Interventions for hand eczema

Wietske Andrea Christoffers; Pieter‐Jan Coenraads; Åke Svensson; Thomas L Diepgen; Janine L Dickinson‐Blok; Jun Xia; Hywel C Williams

doi:10.1002/14651858.CD004055.pub2

. 2019 Apr 26;2019(4):CD004055. doi: 10.1002/14651858.CD004055.pub2

Interventions for hand eczema

Wietske Andrea Christoffers ^1,^✉, Pieter‐Jan Coenraads ¹, Åke Svensson ², Thomas L Diepgen ³, Janine L Dickinson‐Blok ⁴, Jun Xia ⁵, Hywel C Williams ⁶

Editor: Cochrane Skin Group

PMCID: PMC6484375 PMID: 31025714

Abstract

Background

Hand eczema is an inflammation of the skin of the hands that tends to run a chronic, relapsing course. This common condition is often associated with itch, social stigma, and impairment in employment. Many different interventions of unknown effectiveness are used to treat hand eczema.

Objectives

To assess the effects of topical and systemic interventions for hand eczema in adults and children.

Search methods

We searched the following up to April 2018: Cochrane Skin Group Specialised Register, CENTRAL, MEDLINE, Embase, AMED, LILACS, GREAT, and four trials registries. We checked the reference lists of included studies for further references to relevant trials.

Selection criteria

We included randomised controlled trials (RCTs) that compared interventions for hand eczema, regardless of hand eczema type and other affected sites, versus no treatment, placebo, vehicle, or active treatments.

Data collection and analysis

We used standard methodological procedures expected by Cochrane. Primary outcomes were participant‐ and investigator‐rated good/excellent control of symptoms, and adverse events.

Main results

We included 60 RCTs, conducted in secondary care (5469 participants with mild to severe chronic hand eczema). Most participants were over 18 years old. The duration of treatment was short, generally up to four months. Only 24 studies included a follow‐up period. Clinical heterogeneity in treatments and outcome measures was evident. Few studies performed head‐to‐head comparisons of different interventions. Risk of bias varied considerably, with only five studies at low risk in all domains. Twenty‐two studies were industry‐funded.

Eighteen trials studied topical corticosteroids or calcineurin inhibitors; 10 studies, phototherapy; three studies, systemic immunosuppressives; and five studies, oral retinoids. Most studies compared an active intervention against no treatment, variants of the same medication, or placebo (or vehicle). Below, we present results from the main comparisons.

Corticosteroid creams/ointments: when assessed 15 days after the start of treatment, clobetasol propionate 0.05% foam probably improves participant‐rated control of symptoms compared to vehicle (risk ratio (RR) 2.32, 95% confidence interval (CI) 1.38 to 3.91; number needed to treat for an additional beneficial outcome (NNTB) 3, 95% CI 2 to 8; 1 study, 125 participants); the effect of clobetasol compared to vehicle for investigator‐rated improvement is less clear (RR 1.43, 95% CI 0.86 to 2.40). More participants had at least one adverse event with clobetasol (11/62 versus 5/63; RR 2.24, 95% CI 0.82 to 6.06), including application site burning/pruritus. This evidence was rated as moderate certainty.

When assessed 36 weeks after the start of treatment, mometasone furoate cream used thrice weekly may slightly improve investigator‐rated symptom control compared to twice weekly (RR 1.23, 95% CI 0.94 to 1.61; 1 study, 72 participants) after remission is reached. Participant‐rated symptoms were not measured. Some mild atrophy was reported in both groups (RR 1.76, 95% CI 0.45 to 6.83; 5/35 versus 3/37). This evidence was rated as low certainty.

Irradiation with ultraviolet (UV) light: local combination ultraviolet light therapy (PUVA) may lead to improvement in investigator‐rated symptom control when compared to local narrow‐band UVB after 12 weeks of treatment (RR 0.50, 95% CI 0.22 to 1.16; 1 study, 60 participants). However, the 95% CI indicates that PUVA might make little or no difference. Participant‐rated symptoms were not measured. Adverse events (mainly erythema) were reported by 9/30 participants in the narrow‐band UVB group versus none in the PUVA group. This evidence was rated as moderate certainty.

Topical calcineurin inhibitors: tacrolimus 0.1% over two weeks probably improves investigator‐rated symptom control measured after three weeks compared to vehicle (14/14 tacrolimus versus 0/14 vehicle; 1 study). Participant‐rated symptoms were not measured. Four of 14 people in the tacrolimus group versus zero in the vehicle group had well‐tolerated application site burning/itching.

A within‐participant study in 16 participants compared 0.1% tacrolimus to 0.1% mometasone furoate but did not measure investigator‐ or participant‐rated symptoms. Both treatments were well tolerated when assessed at two weeks during four weeks of treatment.

Evidence from these studies was rated as moderate certainty.

Oral interventions: oral cyclosporin 3 mg/kg/d probably slightly improves investigator‐rated (RR 1.88, 95% CI 0.88 to 3.99; 1 study, 34 participants) or participant‐rated (RR 1.25, 95% CI 0.69 to 2.27) control of symptoms compared to topical betamethasone dipropionate 0.05% after six weeks of treatment. The risk of adverse events such as dizziness was similar between groups (up to 36 weeks; RR 1.22, 95% CI 0.80 to 1.86, n = 55; 15/27 betamethasone versus 19/28 cyclosporin). The evidence was rated as moderate certainty.

Alitretinoin 10 mg improves investigator‐rated symptom control compared with placebo (RR 1.58, 95% CI 1.20 to 2.07; NNTB 11, 95% CI 6.3 to 26.5; 2 studies, n = 781) and alitretinoin 30 mg also improves this outcome compared with placebo (RR 2.75, 95% CI 2.20 to 3.43; NNTB 4, 95% CI 3 to 5; 2 studies, n = 1210). Similar results were found for participant‐rated symptom control: alitretinoin 10 mg RR 1.73 (95% CI 1.25 to 2.40) and 30 mg RR 2.75 (95% CI 2.18 to 3.48). Evidence was rated as high certainty. The number of adverse events (including headache) probably did not differ between alitretinoin 10 mg and placebo (RR 1.01, 95% CI 0.66 to 1.55; 1 study, n = 158; moderate‐certainty evidence), but the risk of headache increased with alitretinoin 30 mg (RR 3.43, 95% CI 2.45 to 4.81; 2 studies, n = 1210; high‐certainty evidence). Outcomes were assessed between 48 and 72 weeks.

Authors' conclusions

Most findings were from single studies with low precision, so they should be interpreted with caution. Topical corticosteroids and UV phototherapy were two of the major standard treatments, but evidence is insufficient to support one specific treatment over another. The effect of topical calcineurin inhibitors is not certain. Alitretinoin is more effective than placebo in controlling symptoms, but advantages over other treatments need evaluating.

Well‐designed and well‐reported, long‐term (more than three months), head‐to‐head studies comparing different treatments are needed. Consensus is required regarding the definition of hand eczema and its subtypes, and a standard severity scale should be established.

The main limitation was heterogeneity between studies. Small sample size impacted our ability to detect differences between treatments.

Plain language summary

Treatments for hand eczema

Review question

We reviewed evidence on the effects of topical and systemic (oral or injected medicines that work throughout the entire body) treatments for hand eczema when compared against placebo (an identical but inactive treatment), no treatment, vehicle (inactive ingredients that help deliver an active treatment), or another treatment. We included 60 randomised trials (5469 participants) published up to April 2018.

Background

Hand eczema is an inflammation of the skin of the hands that can be caused by contact allergens (i.e. substances that cause an allergic reaction) such as rubber chemicals, but other external factors (e.g. irritants such as water or detergents) and atopic predisposition are often important triggers. Hand eczema can cause a reduction in quality of life leading to many work‐related problems. Various types of hand eczema exist, and different topical (creams, ointments, or lotions) and systemic treatments with unknown effectiveness can be used.

Study characteristics

Most participants were hospital outpatients over 18 years of age with mild to severe chronic hand eczema. Treatment was usually given for up to four months, and outcomes were mainly assessed after treatment. A large variety of treatments were studied and compared to no treatment, variants of the same medication, placebo, or vehicle. Twenty‐two studies were funded by pharmaceutical companies.

Key results

Limited data are available to support the best way of managing hand eczema due to varying study quality and inability to pool data from studies with similar interventions. Corticosteroid creams/ointments and phototherapy (irradiation with UV light) are the major treatment options, although comparisons between these options are lacking. Below, we present results for the main comparisons of interest.

Corticosteroid creams/ointments: clobetasol propionate foam probably increases participant‐rated good/excellent control of hand eczema when compared to vehicle (516 versus 222 per 1000), but the difference between groups was less clear for investigator‐rated control, and more adverse events were reported with clobetasol propionate (178 versus 79 per 1000) (all based on moderate‐certainty evidence).

Mometasone furoate cream used thrice weekly may slightly improve investigator‐rated good/excellent control compared to twice weekly treatment, and participant‐rated control was not measured. Mild skin thinning occurred in both groups, but cases were few (all based on low‐certainty evidence).

Irradiation with UV light: various types of irradiation (i.e. exposure to radiation) were compared. Local PUVA may improve investigator‐rated good/excellent control compared to narrow‐band UVB (400 versus 200 per 1000); however, we are uncertain of this finding because results also show that local PUVA may make little or no difference. Participant‐rated symptoms were not measured. Nine out of 30 participants in the narrow‐band UVB group reported adverse events (mainly redness) compared to none in the PUVA group (all based on moderate‐certainty evidence).

Topical calcineurin inhibitors: people receiving tacrolimus are probably more likely to achieve improved investigator‐rated good/excellent symptom control compared to those given vehicle (14/14 participants with tacrolimus compared to none with vehicle), but participant‐rated control of symptoms was not measured. Four of 14 people in the tacrolimus group versus zero in the vehicle group had well‐tolerated application site burning/itching. One small study compared tacrolimus to mometasone furoate, which were both well tolerated, but did not measure investigator‐ or participant‐rated control (all based on moderate‐certainty evidence).

Oral interventions: oral immunosuppressant (a drug that hinders the immune response) cyclosporin probably slightly improves investigator‐ or participant‐rated control of good/excellent symptoms compared to topical betamethasone cream (a corticosteroid). The risk of adverse events such as dizziness was similar between groups (all based on moderate‐certainty evidence).

The oral vitamin A derivative (retinoid) alitretinoin (10 mg) achieved investigator‐rated good/excellent symptom control in 307 compared to 194 participants per 1000 with placebo, and alitretinoin 30 mg achieved investigator‐rated control in 432 compared to 157 participants per 1000 with placebo. Similar results were shown for participant‐rated control (high‐certainty evidence). When the dosage of alitretinoin was increased to 30 mg, risk of headache was higher compared to placebo (74 versus 251 per 1000; high‐certainty evidence), but this probably does not differ between alitretinoin 10 mg and placebo (based on moderate‐certainty evidence).

Quality of the evidence

The quality of evidence was mainly moderate, with most analyses based on single studies that had small sample sizes; therefore, some results should be interpreted with care.

Summary of findings

Background

Please note that unfamiliar terms may be listed in Appendix 1 ('Glossary of medical terms').

