety findings apply to all approved uses of JAK inhibitors in chronic inflammatory disorders (rheumatoid arthritis, psoriatic arthritis, juvenile idiopathic arthritis, axial spondyloarthritis, ulcerative colitis, atopic dermatitis and alopecia areata).」「These medicines should be used in the following patients only if no suitable treatment alternatives are available: those aged 65 years or above, those at increased risk of major cardiovascular problems (such as heart attack or stroke), those who smoke or have done so for a long time in the past and those at increased risk of cancer.」 The public explanation also carries: 「Examine your skin periodically and let your doctor know if you notice any new growths on the skin.」

[Note 27] MHRA Drug Safety Update (26 April 2023), therapeutic area label: 「Therapeutic area: Dermatology, GI, hepatology and pancreatic disorders and Rheumatology」; the first five bullets of the patient explanation: 「JAK inhibitors are effective medicines for treating chronic inflammatory disorders such as severe arthritis, psoriatic arthritis, inflammatory bowel diseases, and atopic eczema (atopic dermatitis)／however, these medicines can increase a patient's risk of developing major cardiovascular problems (such as heart attack or stroke), cancer, blood clots in the lungs and in the deep veins of the body, serious infections, and death when compared with other treatments for these conditions called TNF-alpha inhibitors／the risks are greater if you have risk factors, such as being age 65 years or older, have an already increased risk of major cardiovascular problems or cancer, or if you smoke or have smoked in the past for a long time／if you have a risk factor, your doctor will consider alternative treatment options and only offer you treatment with a JAK inhibitor if nothing else is suitable for you／your doctor may recommend a lower dose of the JAK inhibitor medicine, to minimise possible risks」; the class effect qualification: 「The review concluded that the effects could be considered a class effect, while acknowledging that the extent to which the findings of study A3921133 applied to all JAK inhibitors was dependent on the similarities of each treated population in terms of the presence of risk factors.」

[Note 28] Hong Kong College of Paediatricians review of 2020: 「Other novel topical anti-inflammatory agents under investigations including lipoxins and janus kinase inhibitors (JAKi) e.g. tofacitinib require further studies to prove the efficacy and safety in paediatric population.」 The SCORAD >50 tier of the severity ladder: 「- Hospitalisation - Potent to very potent TCS - Systemic immunosuppressants (e.g. Cyclosporin A, Azathioprine, Methotrexate) - Dupilumab if age appropriate」

[Note 29] NICE TA534 (1 August 2018): 「Dupilumab is recommended as an option for treating moderate to severe atopic dermatitis in adults, only if: the disease has not responded to at least 1 other systemic therapy, such as ciclosporin, methotrexate, azathioprine and mycophenolate mofetil, or these are contraindicated or not tolerated」「Stop dupilumab at 16 weeks if the atopic dermatitis has not responded adequately. An adequate response is: at least a 50% reduction in the Eczema Area and Severity Index score (EASI 50) … and at least a 4‑point reduction in the Dermatology Life Quality Index (DLQI)」 NICE TA814 (3 August 2022): 「Abrocitinib and upadacitinib were indirectly compared with ciclosporin, but the results are highly uncertain.」

[Note 30] Legislative Council Finance Committee, Examination of Estimates of Expenditure 2024-25, written reply HHB094: 「根據備存的資料，在2022年和2023年，所有嚴重皮膚病新症都已根據服務承諾，被安排於8個星期內到診，輪候時間中位數分別為2.7個和2.9個星期。至於其他穩定新症，輪候時間中位數為90個和93個星期，整體最長輪候時間分別為203個和183個星期。」

[Note 31] Department of Health clinic page (the page's own revision date: 「二零二五年六月二日」): 「轉介 : 所有新症均需醫生介紹信」「備註 : 所有新 / 舊症均需預約。就新症預約，請携同身份証明文件及香港註册西醫發出的轉介信親身到診所預約，或可以郵寄方式寄到指定診所⋯⋯此外，診所亦接受經香港政府一站通遞交的預約 / 改期申請」 Department of Health page (「二零一九年十月一日」): 「社會衞生服務屬下有性病診所和皮膚科診所，負責治療護理、預防及控制性病和皮膚疾病。」

[Note 32] The Chinese definition of the third cell in version 22.1 of the Hospital Authority Drug Formulary: 「獲安全網資助的自費購買藥物(自費藥物) — 經證實有顯著療效，但超出醫管局一般資助服務範圍所能提供的非常昂貴藥物。這些藥物不屬公立醫院和診所標準收費提供的項目。需要使用這些藥物而有能力負擔費用的病人須自費購買。然而，相關基金會為需要這些藥物而經濟上有困難的病人提供安全網。」

