Introduction Chronic pruritus (CP) is a symptom of dermatologic, neurologic, systemic and psychosomatic diseases. CP has a prevalence of ~20% in the general population and is therefore a significant burden on society, but the transition from acute pruritus to CP is not well understood. It probably involves interactions between biological and psychosocial factors and pruritus-specific risk factors as well as mechanisms shared with other persistent somatic symptoms addressed in other projects of the SOMACROSS Research Unit (RU). Here we aim to identify psychosocial and biological factors and their interactions which might be associated with the persistence of CP with and without immunologic/inflammatory origin, that is, atopic dermatitis and pruritus on non-inflamed skin. We expect that psychosocial factors relevant to the persistence of symptoms such as fatigue and pain may also show associations to CP.
Methods and analysis In this prospective, exploratory observational study situated in Germany, three cohorts of 40 patients each with acute exacerbation of atopic dermatitis and chronic atopic dermatitis and 40 CP patients with unaffected skin will be recruited for a comprehensive translational investigation including pruritus-specific and the shared psychosocial assessments of the RU SOMACROSS. Pruritus-specific measures will include questionnaires, quantitative sensory testing, cutaneous nerve fibre morphology, skin barrier morphology, epidermal metabolism and pruritogen blood levels. Within 1 year, patients and 80 age-matched and sex-matched healthy controls will be examined at three time points, allowing cross-sectional comparison and a longitudinal investigation of predictive outcome factors in patients under treatment according to existing guidelines.
Ethics and dissemination The study has been approved by the ethics committees of Hamburg (2020-10200-BO-ff) and Münster (2020-676 f-S), Germany. All participants are required to provide written informed consent. Findings will be disseminated through peer-reviewed publications, scientific conferences and involvement of relevant stakeholders, patients and the lay public.
Trial registration number DRKS00026646.
- Adult dermatology
- MENTAL HEALTH
- Adult psychiatry
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.
Strengths and limitations of this study
A strength of this study is the comprehensive methodological approach (including biological and psychosocial aspects) in three distinct groups of pruritus patients (acute and chronic atopic dermatitis and pruritus on non-lesional skin), which has not been attempted before.
The novel techniques available in this project (atomic force microscopy, hyperspectral imaging and multiphoton tomography equipped with a fluorescence lifetime imaging module), which have not been applied to patients with atopic dermatitis and chronic pruritus on non-lesional skin before, are another strength of the project.
A further strength is the follow-up over 12 months and the comparison with a control group of healthy participants.
The methodological crosslink with the other projects of the RU SOMACROSS and the use of a common set of instruments in all projects is a methodological strength, as it is a novel strategy that allows us to compare results between projects and to distinguish between symptom-specific and generic mechanisms.
A limitation is the fact that, despite investigating multiple potential risk factors and mechanisms of the persistence of chronic pruritus, other variables, which are not addressed in this project, might be relevant.
Chronic pruritus (CP), also known as chronic itch, is defined as the presence of itching that lasts for 6 weeks or more. CP is a symptom in many dermatologic, systemic, neurologic and psychosomatic diseases. It has a high burden on society, with a prevalence of ~20% in the general population and across various medical specialities.1 2 The onset and course of CP can be influenced by different psychosomatic aspects and psychologic comorbidities.3 4
CP therefore is classified as a persistent somatic symptom (PSS) and fits conceptual models of PSS aetiology.5 Like other forms of PSS, CP is thought to arise from complex interactions between the brain, body and environment (figure 1). Thus, this fits well into the interdisciplinary Research Unit SOMACROSS, which has been presented by Löwe et al6 in this journal.
Not all of the biological, psychological and social factors and their interactions that are associated with the transition from acute pruritus to CP are well understood. Persistent CP is likely to be associated with pruritus-specific pathophysiological, psychological and social factors and also with mechanisms that are shared with other forms of PSS. A comprehensive model of the aetiology and persistence of CP still needs to be validated.
