Systemic Therapy for Paediatric Stevens–Johnson Syndrome and Toxic Epidermal Necrolysis: A Systematic Review of Current Evidence and Limitations

ABSTRACT

Aim

The role of systemic therapy in paediatric Stevens–Johnson Syndrome (SJS) and Toxic Epidermal Necrolysis (TEN) remains undefined. Management is primarily guided by observational data and variable practice patterns. This systematic review aims to integrate recent evidence and evaluate clinical outcomes associated with systemic therapies in paediatric SJS/TEN.

Methods

A systematic search of PubMed, Embase and Scopus was conducted from January 1990 to February 2025. Studies were eligible if they included patients (< 18 years) with SJS, SJS‐TEN overlap or TEN managed with supportive care alone or systemic therapy (corticosteroids, intravenous immunoglobulin (IVIg), cyclosporin, TNF‐α inhibitors or plasmapheresis). Studies without defined diagnoses or outcomes were excluded. Primary outcomes were mortality and length of hospital stay. Risk of bias was assessed using CARE guidelines for case reports and the Joanna‐Briggs checklist for observational studies. Studies providing individual‐level data were synthesised quantitatively, while aggregated data were described narratively.

Results

A total of 260 studies comprising 1608 paediatric patients met inclusion criteria. Most patients received systemic therapy, most commonly corticosteroids or IVIg. Systemic therapy was not associated with a shorter length of stay or reduced mortality rate compared to supportive care alone. Treatment with cyclosporin and TNF‐α inhibitors trended towards lower mortality rates, and for cyclosporin, shorter length of stay. However, these findings did not reach statistical significance.

Conclusions

Evidence for systemic therapy remains limited by study heterogeneity and small sample sizes. Although emerging targeted therapies show early promise, firm conclusions regarding efficacy cannot be drawn. This highlights the need for further collaborative research.

Key Points

1. Supportive care and early withdrawal of the culprit drug remain the only interventions consistently associated with improved outcomes in SJS/TEN.

2. Although cyclosporin and TNF‐α inhibitors showed signals of lower mortality and shorter length of stay in TEN, robust paediatric evidence remains limited, highlighting the need for prospective multi‐centre collaboration.

3. JAK inhibitors have demonstrated promising clinical signals in adult SJS/TEN, aligned with emerging insights into interferon‐driven JAK/STAT pathway activation. However, they have not been studied in paediatric populations yet and thus warrant careful evaluation in this setting.

1. Introduction

Stevens–Johnson Syndrome (SJS) and Toxic Epidermal Necrolysis (TEN) are life‐threatening cutaneous reactions characterised by the extensive detachment of skin and mucosa [1]. These conditions exist on a spectrum of severity, defined by the degree of body surface area (BSA) affected by epidermal detachment. Stevens–Johnson Syndrome involves < 10% BSA, SJS‐TEN overlap 10%–30%, and TEN affects > 30% BSA [2]. The disease is often associated with significant morbidity and long‐term complications which may be more severe in children than in adults [3, 4].

In the paediatric setting, the incidence of SJS, SJS‐TEN, and TEN is estimated to be 5.3, 0.8, and 0.4 cases per million children per year respectively, with mortality rates ranging from 0% to 3.5% in SJS, and 4.6% to 16% in TEN [5]. The most common triggers of SJS/TEN in the paediatric population are medications such as antibiotics, anti‐epileptics, and non‐steroidal anti‐inflammatories, as well as infections [6].