[Note 33] NICE CG57 1.5.1.21: 「Topical tacrolimus and pimecrolimus are not recommended in NICE technology appraisal guidance for treating mild atopic eczema, or as first-line treatments for atopic eczema of any severity.」 1.5.1.22: 「Topical tacrolimus is recommended as an option in NICE technology appraisal guidance for the second-line treatment of moderate to severe atopic eczema in children aged 2 years and older that has not been controlled by topical corticosteroids, where there is a serious risk of important adverse effects from further topical corticosteroid use, particularly irreversible skin atrophy.」 1.5.1.24: 「Start treatment with tacrolimus or pimecrolimus only with specialist dermatological advice, and only after careful discussion with the child and their parents or carers about the potential risks and benefits of all appropriate second-line treatment options.」

[Note 34] Samaritan Fund list of subsidised drugs (effective from 25 July 2026): 「5 阿布昔替尼 (Abrocitinib) 皮膚病學 異位性皮膚炎」「42a 度匹蘆人單抗 (Dupilumab) 皮膚病學/兒科 異位性皮膚炎」「111d 烏帕替尼 (Upadacitinib) 皮膚病學/兒科 異位性皮膚炎」「65 來瑞奇珠單抗 (Lebrikizumab)^ 皮膚病學 異位性皮膚炎^」; the document's note 「^ With effect from 25 July 2026.」; and also 「21 Baricitinib Rheumatology Rheumatoid arthritis」.

[Note 35] Samaritan Fund financial assessment page (Chinese version): 「在公營醫療收費改革下，撒瑪利亞基金的申請資格於2026年1月1日起放寬。」「每年可動用的財務資源 的計算模式為家庭每年的可動用收入的百分之八十，再加以家庭的可動用資產淨值（即可動用資產扣除可扣減豁免額）的百分之五十。」「如病人屬於持續申請人 （即病人在過去十八個月內曾申請撒瑪利亞基金或關愛基金醫療援助項目藥物資助並已獲批核） ，其家庭每年的可動用收入淨值只計算百分之六十。」 The same item in the English version: 「50% of patients' household net assets」.

[Note 36] Hospital Authority "Annual Fee Cap" page: 「a HK$10,000 cap on an eligible patient's annual spending for specified public medical fees and charges without requiring financial assessment with effect from 1 January 2026… Charges for self-financed drugs and medical devices are excluded.」 The four eligibility items in the Chinese version: 「必須為符合資格人士*」「病人於年度內累計繳付的「合資格醫療費用及收費」達到一萬港元」「在申請「全年收費上限」時，在醫管局轄下無任何拖欠的「合資格醫療費用及收費」」「病人接受之醫院服務不得被醫管局裁定為「無臨床需要」」; the application arrangements: 「年度接受申請期由每年的1月1日開始，至次年的3月31日截止。遲交的申請將不獲接納，合資格醫療費用只包括在每年1月1日至12月31日所發出的醫療賬單，並在提交申請時已全數繳付。」「每一個年度需要重新申請「全年收費上限」」 The English version: 「Patients who met the eligibility criteria may submit their applications via HA Go or at any hospitals' shroff offices once their cumulative valid annual spending reached $10,000.」 The Department of Health fee table: 「上述(a)項的收費不包括有關的配藥、放射造影服務及化驗費用。 上述(e)項的收費不包括自費藥物。」

[Note 37] Samaritan Fund "eligibility" page: 「獲基金資助的病人須為醫院管理局（醫管局）病人並符合下列條件：」 English version: 「To be eligible for financial assistance, the patient must be an Hospital Authority (HA) patient and fulfill all of the following requirements.」 The drug item condition: 「藥物項目：根據醫管局現行的臨床指引，有關病人的治療計劃必須經由一名指定的醫管局醫生簽發。」

[Note 38] Official Record of Proceedings of the Legislative Council (28 November 2018), the Secretary for Food and Health's own words: 「在現有的機制下，如果病人需要某種藥物，而衞生署藥物名冊並無這類藥物，可以將病人轉介到醫管局，最重要是在臨床診斷時，醫生認為病人有這樣的需要。」「如果醫生在臨床診斷上，認為有需要採用一些不在衞生署藥物名冊內的藥物，需要向醫管局索取，其實現時已有機制可以轉介患者到醫管局轄下的醫院接受治療或使用該等藥物。」「根據衞生署有關生物製劑的資料，現時有一種已在本港註冊的生物製劑可以醫治嚴重的異位性濕疹。衞生署轄下的社會衞生科會一直密切留意該生物製劑的臨床及科研實證的最新發展，亦會適時透過現有機制，與醫管局協作，將這些病人轉介醫管局，使他們得到適當的公營醫療服務。」 "The Department of Health's Collaboration and Referral Mechanism with the Hospital Authority on Dermatological Services" (17 December 2018): 「現時協作轉介機制只適用於銀屑病患者。」 English version: 「Currently, the existing collaboration-referral mechanism is only applicable to patients with psoriasis.」