Acute pruritus may be triggered by chemical, physical or mechanical stimuli. Pruritus may be induced by pruritogens released from inflammatory cells and keratinocytes, as well as nerve fibres of the peripheral or central nervous system. Cutaneous group C and group A nerve fibres carry corresponding receptors for interleukins (ILs) such as IL31 and IL4, histamine, or neuropeptides such as substance P and somatostatin.1 7
Atopic dermatitis (AD) is the most prevalent pruritic dermatologic disease8 and the most widely studied model of immunologic and skin-barrier related pruritus mechanisms. AD can occur in acute phases, and chronic phases with persistent pruritus and subacute eczema. At all stages, patients suffer from intense pruritus regardless of the severity of the eczema.9 On the other hand, CP may also occur on non-inflamed, non-lesional skin (CPNL), for example, due to systemic, neurologic or other diseases.
Skin barrier dysfunction is widely considered to be the primary event in AD and is pivotal for pruritus.10 The outermost layer of the epidermis (stratum corneum) is the most relevant skin layer for the cutaneous barrier function. In AD skin lesions, the expression of tight junction proteins such as claudin-1 is reduced, and the stratum corneum becomes ‘leaky’ and shows dramatic morphological and structural changes shown by atomic force microscopy (AFM). Furthermore, the skin barrier is also impaired in clinically non-lesional AD skin. The skin barrier dysfunction, along with a biphasic Th2/Th1 response, the release of pruritogenic cytokines, epidermal hypoinnervation and neuronal hypersensitivity, are considered essential for the induction and most likely also the persistence of pruritus, while the associations with psychosocial factors are unclear.11
Neural hypersensitivity of the peripheral and/or central nervous system has been demonstrated in AD and CP of other origin.1 4 12 In this state, nerve fibres are more vulnerable to pruritogens, the corresponding receptors/ion channels have a lower activation threshold and neuromediator production in nerve fibres is enhanced.4 For example, cutaneously applied pruritogens evoke pruritus that is more intense and longer lasting than usual (hyperknesis).1 13 In contrast, alloknesis describes the induction of pruritus by non-pruritogenic stimuli such as mechanical triggers.
Mechanisms associated with neuronal hypersensitivity are largely unknown but may include chronic stimulation with inflammatory pruritogens such as IL31 and IL4,4 permanent scratching,14 15 the depletion of cutaneous Merkel cells,16 psychological and/or social factors.17
Furthermore, several previous studies have shown that social and psychological factors are related to an enhanced perception of pruritus (figure 2).
Positive and negative expectations, which can also be induced by corresponding placebo and nocebo instructions, have been shown to influence the intensity of pruritus, scratching behaviour and even the physiologic skin reactions after experimentally induced pruritus in healthy controls18 19 and in patients with AD.20 Anxiety, depression21 and personality traits such as neuroticism (ie, emotional instability with the tendency to experience more negative emotions), low agreeableness21 22 and low self-efficacy23 were associated with a stronger perception of pruritus in patients with AD, psoriasis and in healthy controls. Self-reported pruritus has been repeatedly linked to self-reported perceptions of stress and stressful life events in population samples20 23 and also in AD patients.24–27
Biologically, patients with AD may show a response to stress in terms of T-cell and mast cell activation, expression and release of neurogenic inflammatory mediators and disruption of normal epidermal barrier function.27 Stress has been shown to deteriorate the skin barrier by activating 11beta-hydroxysteroid dehydrogenase 1 and the hypothalamic-pituitary-adrenocortical (HPA) axis.28 The association between stress and pruritus can also be mediated by negative pruritus-associated cognitions and expectations in AD patients (figure 2).26
Taken together, the results of multiple studies confirm psychological factors to be risk factors for an enhanced perception of pruritus, but the extent to which these factors are associated with the persistence and chronicity of CP remains unclear (figure 1).
Novelty of the study
CP has biological, psychological and social dimensions and the interactions between these factors are poorly understood. Data gathered so far were mostly derived from cross-sectional studies not set out to identify such interactions. We will therefore apply a comprehensive set of instruments to analyse biomarkers of the skin and blood, the function and anatomical structure of the skin barrier, the peripheral nervous system and its interaction with the central nervous system, as well as psychological and social factors in two key forms of CP (inflammatory and non-inflammatory). In addition, we will follow the course of pruritus over 12 months and relate the progression of the disease to the above parameters in order to identify psychosocial and biological factors that are associated with the persistence of symptoms. Such a comprehensive approach has not been attempted before, and we will increase the value of the work by studying three distinct groups (two inflammatory and one non-inflammatory) of pruritus patients. The crosslink with the other projects of the RU SOMACROSS and the use of one common set of instruments in all projects is a completely new strategy that allows us to compare results between projects and to distinguish between symptom-specific and generic mechanisms. The novel techniques available in this project have not been applied to patients with AD or CPNL before.