Future research would involve comparing different treatment groups. Focus on subgroups would provide reliable evidence for informed decisions about which treatment is effective in managing hand eczema.

The overall quality of evidence was very low. Many trials included in this review, particularly older ones, were of low quality with methodological weaknesses in design (small studies, short duration) or were biased (not blinded, sponsored by pharmaceuticals). Most included participants with chronic hand eczema in secondary care settings; some included only specific subtypes of hand eczema, thereby limiting direct application of study findings. Most analyses were based on single studies of small sample size and imprecise results.

Description of the condition

Definition and epidemiology

Hand eczema is an inflammation of the skin (dermatitis) that is confined to the hands. Hand eczema is a common condition with a point prevalence varying between 1% and 5% in the general population. When mild cases are included, one‐year prevalence can reach 10% (Meding 2004; Thyssen 2010; Yngveson 2000). Thyssen et al conducted a review of seven epidemiological studies on hand eczema and concluded the median incidence rate of hand eczema was 5.5 cases/1000 person‐years. For women, the incidence rate of hand eczema was 9.6 cases/1000 person‐years, and for men, 4.0 cases/1000 person‐years (Thyssen 2010). A possible explanation for this sex difference is greater exposure of women to wet work, such as cleaning, nursing, and hair dressing, for example (Mollerup 2014; Nilsson 1985). The incidence of notified (i.e. usually more severe) occupation‐related cases is estimated to be above 0.7 per 1000 people per year, with much higher incidences (up to 1 in 100) in high‐risk populations such as hairdressers (Diepgen 2003). Decreased prevalence has been observed in Swedish adults and was attributed to a decline in occupational exposure to irritants (Meding 2002).

Over the years, several authors have proposed a workable definition of hand eczema, whereby different subtypes have been recognised (Menné 2000). Hand eczema can be classified according to aetiological (causative) factors, clinical‐morphological typology, or a combination of both (Coenraads 2012; Diepgen 2009a). However, due to multi‐causality, it is difficult to assess the influence of each causative factor; therefore only one aetiological diagnosis might be insufficient. The Danish Contact Dermatitis Group developed a classification system based on morphology with clear definitions for each classification and one or more aetiological diagnoses (Menné 2011). This might facilitate the classification of hand eczema and was demonstrated to be a useful tool in general practice (Johansen 2011). However, there is an obvious need for international consensus regarding the classification of subgroups of hand eczema.

In the current literature, different names can be used for the same subgroups, or the same name can be used for different subgroups. An example of this is vesicular hand eczema (Veien 2009): this might be called pompholyx, dyshidrotic eczema, dyshidrosis, or vesicular eczema; no consensus has been reached regarding the definition. The original definition of 'pompholyx' states "an eruption of vesicles and bullae on the palms, which is accompanied by pain and severe itching". Fox 1873 hypothesised that pompholyx was caused by sweating of the palms and introduced the term 'dyshidrosis' (hydrosis from sweating); both terms were used for the same clinical vesicular type. Later, Kutzner 1986 demonstrated that sweat glands are not altered in vesicular hand eczema and discussed the histological features of eczema. However, despite this evidence, the term 'dyshidrosis' is still used in current literature.

Hand eczema may be accompanied by similar skin changes on the feet.

Causes

In many people, hand eczema has more than one cause and both predisposing and external factors play a part. Being atopic (a tendency to develop asthma, hay fever, or eczema) is a major predisposing factor responsible for hand eczema; one‐third to one‐half of people with hand eczema can be considered atopic (Coenraads 1998; Meding 1990; Svensson 1988). The role of genetic factors, especially the association between filaggrin (FLG) mutations and hand eczema, is still under investigation (Heede 2016; Kaae 2012; Molin 2015).

The most common external causes of hand eczema include contact with mild toxic agents or irritants (for instance, water and soaps). The resulting irritant contact dermatitis can be distinguished from allergic contact dermatitis, which is caused by skin contact with allergens. Allergic contact dermatitis is less common than irritant contact dermatitis, and it occurs only in persons who have developed a contact allergy to a specific substance such as rubber, nickel, or perfumes. Ingested allergens (e.g. nickel) may occasionally provoke hand eczema (Jensen 2006). Little evidence suggests that inhalation of house dust mites may increase the severity of vesicular hand eczema (Schuttelaar 2013). The relevance of psychosomatic factors remains speculative (Menné 2000). In many people with chronic hand eczema, a combination of the above‐mentioned factors plays a role. In addition, for several types of hand eczema, the cause is still unknown.

Impact

Itch is common among those with hand eczema. The itch caused by hand eczema can be intense, leading to sleep loss in the sufferer and in other family members. A vicious cycle of symptoms causing skin damage can develop, the so‐called itch/scratch/itch cycle. Cracks and blisters can be painful. Cracking, hyperkeratosis (callus‐like thickening), and inflexibility of the hands are also problematic and may limit mobility of the hands.

A visible skin disease can be a great burden and can lead to a social stigma. The hands are important organs of communication and expression; therefore any visible skin disease on the hands may result in major psychosocial problems (e.g. anxiety, low self‐esteem, social phobia).

Painful cracks and blisters, besides their negative effects on daily life outside work, can impede an individual's ability to carry out manual work, leading to significant disability and huge economic losses for both individuals and society. A systematic review estimated the mean annual total cost per hand eczema patient at between €1712 and €9792 (Politiek 2016). Hand eczema accounts for an estimated 90% of occupational skin disease. Patients have substantial use of sick leave due to their hand eczema. Studies in patients with chronic severe hand eczema have reported job loss up to 20% (Cvetkovski 2005). Quality of life assessments have shown an impact on daily life and on employment (Agner 2008; Moberg 2009). A comparison between the generic quality of life instrument Short Form Health Survey (SF‐36) and the skin‐related Dermatology Life Quality Index (DLQI) revealed slightly higher impact of hand eczema on women compared to men for specific sub‐items (Wallenhammar 2004). A comparison of physician‐rated versus participant‐rated assessments of severity showed a poor correlation, indicating that patients may evaluate several aspects of their hand eczema (including degree of erythema, vesicles, and fissures) differently from physicians (van Coevorden 2006).

Prognosis

Previous studies have suggested that hand eczema tends to run a chronic relapsing course, with the vast majority of people experiencing negative psychosocial consequences (Hald 2009; Meding 2005; Petersen 2014; Veien 2008).

Description of the intervention

Many diverse therapies are used to control the disease, such as:

skin protection measures, including gloves;
topical treatments (bland emollients, corticosteroid creams/ointments, calcineurin inhibitors, coal tar and derivatives, irradiation with ultraviolet (UV) light or X‐rays); and
systemic treatments (oral corticosteroids, oral retinoids, or other immunosuppressants such as cyclosporin).

The main groups of interventions covered by this review are topical corticosteroids, topical calcineurin inhibitors (immunomodulators), irradiation with UV light, and oral retinoids or systemic immunosuppressants.

Overall, after proper education and counselling, including the recommendation of emollients, application of topical corticosteroids remains the mainstream treatment for hand eczema (nationaleczema.org).

How the intervention might work

Theoretically, identifying and eliminating an allergic contact factor (e.g. nickel or rubber allergy) could result in cure of hand eczema, provided this is the sole cause. In clinical practice, however, such cases are rare, as hand eczema is often due to a combination of irritant and allergic contact exposure, as well as to endogenous factors.

This review deals with a great variety of interventions. Major types of interventions are topical corticosteroids, topical immunomodulators, irradiation with UV light, and oral retinoids.

Topical corticosteroids are the most frequently prescribed treatments for hand eczema (Soost 2012). They have overlapping mechanisms of action: like oral immunosuppressants (e.g. corticosteroids), they inhibit inflammation (anti‐inflammatory) and production of inflammatory substances (immunosuppressive) (Ahluwalia 1998; Sakuma 2001; Schleimer 1993).

Topical immunomodulators, such as tacrolimus and pimecrolimus, are non‐steroidal immunosuppressants that are more selective in their mode of action than corticosteroids. They inhibit the production of inﬂammatory substances in the body (such as synthesis and release of inﬂammatory cytokines from T‐lymphocytes, and release of inﬂammatory mediators from mast cells). Calcineurin is present during activation of T‐lymphocytes, and since tacrolimus and pimecrolimus block this step, they are called 'calcineurin inhibitors' (de Paulis 1992; Sakuma 2001).

Topical moisturisers or emollients can relieve dryness of the skin, can improve the skin barrier function, and can influence transepidermal water loss (depending on the composition of the emollient) (Lodén 2012b; Rawlings 2004). Moisturisers are available in various compositions such as oil‐in‐water, water‐in‐oil, lotions, gels, and emulsions, among others, and various adjuvants such as urea or salicylic acid can be added to reduce thickness and scaling of the skin.

Coal tar has been used to treat eczema since ancient times. It is claimed to increase epidermal differentiation and to up‐regulate various key barrier proteins such as filaggrin, thus improving the skin barrier function (McLean 2013; van den Bogaard 2013). Moreover coal tar suppresses the Th2 cytokine response (McLean 2013; van den Bogaard 2013).

Irradiation with UV light can be performed with different types of UVA and UVB, depending on the wavelength. UVA treatment overall is combined with a topical or oral agent (psoralen) to make the skin more sensitive to UVA. Examples of different types of phototherapy include broad‐spectrum UVB (280 to 315 nm), small‐spectrum UVB (311 to 313 nm, also known as TL‐01 or narrow‐band UVB), UVA‐1 (340 to 400 nm), and topical and oral psoralen combined with UVA (PUVA; 315 to 400 nm). UVA‐1 phototherapy can be used at high (HD; 130 J/cm²), medium (MD; 50 J/cm²), and low doses (LD; 10 J/cm²) (Hönigsmann 2003). The mechanism of photo(chemo)therapy is multi‐factorial. In general, UV light locally decreases the activity of the immune system and inhibits the quantity of inflammatory cells. It suppresses the antigen‐presenting function of the Langerhans cells and induction of apoptosis of T‐cells (Majoie 2009). In addition, photo(chemo)therapy results in an increase in the amount of stratum corneum; in other words, the skin gets thicker (Jekler 1990). Finally, UVB reduces the number of microbes on the skin, including Staphylococcus aureus (Faergemann 1987).

Oral retinoids are vitamin A derivatives. Retinoids are thought to interfere at different steps in the inflammatory process. They have immunomodulatory properties and interfere with the epidermal differentiation process in various ways (Blair 2016; Kislat 2014; Schmitt‐Hoffmann 2012). Both alitretinoin and acitretin are retinoids, although their mechanism of action is slightly different. Alitretinoin is thought to have anti‐inflammatory and immunomodulatory effects on the skin. Alitretinoin binds with high affinity to both retinoic acid receptor (RAR) and retinoid X receptor (RXR) and presents anti‐inflammatory and immunomodulatory activity, and acitretin binds only selectively to RAR, although both retinoids are thought to reduce inflammation (Blair 2016; Kislat 2014; Schmitt‐Hoffmann 2012).

Hand eczema is a chronic condition that might be accompanied by flares and might improve as a result of the natural course; therefore, we believe a minimum treatment duration of three months is required to document important data such as duration and frequency of disease relapse.