[Note 39] Hong Kong Economic Journal health column report of 30 July 2025: 「一針濕疹生物製劑價錢約$7,500至$10,000」「若每兩星期注射一次，每月藥費約$15,000至$20,000」

[Note 40] The sixth to eighth bullets of the patient explanation in the MHRA update of 26 April 2023: 「contact your doctor or seek urgent medical advice if, during treatment, you experience chest pain or tightness (which may spread to arms, jaw, neck and back), shortness of breath, cold sweats, light headedness, sudden dizziness, weakness in arms and legs or slurred speech – these are signs of a medical emergency／examine your skin periodically and let your doctor or nurse know if you notice any new growths on your skin or changes to moles (including itching, shape and discharge, which may not be as obvious on darker skin tones); these could require investigation for possible skin cancer／always read the leaflet that accompanies your medicines and talk to your doctor, nurse, or pharmacist if you are concerned about side effects」

#What this article does not state

• "80% of Hong Kong children with eczema are allergic to dust mites" — this article does not say that. No source among the material cited here supports that figure.
• "About 60% of children with eczema remit spontaneously before adulthood" — this article does not say that. Likewise, no source. Nor does this article make any statement about the proportions of the natural course.
• "About 300,000 people in Hong Kong are affected by moderate-to-severe eczema", "about 10% of adults are affected" — this article does not say that. There is no adult eczema incidence figure for Hong Kong, a point each of two first-hand documents states for itself — the Hong Kong College of Paediatricians' and the HKU/PolyU study protocol (originals and renderings in the section "How many people in Hong Kong actually have eczema?" above).
• Specific figures for bath water temperature, duration, room temperature and humidity — this article does not give them. The popularly circulated "water below 38°C, under 10 minutes, room temperature 22–25°C, humidity 50%–60%" appears in none of the guidelines or official documents cited here. On emollients, the material cited here contains no comparison of whether ointment, cream or lotion is better. (⚠️ Version 3 of this article stated at this item that "the original NICE CG57 entry on emollient quantity was not obtained" — that statement was wrong and was corrected in version 4: CG57 1.5.1.4 does give a quantity, and the original and the rendering have been added to the "Emollients" passage in the body. The erroneous statement is kept here so that it can be traced.)
• The role of probiotics in eczema — this article makes no statement. The material cited here contains no study of probiotics in treating eczema.
• How to choose between dupilumab and a JAK inhibitor — this article draws no comparative conclusion. The closest thing to Hong Kong expert opinion this article found is a conference report (Hong Kong Journal of Dermatology & Venereology 2025;32:81-83) recording a lecture by a private dermatologist, which is not any college's position paper; and that record itself writes 「The decision regarding targeted therapy for atopic dermatitis should be personalised for each individual」.
• The full text of the American Academy of Dermatology's individual recommendations is not cited here. This article cites only the statements marked as strong recommendations in the abstracts of that academy's two guidelines; the complete wording of each recommendation and its evidence grading could not be obtained.
• Whether Hong Kong's drug regulator has issued a safety notice of its own on JAK inhibitors — this article makes no statement. No such record appears in the material cited here, and there is no evidence that none exists.
• A subsidised drug purchase scheme once reported on — this article does not describe its current state. A report in 2021 mentioned a subsidised drug purchase arrangement run by a social welfare organisation involving an eczema biologic; that organisation's website no longer lists the drug, so this article does not describe it as a current scheme.
• This article has not converted the public dermatology waiting position into weeks or months. What the Department of Health publishes is the month new case appointments are running to, not a duration; converting would require an assumption the document does not supply.
• A patient attending a Department of Health dermatology clinic cannot apply to the Samaritan Fund as a Department of Health patient (the fund's eligibility page: 「獲基金資助的病人須為醫院管理局（醫管局）病人」). There is a referral route, which the government spoke about in the Legislative Council in 2018; but this article will not say how long that route takes, whether there are quotas, or what the clinical threshold is — no published document explains any of those three, and referral for eczema rests on a general rule plus one oral undertaking, not on a published referral guideline naming atopic dermatitis (the formal mechanism document states for itself that it applies to psoriasis patients only).

