Infection
Synthesised from Infected Eczema, Staphylococcus aureus, MRSA, Impetigo, Eczema Herpeticum, Herpes Simplex Virus, Aciclovir, Flucloxacillin, Clarithromycin, Erythromycin, Tetracyclines, Fusidic Acid, Fucidin H, Mupirocin, Hydrogen Peroxide Cream, Antimicrobial Resistance, Decolonisation, Bleach Baths, Skin Microbiome, Bacteriotherapy, Roseomonas mucosa, Molluscum Dermatitis, NICE NG153 and NICE NG190 (see MOC-infection). Corresponds to infranodus cluster 4 (
nice,infection,herpes,aureus,bleach).
Overview
Infection is the main acute complication of Atopic Dermatitis in toddlers. Several factors make eczema skin easy to infect: a leaky barrier (Skin Barrier Dysfunction), constant scratching (Itch-Scratch Cycle) and a shifted Skin Microbiome. Staphylococcus aureus lives on about 70% of eczema lesions (Totté 2016), and its share of skin bacteria rises to about 65% in untreated flares (Kong 2012). The central clinical point, stressed by NICE NG190, is that colonised is not infected. Weepy, crusted eczema is often not bacterial, and in a well child antibiotics add little to Topical Corticosteroids and emollients.
Two infections need to be recognised. Bacterial impetiginisation (Infected Eczema, Impetigo) shows as golden crusts, pus spots and weeping, and is usually managed in primary care. Eczema Herpeticum is caused by Herpes Simplex Virus, often during a child’s first infection, picked up from a kiss. It shows as uniform painful blisters and punched-out erosions, often with fever, and is an emergency needing same-day Aciclovir. Each day of delay lengthened hospital stay in a 1,331-child cohort. The two can coexist: 30% of children hospitalised with eczema herpeticum also had staph.
Stewardship shapes the treatment ladder. Children with eczema carry more resistant staph (Antimicrobial Resistance), and resistance tracks prior use of Fusidic Acid and Mupirocin. NICE NG153 therefore puts antiseptic Hydrogen Peroxide Cream first for small impetigo patches, and NG190 limits fusidic acid to 5–7 days. Anti-staph add-ons such as Bleach Baths, Decolonisation and experimental Bacteriotherapy have weak or contested evidence. The Cochrane review (George 2019; 41 RCTs) found insufficient evidence that reducing S. aureus improves eczema.
Key Entities
Infected Eczema
evidence: moderate. NG190 lists the signs of bacterial infection: weeping, pustules, crusts, rapid worsening, failure to respond to treatment, fever and malaise. A child who is not systemically unwell usually gets no antibiotic and continues emollient and steroid. If one is needed, fusidic acid 2% for 5–7 days is used for small patches, oral Flucloxacillin for widespread infection or an unwell child, and Clarithromycin for penicillin allergy. Old emollient tubs should be replaced after infection. Any “infected eczema” that does not improve on antibiotics should be re-examined for herpes. Red flags include Cellulitis and spreading redness with fever.
Emergency red flags
Rapidly worsening painful eczema, clusters of identical small blisters or punched-out sores, fever or lethargy, or any spots near the eyes. Get same-day medical care (NHS: urgent GP or 111). Untreated eczema herpeticum can spread to the eyes, brain and other organs.
Staphylococcus aureus and MRSA
evidence: weak for anti-staph treatment benefit. People with AD are about 20 times more likely than controls to carry S. aureus on lesional skin. The nose is a reservoir for recolonisation. Toxin-producing strains (enterotoxin B, TSST-1) fuel itch and Th2 inflammation even without visible infection. MRSA makes up a 4–13 times larger share of AD isolates than in healthy people, especially in children, although one San Diego report found the opposite, an unresolved conflict. In a New York clinic, 68.2% of MRSA isolates from children with AD were also mupirocin-resistant. MRSA needs swab-guided treatment and microbiologist advice. Oral Clindamycin is one outpatient option for minor MRSA infection.
Impetigo and NICE NG153
evidence: strong. Impetigo is a contagious staph or strep infection with “cornflake” golden crusts, usually around the nose, mouth and hands. On eczema it is described as impetiginised. NG153 (2020) recommends hydrogen peroxide 1% for 5 days first. If that is unsuitable or fails, fusidic acid follows, then mupirocin if fusidic resistance is suspected. Oral flucloxacillin or clarithromycin is for widespread or bullous impetigo or an unwell child, and topical and oral antibiotics should not be combined. The NHS says impetigo is non-contagious 48 hours after starting treatment. UKHSA counts only antibiotics, so a toddler treated with peroxide alone faces an unresolved conflict over returning to nursery.
