Disease-specific health related quality of life patient reported outcome measures in Genodermatoses: a systematic review and critical evaluation

Health related Quality of life (HR-QoL) refers to the physical, psychological and social domains of health that are influenced by a person’s experiences, beliefs, expectations and perceptions. [1, 2]. Patient reported outcome measures (PROMs) assessing HR-QoL are increasingly used in dermatology to assess the impact of disease upon individuals and carers and has found utility as a method of understanding chronic lifelong disease and impairment [2, 3]. The significant psychosocial impact of dermatological disease can also result in disconnect between clinical manifestations and burden of disease [2, 3]. HR-QoL PROMs are responsive to changes in disease activity [4] and hence have high utility in assessing new therapeutic interventions, particularly when objective biochemical markers of disease activity are unknown. However, a lack of awareness of the standardised methodologies in the development of PROMs hampers widespread adoption [4‐6].

Methodologies and nomenclature regarding the development and validation of HR-QoL PROMs are well established and standardised [4‐6] (Table 1), with ‘COnsensus based Standards for selection of health Measurement INstruments’ (COSMIN) checklist [7, 8] and ‘quality criteria for measurement properties of health standard questionnaires’ [9] enabling objective evaluation of existing HR-QoL PROMs.

Disease-specific PROMs are useful in complex conditions such as genodermatoses where the content validity of generic PROMs is questionable [4, 5]. Sub-optimal methodologies in development of disease-specific PROMs may cause cascading validity issues surrounding the use of PROMs as outcomes in clinical trials, with consequent risk of recommending the use of an intervention with no significant benefit, or denying recommendation to a treatment with significant benefit to a subgroup of individuals. This is especially pertinent in genodermatoses where interventions and therapies (aside from supportive measures) are often limited [2]. Whilst generic PROMs are more accessible for use in patients with genodermatoses, their validity is questionable, particularly in the setting of severe genetic skin disease [2]. As generic dermatology PROMs are validated in the general dermatology population, when they are employed in genodermatoses cohorts, they suffer from suboptimal content validity, floor and ceiling effects, limited responsiveness and variability in utility between different genodermatoses [2, 5, 6]. Critical evaluation and recommendation of generic HR-QOL PROMs for use in genodermatoses cohorts is beyond the scope of this current review.

Currently, no critical evaluation of existing disease-specific PROMs in genodermatoses has been undertaken. Critical evaluation and recommendation of disease-specific HR-QOL PROMs in genodermatoses would aid in upholding quality standards in HR-QOL measurement in genodermatoses through identifying appropriate measures and increasing awareness of existing validity and reliability standards.

This systematic review aims to identify all published disease-specific HR-QoL PROMs in genodermatoses. Each PROM will be critically evaluated using the COSMIN Checklist [7, 8] and quality criteria for measurement properties of health standard questionnaires [9] to assess adherence to the current standards of development and validation. A brief overview of previous use of these PROMs and general recommendations for their use will be made adapting the published criteria for HR-QoL PROMs in Atopic Dermatitis by Schmitt et al. [10] providing clear recommendation for the clinical use of HR-QOL in genodermatoses.

This systematic review was registered with PROSPERO (CRD42016053301) [11].

Ovid Medline, ‘Epub ahead of print and Non-indexed citations’ Embase and PsycINFO databases were searched to identify HR-QoL PROMs in genodermatoses. A full search strategy is included in Additional file 1: Figure S1. All articles were screened for eligibility with pre-determined inclusion and exclusion criteria (Table 2) by 2 independent authors (JF, MD) with any disagreements mediated by a third author (DM) with inclusion or exclusion decided by simple majority.

Evaluation of the studies describing the identified PROMs as well as their measurement properties were performed using the COSMIN Checklist [7, 8] and the ‘quality criteria for measurement properties of health standard questionnaires’ by Terwee et al. [9] respectively. The ‘Worst score counts’ method was utilised for overall assessment of each methodological quality [8]. Selection bias and recall bias of included studies was addressed in relevant sections of the COSMIN checklist [7, 8]. The evaluation of measurement properties was summarised into an overall recommendation, (graded A-D) derived from published criteria by Schmitt et al. [10] based upon OMERACT filter 2.0. Assessment was undertaken by two independent authors (JF, MD) evaluated each PROM, collated and compared results, with any disagreements mediated by referral to a third author (DF) with final decision made by simple majority. Author JF was excluded from evaluation of one of the PROMs (QOLEB [12]) due to a conflict of interest in being involved in the development of this PROM. As per the COSMIN manual [7, 8], criterion validity does not usually apply to HR-QOL related PROMs. The only exception would occur when “a shortened instrument is compared to the original long version” [7, 8]. Hence COSMIN assessment of criterion validity was not considered appropriate.

All results and discussion pertaining to the COSMIN checklist will adhere to definitions as defined by the COSMIN initiative [7, 8]. All results and discussions pertaining to the ‘quality criteria for measurement properties of health standard questionnaires’ by Terwee et al. [9] will adhere to definitions as per Terwee et al. [9]. Comparison of these definitions is provided in Table 1.