The overall aim of the project is to identify psychosocial and biological factors (and their interactions) that are associated with the persistence of symptoms in CP. The psychosocial factors relevant to the persistence of symptoms in conditions such as fatigue and pain may also be associated with CP. Therefore, the project has the following main objectives:
To characterise psychosocial factors: (i) associated with the persistence of CP in general, and (ii) the persistence of CP of cutaneous inflammatory and non-inflammatory origin in particular.
To determine whether there are associations between psychosocial risk factors and specific peripheral and/or central neuronal sensitisation patterns in CP.
To determine covariation of modifiable psychosocial risk factors and biological factors (eg, cellular metabolism and skin barrier function).
Psychosocial (personality, stress, expectations, catastrophising, illness perceptions, anxiety and depression) and biological factors (central sensitisation processes, cutaneous neuroanatomy, cellular metabolism and skin barrier function) interact in promoting the persistence of pruritus.
Methods and analysis
Study design and participants
Patients with AD will be included as a model of cutaneous inflammatory pruritus, whereas patients with CPNL (CP with pruritus on non-lesional skin) will be included as a model of non-inflammatory pruritus. AD starts in childhood and we cannot capture the long-term follow-up starting at the onset of the disease, so we will recruit two stages of AD: a group of adult patients (group 1; n=40) presenting with acute AD (after a manifestation-free interval of at least 6 months) and a group of adult patients (group 2; n=40) presenting with chronic AD. These groups will be subject to cross-sectional and longitudinal comparisons. We will also recruit CPNL patients (group 3; n=40) in which CP has typically developed over the course of several years. We will compare the patient groups cross-sectionally with age/sex-matched skin-healthy controls (group 4; n=80) (figure 3).
All projects of RU SOMACROSS share common basic inclusion criteria, that is, age≥18 years, sufficient oral and written German language proficiency to complete self-report questionnaires and interviews and written informed consent. In addition to these common criteria, the project-specific inclusion criteria are listed below.
Group 1: CP in acute AD (n=40)
Acute eczema for ≤4 weeks at inclusion, no eczema in the previous 6 months
Group 2: CP in chronic AD (n=40)
AD and CP for at least 6 weeks, chronic eczema
Group 3: CP in CPNL (n=40)
CP for at least 6 weeks, clinically no dermatosis, no prurigo
Groups 1–3: pruritus intensity ≥5, worst intensity in last 24 hours on numeric rating scale (NRS, range 0–10).
Group 4: Healthy controls (n=80)
No pruritus (NRS in last 24 hours=0), clinically normal skin, no atopic predisposition
All projects of RU SOMACROSS share common basic exclusion criteria including serious illness requiring immediate intervention, florid psychosis or substance abuse disorder, and acute suicidality.
In addition, the project-specific exclusion criteria are the following:
Chronic pain syndrome, diagnosis of fibromyalgia, chronic intake of analgesics
Intake of systemic steroids, systemic immunosuppression in the past 4 weeks
Topical therapy of target area with steroids in the past 2 weeks
Allergy against substance used for local anaesthesia.
We included chronic pain conditions including fibromyalgia, as well as intake of analgesics as exclusion criteria, since these conditions and drugs may affect functional assessments of quantitative sensory testing and alloknesis/allodynia. Additionally, we included certain topical and systemic agents (namely systemic steroids, systemic immunosuppressants and topical steroids) as exclusion criteria, as these drugs may affect the expression of relevant markers/genes in the skin and make difficult the interpretation of findings obtained from skin biopsies. Allergy to local anaesthetics is also an exclusion criterion, since local anaesthesia is needed in order to perform the skin biopsies
The ideal design to identify chronicity factors for the symptom pruritus would be a prospective longitudinal investigation starting at the first manifestation of the pruritic condition, which is known to persist in a certain percentage of cases (approximately 60%–70% of patients; analysed by our own cohort). However, this is not feasible in pruritus for the reasons stated above. We therefore propose a combined design of patients showing acute inflammatory pruritus and a later stage of chronic inflammatory pruritus together mimicking the long-term course of the disease. These groups will be contrasted to patients with CP in a non-inflammatory background and followed up for 1 year. Therefore, the study design is prospective, exploratory and observational.