Hydrogen Peroxide Cream
evidence: moderate. Crystacide 1% kills both Gram-positive and Gram-negative bacteria, and no resistance is known, which is the stewardship case for using it first. NHS Pharmacy First allows pharmacists to supply it from age 1. However, the protocol excludes children with uncontrolled eczema, recurrent or extensive impetigo, and eczema herpeticum, so a toddler with flaring eczema should see the GP. It contains salicylic acid and propylene glycol, both possible irritants on inflamed skin, and is not an eczema treatment.
Eczema Herpeticum, Herpes Simplex Virus and Aciclovir
Eczema herpeticum (Kaposi varicelliform eruption) usually occurs 5–12 days after contact with someone who may have no visible cold sore. Early-onset, extensive or head-and-neck eczema and high IgE raise the risk. A toddler whose eczema usually itches but is now sore is a key clue. NICE QS44 says to start oral or IV aciclovir immediately on suspicion and get same-day specialist advice. Hospital stay rose 41% if aciclovir started on day 3 and 98% if it started on days 4–7 (Aronson 2011). Sources disagree on whether every suspected case needs admission (CKS) or can be managed with oral treatment and same-day advice (QS44, DermNet). No RCT compares the oral and IV routes. To prevent infection, carers with a cold sore or its tingle should not kiss the child’s face or broken skin. The first HSV infection in a toddler is often Herpetic Gingivostomatitis, and kissing a baby with a cold sore risks Neonatal Herpes.
Fusidic Acid, Fucidin H and Mupirocin
evidence: weak. Fusidic acid is NG190’s only named topical antibiotic, limited to 5–7 days and localised infection. Resistance was about 2.5 times more frequent in staph from UK and Irish children with AD. It came from repeated fusA mutations that point to prior exposure, much of it from steroid–fusidic combinations. Fucidin H (fusidic acid plus 1% hydrocortisone) has a leaflet course of “up to 2 weeks”, which conflicts with NICE’s 5–7 days. Cochrane found combination creams give at most a slight short-term gain (RR 1.10), mostly from the steroid. A tube kept at home “for flares” is exactly the repeated exposure NICE warns against. Mupirocin is the third-line impetigo option and the usual nasal decolonisation agent. Prior mupirocin use was the only independent risk factor for resistance (OR 19–27), and New Zealand moved it back to prescription-only after resistance rose during its over-the-counter years.
Oral antibiotics: Flucloxacillin, Clarithromycin, Erythromycin, Tetracyclines
Flucloxacillin is dosed 125–250 mg four times daily for ages 2–9. Its bitter liquid and four before-food doses are hard to manage around nursery. Clarithromycin (twice daily, weight-banded) is the NICE alternative if the child is allergic or refuses the flucloxacillin liquid, though macrolide resistance among staph is high. Cochrane found no clear benefit of oral antibiotics over placebo in clinically infected childhood eczema (low quality), but the trials largely did not test unwell children, which is an unresolved conflict with NG190. Erythromycin is NICE’s choice for infection only in pregnancy. Its toddler role is oral or topical treatment of Perioral Dermatitis, where it cleared 86% of a 106-child series by 4 weeks. Tetracyclines are avoided under about 8–12 years because of tooth staining. Short doxycycline courses look safe, but no source sets a safe skin-disease course length for under-8s.
Antimicrobial Resistance and Decolonisation
Resistance evidence comes from observational surveillance, not RCTs. The stewardship message for parents is: no leftover antibiotic tubes, no antibiotic cream for flares without signs of infection, finish the course, and control the eczema. Decolonisation (nasal mupirocin or fusidic acid plus antiseptic washes, sometimes swabbing household members) is reserved for frequently recurring infection after swabs. Cochrane rates its eczema benefit as unproven (evidence: weak), and resistance can defeat it, so susceptibility testing should come first.
Bleach Baths
evidence: contested. Dilute sodium hypochlorite (about 0.005%, roughly swimming-pool strength) is used 2–3 times a week for moderate-to-severe AD. Chopra 2017 found bleach baths no better than plain water baths (“the water baths appear to be doing most of the heavy lifting”), and Cochrane 2019 agreed (low quality). Bakaa 2022 (10 RCTs, GRADE) found a probable 22% relative EASI improvement, with about 1 in 10 patients improving by 50%. The National Eczema Association supports them. They are a doctor-recommended Track B measure carried out at home.