The PRIMSA flow diagram is presented in Fig. 1. Of 693 non-duplicate citations, 303 were removed after screening with eligibility criteria (Table 2). 390 full text articles were assessed, with 309 articles excluded in line with eligibility criteria. 5 articles pertaining to 2 conditions (Osteogenesis Imperfecta and Chronic Granulomatous Disease) were excluded during data extraction due to these conditions not being primarily dermatological in nature. 76 articles (listed in Additional file 2: Figure S2) underwent full text screening by 2 independent authors (JF and MD) to identify 15 papers reporting the development of 15 disease-specific PROMs across 10 discrete genodermatoses. Genodermatoses included congenital ichthyoses [13, 14], pachyonychia congenita [15‐17], darier’s disease [18], neurofibromatosis type 2 [19], hereditary haemorrhagic telangiectasia [20, 21], hereditary angioedema [22, 23], epidermolysis bullosa [12, 24], fabry disease [25], basal cell naevus syndrome [26, 27], and peutz -jeagher syndrome [28]. The remaining 61 papers employed generic PROMs or evaluated HR-QoL using the PROMs in the specific condition(s) as opposed to describing development and/or validation. Results of the COSMIN Checklist evaluation are presented in Table 3, with the results of evaluation using ‘quality criteria for measurement properties of health standard questionnaires’ are presented in Table 4.

Two disease-specific PROMs exist for QoL evaluation in inherited ichthyoses: The IQoL-32 [13] which is designed for evaluation of HR-QoL in individuals with ichthyoses; and the Family Burden Ichthyosis (FBI) Questionnaire [14] designed for assessment of HR-QoL impact upon those caring for an individual with ichthyosis. To date no interventional trials have utilised these PROM. Assessment of QoL in Pachyonychia congenital (PC) has involved the use of the IPCRR Questionnaire [15, 16], which combines questions pertaining to PC phenotype with HR-QoL evaluation [15]. The IPCRR has not been used thus far to evaluate HR-QoL in interventional trials [15, 16]. The darier disease health related quality of life questionnaire (DD-QOL) has been developed [18] but is yet to be utilised in assessment of therapeutic interventions. Validated HR-QoL PROMs (including the NFTI-QoL) [19] exist for visual and neurological impairment associated with neurofibromatosis (NF) [19]. The NFTI-QoL has been used for measuring longitudinal changes in HR-QoL in NF patients as well as evaluating the impact of auditory brainstem implants [29, 30]. The EQ-QOL is an epistaxis-specific HR-QoL questionnaire (EQ-QOL) [20] for use in hereditary haemorrhagic telangectasia (HHT) and has highlighted the frequency of epistaxis, duration of epistaxis and history of prophylactic embolisation as predictive of HR-QOL impact in HHT via EQQOL scores [20, 21]. It has been used in 1 randomised control trial of iron supplementation in HHT [31] as well as two observational studies [32, 33]. The IHAE-QOL [22] (Impact of Hereditary Angioedema) is a validated disease-specific HR-QOL PROM developed for hereditary angioedema patients [28]. The HAE-PRO [23] is similarly designed for hereditary angioedema patients and is currently undergoing further validation. Both PROMs are yet to be used in observational or interventional studies [22, 23]. In Epidermolysis Bullosa, the QOLEB [12] is a HR-QoL PROM for individuals, whereas the EB-BoD [24] is designed to evaluate the burden of disease upon family members. The QOLEB has been utilised in observational [33‐36] as well as interventional studies [37‐39]. The fabry-specific FPHPQ [25] BCCNS QoL Questionnaire (aBCCdex) [26, 27] and Peutz-Jeagher Syndrome (PJS) Questionnaire [28] all measure HR-QOL in individuals affected by respective conditions and have not yet been utilised in published interventional studies.

Of the fifteen disease-specific PROM in genodermatoses, none could be recommended for current use, with an additional seven appropriate for use, but only recommended if deficiencies in the domains in Tables 3 and 4 are addressed in future studies.

Deficiencies identified in development included internal consistency (3 ‘Poor’ ratings and 3 PROMS not evaluated), reliability (two ‘Poor’ ratings and 5 PROMs not evaluated) Content validity was the strongest domain evaluated with seven ratings of ‘Excellent’. Cross cultural validity was largely unreported, although, when completed, the standard was generally acceptable. Deficiencies in measurement properties included floor and ceiling effects (1 negative rating and 2 indeterminate ratings) and agreement (9 indeterminate ratings). No data was available for responsiveness. Some confusion was evident in the variable nomenclature of measurement properties and we would suggest adherence to a pre-defined set of defintions (eg COSMIN definitions) for consistency. To maximise generalisability, PROMs require validation in ethnically diverse contexts [41] as well as adolescent and paediatric populations [42]. Tailoring PROMs to age-specific ranges maximises patient understanding, as well as eliminating concern over parental proxy over-reporting of disease burden [43].

The strengths of this review include the evaluation of both the development (using COSMIN criteria) as well as the measurement properties of the identified PROMs. Providing overall recommendations regarding the quality of PROMs in genodermatoses will provide researchers and practicing dermatologists (including those with little knowledge of outcomes research methods) with the evidence to select the most appropriate PROM for assessment of a genodermatosis of interest. The limitations of this review include the late registration of this study with PROSPERO, after the commencement of formal screening of studies. Despite this late registration all methodology for this systematic review were pre-specified prior to the commencement of review. However, the delay in registration may potentially contribute to review bias in study eligibility criteria as defined by the ROBIS [44] Group.

HR-QoL assessment in genodermatoses with generic PROMs demonstrates deficiencies in content validity and sensitivity. There is a need for a profound paradigm shift in the standard to which HR-QoL PROMs in dermatology are held to account. Of the fifteen PROMs identified in none met current COSMIN standards for recommendation. Valid and reliable disease-specific PROMs do exist however many existing PROMs require additional psychometric analysis. Dermatologists need to be aware of the deficiencies in existing PROMs in order to support high quality data in clinical trials involving genodermatoses.