#Sources

• Hong Kong's only government eczema page for the public; the list of possible causes; the description of topical and oral steroids; and dosing instructions: Department of Health Student Health Service, "Eczema", the page's own revision date June 2022, Chinese https://www.studenthealth.gov.hk/tc_chi/health/health_ophp/health_ophp_tet.html ; English https://www.studenthealth.gov.hk/english/health/health_ophp/health_ophp_tet.html (retrieved 2 August 2026)
• Public dermatology belonging to the Department of Health's Social Hygiene Service: Department of Health, "Sexually Transmitted Infections and Skin Diseases", the page's own revision date 「二零一九年十月一日」, Chinese https://www.dh.gov.hk/tc_chi/main/main_chp/sexually.html ; English https://www.dh.gov.hk/english/main/main_chp/sexually.html (retrieved 2 August 2026)
• The Samaritan Fund's 「獲基金資助的病人須為醫院管理局（醫管局）病人」 and the requirement that drug items be issued by a designated Hospital Authority doctor: Hospital Authority, "Eligibility", Chinese https://www.ha.org.hk/haho/ho/sf/serviceguide_samaritan_link3-b5.htm ; English https://www.ha.org.hk/haho/ho/sf/eligibility_en.htm (retrieved 2 August 2026)
• The Department of Health having a drug formulary of its own; a drug outside that formulary leading to a referral to the Hospital Authority; and the referral undertaking on the biologic for severe atopic eczema: Official Record of Proceedings of the Legislative Council, 28 November 2018, question four, "Treating eczema patients" (asked by the Hon Ann Chiang, answered in Cantonese by the Secretary for Food and Health), Chinese https://www.legco.gov.hk/yr18-19/chinese/counmtg/hansard/cm20181128-translate-c.pdf ; English https://www.legco.gov.hk/yr18-19/english/counmtg/hansard/cm20181128-translate-e.pdf (retrieved 2 August 2026)
• The formal collaboration and referral mechanism applying 「只適用於銀屑病患者」: Food and Health Bureau, Department of Health and Hospital Authority, "The Department of Health's Collaboration and Referral Mechanism with the Hospital Authority on Dermatological Services", Legislative Council paper CB(2)423/18-19(06), 17 December 2018, Chinese https://www.legco.gov.hk/yr18-19/chinese/panels/hs/papers/hs20181217cb2-423-6-c.pdf ; English https://www.legco.gov.hk/yr18-19/english/panels/hs/papers/hs20181217cb2-423-6-e.pdf (retrieved 2 August 2026)
• New dermatology appointments running to December 2027 through February 2029 (the complete list of 9 clinics): "Department of Health Social Hygiene Service — New Skin Case Appointment Status", the document's own update date 30 June 2026, Chinese https://www.dh.gov.hk/tc_chi/tele/tele_chc/files/New_Skin_Case_Appointment_Status_chi.pdf ; English https://www.dh.gov.hk/english/tele/tele_chc/files/New_Skin_Case_Appointment_Status_en.pdf (retrieved 2 August 2026)
• All new cases requiring a referral letter from a Hong Kong registered western medicine practitioner; the appointment methods; and the list of 9 clinics: Department of Health, "Clinics Providing Dermatological Services", the page's own revision date 2 June 2025, Chinese https://www.dh.gov.hk/tc_chi/tele/tele_chc/tele_chc_dc.html ; English https://www.dh.gov.hk/english/tele/tele_chc/tele_chc_dc.html (retrieved 2 August 2026)
• Dermatology clinic fees (eligible persons 250 dollars per attendance and 20 dollars per drug item for up to 4 weeks; non-eligible persons 850 dollars and 90 dollars; and "not including self-financed drugs"): Department of Health, "Fees for Outpatient Services", the page's own date 「二零二六年一月一日」, Chinese https://www.dh.gov.hk/tc_chi/useful/useful_fee/useful_fee_os.html ; English https://www.dh.gov.hk/english/useful/useful_fee/useful_fee_os.html (retrieved 2 August 2026)
• The Hospital Authority's open data on specialist outpatient waiting times covering eight specialties only, with no dermatology: Hospital Authority open data https://www.ha.org.hk/opendata/sop/sop-waiting-time-en.json (covering 1 July 2025 to 30 June 2026; retrieved 2 August 2026)
• The definitions of the formulary's four categories (Chinese and English versions); and 「醫院管理局藥物名冊 [只備英文版]」: Hospital Authority, "Drug Formulary Categories and Charges", Chinese https://www.ha.org.hk/hadf/tc/Drug-Formulary/Drug-Formulary-Categories-And-Charges.html ; English https://www.ha.org.hk/hadf/en/Drug-Formulary/Drug-Formulary-Categories-And-Charges.html (retrieved 2 August 2026)
• The complete drug lists and classifications of the "PREPARATIONS FOR ECZEMA AND PSORIASIS" and "TOPICAL CORTICOSTEROIDS" sections: Hospital Authority Drug Formulary version 22.1, the page's own version statement "with effect from 25 July 2026", https://www.ha.org.hk/hadf/en/Others/Search-Result.html (retrieved 2 August 2026)