We expect that each group will include patients whose itch symptoms improve. This is considered clinically relevant following a decrease of at least two points on the NRS (range 0–10). We also expect that some patients will show less than a 2-point NRS improvement, some will not improve or the symptoms may get worse. We will divide the patients at 12 months into a PSS group (0–1 points change on the pruritus NRS) and a non-PSS group (decrease of ≥2 pruritus NRS points).
This design enables us to identify the similarities and differences between patient groups with acute pruritus and CP and between those with and without PSS:
A cross-sectional comparison between acute AD, chronic AD, CPNL and healthy controls.
A longitudinal investigation of predictive outcome factors in AD and CPNL under treatment according to existing guidelines.
Patients will be followed up for 12 months with three visits (baseline, 6 months and 12 months). During each visit, we will obtain patient-reported outcomes (RU SOMACROSS core set of questionnaires, pruritus characteristics and course, quality of life, itch questionnaires and scratch pleasure rating questionnaires). Structured clinical interviews, a clinical examination, quantitative sensory testing (QST) and psychophysical/alloknesis testing, biopsies, blood collection and, in a subset of 10 patients from each group, non-invasive multiphoton tomography equipped with a fluorescence lifetime imaging module (MPT-FLIM), hyperspectral imaging (HSI), AFM and trans-epidermal water loss measurement (TEWL) will be performed at each visit (summarised in figure 4 and table 1).
Data collection and analyses
Sociodemographic and psychosocial variables
In order to assure comparability of the projects within the Research unit, this study employs the same questionnaires to assess sociodemographic and psychosocial variables as the other projects of the Research Unit Somacross—the so-called ‘core set of instruments’. These have been listed and described in detail in the publication of Löwe et al6 in this journal. The ‘core set’ of instruments assesses sociodemographic and psychosocial factors, cognitive-perceptual and emotional mechanisms and behavioural factors by well established and validated self-assessment questionnaires. For details, please consult Löwe et al6 in this journal.
Quantitative sensory testing (QST) alloknesis
Sensitisation profiles of patients and healthy volunteers will be analysed using a comprehensive battery for QST and tests for mechanically induced pruritus (alloknesis). QST is a validated psychophysical test protocol established in both centres. We will determine thermal and mechanical detection and pain thresholds as well as responses to supra-threshold stimuli. QST reports the gain or loss of function among peripheral nerve fibre populations and may indicate the presence of central sensitisation. In alloknesis testing, stimuli that usually do not induce itching (eg, heat or touch) will be tested for their ability to evoke abnormal itching in sensitised participants.
After quantitative sensory testing and alloknesis testing, 4 mm biopsies will be obtained from each participant. In CPNL patients, biopsies will be obtained from non-lesional pruritic skin, whereas in acute and chronic AD patients, biopsies will be taken from lesional pruritic skin. Biopsies from non-lesional non-pruritic skin will be obtained in all patients and in the healthy controls. An additional biopsy will be obtained from healed non-pruritic skin from matched locations at month 12 from the patients. The biopsies at month 12 will be performed in the same anatomical area to avoid inter patient differences.
Biopsies will be divided in two, one part for RNA/protein expression analysis and the other for histopathology.
The fixed biopsy samples will be used to analyse the epidermal and dermal nerve fibre density and anatomy by staining 30 µm sections for the pan-axonal marker PGP9.5. The basement membrane will be visualised by staining for collagen IV in order to determine the precise point at which nerve fibres invade the epidermis.
The unfixed biopsy samples will be used for the detection of relevant pruritogens and mediators (NGF, artemin, semaphorin 3A, IL1β, IL6 and IL31) at the RNA and protein levels. Samples stored in RNA later will be processed by company instructions.
Finally, blood samples from each participant at baseline and from the CP patients at month 12 will be used for the analysis of serum levels of CRP, IL1β, IL6, IL31 and tumour necrosis factor α.