Why it is important to do this review

The high prevalence of hand eczema, along with its poor prognosis and associated disability with economic losses and impairment of quality of life, makes hand eczema an important disease to study from an individual and a societal perspective. This, coupled with the long list of diverse treatments of unknown effectiveness and several conflicting studies (Diepgen 2007; van Coevorden 2004b), suggests that a systematic review is needed. Even if methodological constraints do not permit sufficient clarification of existing conflicts to provide clear guidance in clinical practice, this review will be an important step in identifying research gaps and consequently providing directions for future research.

The plans for this review were published as a protocol "Interventions for hand eczema" (van Coevorden 2009). Differences between the review and the protocol are stated in the section Differences between protocol and review.

Objectives

To assess the effects of topical and systemic interventions for hand eczema in adults and children.

Methods

Criteria for considering studies for this review

Types of studies

We included randomised controlled trials (RCTs) of interventions for hand eczema regardless of hand eczema type and other affected localisations.

Types of participants

People (adults and children, occupational and non‐occupational) with the diagnosis of hand eczema, regardless of the underlying assumed cause, were eligible. We also included participants with other parts of the body affected in addition to the hand. The terms 'eczema' and 'dermatitis' were acceptable whenever they referred to the hands. Other terms such as 'pompholyx', 'dyshidrosis', and 'pulpitis' were also deemed acceptable. We included participants with different types of hand eczema, for example, chronic hand eczema, hyperkeratotic palmar (also know as tylotic) hand eczema, and vesicular eczema (also known as dyshidrotic hand eczema or pompholyx).

We included in this review studies that included participants with other diagnoses besides hand eczema only when we were able to obtain separate data for hand eczema participants.

Types of interventions

We included only studies comparing the intervention versus no treatment, placebo, vehicle, or other active treatments. We considered all types of interventions, except interventions to prevent hand eczema (primary prevention). We excluded studies that focused on prevention of hand eczema and studies that investigated integrated care programmes or educational programmes (non‐pharmacological interventions).

We considered studies comparing different interventions, for example, topical corticosteroids versus topical calcineurin inhibitors or oral cyclosporin versus topical corticosteroids, as most clinically relevant. For 'Summary of findings' tables, we included the following comparisons.

Mometasone furoate cream on different treatment schedules.
Local narrow‐band UVB versus local PUVA.
Tacrolimus 0.1% ointment versus vehicle.
Tacrolimus 0.1% ointment versus mometasone furoate ointment.
Oral cyclosporin versus topical betamethasone dipropionate.
Oral alitretinoin at 10 mg and 30 mg a day versus placebo.

When a study reported on treatment during a remission‐ or clearance‐induction phase for participants before they were randomised to a follow‐up or maintenance phase, we considered only the latter (randomised) phase for this review.

Types of outcome measures

We extracted the following primary and secondary outcomes from the included studies.

Primary outcomes

Percentage of participants with self‐rated good/excellent control of symptoms.
Percentage of participants with investigator‐rated good/excellent control of symptoms.
Adverse events: adverse effects (long‐ and short‐term) of the intervention. Long‐term adverse events are defined as adverse events occurring after completion of the treatment phase; short‐term adverse events occur during the treatment phase.

Secondary outcomes

Reduction in severity (participant‐rated).
Reduction in severity (investigator‐rated).
Time until relapse, defined as the number of days/weeks until the participant reported worsening of symptoms after initial response.
Dose reduction: reduction in treatment dose per time unit or cumulative prescribed treatment dose. For example, a decrease in daily topical medication, or a decrease in weekly photo irradiation.

We did not exclude studies from the review that did not include these outcomes.

We believe that three months is the minimum study duration required to document important data such as duration and frequency of disease relapse.

Search methods for identification of studies

We aimed to identify all relevant RCTs regardless of language or publication status (published, unpublished, in press, or in progress).

Electronic searches

The Cochrane Skin Information Specialist searched the following databases up to 19 April 2018, using strategies based on the draft strategy for MEDLINE presented in our published protocol (van Coevorden 2009).

Cochrane Skin Group Specialised Register (search strategy in Appendix 2).
Cochrane Central Register of Controlled Trials (CENTRAL; 2018, Issue 3), in the Cochrane Library (search strategy in Appendix 3).
MEDLINE via Ovid (from 1946) (search strategy in Appendix 4).
Embase via Ovid (from 1974) (search strategy in Appendix 5).
Allied and Complementary Medicine (AMED) via Ovid (from 1985) (search strategy in Appendix 6).
Latin American and Caribbean Health Science Information database (LILACS) (from 1982) (search strategy in Appendix 7).
Global Resource of Eczema Trials. Centre of Evidence Based Dermatology (accessed at http://www.greatdatabase.org.uk on 19 April 2018), using the following terms in the title of the records: hand* or finger* or palm or palms.

Trials registries

We (WAC and PJC) searched the following trials registries up to 21 April 2018, using the following search terms: hand and (eczema or dermatitis).

International Standard Randomized Controlled Trials Number (ISRCTN) registry (www.isrctn.com).
ClinicalTrials.gov (www.clinicaltrials.gov).
Australian New Zealand Clinical Trials Registry (www.anzctr.org.au).
World Health Organization International Clinical Trials Registry Platform (ICTRP) (apps.who.int/trialsearch/).

Searching other resources

Correspondence with authors

If we needed clarification regarding studies, we contacted study authors using the correspondence options stated in their papers (for studies published since 1999). If email addresses did not work, we tried to find recent publications by the same corresponding author with more recent contact data, or we searched Facebook, LinkedIn, and the Internet to connect with these authors. In addition, we tried to contact all authors of studies that included other dermatoses among hand eczema, to obtain separate data for hand eczema participants. We listed in the 'notes' section of the Characteristics of included studies tables whether we contacted study authors, and if they responded. We have not included in the review complete correspondence with all studies, but we have shown the relevant citations in the Characteristics of included studies tables. The full correspondence with study authors is available upon request.

References from published studies

We checked the bibliographies of included studies for further references to relevant trials.

Adverse events

We did not perform a separate search for adverse events. However, we did examine data on adverse events from the included studies.

Unpublished literature

We contacted authors and pharmaceutical companies in relation to ongoing trials that were recently completed according to the trial registries mentioned under Electronic searches. When results were published on the trial register websites, we included these in the results, and we tried to contact study authors for additional information if necessary.

Conference proceedings

We searched the conference proceedings of annual conferences of the European Academy of Dermatology and Venereology (EADV) from 2000 to 2011 for further relevant RCTs. Some were available from the JEADV; however, some others had to be obtained from the organisation itself, from which we requested the material on CD‐ROM.

Handsearching

We handsearched using the terms 'eczema', 'dermatitis', 'hand(s)', 'palmoplantar', and 'inflammatory' in 16 English, two German, one Italian, one French, and one Dutch dermatology journal (all journals 1977 through 2003). We searched the journals listed in Appendix 8.

Data collection and analysis

Selection of studies

Three review authors (PJC, JLB, and WAC) independently checked titles and abstracts identified from the searches. Three review authors (PJC, TD, and ÅS) conducted an additional handsearch. If it was clear that the study did not refer to a randomised controlled trial on hand eczema, we excluded it. We retrieved all potential trials as full‐text articles for further independent examination by two review authors (TD and ÅS). These two review authors decided which trials conformed to the inclusion criteria and resolved discrepancies by discussion in consensus meetings. We obtained missing data from the trial authors when possible. Whenever we found duplicate publications of the same trial, we used the paper with the most relevant data (usually we had a conference abstract and a full article) as the primary reference and listed the other publication in the additional references following the reference section.

Data extraction and management

Three review authors (PJC, TD, and ÅS) extracted data independently, using a standardised data extraction form. These review authors and future reviewers piloted the data extraction form during a meeting of the European Dermato‐Epidemiology Network, in July 2000. This form was based on a preceding systematic review of psoriasis interventions and was later updated according to Cochrane recommendations. We resolved discrepancies and uncertainties in a series of consensus meetings, which were led by one review author (PJC).

Two other review authors (JLB and WAC) entered into Review Manager 5.3 and checked the outcome data extracted from the included studies (RevMan).

Assessment of risk of bias in included studies

Two review authors (ÅS and TD) independently assessed the risk of bias in included studies following the domain‐based evaluation described in Chapter 8 of the Cochrane Handbook for Systematic Reviews of Interventions and, using the Cochrane risk of bias tool, assessed all included studies from the following aspects for potential risk of bias (Higgins 2011b).

Random sequence generation, which refers to selection bias due to inadequate generation of a randomised sequence.
Allocation concealment, which also refers to selection bias but due to inadequate concealment of the allocation sequence before assignment.
Blinding of participants and personnel, which refers to performance bias due to knowledge of intervention allocation by participants or personnel.
Blinding of outcome assessment, which refers to detection bias due to knowledge of intervention allocation by the outcome assessor.
Incomplete outcome data, which refers to the quantity, nature, or manner in which incomplete outcome data were handled.
Selective reporting, which refers to reporting bias due to selective reporting.
Other source of bias, which refers to any other types of bias not covered above, including inclusion of baseline comparisons, certainty of the diagnosis, and premature ending.

Whenever we encountered disagreement regarding assessment of risk of bias, we resolved this in a consensus meeting with a third review author (PJC or HW). Two review authors (JLB and WAC) assessed completed 'Risk of bias' forms and entered the data into RevMan.

Measures of treatment effect

We employed risk ratios (RR) with 95% confidence intervals (CIs) to measure the effect of a treatment for dichotomous outcomes. We expressed results as number needed to treat for an additional beneficial outcome (NNTB) when appropriate, along with different rates of baseline risk. We expressed results from analyses of continuous data as mean differences (MDs), along with CIs and respective P values. Whenever a small study (fewer than 30 participants) included zero events in one arm, we used Fisher's exact test to calculate the P value, and we provided numerical data for the numerator/denominator for each treatment (Grainge 2013). We calculated Fisher's exact test using GraphPad software (GraphPad).

We interpreted numerical data in charts and tables when possible. We tried to extract numerical data from graphical presentations by using a ruler, or we contacted study authors for recent trials if the data were unclear. For data that had been extracted from a graph, we added remarks.

For studies that exclusively presented median values for a particular outcome, we substituted the median for the mean, provided that data were not too skewed. When standard deviations were not available from a paper, we tried to calculate these from other available data. When confidence intervals were provided, we used the formula given in Chapter 7.7.3.2 of the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011a).

For multi‐arm studies, we analysed each arm in comparison with placebo when possible.

Unit of analysis issues

Cluster randomised trials

We checked cluster randomised trials (groups of individuals instead of individuals randomised to intervention or control) for unit of analysis errors based on advice provided in Section 16.3.4 of the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011d).

Cross‐over studies

In cross‐over studies (with each participant allocated to a sequence of interventions, instead of to only one intervention), unit of analysis issues can arise when participants have been randomised to multiple treatments over multiple periods, or when there has been an inadequate washout period. We dealt with cross‐over studies by analysing only the first treatment period as a simple parallel‐group study.