• The Samaritan Fund subsidising abrocitinib, dupilumab, upadacitinib and lebrikizumab against "atopic dermatitis", and baricitinib against rheumatology and rheumatoid arthritis: Samaritan Fund List of Subsidised Drugs, the document's own effective date 25 July 2026, Chinese https://www.ha.org.hk/haho/ho/sf/SF_Items_tc.pdf ; English https://www.ha.org.hk/haho/ho/sf/SF_Items_en.pdf (retrieved 2 August 2026)
• The Community Care Fund's first phase being for specified self-financed cancer drugs, its item list containing no atopic dermatitis drug: Community Care Fund medical assistance programme item list, https://www.ha.org.hk/haho/ho/ccf/CCF_items_tc.pdf and https://www.ha.org.hk/haho/ho/ccf/CCF_items_en.pdf (retrieved 2 August 2026)
• The Samaritan Fund's financial assessment formula (80% of disposable income plus 50% of net disposable assets), the provision counting only 60% of income for continued applicants, and the eligibility criteria relaxed from 1 January 2026; eligibility to be assessed by a medical social worker on a household basis; and the clinical plan to be issued by a designated Hospital Authority doctor under the "prevailing clinical guidelines": Hospital Authority Samaritan Fund financial assessment page, Chinese version https://www.ha.org.hk/haho/ho/sf/sf_fa_b5.htm ; English version https://www.ha.org.hk/haho/ho/sf/sf_fa_en.htm (the Chinese and English versions word the assets item differently, and this article takes the Chinese as authoritative; retrieved 3 August 2026) and the eligibility page https://www.ha.org.hk/visitor/ha_view_content.asp?Content_ID=212018&Lang=ENG&Dimension=100&Ver=HTML (retrieved 2 August 2026)
• Specialist outpatient 250 dollars per attendance and 20 dollars per drug item (4 weeks); a self-financed drug administration fee of 130 dollars per item; and self-financed drugs being paid in full: Hospital Authority, "Fees and Charges", the charges effective from 1 January 2026, https://www.ha.org.hk/visitor/ha_view_content.asp?Content_ID=10045 (retrieved 2 August 2026)
• The ten-thousand-dollar annual fee cap excluding self-financed drugs and medical devices; the four eligibility items, cumulative payment reaching ten thousand dollars, no outstanding amount at the time of application, self-application through HA Go or a hospital shroff office, a fresh application each year, and a deadline of 31 March of the following year with late submissions not accepted: Hospital Authority, "Annual Fee Cap" (for eligible persons), Content_ID=281820, Chinese https://www.ha.org.hk/visitor/ha_view_content.asp?Content_ID=281820&Lang=CHIB5&Dimension=100&Ver=HTML ; English https://www.ha.org.hk/visitor/ha_view_content.asp?Content_ID=281820&Lang=ENG&Dimension=100&Ver=HTML (effective 1 January 2026; retrieved 2 August 2026)
• Hong Kong registration status (dupilumab, abrocitinib, upadacitinib, baricitinib and lebrikizumab registered; tralokinumab not registered); and the database having no indication field: Department of Health Drug Office Hong Kong registered pharmaceutical products database, the page's own marking "Last Updated: 31-Jul-2026", https://www.drugoffice.gov.hk/eps/do/en/consumer/search_drug_database.html (retrieved 2 August 2026)
• The communication requirement that the benefits of topical corticosteroids outweigh the harms; the whole of 1.5.1.13 (potency tailored to severity, with site as a modifying condition); the 48-hour rule; the whole three steps of 1.5.1.16 (exclude secondary infection / short-term potent in those 12 months and over, not on face or neck / review the diagnosis and refer if still uncontrolled); 1.5.1.17 on infants under twelve months; proactive maintenance two days a week; the handling of tachyphylaxis; 1.5.1.21, 1.5.1.22 and 1.5.1.24 on topical calcineurin inhibitors not being first line, being a second-line option in those aged 2 and above, and being started on specialist dermatological advice; 1.5.1.4 on emollients at 250–500 g weekly and the MHRA fire hazard warning in the same section: NICE clinical guideline CG57 Atopic eczema in under 12s: diagnosis and management, the guideline self-dated Published 12 December 2007 / Last updated 22 September 2025, https://www.nice.org.uk/guidance/cg57/chapter/Recommendations (retrieved 2 August 2026)