An important consideration when assessing triggers is the evolving classification of paediatric SJS/TEN and related mucocutaneous syndromes. Infection‐associated entities such as reactive infectious mucocutaneous eruptions (RIME) and mycoplasma‐induced rash and mucositis (MIRM) are characterised by prominent mucositis with limited skin involvement and a more favourable prognosis [7, 8, 9]. In contrast, drug‐induced epidermal necrolysis (DEN) has been proposed for medication‐triggered cases and is generally associated with more extensive epidermal detachment [10]. This distinction is particularly relevant in children, where infection‐associated presentations are more common and misclassification may have implications for management [7, 8]. However, contemporary guidelines and systematic evaluations demonstrate variability in diagnostic frameworks and a lack of universally accepted criteria to distinguish these entities within the SJS/TEN spectrum [7, 11]. Infection‐associated cases therefore represent a heterogenous group, and their classification remains an area of ongoing refinement.

Current guidelines emphasise multidisciplinary care in specialist centres and include prompt withdrawal of culprit drugs, early assessment of respiratory status, fluid and electrolyte replacement, nutritional support and meticulous wound care [7]. Stopping the culprit drug as early as possible is associated with improved prognosis, and each day of delay increases mortality risk [12]. Beyond this, acute immunomodulatory treatment is controversial as there is no convincing data demonstrating that systemic therapies are superior to supportive care alone [13]. Cyclosporin and TNF‐α inhibitors have shown promise among adults in randomised controlled trials and observational studies, but their role in children where infectious and non‐drug triggers are proportionally more common remains undefined [13, 14, 15, 16]. Emerging evidence in adult populations has also reported the use of JAK inhibitors, supported by spatial proteomic analyses [17, 18, 19, 20, 21, 22]. However, these data are limited to small, uncontrolled adult cohorts, and paediatric evidence is lacking. Therefore, the aim of this study is to integrate and examine the current evidence in the literature relating to the outcomes of systemic treatment in the paediatric population compared to supportive care alone.

2. Methods

2.1. Identification of Studies

A systematic search was conducted in February 2025 across PubMed, Embase and Scopus for articles published from January 1990 to February 2025. The search focused on paediatric patients (< 18 years), with SJS, SJS‐TEN overlap, or TEN (Supporting Information Appendix 1). Studies before 1990 were excluded due to the lack of a consensus definition for SJS/TEN [2]. Systemic treatment included corticosteroids, IVIg, cyclosporin, TNF‐α inhibitors and plasmapheresis. Supportive treatment encompassed fluid and electrolyte replacement, nutrition, antibiotics, and wound management. All patients receiving systemic therapy also received supportive treatment. Studies that reported either systemic or supportive treatment of paediatric SJS/TEN with key outcomes such as mortality and time to re‐epithelialisation were included. Exclusions included undefined diagnoses, predominantly mucosal presentations, and studies lacking outcome data. Reference lists of included studies were hand‐searched.

All search results were populated in Covidence, and duplicates were removed. Titles and abstracts of relevant studies were screened by two independent reviewers (ST and KRC). Full texts were reviewed by the same authors, with discrepancies resolved by consensus.

Risk of bias was assessed using the CARE (Case Report) guidelines for case reports [23]. The Joanna Briggs Institute (JBI) critical appraisal checklist was used for case series and observational studies [24]. Studies were scored according to the proportion of checklist items fulfilled (Supporting Information Appendix 2). Certainty of evidence and risk of reporting bias were not formally graded due to the absence of quantitative synthesis.

Studies providing individual‐level patient data were classified as Category 1 for quantitative analysis. Studies reporting aggregated outcomes were classified as Category 2 and synthesised qualitatively to provide broader epidemiological context. Given the exploratory nature and inclusion of heterogenous studies, this review was not registered and no protocol was prepared.

2.2. Data Extraction

Extracted data included study characteristics, baseline patient characteristics (age, gender, diagnosis, trigger, and SCORTEN), treatment type, and outcome measures (mortality, length of stay, acute complications, and cause of death).

2.3. Statistical Analysis

Analyses were performed in R (version 4.3.2, Boston, MA). Data were summarised using medians with interquartile ranges and counts with percentages. Meta‐regression was not performed due to limited comparable outcome data.