Within‐participant studies (self‐controlled, left‐right designs)

Given that analysis of paired data was not possible with RevMan, we summarised the data from within‐participant studies in the text. The unit of analysis in within‐participant studies was one hand per participant, whereas in parallel‐group studies, the unit of analysis was per participant. Relevant data were presented in the analysis as "other data", in table format.

Studies with multiple arms

For studies including multiple arms (more than two) in the analyses, we plotted the different comparisons in different forest plots when possible.

Dealing with missing data

For trials published from 1999 onwards and with uncertainty, we tried to contact trial authors if we felt that this may yield essential additional information. In these cases, we contacted the first author or, when stated, the corresponding author of the article. For the current review, we did not make any assumption or imputation to missing data. We extracted all outcome data as they were reported in the original studies. We stated when authors were contacted and whether additional information was provided under Characteristics of included studies.

Assessment of heterogeneity

We had planned to explore reasons for heterogeneity amongst studies and, if necessary, to carry out sensitivity analyses to examine the effects of excluding study subgroups (e.g. children versus adults, atopic versus allergic contact hand eczema) or studies with high risk of bias.

Clinical heterogeneity (or clinical diversity) is considered as variability among participants, interventions, and outcomes. In future updates of this review, we plan to assess clinical heterogeneity by examining characteristics of the studies and similarity between types of participants, interventions, and outcomes. If studies were sufficiently similar, we achieved statistical pooling by using a weighted treatment effect.

We used random‐effects model meta‐analysis because of anticipated differences across studies in, amongst other things, the participant base included. Statistical heterogeneity was investigated with the I² test. If the I² statistic had been greater than 50%, reasons for heterogeneity in studies would have been explored.

Assessment of reporting biases

We planned on including statistical methods for detecting publication bias (e.g. Begg's funnel plots). However, funnel plots are recommended by the Cochrane Handbook for Systematic Reviews of Interventions, Section 10.4 (Higgins 2011c), when at least a substantial number of studies (10 or more) are included in the meta‐analysis. This was not feasible due to the heterogeneity of included studies. For reporting bias, we studied the study authors and institutions involved (pharmaceutical companies or not), funding, sponsorship of commercially available supplements, and, finally, conflicts of interest.

Data synthesis

When data permitted, we had planned to conduct statistical pooling, using a random‐effects model whenever studies appeared sufficiently similar.

Subgroup analysis and investigation of heterogeneity

We conducted no pre‐planned subgroup analyses in the current review, but in future updates, we will carry out analyses, if data permit, to examine the effects of including specific study subgroups (e.g. children versus adults, recurrent vesicular versus hyperkeratotic hand eczema).

Sensitivity analysis

We conducted no pre‐planned sensitivity analyses in this review, but for future updates, we will consider performing sensitivity analyses for pooled analysis involving only studies at low risk of bias.

'Summary of findings' tables

We included in the 'Summary of findings' tables all primary outcomes and the secondary outcome 'investigator‐rated reduction in severity' for the clinically most relevant studies (Ryan 2016). We assessed clinical relevance based on the clinical experiences of study authors. We tried to include studies from every group of interventions (topical corticosteroids, topical calcineurin inhibitors, UV therapy, and systemic treatments), and to keep the total number of included studies to a minimum. We therefore aimed to include studies that compared different groups of comparisons or studies that answered the questions that authors ask themselves on a regular basis in everyday practice. The 'Summary of findings' tables are based on the GRADE principles (GRADEPro, version 3.6.1). The GRADE approach is a sequential process that evaluates the quality of a body of evidence by considering the following domains.

Study limitations, which refers to risk of bias in either study design or conduct that could lead to biased estimation of treatment effect.
Inconsistency of results, which refers to unexplained heterogeneity of results.
Indirectness of evidence, which refers to directness of comparisons of target populations, interventions, comparators, and outcomes of the included studies compared to those of the planned PICO of the systematic review.
Imprecision, because results are generally imprecise when the study includes few participants, few events, or a wide confidence interval of the effect estimate.
Publication bias.

Randomised controlled trials (RCTs) began as high‐quality/certainty evidence, but If concerns were identified in the above domains, certainty was rated down by one or two levels depending on the severity of the concern. The GRADE approach completes assessments of the body of evidence by grading it in the high, moderate, low, or very low certainty category.

A duration of longer than three months was preferred for evaluating a clinically relevant effect. We considered interventions comparing different groups of interventions and studies providing different treatment regimens with regards to dosages or frequency as clinically most relevant. Therefore we included the following comparisons in the 'Summary of findings' tables.

Clobetasol foam compared to vehicle (Table 1).
Mometasone furoate cream in different treatment schedules (Table 2).
Local narrow‐band UVB compared to local PUVA (Table 3).
Tacrolimus 0.1% ointment compared to mometasone furoate ointment (Table 4).
Tacrolimus 0.1% ointment compared to vehicle (Table 5).
Oral cyclosporin compared to topical betamethasone dipropionate (Table 6).
Oral alitretinoin compared to placebo at 10 mg and 30 mg a day (Table 7; Table 8).

Summary of findings for the main comparison. Corticosteroid creams/ointments: clobetasol propionate foam compared to vehicle foam for hand eczema.

Corticosteroid creams/ointments: clobetasol propionate foam compared to vehicle foam for hand eczema
Patient or population: participants with moderate to severe hand eczema  Setting: secondary care with outpatients in Northern America  Intervention: clobetasol propionate 0.05% foam twice a day for 14 days  Comparison: vehicle/placebo foam twice a day for 14 days
Outcomes	*Anticipated absolute effects (95% CI)**		Relative effect  (95% CI)	No. of participants  (studies)	Certainty of the evidence  (GRADE)	Comments
	Assumed risk^a	Corresponding risk
	Risk with vehicle foam	Risk with clobetasol propionate foam
Primary: investigator‐rated good/excellent control of symptoms Follow‐up: day 15	Study population		RR 1.43  (0.86 to 2.40)	125  (1 RCT)	⊕⊕⊕⊝  Moderate^b	‐
	270 per 1000	386 per 1000  (232 to 648)
Primary: participants with self‐rated good/excellent control of symptoms Follow‐up: day 15	Study population		RR 2.32  (1.38 to 3.91)	125  (1 RCT)	⊕⊕⊕⊝  Moderate^c	NNTB 3 (95% CI 2 to 8)
	222 per 1000	516 per 1000  (307 to 869)
Primary: adverse events ‐ at least 1 adverse event Follow‐up: day 15	Study population		RR 2.24  (0.82 to 6.06)	125  (1 RCT)	⊕⊕⊕⊝  Moderate^c	‐
	79 per 1000	178 per 1000  (65 to 481)
Primary: adverse events ‐ any adverse event treatment‐related (application site pruritus) Follow‐up: day 15	Study population		RR 1.02  (0.06 to 15.89)	125  (1 RCT)	⊕⊕⊕⊝  Moderate^d	‐
	16 per 1000	16 per 1000  (1 to 252)
Secondary: reduction in severity, participant‐rated scoring Follow‐up: day 15	Study population		RR 1.57  (1.21 to 2.04)	125  (1 RCT)	⊕⊕⊕⊝  Moderate^b	NNTB 3 (95% CI 2 to 7)
	524 per 1000	822 per 1000  (634 to 1000)
Secondary: reduction in severity, investigator‐rated scoring ‐ improvement at least 2 grades Follow‐up: day 15	Study population		RR 1.47  (0.90 to 2.39)	125  (1 RCT)	⊕⊕⊕⊝  Moderate^b	‐
	286 per 1000	420 per 1000  (257 to 683)
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).    CI: confidence interval; NNTB: number needed to treat for an additional beneficial outcome; RCT: randomised controlled trial; RR: risk ratio. Kircik 2013
GRADE Working Group grades of evidence.  High certainty: we are very confident that the true effect lies close to that of the estimate of the effect.  Moderate certainty: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different.  Low certainty: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect.  Very low certainty: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.

Corticosteroid creams/ointments: mometasone furoate cream thrice a week versus twice a week
Patient or population: people (all patch‐tested) with hand eczema > 6 months that had cleared upon daily treatment for a maximum of 9 weeks with mometasone furoate cream  Settings: secondary care with outpatients from hospitals in Denmark  Intervention: mometasone furoate cream 3 times/week up to 36 weeks Comparision: mometasone furoate cream 2 times/week up to 36 weeks
Outcomes	*Anticipated absolute effects (95% CI)**		Relative effect  (95% CI)	No. of participants  (studies)	Certainty of the evidence  (GRADE)	Comments
	Assumed risk^a	Corresponding risk
	Risk with mometasone furoatetwice a week	Risk with mometasone furoate thrice a week
Primary: investigator‐rated good/excellent control of symptoms  Follow‐up: 36 weeks	Study population		RR 1.23   (0.94 to 1.61)	72  (1 RCT)	⊕⊕⊝⊝  Low^b	‐
	676 per 1000	831 per 1000  (635 to 1000)
Primary: participant‐rated good/excellent control of symptoms  Not measured	See comment	See comment	Not estimable	‐	See comment	No data available
Primary: adverse events  Follow‐up: 36 weeks	Study population		RR 1.76   (0.45 to 6.83)	72  (1 RCT)	⊕⊕⊝⊝  Low^c	‐
	81 per 1000	143 per 1000  (36 to 554)
Secondary: investigator‐rated reduction in severity Not measured	See comment	See comment	Not estimable	‐	See comment	No data available
The basis for the assumed risk* (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).  CI: confidence interval; RCT: randomised controlled trial; RR: risk ratio. Veien 1999
GRADE Working Group grades of evidence.  High certainty: further research is very unlikely to change our confidence in the estimate of effect.  Moderate certainty: further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.  Low certainty: further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.  Very low certainty: we are very uncertain about the estimate.

Irradiation with UV light: local narrow‐band UVB compared to local PUVA for hand eczema
Patient or population: people with hand eczema unresponsive to clobetasol propionate  Setting: secondary care with outpatients in the United Kingdom.  Intervention: local narrow‐band UVB twice weekly for 12 weeks  Comparison: immersion PUVA twice weekly for 12 weeks
Outcomes	*Anticipated absolute effects (95% CI)**		Relative effect  (95% CI)	No. of participants  (studies)	Certainty of the evidence  (GRADE)	Comments
	Assumed risk^a	Corresponding risk
	Risk with local PUVA	Risk with local narrow‐band UVB
Primary: investigator‐rated good/excellent control of symptoms Follow‐up: 12 weeks	Study population		RR 0.50  (0.22 to 1.16)	60  (1 RCT)	⊕⊕⊕⊝  Moderate^b	‐
	400 per 1000	200 per 1000  (88 to 464)
Primary: participant‐rated good/excellent control of symptoms Not measured	See comment	See comment	Unable to estimate treatment effect	‐	See comment	No data reported
Primary: adverse events ‐ reported adverse events, mainly erythema Follow‐up: 12 weeks	See comment	See comment	RR 19.00  (1.16 to 312.42)	60  (1 RCT)	⊕⊕⊕⊝  Moderate^c	PUVA: No adverse events reported (0/30) Narrow‐band UVB: 9 out of 30 participants reported an adverse event, mainly erythema Fisher's exact test P = 0.0019
Secondary: investigator‐rated reduction in severity in mTLSS^d Follow‐up: 12 weeks	‐	‐	Unable to estimate treatment effect	43 (1 RCT)	⊕⊕⊕⊝  Moderate^e	Reduction in mTLSS PUVA: Median mTLSS of 8.5 (range 0 to 16) and 8 (range 3 to 15) for the left and right hand, to a median mTLSS 3 (range 0 to 13) and 3 (range 0 to 14) (n = 23) Reduction mTLSS local narrow‐band UVB group: Median mTLSS of 7 (range 0 to 16) and 8.5 (range 1 to 15) to a median mTLSS5 (range 0 to 11) and 4.5 (range 0 to 11) (n = 20)
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).    CI: confidence interval; mTLSS: modified total lesion symptom score; PUVA: oral psoralen combined with UVA; RCT: randomised controlled trial; RR: risk ratio; UV: ultraviolet. 2015
GRADE Working Group grades of evidence.  High certainty: we are very confident that the true effect lies close to that of the estimate of the effect.  Moderate certainty: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different.  Low certainty: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect.  Very low certainty: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.