Skin Microbiome, Bacteriotherapy and Roseomonas mucosa
In flares, bacterial diversity falls and S. aureus rises. Diversity recovers with even intermittent treatment, before the eczema itself improves. Protective Coagulase-Negative Staphylococci (S. hominis, S. epidermidis) that kill S. aureus are rarer on eczema skin. Whether S. epidermidis rises or stays flat during flares is an unresolved conflict. Bacteriotherapy tries to restore these commensals. S. hominis A9 reduced staph in adults but did not improve severity overall. R. mucosa looked impressive open-label (4 of 5 children improved by more than 50% on SCORAD), but its FB-401 RCT in 154 patients aged 2 and over failed: 58% vs 60% reached EASI-50 on product vs placebo. No product is licensed. evidence: weak.
Molluscum Dermatitis
Molluscum contagiosum is a poxvirus infection of young children: pearly, dimpled papules in the folds and on the trunk, spread by scratching. In children with AD it is more numerous and lasts longer, possibly linked to Filaggrin mutations, although whether the association is real is disputed. The eczema patch around the spots is often misdiagnosed as a plain AD flare. Inflamed red spots (the BOTE sign) or a distant id eruption usually mean the infection is clearing.
Connections Within This Community
Staphylococcus aureus [causes] Infected Eczema Staphylococcus aureus [causes] Impetigo Herpes Simplex Virus [causes] Eczema Herpeticum Aciclovir [treats] Eczema Herpeticum NICE NG190 [regulates] Flucloxacillin NICE NG190 [regulates] Fusidic Acid NICE NG153 [regulates] Hydrogen Peroxide Cream Hydrogen Peroxide Cream [treats] Impetigo Clarithromycin [relates] Flucloxacillin Fucidin H [part-of] Fusidic Acid Fusidic Acid [causes] Antimicrobial Resistance Mupirocin [causes] Antimicrobial Resistance MRSA [part-of] Staphylococcus aureus Decolonisation [treats] MRSA Bleach Baths [opposes] Staphylococcus aureus Roseomonas mucosa [part-of] Bacteriotherapy Staphylococcus aureus [part-of] Skin Microbiome Erythromycin [relates] Tetracyclines
Bridges to Other Communities
- MOC-disease-and-mechanisms: Skin Barrier Dysfunction, raised Skin pH, low Filaggrin and the Th2 Immune Response favour staph and herpes. The Skin Microbiome belongs to both communities.
- MOC-prescription-anti-inflammatory-therapies: NG190’s core message is to keep Topical Corticosteroids going during infection. Fucidin H combines both approaches, TCIs are avoided on infected skin, and Systemic Immunosuppressants were linked to eczema herpeticum admissions in the TREAT Trial. Erythromycin and Tetracyclines treat steroid-induced Perioral Dermatitis.
- MOC-emollients-and-skin-care: contaminated emollient tubs, BEEP’s excess skin infections (BEEP Trial) and soak practice link to Bleach Baths. The infranodus bridge candidate
soakconnects these clusters. - MOC-diagnosis-and-differentials: impetigo, herpeticum and molluscum all mimic a flare. When to Refer sets the same-day herpeticum rule, and Scabies is another contagious mimic.
- MOC-home-and-herbal-remedies: Coconut Oil (lauric acid) and Manuka Honey are promoted as anti-staph remedies. Probiotics are distinct from topical bacteriotherapy.
- MOC-evidence-base-and-guidance: NICE NG153, NICE NG190, NICE CG57 and Cochrane (George 2019) are the main sources. Infranodus Gaps 1–2 (
cochrane↔nice,definition↔nice) mark a weak link between the Cochrane anti-staph evidence and NICE’s prescribing rules.
Open Questions
- eczema-infection-questions
- eczema-infection-conflicts
- infection-treatment-questions
- infection-treatment-conflicts
- antimicrobials-questions
- antimicrobials-conflicts
- water-therapies-questions
- water-therapies-conflicts
- Gap-derived (cochrane ↔ nice): Cochrane found no clear benefit of oral antibiotics in clinically infected childhood eczema, yet NICE NG190 makes flucloxacillin first-line for unwell children. Is there any trial evidence specific to systemically unwell 1–4-year-olds, or is that recommendation consensus-only?
- Gap-derived (definition ↔ nice): Is there a validated clinical definition that separates “colonised” from “infected” toddler eczema well enough for parents and pharmacists to apply the Pharmacy First exclusions consistently?