• Hong Kong's potency and day limits by body site; the 2013 version of the fingertip unit description; at least 30 minutes between emollient and steroid; and the lists of cutaneous and systemic complications: Leung TNH, Chow CM, Chow MPY, et al. Clinical Guidelines on Management of Atopic Dermatitis in Children. HK J Paediatr 2013;18:96-104, the document self-dated as endorsed 14 December 2012, https://www.hkjpaed.org/pdf/2013;18;96-104.pdf (retrieved 2 August 2026)
• Subclinical inflammation and tapering too early; the monthly maintenance quantities (15 g in infants, 30 g in children, 90 g in adolescents); the 2021 version of the fingertip unit description; at least 15 minutes between applications; dilution not reducing potency or side effects; tachyphylaxis being unproven; the steroid phobia passage; local prevalence data missing for more than a decade; the severity ladder; and JAK inhibitors still being under investigation: Leung TNH, Cheng JWCH, Chan SCW, et al. Management of Atopic Dermatitis in Children: 2020 Review by the Guidelines Development Panel of Hong Kong College of Paediatricians. HK J Paediatr 2021;26:42-57, https://www.hkjpaed.org/pdf/2021;26;42-57.pdf (retrieved 2 August 2026)
• The original fingertip unit measurement (286 square centimetres, standard deviation ±80, n=30; and the number of FTUs by site): Long CC, Finlay AY. The finger-tip unit—a new practical measure. Clin Exp Dermatol 1991;16:444-7 (PMID 1806320; doi:10.1111/j.1365-2230.1991.tb01232.x)
• Using too little prolonging treatment; the framing that they are safe and effective when used correctly; the risk factors for withdrawal reactions; the 55 Yellow Card reports and the inability to estimate incidence; and the difficulty of distinguishing a flare from a withdrawal reaction: MHRA Drug Safety Update "Topical steroid withdrawal reactions: a review of the evidence", the page self-dated 15 September 2021, https://www.gov.uk/drug-safety-update/topical-corticosteroids-information-on-the-risk-of-topical-steroid-withdrawal-reactions (retrieved 2 August 2026)
• Skin thinning, adrenal suppression and very rarely Cushing's syndrome; the four phases of withdrawal reactions; and the 267 reports up to August 2023 with still no agreed clinical definition: MHRA Drug Safety Update "Topical steroids: introduction of new labelling and a reminder of the possibility of severe side effects, including Topical Steroid Withdrawal Reactions", the page self-dated 29 May 2024, https://www.gov.uk/drug-safety-update/topical-steroids-introduction-of-new-labelling-and-a-reminder-of-the-possibility-of-severe-side-effects-including-topical-steroid-withdrawal-reactions (retrieved 2 August 2026)
• Steroid phobia prevalence of 21.0%–83.7% (16 cross-sectional studies); the adherence comparison coming from only 2 of them; and quantitative comparison and extrapolation being precluded: Li AW, Yin ES, Antaya RJ. Topical Corticosteroid Phobia in Atopic Dermatitis: A Systematic Review. JAMA Dermatol 2017;153:1036-1042 (PMID 28724128; doi:10.1001/jamadermatol.2017.2437)
• Dupilumab's efficacy across the three trials (both arms with denominators); the original §1.1 indication (6 months and older, not adequately controlled with topical prescription therapies or those being inadvisable); the trial entry criteria; the adverse reactions table; the definition of the pooled conjunctivitis term; baseline comorbidity; and the discontinuation rates: DUPIXENT (dupilumab) United States FDA-approved prescribing information, obtained through DailyMed (set id 595f437d-2729-40bb-9c62-c8ece1f82780), the DailyMed record published 6 July 2026, https://dailymed.nlm.nih.gov/dailymed/services/v2/spls/595f437d-2729-40bb-9c62-c8ece1f82780.xml (retrieved 2 August 2026)
• The original SOLO 1 and SOLO 2 paper: Simpson EL, Bieber T, Guttman-Yassky E, et al. Two Phase 3 Trials of Dupilumab versus Placebo in Atopic Dermatitis. N Engl J Med 2016;375:2335-2348 (PMID 27690741; doi:10.1056/NEJMoa1610020)
• The full text of CIBINQO's black box warning; the original AD indication (refractory / not controlled by other systemic drugs including biologics / 12 and above), the Limitations of Use on combination, and the section 4 antiplatelet contraindication; the six statements that it is not approved for rheumatoid arthritis; the events occurring in its AD trials; the two sentences on smoking and thrombotic risk at sections 5.4 and 5.5, the latent tuberculosis testing requirement at section 5.1, the full blood count and lipid monitoring, and the vaccination requirement before starting: CIBINQO (abrocitinib) United States FDA-approved prescribing information, obtained through DailyMed (set id 16c12a56-4550-414b-ac9d-b785b41fea6b), the DailyMed record published 13 July 2026, https://dailymed.nlm.nih.gov/dailymed/services/v2/spls/16c12a56-4550-414b-ac9d-b785b41fea6b.xml (retrieved 3 August 2026)