Characterisation of CP-related alterations of keratinocyte metabolism and skin barrier integrity
A clinical panel will be enrolled at the study centre UKE for the objective characterisation of CP by HSI and MPT-FLIM. The panel will comprise 12 patients each with acute AD, chronic AD and CPNL and 24 healthy controls. Intravital images will be taken from two lesional or non-lesional skin areas per subject at the volar forearm or antecubital fossa and aligned with photographic documentation and clinical scores gathered by dermatologic consultants. We will evaluate the following outcome parameters: high-resolution images of the epidermal and dermal architecture (eg, showing the thickness of epidermal layers, cell size and degree of spongiosis), peripheral blood perfusion and oxygenation/haemoglobin saturation, metabolic status of epidermal skin layers at the single-cell level based on the free-to-bound Nicotinamid Adenin Dinucleotid Hydrid (NADH) ratio and mitochondrial distribution analysis.29
In parallel to HSI and MPT-FLIM measurements, we will investigate skin areas by AFM30 following the acquisition of stratum corneum samples by tape stripping. Subsequent image analysis will reveal the size of individual corneocytes, intercellular gaps and the number of corneodesmosomes. By combining these data with immunofluorescence microscopy, we will also identify the tight junction protein claudin-1. To supplement the AFM measurements, we will therefore determine TEWL (CE-certified medical device strictly used in its intended purpose/use) values as a common correlate of skin dryness, disease severity and itchiness.
Assessment and study outcomes
Given that the severity of somatic symptoms must be specifically assessed for each symptom in RU SOMACROSS, and that generic instruments are needed for comparisons and common evaluations across projects, somatic symptoms will be measured in two ways: (a) symptom-specific primary outcome: intensity of pruritus, as assessed by worst NRS score in the past 24 hours, (b) generic instruments: generic assessment of overall symptom severity using the well-established patient health questionnaire 15 (PHQ-15)31 and the NRS for symptom intensity recommended by the EURONET-SOMA group.32
Pruritus characteristics (intensity, quality and duration) and course (Itch Controlled Days)
Pruritus-related quality of life assessed by ItchyQoL
Scratch pleasure rating (NRS)
Function of group C and group A nerve fibres (QST, psychophysical results)
Cutaneous nerve fibre anatomy, biomarker (PCR, immunohistochemistry, blood markers)
Nanostructure of the skin barrier, metabolic changes in keratinocytes
Generic secondary outcome measures:
Interference by additional symptoms, disability and quality of life.
Comorbidity of somatic symptom disorder
Pruritus-specific: Pruritus characteristics (intensity, quality and duration) and course (Itch Controlled Days).
Generic predictors: The set of core instruments reflects the most relevant variables for the development of persistent symptoms, as proposed by the RU’s working model.
In addition to the project-specific outcome variables (table 1), we will apply all additional joint core instruments of the RU SOMACROSS (as presented by Löwe et al6 in this journal) in our project to contribute to the main research aim of RU SOMACROSS, that is, to identify risk factors and mechanisms for the persistence of somatic symptoms across diseases.
Validated itch questionnaires (including pruritus-specific quality of life) and the core instruments of the research unit will be presented to all participants at baseline, 6 months and 12 months.
Participants will be tested for peripheral and central neuronal sensitisation by QST, alloknesis testing and the detection of laboratory inflammation markers (baseline, 6 months and 12 months). Sensitisation patterns will be correlated with findings from (a).
Sample size estimation
Previous data are not available for our outcome measures, so we will conduct an exploratory study. However, we based the sample size on our previous work. Given our previous results, we anticipate normally distributed data for intraepidermal nerve fibre (IENF) density in CP patients with AD (A) and healthy controls (A*).12 Conservative corrected values predicting the reliability of µA=10 and µA*=14.5 (σA=5, σA*=6) lead to a difference (µA – µA*) of –4.5 and the conservative corrected SD σ = (σA2 + σA*2 ½=7.8). Taking this into account, we propose to reject the hypothesis H0 (There is no difference between the groups in IENF; two-sided testing, significance level=5%) with a power of ~84% and a sample size of 30 per group. We assume similar differences for other comparisons of cutaneous nerve fibre characteristics between the groups, but a valid power calculation is not possible due to the lack of data. These analyses will therefore be interpreted as exploratory (hypothesis generating) rather than confirmatory. To avoid loss of power due to overoptimistic assumptions, we will analyse larger samples (n=40 per group) in the proposed study. For single-cell MPT-FLIM and AFM experiments, we expect (based on our preliminary data) a minimal effect size of f=1.4 for the comparison of lesional AD and healthy skin,33 which leads to nine patients per group (power: 1–β=0.8, α=0.05).