The primary analysis compared any systemic treatment with supportive care, stratified by diagnosis, age group (< 12 vs. ≥ 12 years), and disease severity (SCORTEN). A sensitivity analysis excluded cases with infection‐related triggers. Mortality was compared using Fisher's exact or chi‐squared tests, and length of stay was compared using Wilcoxon rank‐sum tests. Multivariable logistic regression adjusted for SCORTEN was used to assess treatment effect on mortality (α = 0.05).

3. Results

The search yielded 1107 unique studies (Figure 1). Two hundred and sixty studies reporting 1608 patients were included and divided into Category 1, where individual‐level data was available, and Category 2, which consisted of cohort studies with aggregated data. Most studies were judged to be at moderate risk of bias (Supporting Information Appendix 2).

FIGURE 1.

PRISMA flow diagram of study selection. A total of 1551 records were identified through database searches (Embase, PubMed, Scopus) and citation searching. After removal of 444 duplicates, 1107 studies were screened, of which 690 were excluded at title and abstract screening. A further 157 studies were excluded at full‐text review for predefined reasons. In total, 260 studies were included in the final review.

3.1. Category 1 Descriptive Statistics

Category 1 included 251 studies reporting 450 patients (182 case reports, 63 case series, 5 cohort studies, and 1 prospective open trial). Among 450 patients in Category 1, 159 patients had SJS, 41 patients had SJS‐TEN overlap, and 250 patients had TEN. Median age was 9 years (IQR 5–13 years), with 305 (67.8%) patients under 12 years of age, and 144 (32.0%) patients 12 years and above. Females made up 47.4% (n = 205/432) of the cohort. Overall, 15.3% (n = 69) patients were treated with supportive care alone, and the remaining 84.7% (n = 381) were treated with systemic therapy. Across all diagnostic subtypes, IVIg and systemic corticosteroids were the most common systemic treatments (Table 1). A breakdown of treatment type is detailed in (Supporting Information Table 1). The overall mortality rate was 8.4% (n = 38/450).

TABLE 1.

Mortality by treatment and diagnostic subtype.

Diagnosis	Treatment (n)	Deaths	Mortality %	OR (95% CI)	p
SJS	Cyclosporin (6)	1	16.7	4.20 (0.22–79.32)	0.39
	IVIg (65)	3	4.6	1.02 (0.10–10.31)	1.00
	Plasmapheresis (3)	0	0	2.05 (0.07–61.00)	1.00
	Steroid (106)	2	1.9	0.40 (0.03–4.66)	0.44
	Supportive (22)	1	4.5	1.0 (ref)	—
SJS–TEN	Cyclosporin (7)	0	0	0.60 (0.01–35.86)	1.00
	IVIg (23)	1	4.3	0.60 (0.02–17.22)	1.00
	Plasmapheresis (2)	0	0	1.80 (0.03–121.71)	1.00
	Steroid (20)	2	10.0	1.22 (0.05–30.03)	1.00
	TNF inhibitor (2)	0	0	1.80 (0.03–121.71)	1.00
	Supportive (4)	0	0	1.0 (ref)	—
TEN	Cyclosporin (18)	0	0	0.13 (0.01–2.43)	0.09
	IVIg (138)	16	11.6	0.67 (0.26–1.77)	0.59
	Plasmapheresis (19)	2	10.5	0.61 (0.11–3.23)	0.71
	Steroid (123)	18	14.6	0.88 (0.34–2.28)	0.99
	TNF inhibitor (22)	1	4.5	0.24 (0.03–2.13)	0.25
	Supportive (43)	7	16.3	1.0 (ref)	—

SCORTEN was reported in a limited subset of patients (n = 75). Patients receiving systemic therapy had a higher median SCORTEN (2.0 vs. 1.5); although this was not statistically significant.

3.1.1. Triggers

Triggers were predominantly medication‐related. Antibiotics were implicated in 28.9% (n = 130) of triggers, antiepileptic drugs (AEDs) in 27.6% (n = 124), and other medications in 28.4% (n = 128). Infections accounted for 6.9% (n = 31) of triggers.