Topical calcineurin inhibitors: tacrolimus 0.1% ointment compared to mometasone furoate ointment for vesicular hand eczema
Patient or population: people with moderate to severe chronic relapsing dyshidrotic eczema on hands  Setting: secondary care setting at a single dermatology department in Germany  Intervention: topical calcineurin inhibitors tacrolimus 0.1% ointment twice daily during 4 weeks  Comparison: topical corticosteroid mometasone furoate ointment twice daily during 4 weeks
Outcomes	*Anticipated absolute effects (95% CI)**		Relative effect  (95% CI)	No. of participants  (studies)	Certainty of the evidence  (GRADE)	Comments
	Assumed risk	Corresponding risk
	Risk with mometasone furoate ointment	Risk with topical calcineurin inhibitor tacrolimus 0.1% ointment
Primary: investigator‐rated good/excellent control of symptoms ‐ Not measured	See comment	See comment	Not estimable	‐	See comment	Not measured
Primary: participant‐rated good/excellent control of symptoms ‐ Not measured	See comment	See comment	Not estimable	‐	See comment	Not measured
Primary: adverse events   Follow‐up: 2 weeks	See comment	See comment	Not estimable	16 pairs of hands (1 RCT)	⊕⊕⊕⊝  Moderate^a	Within‐participant design None of the participants dropped out because of adverse events
Secondary: investigator‐rated reduction in severity ‐ DASI^b Follow‐up: 2 weeks	See comment	See comment	Not estimable	16 pairs of hands (1 RCT)	⊕⊕⊕⊝  Moderate^a	Within‐participant design Tacrolimus group: Mean DASI from 18 (SD 12.68) to 6.6 (SD 6.18) Mometasone furoate group: Mean DASI from 18.5 (SD 14.09) to 6.9 (SD 7.7)
The basis for the assumed risk* (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).  CI: confidence interval; DASI: Dyshydrotic Eczema Area and Severity Index; RCT: randomised controlled trial; RR: risk ratio; SD: standard deviation. Schnopp 2002
GRADE Working Group grades of evidence.  High certainty: further research is very unlikely to change our confidence in the estimate of effect.  Moderate certainty: further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.  Low certainty: further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.  Very low certainty: we are very uncertain about the estimate.

Topical calcineurin inhibitor tacrolimus 0.1% ointment compared to vehicle for hand eczema
Patient or population: people with moderate to severe nickel sulphate‐induced allergic contact dermatitis based on clinical history (hand eczema) and proven by patch testing, resistant to topical corticosteroids  Settings: secondary care setting in a single‐centre study in Italy  Intervention: topical calcineurin inhibitor tacrolimus 0.1% ointment twice daily for 2 weeks Comparison: vehicle twice daily for 2 weeks
Outcomes	*Anticipated absolute effects (95% CI)**		Relative effect  (95% CI)	No. of participants  (studies)	Certainty of the evidence  (GRADE)	Comments
	Assumed risk	Corresponding risk
	Risk with vehicle	Risk with tacrolimus 0.1% ointment
Primary: investigator‐rated good/excellent control of symptoms Follow‐up: 3 weeks	See comment	See comment	RR 29.00  (1.90 to 443.25)	28  (1 RCT)	⊕⊕⊕⊝  Moderate^a	Unable to calculate assumed risk as no events in the control group ‐ 14/14 participants in the tacrolimus group had good/excellent control of symptoms. Fisher's exact test P = 0.0001, NNTB 1, 95% CI 1 to 1
Primary: participant‐rated good/excellent control of symptoms   Not measured	See comment	See comment	Not estimable	‐	See comment	No data reported
Primary: adverse events ‐ burning/itching at application site Follow‐up: 3 weeks	See comment	See comment	RR 9.00  (0.53 to 152.93)	28  (1 RCT)	⊕⊕⊕⊝  Moderate^a	Unable to calculate assumed risk as no events in the control group ‐ 4/14 participants in the tacrolimus group had burning/itching at the application site. Fisher's exact test P = 0.1129, RR 9.00, 95% CI 0.53 to 152.93 No data on "all adverse events"
Secondary: investigator‐rated reduction in severity ‐ Not measured	See comment	See comment	Not estimable	‐	See comment	No data reported
The basis for the assumed risk* (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).  CI: confidence interval; NNTB: number needed to treat for an additional beneficial outcome; RCT: randomised controlled trial; RR: risk ratio. Pacor 2006
GRADE Working Group grades of evidence.  High certainty: further research is very unlikely to change our confidence in the estimate of effect.  Moderate certainty: further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.  Low certainty: further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.  Very low certainty: we are very uncertain about the estimate.

Oral cyclosporin compared to topical betamethasone for patient with hand eczema
Patient or population: people with hand eczema, continuously for 6 months, significant disability, inadequate response to conventional treatment, confirmation by histopathology  Setting: secondary care setting at a single centre in Finland  Intervention: oral cyclosporin 3 mg/kg/d and placebo cream for 6 weeks  Comparison: topical betamethasone dipropionate 0.05% cream and placebo capsules for 6 weeks
Outcomes	*Anticipated absolute effects (95% CI)**		Relative effect  (95% CI)	No. of participants  (studies)	Certainty of the evidence  (GRADE)	Comments
	Assumed risk^a	Corresponding risk
	Risk with topical betamethasone	Risk with oral cyclosporin
Primary: investigator‐ rated good/excellent control of symptoms^b Follow‐up: 6 weeks	Study population		RR 1.88  (0.88 to 3.99)	34  (1 RCT)	⊕⊕⊕⊝  Moderate^c	‐
	333 per 1000	627 per 1000  (293 to 1000)
Primary: participant‐rated good/excellent control of symptoms^b Follow‐up: 6 weeks	Study population		RR 1.25  (0.69 to 2.27)	34  (1 RCT)	⊕⊕⊕⊝  Moderate^c	‐
	500 per 1000	625 per 1000  (345 to 1000)
Primary: adverse events ‐ at least 1 adverse event Follow‐up: 36 weeks	Study population		RR 1.22  (0.80 to 1.86)	55^d  (1 RCT)	⊕⊕⊕⊝  Moderate^c	Because of partial cross‐over design, a different number of participants is given for this outcome
	556 per 1000	678 per 1000  (444 to 1000)
Secondary: investigator‐rated reduction in severity^b Follow‐up: 6 weeks	Mean investigator‐rated reduction in severity in total disease activity score after 6 weeks of treatment was 5.7	MD 0.30 higher  (2.50 lower to 3.10 higher)	‐	34  (1 RCT)	⊕⊕⊕⊝  Moderate^c	‐
The basis for the assumed risk* (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).  CI: confidence interval; MD: mean difference; RCT: randomised controlled trial; RR: risk ratio. Granlund 1996
GRADE Working Group grades of evidence.  High certainty: further research is very unlikely to change our confidence in the estimate of effect.  Moderate certainty: further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.  Low certainty: further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.  Very low certainty: we are very uncertain about the estimate.

Oral retinoids: alitretinoin 30 mg versus placebo for hand eczema
Patient or population: people with moderate to severe chronic hand eczema  Settings: secondary care with outpatients in an international multi‐centre setting  Intervention: oral retinoid alitretinoin 30 mg for 12 to 24 weeks Comparison: oral placebo for 12 to 24 weeks
Outcomes	*Anticipated absolute effects (95% CI)**		Relative effect  (95% CI)	No. of participants  (studies)	Certainty of the evidence  (GRADE)	Comments
	Assumed risk^a	Corresponding risk
	Risk with placebo	Risk with alitretinoin 30 mg
Primary: investigator‐rated good/excellent control of symptoms Follow‐up: 48 weeks to 72 weeks	Study population		RR 2.75 (2.20 to 3.43)	1210  (2 RCTs)	⊕⊕⊕⊕  High^b	NNTB 4, 95% CI 3 to 5
	157 per 1000	432 per 1000  (346 to 539)
Primary: participant‐rated good/excellent control of symptoms Folluw‐up: 48 weeks to 72 weeks	Study population		RR 2.75  (2.18 to 3.48)	1210  (2 RCTs)	⊕⊕⊕⊕  High^b	NNTB 4, 95% CI 3 to 5
	143 per 1000	394 per 1000  (312 to 498)
Primary: adverse events ‐ headache Folluw‐up: 48 weeks to 72 weeks	Study population		RR 3.43  (2.45 to 4.81)	1210  (2 RCTs)	⊕⊕⊕⊕  High^b	All adverse events not stated in Ruzicka 2008 NNTH 6, 95% CI 4 to 11
	74 per 1000	251 per 1000  (179 to 352)
Secondary: investigator‐rated reduction in severity in TLSS^c and mTLSS^d	See comment	See comment	Not estimable	‐	See comment	Only incomplete data reported; therefore we were unable to extract these data
The basis for the assumed risk* (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).  CI: confidence interval; NNTB: number needed to treat for an additional beneficial outcome; NNTH: number needed to treat for an additional harmful outcome; RCT: randomised controlled trial; RR: risk ratio. Ruzicka 2008; Fowler 2014
GRADE Working Group grades of evidence.  High certainty: further research is very unlikely to change our confidence in the estimate of effect.  Moderate certainty: further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.  Low certainty: further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.  Very low certainty: we are very uncertain about the estimate.