• The full text of RINVOQ's black box warning; the original AD indication (§1.3), the Limitations of Use, and the section 4 hypersensitivity contraindication; the skin examination recommendation for non-melanoma skin cancer; the laboratory monitoring thresholds; and gastrointestinal perforation: RINVOQ (upadacitinib) United States FDA-approved prescribing information, obtained through DailyMed (set id 2966aec7-2ef0-923c-d8ff-fe1a957bf095), the DailyMed record published 15 July 2026, https://dailymed.nlm.nih.gov/dailymed/services/v2/spls/2966aec7-2ef0-923c-d8ff-fe1a957bf095.xml (retrieved 3 August 2026)
• The complete list of OLUMIANT's United States indications (containing no atopic dermatitis; the rheumatoid arthritis one limited to inadequate response to one or more TNF blockers): OLUMIANT (baricitinib) United States FDA-approved prescribing information, obtained through DailyMed (set id 866e9f35-9035-4581-a4b1-75a621ab55cf), the DailyMed record published 14 July 2026 (retrieved 3 August 2026)
• The ORAL Surveillance trial population, the three groups' denominators, and the MACE and cancer hazard ratios with confidence intervals — non-inferiority not established: Ytterberg SR, Bhatt DL, Mikuls TR, et al. Cardiovascular and Cancer Risk with Tofacitinib in Rheumatoid Arthritis. N Engl J Med 2022;386:316-326 (PMID 35081280; doi:10.1056/NEJMoa2109927)
• The safety findings applying to all approved uses including atopic dermatitis; the restriction by risk factor; and the prompt to examine the skin periodically: European Medicines Agency, Janus kinase inhibitors (JAKi) – referral, the CHMP endorsement date on the page 23 January 2023, https://www.ema.europa.eu/en/medicines/human/referrals/janus-kinase-inhibitors-jaki (retrieved 2 August 2026)
• The therapeutic area label listing "Dermatology"; how it names atopic eczema in terms; the qualification that the class effect depends on the similarity of the treated populations' risk factors; the list of emergency symptoms for patients; and the effect estimates from the baricitinib observational study: MHRA Drug Safety Update "Janus kinase (JAK) inhibitors: new measures to reduce risks…", the page self-dated 26 April 2023, https://www.gov.uk/drug-safety-update/janus-kinase-jak-inhibitors-new-measures-to-reduce-risks-of-major-cardiovascular-events-malignancy-venous-thromboembolism-serious-infections-and-increased-mortality (retrieved 2 August 2026)
• The EU Olumiant atopic dermatitis indication covering those aged 2 and above: European Medicines Agency, Olumiant EPAR (EMEA/H/C/004085, Revision 26), https://www.ema.europa.eu/en/medicines/human/EPAR/olumiant (retrieved 2 August 2026)
• The four groups' denominators in the head-to-head trial of abrocitinib against dupilumab, the week 12 results, and that trial's own comparative conclusion: Bieber T, Simpson EL, Silverberg JI, et al. Abrocitinib versus Placebo or Dupilumab for Atopic Dermatitis. N Engl J Med 2021;384:1101-1112 (PMID 33761207; doi:10.1056/NEJMoa2019380)
• The three groups' denominators and week 16 EASI-75 in upadacitinib's two trials: Guttman-Yassky E, Teixeira HD, Simpson EL, et al. Lancet 2021;397:2151-2168 (PMID 34023008; doi:10.1016/S0140-6736(21)00588-2; and see the erratum in the same volume, Lancet 2021;397:2150)
• England's access conditions and the 16-week stopping rule (dupilumab): NICE technology appraisal guidance TA534, the guidance self-dated 1 August 2018, https://www.nice.org.uk/guidance/ta534/chapter/1-Recommendations (retrieved 2 August 2026)
• England's access conditions and the 16-week stopping rule (abrocitinib, upadacitinib); and the indirect comparison being highly uncertain: NICE technology appraisal guidance TA814, the guidance self-dated 3 August 2022, https://www.nice.org.uk/guidance/ta814/chapter/1-Recommendations (retrieved 2 August 2026)
• The American Academy of Dermatology's strong recommendations for dupilumab, tralokinumab, abrocitinib, baricitinib and upadacitinib: Davis DMR, Drucker AM, Alikhan A, et al. J Am Acad Dermatol 2024;90:e43-e56 (PMID 37943240; doi:10.1016/j.jaad.2023.08.102); and the topical treatment guideline: Sidbury R, Alikhan A, Bercovitch L, et al. J Am Acad Dermatol 2023;89:e1-e20 (PMID 36641009)
• Filaggrin loss-of-function variants and atopic dermatitis, with the 9% being a figure for people of European descent: Palmer CN, Irvine AD, Terron-Kwiatkowski A, et al. Nat Genet 2006;38:441-6 (PMID 16550169; doi:10.1038/ng1767)