3.1.2. Mortality

Across all diagnostic subtypes, mortality did not differ between systemic treatment and supportive‐only groups (7.9% vs. 11.6%, p = 0.431), including after exclusion of infection triggered cases (8.3% vs. 11.3%, p = 0.463) (Supporting Information Table 2). Age‐stratified analysis also demonstrated no significant difference in mortality between treatment groups (Supporting Information Table 3). In SJS and SJS‐TEN, deaths were infrequent (< 5%), precluding meaningful comparison. In TEN which comprised 84.2% (n = 32/38) of all deaths, mortality appeared lower with cyclosporin and TNF‐α inhibitors (Table 1). However, this difference was not statistically significant.

Among the 38 recorded fatalities, most occurred in patients with TEN (n = 32). The leading causes of death were TEN‐related complications (26.3%), sepsis (21.0%) and multiorgan or respiratory failure (26.3%) (Table 2). Mortality increased with SCORTEN (Supporting Information Table 4). In multivariable logistic regression, SCORTEN independently predicted mortality (OR 2.42 per point increase, p = 0.037), whereas treatment group did not (OR 0.83, p = 0.88).

TABLE 2.

Recorded causes of death by diagnostic subtype in paediatric SJS/TEN.

Diagnosis	Recorded cause of death	n
SJS	Sepsis	1
	Multiorgan failure	1
	Respiratory failure	1
	Not specified	1
SJS–TEN	Sepsis	1
	Not specified	1
TEN	Toxic epidermal necrolysis	10
	Sepsis	6
	Cardiac arrest	1
	Multiorgan failure	5
	Pulmonary embolus	1
	Respiratory failure	3
	Unknown	1
	Not specified	5

3.1.3. Length of Stay

Across all patients, the median length of stay was longer among patients treated with systemic therapy compared with supportive care alone (18.0 vs. 13.0 days, p = 0.009) (Supporting Information Tables 5 and 5a), including after excluding infection‐triggered cases (Supporting Information Table 6). This difference remained significant among older children (Supporting Information Table 7).

However, when systemic agents were analysed individually, differences in length of stay emerged. Patients on cyclosporin generally demonstrated the shortest median length of stay compared to both supportive treatment and other systemic therapies (Table 3), although sample sizes were insufficient for a meaningful statistical comparison. In contrast, patients treated with steroids consistently demonstrated the longest median length of stay across all disease severities (19.0, 19.0, 18.0 days for SJS, SJS‐TEN overlap and TEN respectively).

TABLE 3.

Length of stay and mortality by monotherapy and diagnostic subtype.

	Treatment a	Median LOS c	Lower quartile LOS	Upper quartile LOS	Mortality (%)	p b
SJS	Steroid (n = 58)	19.0	12.5	24.8	0 (0)	< 0.01
SJS	IVIg (n = 28)	13.0	10.0	15.0	1 (3.6)	0.05
SJS	Cyclosporin (n = 3)	8.0	6.0	10.0	0 (0)	0.57
SJS	Supportive (n = 22)	10.0	6.0	14.0	1 (4.5)	NA
SJS‐TEN	Steroid (n = 7)	19.0	17.3	21.5	1 (14.3)	0.34
SJS‐TEN	IVIg (n = 12)	17.0	12.0	20.0	0 (0)	0.64
SJS‐TEN	Cyclosporin (n = 2)	9.0	9.0	9.0	0 (0)	> 0.99
SJS‐TEN	Plasmapheresis (n = 1)	13.0	13.0	13.0	0 (0)	0.8
SJS‐TEN	Supportive (n = 4)	10.5	8.3	31.5	0 (0)	NA
TEN	Steroid (n = 43)	18.0	11.3	27.0	7 (16.3)	> 0.99
TEN	IVIg (n = 67)	14.4	12.0	19.0	7 (10.4)	0.53
TEN	Cyclosporin (n = 2)	14.0	11.5	16.5	0 (0)	0.62
TEN	Plasmapheresis (n = 4)	14.5	12.3	18.3	0 (0)	0.96
TEN	TNF (n = 5)	17.0	15.0	20.0	0 (0)	0.91
TEN	Supportive (n = 43)	15.0	10.0	24.0	7 (16.3)	NA

^a Treatment groups represent monotherapy only.