Oral retinoids: alitretinoin 10 mg versus placebo for hand eczema
Patient or population: people with moderate to severe chronic hand eczema  Settings: secondary care with outpatients in an international multi‐centre setting  Intervention: oral retinoid alitretinoin 10 mg for 12 to 24 weeks Comparison: oral placebo for 12 to 24 weeks
Outcomes	*Anticipated absolute effects (95% CI)**		Relative effect  (95% CI)	No. of participants  (studies)	Certainty of the evidence  (GRADE)	Comments
	Assumed risk^a	Corresponding risk
	Risk with placebo	Risk with alitretinoin
Primary: investigator‐rated good/excellent control of symptoms Follow‐up: up to 48 weeks	Study population		RR 1.58 (1.20 to 2.07)	781  (2 RCTs)	⊕⊕⊕⊕  High^b	NNTB 11, 95% CI 6.3 to 26.5
	194 per 1000	307 per 1000  (233 to 402)
Primary: participant‐rated good/excellent control of symptoms Follow‐up: up to 48 weeks	Study population		RR 1.73  (1.25 to 2.40)	765  (2 RCTs)	⊕⊕⊕⊕  High^b	NNTB 9, 95% CI 6 to 20
	144 per 1000	249 per 1000  (180 to 345)
Primary: all adverse events Follow‐up: up to 48 weeks	Study population		RR 1.01 (0.66 to 1.55)	158  (1 RCT)	⊕⊕⊕⊝  Moderate^c	NNTH 260, 95% CI ‐14.47 to 15.24
	346 per 1000	350 per 1000 (228 to 537)
Secondary: investigator‐rated reduction in severity of TLSS^d Follow‐up: up to 48 weeks	See comment	See comment	‐	158  (1 RCT)	⊕⊕⊕⊝  Moderate^c	Median % change in score from baseline (95% CI) Placebo group: ‐25% (95% CI ‐42 to ‐14) Aitretinoin 10 mg: –59 (95% CI –73 to –33)
The basis for the assumed risk* (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).  CI: confidence interval; NNTB: number needed to treat for an additional beneficial outcome; NNTH: number needed to treat for an additional harmful outcome; RCT: randomised controlled trial; RR: risk ratio; TLSS: total lesion symptom score. Ruzicka 2004; Ruzicka 2008
GRADE Working Group grades of evidence.  High certainty: further research is very unlikely to change our confidence in the estimate of effect.  Moderate certainty: further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.  Low certainty: further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.  Very low certainty: we are very uncertain about the estimate.

Study	Comparison	Investigator‐rated good/excellent control in RR (95% CI)	Comment
Kucharekova 2003	Ceramide‐containing emollient versus regular petrolatum‐based emollient	No data regarding the primary outcome investigator‐rated good/excellent control	‐
Chu 2009	E‐DO lotion versus vehicle lotion	E‐DO: 37 responders, 12 of whom responded to E‐DO only (19%), and 25 to both (39.7%) Vehicle: 36 responders, 11 of whom responded to vehicle only (17.5%), and 25 to both (39.7%) Investigator‐rated RR 1.06 (95% CI 0.54 to 2.10)	Within‐patient study

Study	Comparison of topical corticosteroids	Investigator‐rated good/excellent control in RR (95% CI)	Comment
Möller 1983	Intermittent clobetasol versus fluprednidene	Clobetasol better since 32/46 versus 14/46 hands remained in remission; investigator‐rated	Within‐patient study
Uggeldahl 1986	Desonide 0.1% versus desonide 0.05%	No data regarding the primary outcome investigator‐rated good/excellent control	Within‐patient study
Bleeker 1989	Fluprednidene versus betamethasone	RR 0.59 (95% CI 0.28 to 1.23); investigator‐rated	‐
Gupta 1993	Betamethasone film versus betamethasone lotion	RR 10.24 (95% CI 0.59 to 176.56); investigator‐rated	‐
Veien 1999	Mometasone 3 times/week versus mometasone 2 times/week	RR 1.23 (95% CI 0.94 to 1.61); investigator‐rated	‐
Fowler 2005	Hydrocortisone butyrate versus fluticasone propionate twice daily  Hydrocortisone butyrate versus prednicarbate emollient twice daily Hydrocortisone butyrate 0.1% cream versus mometasone furoate twice daily	No data regarding the primary outcome investigator‐rated good/excellent control	Three parallel treatment groups Each group separately within‐patient
Faghihi 2008	0.05% clobetasol cream versus 0.05% clobetasol + zinc sulphate cream	Clobetasol + zinc sulphate better in terms of respectively scaling (25/47 versus 3/47), redness (41/47 versus 1/41), and lichenification (24/47 versus 7/47); investigator‐rated	Within‐patient study
Lodén 2012a	Betamethasone‐valerate 0.1% cream twice daily versus betamethasone‐valerate 0.1% cream once daily + urea 5% cream once daily	RR 0.75 (0.55 to 1.03); investigator‐rated	‐
Kircik 2013	Clobetasol propionate 0.05% foam twice daily versus vehicle foam twice daily	RR 1.43 (0.86 to 2.40); investigator‐rated	‐

Study	Comparison	Investigator‐ or participant‐rated good/excellent control in RR (95% CI)	Comments
Thestrup‐Pedersen 2001	Oral acitretin versus placebo	No data regarding the primary outcome investigator‐rated good/excellent control	‐
Ruzicka 2004	Oral alitretinoin (20 mg and 40 mg) versus placebo	40 mg  Participant‐rated  RR 3.51 (95% CI 1.80 to 6.82)  Investigator‐rated  RR 1.97 (95% CI 1.3 to 3.0) 20 mg  Participant rated  RR 2.74 (95% CI 1.37 to 5.46)  Investigator‐rated  RR 1.49 (95% CI 0.94 to 2.34)	‐
Ruzicka 2008; Fowler 2014	Oral alitretinoin 30 mg versus placebo	30 mg  Participant‐rated  RR 2.75 (95% CI 2.18 to 3.48)  Investigator‐rated  RR 2.75 (95% CI 2.20 to 3.43)	‐
Ruzicka 2004; Ruzicka 2008	Oral alitretinoin 10 mg versus placebo	10 mg Participant‐rated  RR 1.73 (95% CI 1.25 to 2.40) Investigator‐rated  RR 1.58 (95% CI 1.20 to 2.07)	‐
Bissonnette 2010	Re‐treatment with oral alitretinoin (30 mg and 10 mg) versus placebo	30 mg  Investigator‐rated  RR 9.55 (95% CI 2.51 to 36.27) 10 mg  Investigator‐rated  RR 4.76 (95% CI 0.70 to 32.25)	‐

Primary outcome: adverse events
Study	Group ‐ within‐participant study	Event number	Total number of pairs of hands analysed
Kemper 1998	Betamethasone valerate 0.1% cream	0	19
Kemper 1998	Coal tar paste	1	19

Primary: investigator‐rated good/excellent control of symptoms in UVB vs PUVA
Study	Group ‐ within‐participant study	Event number	Total number of pairs of hands analysed
Sezer 2007	NB‐UVB	2	12
Sezer 2007	Local PUVA	1	12

Primary: adverse events
Study	Group ‐ within‐participant study	Event number	Total number of pairs of hands analysed
Palmar hyperpigmentation
Sezer 2007	NB‐UVB	0	12
Sezer 2007	Local PUVA	3	12

Primary: adverse events
Study	Group ‐ within‐participant study	Event number	Total number of pairs of hands analysed
Discontinuation due to adverse events
Grattan 1991	Topical PUVA	1	15
Grattan 1991	UVA	1	15
Burning
Grattan 1991	Topical PUVA	1	15
Grattan 1991	UVA	0	15
Exacerbation of eczema
Grattan 1991	Topical PUVA	1	15
Grattan 1991	UVA	0	15

Primary: percentage of participants with self‐rated good/excellent control at week 4
Study	Group ‐ within‐participant study	Total number of events	Total number of pairs of hands analysed
Chu 2009	Emollient E‐DO	22	67
Chu 2009	Vehicle	23	67

Primary: percentage of participants with investigator‐rated good/excellent control at week 4
Study	Group‐ within‐participant study	Event number	Total number of pairs of hands randomised
Chu 2009	Emollient E‐DO	37	67
Chu 2009	Vehicle	36	67

Primary: investigator‐rated good/excellent control of symptoms at day 7
Study	Group	Number of participants with good/excellent control	Total number of participants
Gupta 1993	B‐film forming lotion	5	28
Gupta 1993	B‐thick lotion	0	26

Secondary: reduction in investigator‐rated severity ‐ DASI
Study	Group ‐ within‐participant study	Mean	SD	Total number of pairs of hands analysed
Schnopp 2002	Tacrolimus	6.6	6.18	8
Schnopp 2002	Mometasone	6.9	7.7	8

Secondary: investigator‐rated reduction in severity in total lesion symptom score
Study
Alitretinoin 40 mg
Ruzicka 2004	Median of % change from baseline: ‐70.5% (95% CI ‐44 to ‐80).
Ruzicka 2004	Significant more reduction than placebo (P < 0.001; Kruskal‐Wallis test).
Alitretinoin 30 mg
Ruzicka 2008	Median of % change from baseline: ‐75%
Ruzicka 2008	Significant more reduction than placebo (P < 0.001; Kruskal‐Wallis test).
Alitretinoin 20 mg
Ruzicka 2004	Median of % change from baseline: ‐52 (95% CI ‐42 to ‐73)
Ruzicka 2004	Significant more reduction than placebo (P < 0.01; Kruskal‐Wallis test).
Alitretinoin 10 mg
Ruzicka 2004	Median of % change from baseline: ‐25% (95% CI ‐14 to ‐42)
Ruzicka 2004	Significant more reduction than placebo (P < 0.01; Kruskal‐Wallis test).
Ruzicka 2008	Median of % change from baseline: ‐56%
Ruzicka 2008	Significant more reduction than placebo (P < 0.01; Kruskal‐Wallis test).

Secondary: investigator‐rated reduction in severity in total lesion symptom score
Study	Group	Median	SD	N
Alitretinoin 40 mg vs placebo
Ruzicka 2004	Alitretinoin 40 mg	70.5	81.407	81
Ruzicka 2004	Placebo	25.0	62.13	78
Alitretinoin 20 mg vs placebo
Ruzicka 2004	Alitretinoin 20 mg	52.0	80.9	80
Ruzicka 2004	Placebo	25.0	62.13	78
Alitretinoin 10 mg vs placebo
Ruzicka 2004	Alitretinoin 10 mg	59.0	89.892	80
Ruzicka 2004	Placebo	25.0	62.13	78

Primary: investigator‐ and/or participant‐rated improvement
Study	Group ‐ within‐participant study	Event	Total number of participants analysed
Burrows 1986	Trientine	6	20
Burrows 1986	Placebo	10	20