During the first funding phase, we will test paths and associations between psychosocial and biological predisposing and triggering factors with the primary and secondary outcome measures (pruritus-specific as well as generic) using an exploratory approach. Pruritus data are scarce, as is the case for most other diseases investigated by our research group. We therefore included a large number of variables in the first funding phase, enabling us to generate hypotheses for rigorous testing during the second funding phase. The statistical evaluation will be carried out by the biometric experts of the RU SOMACROSS (Z-Project) using cross-sectional (comparisons between the three patient groups and healthy controls) and longitudinal (comparisons across the measurement points) as well as combined approaches, as appropriate. Missing data will be imputed if more than 5% of the data are missing. In accordance to White et al,34 the number of imputations will be chosen dependent on the proportion of missing data. To evaluate subgroup effects, interaction tests will be performed. The statistical analysis of data generated during the first funding phase will reduce the number of predictors and confounders of symptom persistence by removing irrelevant pathways, which will make confirmatory analysis possible during the second funding phase. In the second funding phase, the predictors and confounders identified in the first funding phase will be examined using confirmatory designs. Given that the first funding phase is exploratory, we have not adjusted our statistical methods for multiple testing in order to avoid a loss of power.
Current status and timeline of the study
Patient recruitment started in October 2021 and the first patients have been included. We expect to complete the study within 4 years. The work schedule and milestones are reported in figure 5.
Patient and public involvement
It was not appropriate to involve patients or the public in the design of our study. However, the main findings will be published in peer-reviewed journals and made publicly available. In addition, we will communicate scientific results in lay language via press releases, social media, and patient forums.
Impact and contribution to overall project objectives
In this first exploratory phase of the project, we expect to identify biomedical and psychosocial factors associated with the persistence of pruritus, thus generating a predictive model that we intend to replicate and validate for other skin diseases during a planned second funding phase. There are mechanistic and clinical similarities between fatigue, pain and pruritus. Pruritus is also well known in kidney and liver diseases. These are conditions which are examined in other projects of the RU SOMACROSS. The data generated in the first funding phase are relevant for our objectives and will allow us to generate hypotheses on the mechanisms of pruritus persistence, which will be tested during the second funding phase.
In matching our results with those obtained from all the other projects, we expect to identify generic mechanisms associated with the persistence of pruritus, fatigue, pain and other symptoms. Thus, we will contribute to the overall aims of the research group, that is, to differentiate between disease-specific and generic factors affecting the persistence of symptoms and to generate an overarching model.
Ethics and dissemination
The study was approved by the Ethics Committees of the Medical Associations Hamburg (2020-10200-BO-ff) and Westphalia-Lippe/Westphalian Wilhelms University, Münster (2020-676 f-S), Germany. The study will be conducted in accordance with the WMA Declaration of Helsinki, guidelines for Good Clinical Practice, national and local laws. All participants are required to provide written informed consent. Furthermore, the ethical considerations published by Löwe et al6 also apply for this project.
The main findings will be published in peer-reviewed journals and made publicly available. In addition, we will communicate scientific results in lay language via press releases, social media, and patient forums.
Patient consent for publication
GS and SS are joint first authors.
GS and SS contributed equally.
Contributors GS, SWS and SS conceptualised the study, wrote the project protocol and applied for funding. At present, they are supervising the project staff. SK, MPP, GF, CM, VH and KA are involved in the preparation of the study start and in data collection. All authors contributed to the manuscript and critically revised the manuscript draft and its present revision.
Funding The study is funded by the German Research Association (DFG), grant numbers SCHN 657/5-1; STA 1159/7-1; SCHN 474/9-1.
Competing interests None declared.
Patient and public involvement Patients and/or the public were not involved in the design, or conduct, or reporting or dissemination plans of this research.
Provenance and peer review Not commissioned; externally peer reviewed.