^b p‐values compare LOS between each treatment group and supportive care within the same diagnostic subgroup.

^c Length of stay was calculated among patients with available LOS data.

Acute complications were more commonly observed among patients with TEN, but not increased with systemic therapy (Supporting Information Table 8).

3.1.4. Time to Re‐Epithelialisation

Time to initial healing was reported in a limited subset of patients (n = 152). Median time to initial healing was shorter in patients who received systemic therapy compared to supportive care alone (4.0 vs. 5.5 days, p = 0.049). However, the supportive‐only group was small (n = 14). There was no significant difference in time to complete re‐epithelialisation between groups, which both had a median of 14.0 days (p = 0.050). Further information can be found in Supporting Information Table 9.

3.2. Category 2 Descriptive Statistics

We examined nine cohort studies that reported on the treatment outcomes of paediatric patients with SJS/TEN, encompassing a total of 1158 patients. Of these patients, 77.5% were represented in a large study by Antoon et al. [25]. Overall, the median age ranged from 8 to 10.8 years. Patients with SJS predominated in most studies, comprising 78.2% of all patients. Among patients who received systemic treatment, most were treated with either systemic corticosteroids or IVIg, or both.

Across all studies, treatment data was available for 1149 patients. Of these, 45.4% (n = 518) received supportive care only, 24.9% (n = 284) IVIg alone, 20.8% (n = 237) corticosteroids alone, 8.2% (n = 93) combination corticosteroids and IVIg, 0.3% (n = 3) were treated with a combination of a TNF‐α inhibitor and steroid, and 0.1% (n = 1) were treated with cyclosporin alone [3, 25, 26, 27, 28, 29, 30, 31, 32].

Reported mortality rates ranged between 0% and 0.13% for patients with SJS, and 0% and 15% for patients with SJS‐TEN overlap or TEN. Mortality data linked to specific treatments were available for 158 patients across six studies. Ten deaths were reported overall. Three occurred in patients with TEN who received corticosteroids, while treatment details were not recorded for the remaining seven cases (six with SJS‐TEN overlap or TEN, and one with SJS) [25, 27, 31, 32].

Across seven studies encompassing 1083 patients, reported length of stay ranged from 7.0 to 26.1 days, and was consistently longer in patients with TEN compared to those with SJS. Few studies reported the association between specific treatments and length of stay. In the studies that did, systemic treatment was not associated with a reduction in hospitalisation duration. Antoon et al. found that IVIg alone or combined with corticosteroids was associated with a longer length of stay, a finding similarly observed by Chatproedprai et al. [25, 26]. In contrast, Quirke et al. found no significant association between the use of systemic therapy and length of stay [3].

4. Discussion

The treatment of paediatric SJS/TEN remains an area of ongoing investigation. Early recognition, prompt withdrawal of culprit drugs and management of underlying triggers are the only interventions consistently supported by robust evidence [12]. This is the first paediatric‐specific analysis for over a decade [33]. In the intervening period, clinical practice has evolved, with increasing use of immunomodulatory treatments, including TNF‐α inhibitors and JAK inhibitors [18]. This study synthesises current evidence on systemic therapy in children to inform decision‐making in a field where practice varies.

