A review of quality of life themes in Duchenne muscular dystrophy for patients and carers

Searches of MEDLINE (including Medline in Process via OVID), EMBASE (via OVID), and CINAHL (via EBSCO) medical electronic databases were conducted to identify evidence relevant to QoL in people with DMD as well as the effects on family and carers. In order to review the relevant evidence within allocated time constraints, abbreviations to comprehensive review techniques were made including limiting searches to studies published in the English language and those published between 2010 and 2016. In each database, search terms for DMD were combined with four different search statements to identify a variety of research designs. Free text and thesaurus terms were used in combination to maximise sensitivity in the searches using filters. A sample search strategy is provided in Additional file 1.

Titles and abstracts of references retrieved from the searches were assessed for inclusion against a priori established eligibility criteria by one reviewer. Reference lists of included studies were manually searched for further relevant studies that were not captured by the searches. Table 1 outlines the inclusion and exclusion criteria used to screen studies that would be eligible to the review. Studies of interest included primary research studies of people with DMD investigating issues related to the patient, the carer or the family’s QoL. This included research designs such as surveys, questionnaires, clinical trials, qualitative research or other designs of QoL research.

Studies selected for inclusion were read in full, annotated and coded to extract themes relevant to QoL from the results and discussions sections. One reviewer performed data extraction for QoL themes without blinding to study author. Themes were then pooled across studies into a master database and grouped firstly into domains: i) Physical; ii) Psychological; iii) Social; iv) Well-being as these are commonly encountered domains in general QoL research. Sub-domains within each domain were then created to capture the variety of data from the included studies. Themes that did not fall under any of the four initial domains were grouped under an “Other” domain. This framework of themes was reviewed by one further review author to verify agreement of coding. As the review is not an aggregation of data to provide a pooled treatment effect, or to rank findings based on frequency, no quality assessment or risk of bias assessment was deemed necessary.

Table 2 provides the details of the 45 included studies, published from 2010 to 2016. Three studies were conducted in the UK whilst another four were conducted in the UK in combination with other European countries. Nine were conducted in the United States, 16 in European countries and 13 from other countries around the world. The number of participants (N) in the studies ranges from 10 to 770 with a median of 60. Cross-sectional studies of DMD patients or carers were the most common study design (n = 27), followed by case-control (n = 8), cohort/before and after studies (n = 6), tool development only (n = 2). There was one mixed-methods study (audit, interviews and survey) and one qualitative study. The majority of studies focused on children and adolescents with DMD (n = 30) whilst nine studies were of parents/ carers of people with DMD. Six studies were interested in both patient and parent/carer perspectives.

Most studies reported measuring QoL or health-related QoL as the primary outcome of interest. However, some studies were not strictly investigating QoL but did study and report issues that could be of potential relevance to QoL. A variety of named questionnaires and instruments were employed across the studies (n = 36) with 18 studies employing multiple instruments. Five studies employed both generic instruments and a specific instrument for paediatric neuromuscular disorders [17‐21]. A condition-specific instrument for QoL in DMD does not currently exist. Table 3 documents the instruments reported across the 45 included studies.

A total of 36 validated instruments were used across the 45 studies. The most commonly employed instruments were the Pediatric Quality of Life Inventory (PedsQL) (n = 9), the Short Form-36 (SF-36) (n = 6), and the World Health Organization Quality of Life (WHOQoL) (n = 5). A small proportion of studies did not provide specific details for the instruments used (n = 7) [18, 22‐27] or they used non-validated questionnaires or interview schedules (n = 8) [28‐35] or visual analogue scales (n = 2) [20, 36].

Three of the included studies focussed mainly on the development of instruments measuring QoL in the patient population [21, 35, 37]. The remaining 42 studies covered a large variety of topics relevant to people with DMD and their carers’. Themes from all 45 studies were extracted and coded into the following five domains: i) Physical; ii) Psychological; iii) Social; iv) Well-being and; v) Other. The first four domains were derived from an initial a priori framework to encompass the commonly anticipated domains in general QoL research. Based on the number, variety and recurrence of themes identified across the included literature, the domains were further divided into relevant sub-domains as presented in Table 4. An “other” domain was created to capture themes recurring and potentially relevant that may be regarded independently to the four initial domains. Themes were assimilated across studies into one framework to encompass and summarise the diversity of research.