Medical term	Explanation
Acrovesicular eczema	Form of vesicular hand eczema. (Large) vesicle eruptions on the palms that usually tend to recur. Also called dyshidrotic eczema or pompholyx
Betamethasone	Topical corticosteroid, high potency
Clobetasol propionate	Topical corticosteroid, very high potency
Desonide	Topical corticosteroid, low potency
Dyshidrotic hand eczema	Form of vesicular hand eczema. (Large) vesicle eruptions on the palms that usually tend to recur. Also called dyshidrotic eczema, pompholyx or acro vesicular eczema
Fluprednidene acetate	Topical corticosteroid, medium potency
Heterogeneity	Differences in which studies have been undertaken with regard to methods and/or materials
Hydrocortisone butyrate	Topical corticosteroid, low potency
Hyperkeratotic hand eczema	Form of hand eczema with areas of thick scaling on the palms, also called tylotic hand eczema
IGA	Investigator global assessment: global assessment of disease severity usually on a 5‐point scale
Immunomodulator	Drug which changes the immune response such as tacrolimus
Immunosuppressor	Drug which suppresses the immune response such ad topical corticosteroids
Iontophoresis	Treatment by which the skin is soaked in (tap) water through which a weak electric current is passed
Mometasone furoate cream	Potent steroid cream
NB‐UVB	Narrow‐band ultraviolet B
Nummular hand eczema	Round ("coin sized") eczematous patches on the back of the hands
Palmar	Hand palms, the inside surface of the hands
Palmoplantar	Hand palms and foot soles
Phase I clinical trial	A clinical trial of a new drug or therapy. Phase I trials are conducted in small groups of participants
Phase II clinical trial	A clinical trial of a new drug or therapy. Phase II trials are conducted in larger groups of participants than phase I trials
Pimecrolimus	Topical calcineurin inhibitor, also known as "elidel"
Placebo	Simulated or otherwise medically ineffective treatment
Pompholyx	Form of vesicular hand eczema. (Large) vesicle eruptions on the palms that usually tend to recur. Also called dyshidrotic eczema, pompholyx or acro vesicular eczema
Potency	Strength
Prevalence	The proportion of a population having a particular condition or characteristic: e.g. the percentage of people in a city with hand eczema, or the proportion of people who smoke
Primary care	Health care provided at the principal point of consultation for patients within a healthcare system, e.g. GP
Pruritus	Itch
Psychosomatic disorder	A disorder in which physical symptoms originate from mental or emotional causes
Pulpitis	A dry, fissured, scaling dermatitis of the fingertips with occasional episodes of vesicles. Also known as fingertip dermatitis
PUVA	(Topical and oral) psoralen combined with UVA
Randomised control trials	A study in which a number of similar people are randomly assigned to two (or more) groups to test a specific drug, treatment or other intervention. (National Institute for Health and Care Excellence)
RR	Relative risk
Secondary care	Health care provided by medical specialists and other health professionals, including dermatologists, who generally do not have first contact with patients. This contains hospital and out‐patient care
Systemic treatment	Treatment which does not pertain to a certain surface area but might affect the entire body, usually taken by mouth or injection.
Tacrolimus	Topical calcineurin inhibitor, also known as "protopic"
Teratogenicity	Developmental abnormalities in the foetus
Therapy	A treatment that helps someone feel better, grow stronger, etc., especially after an illness (Cambridge dictionary)
Topical treatment	Treatment pertaining to a certain surface area (usually the skin) and only affecting the area to which it is applied
Transepidermal water loss (TEWL)	The amount of water that moves from inside the body to the surrounding atmosphere through the epidermal layer of the skin by means of diffusion and evaporation.
Tylotic hand eczema	Form of hand eczema with areas of thick scaling on the palms, also called hyperkeratotic hand eczema
UVA‐1	Form of UV‐phototherapy which only uses the longer UV wavelengths (340 to 400 nm) and reduces the risk of burning, which is associated with the shorter‐wavelength UVA2 (320 to 340 nm) and UVB (290 to 320 nm).
UVB	Ultraviolet B
VAS (Visual Analogue Scale)	Continious scale to measure a (subjective) response
Vehicle	Something used to transport people or goods (Cambridge dictionary), in this case something to help the treatment get transport in/on the skin, but a vehicle alone (without the active substance) can be used as placebo.

Outcome or subgroup title	Statistical method	Effect size
1 Primary: percentage of participants with self‐rated good/excellent control at week 4	Other data	No numeric data
2 Primary: percentage of participants with investigator‐rated good/excellent control at week 4	Other data	No numeric data
3 Adverse events	Other data	No numeric data
3.1 At least 1 adverse event	Other data	No numeric data
3.2 Pruritus	Other data	No numeric data

Outcome or subgroup title	No. of studies	Statistical method	Effect size
1 Primary: investigator‐rated good/excellent control of symptoms after 3 weeks of treatment	1	Risk Ratio (M‐H, Random, 95% CI)	Totals not selected
2 Primary: number of participants with at least 1 adverse event	1	Risk Ratio (M‐H, Random, 95% CI)	Totals not selected
3 Secondary: investigator‐rated improvement > 50% after 3 weeks	1	Risk Ratio (M‐H, Random, 95% CI)	Totals not selected

PERMALINK

Interventions for hand eczema

Wietske Andrea Christoffers

Pieter‐Jan Coenraads

Åke Svensson

Thomas L Diepgen

Janine L Dickinson‐Blok

Jun Xia

Hywel C Williams

Abstract

Background

Objectives

Search methods

Selection criteria

Data collection and analysis

Main results

Authors' conclusions

Plain language summary

Summary of findings

Background

Description of the condition

Definition and epidemiology

Causes

Impact

Prognosis

Description of the intervention

How the intervention might work

Why it is important to do this review

Objectives

Methods

Criteria for considering studies for this review

Types of studies

Types of participants

Types of interventions

Types of outcome measures

Primary outcomes

Secondary outcomes

Search methods for identification of studies

Electronic searches

Trials registries

Searching other resources

Correspondence with authors

References from published studies

Adverse events

Unpublished literature

Conference proceedings

Handsearching

Data collection and analysis

Selection of studies

Data extraction and management

Assessment of risk of bias in included studies

Measures of treatment effect

Unit of analysis issues

Cluster randomised trials

Cross‐over studies

Within‐participant studies (self‐controlled, left‐right designs)

Studies with multiple arms

Dealing with missing data

Assessment of heterogeneity

Assessment of reporting biases

Data synthesis

Subgroup analysis and investigation of heterogeneity

Sensitivity analysis

'Summary of findings' tables

Summary of findings for the main comparison. Corticosteroid creams/ointments: clobetasol propionate foam compared to vehicle foam for hand eczema.

Summary of findings 2. Corticosteroid creams/ointments: mometasone furoate cream 3 times/week versus 2 times/week for hand eczema.

Summary of findings 3. Irradiation with UV light: local narrow‐band UVB compared to local PUVA for hand eczema.

Summary of findings 4. Topical calcineurin inhibitors: tacrolimus 0.1% ointment compared to mometasone furoate ointment for vesicular hand eczema.

Summary of findings 5. Topical calcineurin inhibitors: tacrolimus 0.1% ointment versus vehicle for hand eczema.

Summary of findings 6. Oral immunosuppressants: oral cyclosporin versus topical betamethasone dipropionate.

Summary of findings 7. Oral retinoids: alitretinoin 30 mg versus placebo for hand eczema.

Summary of findings 8. Oral retinoids: alitretinoin 10 mg versus placebo for hand eczema.

Results

Description of studies

Results of the search

1.

Included studies

Design

Participants

Sample size calculation

Outcome or subgroup title	Statistical method	Effect size
1 Primary outcome: percentage of participants with investigator‐rated good/excellent control	Other data	No numeric data
1.1 Scaling	Other data	No numeric data
1.2 Redness	Other data	No numeric data
1.3 Lichenification	Other data	No numeric data

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Primary: investigator‐rated good/excellent control of symptoms in UVB vs PUVA	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected
2 Primary: adverse events ‐ reported adverse event, mainly erythema	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected

Outcome or subgroup title	No. of studies	Statistical method	Effect size
1 Primary: adverse events	2	Risk Ratio (M‐H, Random, 95% CI)	Totals not selected
1.1 Nausea leading to dropout	1	Risk Ratio (M‐H, Random, 95% CI)	0.0 [0.0, 0.0]
1.2 Burn leading to dropout	1	Risk Ratio (M‐H, Random, 95% CI)	0.0 [0.0, 0.0]
1.3 Erythema	1	Risk Ratio (M‐H, Random, 95% CI)	0.0 [0.0, 0.0]
1.4 Dizziness	1	Risk Ratio (M‐H, Random, 95% CI)	0.0 [0.0, 0.0]
1.5 Nausea	1	Risk Ratio (M‐H, Random, 95% CI)	0.0 [0.0, 0.0]
1.6 Headache	1	Risk Ratio (M‐H, Random, 95% CI)	0.0 [0.0, 0.0]
2 Secondary: investigator‐rated reduction in severity at week 10 (bigger reduction in severity = better outcome)	1	Mean Difference (IV, Random, 95% CI)	Totals not selected
3 Secondary: investigator‐rated reduction in severity at week 18 (bigger reduction in severity = better outcome)	1	Mean Difference (IV, Random, 95% CI)	Totals not selected

Outcome or subgroup title	No. of studies	Statistical method	Effect size
1 Primary: adverse events ‐ discontinuation because of exacerbation	1	Risk Ratio (M‐H, Random, 95% CI)	Totals not selected
2 Secondary: participant‐rated reduction in severity on VAS for itch (week 3, bigger reduction in severity = better outcome)	1	Mean Difference (IV, Random, 95% CI)	Totals not selected
3 Secondary: investigator‐rated reduction in severity on dyshidrotic eczema area and severity index (DASI) (week 3, bigger reduction in severity = better outcome)	1	Mean Difference (IV, Random, 95% CI)	Totals not selected

Outcome or subgroup title	Statistical method	Effect size
1 Primary: investigator‐rated improvement good/excellent control	Other data	No numeric data
1.1 after 1 month	Other data	No numeric data
1.2 after 3 months	Other data	No numeric data
1.3 after 6 months	Other data	No numeric data
1.4 Hyperkeratotic eczema after 6 months	Other data	No numeric data
1.5 Pompholyx after 6 months	Other data	No numeric data
1.6 Chronic palmar eczema after 6 months	Other data	No numeric data
2 Primary: adverse events ‐ hyperpigmentation	Other data	No numeric data

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Primary: investigator‐rated good/excellent control of symptoms pimecrolimus cream vs vehicle	2		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
1.1 Primary: investigator‐rated clear or almost clear pimecrolimus cream vs vehicle 3 weeks	1	294	Risk Ratio (M‐H, Random, 95% CI)	1.53 [0.99, 2.36]
1.2 Primary: investigator‐rated clear or almost clear pimecrolimus cream vs vehicle 6 weeks	1	652	Risk Ratio (M‐H, Random, 95% CI)	1.28 [0.99, 1.66]
1.3 Primary: investigator‐rated clear or almost clear pimecrolimus cream vs vehicle 3 weeks irritant hand eczema	1	185	Risk Ratio (M‐H, Random, 95% CI)	1.7 [0.93, 3.10]
1.4 Primary: investigator‐rated clear or almost clear pimecrolimus cream vs vehicle 3 weeks allergic hand eczema	1	49	Risk Ratio (M‐H, Random, 95% CI)	1.33 [0.30, 5.96]
1.5 Primary: investigator‐rated clear or almost clear pimecrolimus cream vs vehicle 3 weeks endogenous hand eczema	1	134	Risk Ratio (M‐H, Random, 95% CI)	1.32 [0.75, 2.33]
2 Primary: adverse events	3		Risk Ratio (M‐H, Random, 95% CI)	Totals not selected
2.1 Discontinuation because of adverse event	2		Risk Ratio (M‐H, Random, 95% CI)	0.0 [0.0, 0.0]
2.2 Application site reaction	1		Risk Ratio (M‐H, Random, 95% CI)	0.0 [0.0, 0.0]
2.3 At least 1 adverse event	2		Risk Ratio (M‐H, Random, 95% CI)	0.0 [0.0, 0.0]
2.4 Serious adverse event (not related to study)	1		Risk Ratio (M‐H, Random, 95% CI)	0.0 [0.0, 0.0]
2.5 Treatment‐related adverse event	1		Risk Ratio (M‐H, Random, 95% CI)	0.0 [0.0, 0.0]
2.6 Erythema or irritation	1		Risk Ratio (M‐H, Random, 95% CI)	0.0 [0.0, 0.0]
2.7 Itching	1		Risk Ratio (M‐H, Random, 95% CI)	0.0 [0.0, 0.0]
2.8 Warmth, stinging, burning	1		Risk Ratio (M‐H, Random, 95% CI)	0.0 [0.0, 0.0]
2.9 Herpes simplex virus infection	1		Risk Ratio (M‐H, Random, 95% CI)	0.0 [0.0, 0.0]
3 Secondary: participant‐rated reduction in severity pruritus relief between pimecrolimus 1% and vehicle	1		Risk Ratio (M‐H, Random, 95% CI)	Totals not selected