The key finding showed no significant difference in mortality or length of stay between patients receiving systemic therapy and those managed with supportive care alone. This can firstly be explained by treatment heterogeneity across studies, characterised by differences in treatment selection, dosing, and differing combinations of treatments, likely introducing confounding factors that obscured treatment effect. Secondly, patients with more severe disease were more likely to receive systemic therapy, which introduces confounding by indication, potentially masking any true treatment benefit. Consistent with this, length of stay was generally longer in the systemic treatment group, particularly among those who received IVIg or corticosteroids, likely reflecting baseline severity rather than reduced efficacy. Age‐stratified analyses suggested potential differences in length of stay between younger and older children, but small subgroup sizes warrant caution. Furthermore, mortality and length of stay are influenced by multiple factors, such as differences in supportive care practices, timing of treatment initiation, co‐morbidities and varying institutional practices. Therefore, these endpoints may not reflect treatment efficacy alone, and comparisons across studies should be made with caution. Alternative outcomes such as time to re‐epithelialisation may better reflect therapeutic response, but inconsistent reporting limited firm conclusions. In this context, where the therapeutic value of systemic therapy remains equivocal, early recognition and rigorous supportive care remain the most critical determinants of outcome.

When individual therapies were examined, cyclosporin and TNF‐α inhibitors were associated with lower mortality in TEN. However, these findings were derived from small sample sizes and should be interpreted with caution. Similar trends appear in predominantly adult cohorts. Meta‐analyses from 2017 to 2021 which included both adult and paediatric populations suggested a potential survival benefit with cyclosporin, but were constrained by low patient numbers [34, 35]. Consistent with this, a Cochrane review that evaluated systemic therapies for SJS/TEN among adults and children found minimal prospective data on cyclosporin that was limited to a single small cohort study [13]. The Cochrane review also reported a potential survival benefit of etanercept over corticosteroids, although this finding was graded as low‐certainty evidence [13]. More recently, JAK inhibitors represent an emerging therapeutic strategy, supported by growing evidence of their safety and efficacy in adult SJS/TEN. A study published in 2024 by Nordmann et al. utilised deep visual proteomics to demonstrate up‐regulation of type I and type II interferon signalling, and aberrant activation of the JAK/STAT pathway in TEN [18]. These findings were corroborated by in vitro systems, mouse models, and a clinical series of seven adult patients with SJS‐TEN overlap or TEN, in whom JAK inhibitors were associated with a rapid clinical improvement [18]. Since this publication, multiple adult case reports have demonstrated the rapid efficacy and safety of JAK inhibitors in SJS/TEN [17, 19, 20, 21, 22]. However, JAK inhibitors have not been systematically studied in paediatric SJS/TEN. Only isolated reports exist, reporting improvement with treatment of TEN‐like graft versus host disease with JAK inhibitors [36]. However, no published case reports describing the use of JAK inhibitors in paediatric SJS/TEN exist, limiting direct conclusions.

This study addresses a longstanding evidence gap. The most comparable prior systematic review, published in 2011, included only 128 paediatric patients and focused on corticosteroids, IVIg, and supportive care [33]. Since then, therapeutic approaches have evolved with the introduction of agents such as cyclosporin and TNF‐α inhibitors, which are included in the present analysis. More recently, a Cochrane review evaluated systemic therapies across adult and paediatric populations, but was restricted to higher‐level study designs, resulting in the exclusion of several paediatric studies [13]. While methodologically rigorous, this approach has inherent limitations in the context of sparse paediatric data. Thus, by incorporating both cohort studies and case studies, the present study captures the largest and most contemporary dataset to date, including newer therapies that have not previously been systematically evaluated in children.