The domains and sub-domains were derived from either studies measuring self-reported QoL of people with DMD, parent proxy estimations, parent/carer/family specific or general public estimations. Themes are presented by domain and sub-domain for people with DMD in Tables 5 and carers of people with DMD in Table 6.

Particular trends in the data were noted during the course of the review. The tendency for the general public to estimate the HRQoL of people with DMD lower than the patients themselves, or the so-called disability paradox [38], was observed indicating the potential for preference-based measures that necessarily incorporate public views to underestimate QoL in DMD. Two studies documented substantial differences between public preference estimates for QoL in DMD compared to patients [20, 39]. These variations are in line with previous literature [38, 40] on the topic of incorporating different perspectives in health state valuation and highlight the need to fully consider the reasons behind these differences and the potential consequences of low agreement.

Agreement between patients and their carers tended to be consistent across domains. However, as noted by other research [13‐16] a number of studies found that patients with DMD rate their QoL more favourably by self-report compared to their caregivers’ proxy report [19, 39, 41, 42]. A variety of reasons for this are posited across these studies. Parents own fears and worries about their child’s disease may influence their assessment of their child’s HRQoL [19]. Children may have adapted to their illness better than their parents have and parents may not always have the most accurate assessment of their child’s internal state [19]. Also, parents may reflect on their own frame of reference by interpreting the limitations of their sons or they may be more sensitive to negative comments of the environment [42]. Reasons postulated for low agreement between self- and proxy report appear most relevant to differential processes of adaptation between children and their parents.

Alternatively, if the notion that parents underestimate their child’s QoL is rejected, it is conceivable that children may have overestimated their QoL under certain circumstances. For example, one study suggested that when tests were completed with help of the paediatricians who are conducting the assessment, an element of reporting bias may be introduced [24]. Patients were noted as less likely to mention pain as one of their complaints [43]. The issue of pain, and the communication of pain, is a potentially relevant factor to the issue of differences between self and proxy reports of QoL. One study [22] reports that patients were limited in their ability to manage their own pain and may only confide in pain to parents even though these discussions might need prompting. In this respect children are dependent on their parents or carers correctly picking up on their behavioural and emotional indicators of pain. The reluctance of the young men to talk about their pain may be linked to this reliance on other people to take action on their behalf [22]. Therefore, those interested in measuring HRQoL should be aware of potential response biases that may mean that relevant factors to patients’ well-being are overlooked. The consequences of parental underestimation of their sons’ QoL could have a negative effect on decision making when boys are too ill to speak for themselves [8]. One study found that boys reported more worry than appreciated by parents which in turn may interfere with communication and the parent’s understanding of their child’s psychosocial needs [44]. Here the importance of communication is evident as well as the accuracy of parent-proxy estimations in the absence of the child being able to complete a self-report.

There were a number of suggestions for interventions required for people with DMD and their families discussed in the included studies. Whilst not directly relevant to QoL they may provide further information about which QoL domains were regarded as important to patients and their families. These included, but are not limited to: initiatives to relieve family burden and improve carer well-being [18, 20, 33, 45, 46]; social skills training and modified activities inclusive of children with limited mobility [41]; education and care focused on optimising participation [43]; strategies to avoid diagnostic delay [23]; routine screening questionnaires and vigilance for signalling symptoms of pain, depression and anxiety in patients [43]; and improvements to welfare policies to simplify bureaucratic procedures [45].

Suggestions for future primary studies included research investigating the effect on other family members [17, 18, 45]; studies with follow up, rather than cross-sectional designs [17‐19, 44]; studies with more diverse samples with regards to race, educational status and single-parent families [44]; and further validation of the QoL-NMD [37].

This review included studies that used either quantitative and/or qualitative methods to determine the impact of DMD on QoL. However, there were very few qualitative studies identified, with most studies using an instrument to measure the impact of DMD on QoL. The included studies employed a wide range of different instruments that are relevant to QoL. This is common in the investigation of QoL, particularly for children [47]. Given complexities specific to DMD, the difficulties of using a generic HRQoL instrument in a specific disease area were acknowledged. Baiardini et al (2011) [7] suggested it would be impossible to use the same measure of HRQoL for all DMD patients due to the age range of affected patients. Another study which employed the QoL-NMD, an instrument specific to neuromuscular disease, even reported that that were some categories that were covered by a more generic scale (the INQoL) that were not covered by the more disease-specific QoL-NMD [37]. These authors recommended that a questionnaire specific to each neuromuscular disorder should be developed on the basis that whilst genetic neuromuscular disorders share a common pattern (i.e. a progressive loss of physical condition), they are also heterogeneous on several criteria. These include the age of disease onset, muscles affected, and the range of severity between the beginning and the end of the disease. Therefore, DMD can be considered as quite unique within the broader category of neuromuscular dystrophy.