• The common European FLG mutations being rare or absent in Chinese patients, and the Singaporean Chinese cohort denominators: Chen H, Common JE, Haines RL, et al. Br J Dermatol 2011;165:106-14 (PMID 21428977; doi:10.1111/j.1365-2133.2011.10331.x)
• The pathophysiology involving multiple immune pathways: Langan SM, Irvine AD, Weidinger S. Atopic dermatitis. Lancet 2020;396:345-360 (PMID 32738956; doi:10.1016/S0140-6736(20)31286-1)
• The percentages and denominators of each item across the three Hong Kong ISAAC rounds: Lee SL, Lau YL, Wong WH, Tian LW. Int J Environ Res Public Health 2022;19:16503 (PMID 36554390), https://pmc.ncbi.nlm.nih.gov/articles/PMC9779471/ (retrieved 2 August 2026)
• The Chinese ISAAC questionnaire's sensitivity of 23.5%–70.6%, and the low reported Chinese prevalence being partly a product of the translated questionnaire: Chan HH, Pei A, Van Krevel C, Wong GW, Lai CK. Clin Exp Allergy 2001;31:903-7 (PMID 11422155)
• The denominators of the 2021 schoolchildren allergy survey, the 41.6% and 16.3%, the difference in wording between the Chinese and English versions, and the sampling note: University of Hong Kong press release on the results of a study of allergic disease prevalence among Hong Kong primary and secondary school children, 13 January 2022, Chinese https://www.hku.hk/press/press-releases/detail/c_23934.html ; English https://www.hku.hk/press/news_detail_23934.html ; appendix (English only) http://www.hku.hk/f/upload/23935/Appendix_eng.pdf (retrieved 2 August 2026)
• Hong Kong lacking complete epidemiological data on childhood and adult AD, and the international adult range of 0.3%–14%: ClinicalTrials.gov NCT04154839 (record last updated 5 May 2022), https://clinicaltrials.gov/study/NCT04154839 (retrieved 2 August 2026)
• The Centre for Health Protection's non-communicable disease framework not covering eczema: Centre for Health Protection, Non-Communicable Diseases Watch, April 2024, "Burden of Non-communicable Diseases: An Update" (retrieved 2 August 2026)
• 129 million people worldwide with atopic dermatitis and the change in age-standardised prevalence: GBD 2021 Asthma and Allergic Diseases Collaborators. Lancet Respir Med 2025;13:425-446 (PMID 40147466); and the heaviest disease burden among skin diseases: Laughter MR, Maymone MBC, Mashayekhi S, et al. Br J Dermatol 2021;184:304-309 (PMID 33006135)
• The mental health correlations among 432 children and 380 parents at one Hong Kong paediatric dermatology clinic: Lam PH, Hon KL, Loo S, et al. Hong Kong Med J 2024;30:362-70 (doi:10.12809/hkmj219738), https://www.hkmj.org/abstracts/v30n5/362.htm (retrieved 2 August 2026)
• The recommendation of an unrestricted diet during pregnancy and lactation, breastfeeding for the first 6 months, and hydrolysed formula and atopic eczema in high-risk infants (a prevention guideline, not a treatment guideline): Chan AWM, Chan JKC, Tam AYC, Leung TF, Lee TH. Guidelines for allergy prevention in Hong Kong. Hong Kong Med J 2016;22:279-85 (doi:10.12809/hkmj154763), https://www.allergy.org.hk/publications/Guidelines%20for%20Allergy%20Prevention%20in%20Hong%20Kong%20(Published%20version%20on%20HKMJ).pdf (retrieved 2 August 2026)
• About $7,500 to $10,000 per injection and about $15,000 to $20,000 a month (a news report, not an official price list): Hong Kong Economic Journal health column, 30 July 2025, health.hkej.com article 4149165 (retrieved 2 August 2026)
• The closest publicly available material to Hong Kong expert opinion being a conference report: Hong Kong Dermatology Symposium 2024 conference report, Hong Kong J Dermatol Venereol 2025;32:81-83, https://medcomhk.com/hkdvb/details.asp?id=1023&show=1234 (retrieved 2 August 2026)

---

#Further reading

• On this platform: GP vs specialist
• On this platform: Public waiting times too long? Private and PPP alternatives
• The Hong Kong College of Paediatricians' clinical guideline of 2013 on atopic dermatitis in children (the source of the potency and day limits by body site cited here): HK J Paediatr 2013;18:96-104 (PDF)
• The Hong Kong College of Paediatricians' review of 2020 (the source of the maintenance quantities, the 2021 fingertip unit version and the steroid phobia passage cited here): HK J Paediatr 2021;26:42-57 (PDF)
• Hong Kong Medical Journal — the mental health correlations among 432 children and 380 parents at one Hong Kong paediatric dermatology clinic (cited here): Hong Kong Med J 2024;30:362-70 abstract