However, this study has several important limitations. Firstly, the published literature is constrained by a lack of high‐quality data and heterogenous study designs and treatment protocols. Diagnostic inconsistency, particularly misclassification of infection‐triggered eruptions such as RIME as SJS/TEN, is particularly relevant in paediatric populations where such presentations are common and may dilute treatment effect [8, 9]. This is especially relevant given the proposed unifying diagnosis of DEN for medication‐triggered cases [10]. Consequently, earlier studies likely encompassed a broader spectrum of mucocutaneous disorders under SJS/TEN, limiting the ability to draw firm conclusions regarding treatment efficacy. Secondly, publication bias likely resulted in an over‐representation of severe disease. In this study, 55% of individual‐level cases were TEN, despite its lower population incidence compared to SJS [5]. Larger cohort studies were more representative of the disease epidemiology. This illustrates that severe presentations are more likely to be reported, while milder cases may be under‐represented. Such bias may skew observed treatment patterns and contribute to longer length of stay among patients receiving systemic therapy, reflecting higher confounding by indication rather than treatment effect. Inconsistent reporting of SCORTEN and lack of paediatric‐specific severity measures further limited adjustment for baseline severity.

In summary, this study provides the first paediatric‐specific review of systemic treatments for SJS/TEN, incorporating newer agents. Despite evolving practice, evidence supporting their efficacy remains limited.

From a clinical perspective, early recognition, prompt drug withdrawal, and meticulous supportive care remain the most evidence‐based interventions. Although treatments such as cyclosporin and TNF‐α inhibitors show promise, paediatric data is sparse and of low certainty, preventing firm conclusions regarding their efficacy and safety. Treatment decisions should therefore be individualised, taking into account disease severity, co‐morbidities, and institutional expertise.

From a research perspective, advances in understanding SJS/TEN pathogenesis have enabled the identification of key pathways, providing a basis for the development of targeted therapies. Recent analyses have highlighted a central role for cytotoxic CD8⁺ T‐cell‐keratinocyte interactions, characterised by dysregulation of inhibitory immune checkpoint signalling and interferon‐driven JAK/STAT signalling [18, 37]. Accordingly, several therapeutic strategies are under investigation. This includes Fas ligand blockade with PC111, which has demonstrated efficacy in mouse models, and daratumumab, which targets pathogenic T‐cell populations and will be investigated in early phase clinical trials [38, 39]. Among emerging therapies, JAK inhibitors have generated particular interest due to recent pathophysiological and early clinical evidence in adult SJS/TEN [18, 36, 37]. However, robust paediatric data remain lacking, and the safety, optimal dosing, and timing of such agents in children warrants careful evaluation.

Given the rarity and severity of paediatric SJS/TEN, there is a clear and pressing need for collaborative, multi‐centre approaches. Although randomised controlled trials would provide the highest level of evidence, their feasibility is limited due to the condition's low incidence and severity. Thus, prospective multi‐centre registries with standardised severity reporting and consistent outcome measures are likely to be more practical and would enable the systematic evaluation of emerging targeted therapies, including TNF‐α inhibitors and JAK inhibitors in real‐world paediatric populations.

Funding

The authors have nothing to report.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Supporting Information: Table 1. Overall and monotherapy treatment distribution.

Supporting Information: Table 2. Mortality rates among patients treated with systemic treatment vs. supportive care alone with non‐infectious triggers.

Supporting Information: Table 3. Mortality and treatment group by age group.

Supporting Information: Table 4. Mortality by SCORTEN group and treatment.

Supporting Information: Table 5. Length of stay by diagnostic subtype.

Supporting Information: Table 5a. Length of stay (LOS) by treatment group.

Supporting Information: Table 6. Length of stay among patients treated with systemic treatment vs. supportive care alone with non‐infectious triggers.

Supporting Information: Table 7. Length of stay by age group and treatment.

Supporting Information: Table 8. Comparison of complication rates between systemic and supportive treatment in paediatric SJS/TEN.

Supporting Information: Table 9. Time to healing outcomes by treatment group in paediatric SJS/TEN.

Supporting Information: Appendix 1. Search strategy used for the systematic review across PubMed, Embase, and Scopus databases.

Supporting Information: Appendix 2. Quality appraisal of included case reports and case series using CARE and JBI assessment tools.

Data S1: PRISMA 2020 Checklist.

Data Availability Statement

Data collection forms, extracted data, analytic code, and other materials are not publicly available.