Outcome or subgroup title	No. of studies	Statistical method	Effect size
1 Primary: investigator‐rated very good or good efficacy	1	Risk Ratio (M‐H, Random, 95% CI)	Totals not selected
2 Primary: participant‐rated very good or good efficacy	1	Risk Ratio (M‐H, Random, 95% CI)	Totals not selected
3 Primary: adverse events	1	Risk Ratio (M‐H, Random, 95% CI)	Totals not selected
3.1 At least 1 adverse event	1	Risk Ratio (M‐H, Random, 95% CI)	0.0 [0.0, 0.0]
3.2 Discontinuation due to dizziness, vomiting, and facial oedema	1	Risk Ratio (M‐H, Random, 95% CI)	0.0 [0.0, 0.0]
3.3 Discontinuation due to severe insomnia	1	Risk Ratio (M‐H, Random, 95% CI)	0.0 [0.0, 0.0]
3.4 Increase in serum creatinine > 30%	1	Risk Ratio (M‐H, Random, 95% CI)	0.0 [0.0, 0.0]
4 Secondary: investigator‐rated reduction in severity in total disease activity score (6 weeks; bigger reduction in severity = better outcome)	1	Mean Difference (IV, Random, 95% CI)	Totals not selected

Methods	Within‐participant, randomised controlled study (left‐right design). This study was carried out in the secondary care setting; it was a single‐centre study conducted in Germany
Participants	15 participants at least 18 years old suffering from chronic relapsing dyshidrotic hand eczema with a minimal duration of 1 month that was resistant to conventional therapies Dropouts: 4 Inclusion criteria of the trial At least 18 years old Chronic relapsing dyshidrotic hand eczema with a duration of at least 1 month Resistant to conventional therapies Exclusion criteria of the trial Other dermatological diseases Pregnancy Light therapy during the last 4 weeks Topical corticosteroids during the last week > 200 PUVA treatments in the past Medication or alcohol abuse Immune suppressive therapy Study population Gender: 8 female, 3 male Age: median 45.1 years, range 28 to 66 years
Interventions	Intervention • Middle‐dose UVA‐1 irradiation 3 times a week (cumulative dose of 600 J/cm²) in 11/15 hands during 5 weeks • Local 8‐MOP‐cream‐PUVA irradiation 3 times a week during 5 weeks (cumulative dose of 17.4 J/cm²) in 11/15 contralateral hands. 8‐MOP‐crème was applied 30 minutes before the start of irradiation Duration 5 weeks
Outcomes	Primary outcomes of the trial • Observer‐rated assessment of improvement (DASI score) Other outcomes • Adverse events
Notes	Therapeutic efficacy was shown with relatively low cumulative doses of UVA and UVA‐1 The secondary outcome ‐ reduction in severity, investigator‐rated ‐ was included but did not provide reproducible data Study authors were contacted by mail on 6 March 2014 and responded 10 March 2014 Declarations of interest: not stated Funding: not stated Sample size rationale: not stated
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	The article states that randomisation was done by an independent third person. No information regarding random sequence generation appeared in the article; however personal communication clarified that this was done appropriately: "cards with the characterisation "A" and "B" were enclosed in envelopes by a third person, mixed like a card play by a third person, then numbered consecutively by a third person and opened by the study doctor consecutively after informed consent to the study"
Allocation concealment (selection bias)	Low risk	Quote: "randombriefe wurden nach Wűrfeln von einer von der Studie unabhängigen Person erstellt" (free translation: the randomisation letter was created by an independent person after throwing dices) Comment: adequate allocation concealment as randomisation was accomplished by a third party
Blinding of participants and personnel (performance bias)   All outcomes	High risk	Quote: "eine Verblindung der Studie erfolgte nicht" (free translation: the study was not blinded) Comment: no attempts were made to blind participants or personnel
Blinding of outcome assessment (detection bias)   All outcomes	High risk	Quote: "eine Verblindung der Studie erfolgte nicht" (free translation: the study was not blinded) Comment: observers were not blinded
Incomplete outcome data (attrition bias)   All outcomes	High risk	No intention‐to‐treat analysis but per protocol (11 of 15 = less than 80%)
Selective reporting (reporting bias)	Low risk	No trial registration found; however all outcomes listed in Materials and Methods are given in the Results section
Other bias	Low risk	Baseline comparisons revealed no significant differences between groups, as within‐participant study was not applicable Diagnostic certainty: yes The study was completed

Methods	Parallel‐group, randomised controlled trial This study was probably carried out in a secondary care setting as a single‐centre study at a Department of Dermatology in India
Participants	108 participants with clinically diagnosed hand eczema included; 91 completed the study Dropouts: 17 Inclusion criteria of the trial Clinical diagnosis of hand eczema with duration longer than 6 months Exclusion criteria of the trial Pregnancy Lactating mothers Younger than 18 years or older than 65 years Any associated systemic disease (diabetes, hypertension, thyroid disorders, any renal or liver disease, malignancy, etc.) Hypersensitivity to azathioprine Study population Gender: 29 female, 62 male Age: group A mean 36.86 years, SD 11.55 years; group B mean 35.82 years, SD 10.67 years
Interventions	Intervention • Topical clobetasol propionate 0.05% cream twice daily with oral azathioprine 50 mg daily in 46 participants for 24 weeks Control intervention • Topical clobetasol propionate 0.05% cream twice daily alone in 45 participants for 24 weeks Participants were instructed to use the topical clobetasol intermittently and to stop application whenever the signs and symptoms disappeared and restart when the complaints returned Duration 24 weeks
Outcomes	Primary outcomes of the trial Not defined  Other outcomes • Reduction in severity, investigator‐rated scoring measured by the Hand Eczema Scoring Index (HECSI) at 2, 4, 8, 12, and 24 weeks • Reduction in severity of itch, participant‐rated, measured on a numerical scale from 0 to 10 • Number of exacerbations • Adverse events
Notes	The total quantity of corticosteroids used was not registered Study authors were contacted on 27 February 2014 by email and responded 1 March 2014 Declarations of interest: not stated Funding: not stated Sample size rationale: not stated
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "patients were randomised into 2 groups using block randomization" Comment: unclear from the article how these blocks were generated; however contact with study authors clarified that a valid computer‐generated table was used
Allocation concealment (selection bias)	Low risk	No information is provided in the article on how allocation was concealed from participants and investigators, but personal communication revealed that randomisation was done by a third person
Blinding of participants and personnel (performance bias)   All outcomes	High risk	Quote: "patients paid for the medicine themselves" Comment: participants were not blinded because they had to buy their own study drugs, and no placebo was used
Blinding of outcome assessment (detection bias)   All outcomes	Low risk	Quote: "an observer blinded randomized comparative trial" Comment: study authors stated that the study was observer blinded but gave no details as to how this was achieved. Contact with study authors clarified that observers were independent and were not involved in treatment nor in dispensation of study drugs
Incomplete outcome data (attrition bias)   All outcomes	Unclear risk	91 of the 108 (84.3%) included participants were analysed. No intention‐to‐treat analysis. Dropouts were not evenly distributed between the 2 groups
Selective reporting (reporting bias)	Low risk	No trial registration found. However, all mentioned outcomes are described in the Results section
Other bias	Unclear risk	No baseline comparisons were conducted or reported Diagnostic certainty: yes The study was completed

Methods	Within‐participant, placebo‐controlled, randomised controlled trial The study was conducted in the secondary setting at a single centre in Turkey
Participants	25 participants with moderate to severe bilateral hand dermatitis with a minimal duration of 6 months Dropouts: 1 Inclusion criteria of the trial Bilateral hand dermatitis Moderate to severe hand eczema with a minimal duration of 6 months 18 years of age or older Exclusion criteria of the trial Pregnancy Lactation Use of systemic immunosuppressants Other diagnosis such as urticaria, psoriasis, bacterial or fungal infection Illness in the previous 4 weeks Study population Gender: 15 female, 9 male Age: mean 35.8 years, range 18 to 63 years
Interventions	Intervention • Pimecrolimus 1% cream twice daily in 24/25 hands presumably for 8 weeks Control intervention • Placebo cream twice daily presumably for 8 weeks in 24/25 hands • Participants were followed up for the same period Duration 16 weeks (8 weeks active treatment, 8 weeks follow‐up)
Outcomes	Primary outcomes of the trial Not defined Other outcomes • Clinical response to therapy; erythema, desquamation, lichenification, oedema, vesiculation, and fissuring were scored between 0 and 4 and were controlled at 2nd, 4th, 6th, and 8th weeks of therapy • Clinical response to therapy for pruritus • At the end of therapy, participants were followed up for the same period to observe recurrences • Adverse events
Notes	Conference abstract from which only limited information can be extracted Study authors were contacted for additional information, which led to review of an additional full‐text article in Turkish. The secondary outcome ‐ reduction in severity, investigator‐rated ‐ was included but did not provide reproducible data Declarations of interest: not stated Funding: not stated in the paper, but personal communication clarified that study authors did not receive any funding Sample size rationale: not stated
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Study authors stated randomised but gave no clear description of how this was attained. Personal communication with study authors clarified that they used a card drawing system Quote: "a simple randomisation method by card drawing was performed"
Allocation concealment (selection bias)	Low risk	No information on how allocation was concealed from participants and investigators is provided in the abstract Personal communication clarified that one investigator drew the cards after participants gave their consent. The topical drugs were packed in yellow and red boxes, and staff gave participants the corresponding boxes for treatment, without knowledge about the content of those boxes
Blinding of participants and personnel (performance bias)   All outcomes	Low risk	Not clear from the abstract; however personal communication clarified that dispensation of study drug was done by a third person in boxes labelled for each hand. The vehicle and pimecrolimus were packed in similar boxes and were indistinguishable for participants
Blinding of outcome assessment (detection bias)   All outcomes	Low risk	Not clear from the abstract; however personal communication clarified that observation was done by another physician who was not involved in randomisation or drug dispensation
Incomplete outcome data (attrition bias)   All outcomes	Low risk	Unclear whether an intention‐to‐treat analysis was conducted; however 24 of the 25 included participants completed the study, which is more than 80%
Selective reporting (reporting bias)	Low risk	No trial registration found. All relevant clinical signs were scored in the symptom score, and all described outcomes were depicted in the Results section
Other bias	Low risk	No baseline comparisons were conducted or reported, as within‐participant study was not applicable Diagnostic certainty: yes The study was completed