A key benefit of using generic HRQoL instruments is the comparability of findings across clinical areas, for example, to estimate cost-effectiveness of interventions. Commonly employed generic HRQoL instruments such as the SF-36 and WHOQoL-BREF were used with different levels of success in the included studies. Landfeldt et al (2016) found that the SF-12 indicated that DMD patients had normal physical health where the EQ-5D did not [20]. Pangalila et al (2015) found that the occurrence of pain, anxiety and depression were not reflected in the QoL results using the WHO-QoL BREF and that only the ‘social relationships’ domain of QoL was impaired compared to the reference population [43]. Houwen-van Opstal et al (2014) highlights that the PedsQL, which is the most frequently used instrument in this review, was not designed to assess the full range of functioning unlike KIDSCREEN-52 [42]. The relationship between QoL and physical functioning is noted to be a complex, non-linear one with QoL for both children and/or carers potentially improving following the loss of ambulation where the risk of children falling is removed and the use of adaptive medical aids becomes routine. The same authors also highlight that there is no specific outcome measure to assess participation levels in patients. This issue was seen to be particularly lacking for people with DMD and their families. They also note that the use of generic HRQoL instruments sometimes means the inclusion of some inappropriate questions, such as “were you able to run well?” for patients with DMD.

The included literature identified varied aspects of physical, psychological and social implications of DMD. As stated, these were primarily measured using quantitative methods via existing QoL instruments. Table 3 shows that the common instruments used were PedsQL (neuromuscular module), SF-36, and WHOQOL. However, it can be argued that the SF-36 and WHOQOL instruments are not necessarily appropriate for the DMD population, as being that these instruments are generic, they do not incorporate all factors that are relevant to QoL for people with DMD. The PedsQL Neuromuscular Module (PedsQL-NMD) is a specific module of the PedsQL. Existing versions of this questionnaire include Acute version and Standard version, for Toddlers (2–4 years of age), Young Child (5–7 years of age), Child (8–12 years of age), Adolescent (13–18 years of age). The neuromuscular module was developed for neuromuscular disorders such as spinal muscular atrophy (SMA), DMD and other chronic conditions. Whilst this instrument may contain items that are relevant to those with DMD, it should be noted that the construct of the PedsQL-NMD was developed with participants aged from 18 to 80 years.

The literature suggests that many aspects of QoL are important to those with DMD, but the majority of these have been determined in a top-down fashion via existing instruments. It is possible that the incorporation of additional or alternative QoL themes into QoL assessment in DMD using a bottom-up approach with the choice of a wider choice of themes could benefit the study of QoL in this specific condition. Qualitative studies that build upon the themes identified in this review, undertaken with children and adolescents with DMD may identify new themes, or highlight QoL themes that are not relevant.

Age is an important factor to consider when measuring the impact of any condition upon QoL. The degree to which a child can self-complete a questionnaire is dependent upon their ability. Young children may require assistance with reading the questionnaire, and the administrative procedures. The task burden (which incorporates both the complexity of the questionnaire, and the number of items within it) will influence the quality of the data an instrument is able to provide. Instruments that contain a large number of items may mean that younger children are unable to reliably complete them. Matza et al. (2004) [48] identified a number of considerations to be made when designing a paediatric QoL instrument. The first is the age at which children can report their own health. Children appear able to reliably report on their own health between the ages of 4 and 6 years [49‐53].

The issue of age within the context of DMD is of particular importance. With the recognised deterioration of individual’s physical ability, it could be hypothesised that QoL would deteriorate with increasing age; however, this was not confirmed by the literature. There were inconsistent findings with regards to the patients deterioration in QoL over time with some reporting changes in at least one domain according to ambulatory status [54] and others suggesting that QoL remained stable despite declines in physical functioning [24, 44, 55]. However these inferences are often based on the comparison of DMD patients in cross-sectional or case-control study designs and therefore suitably powered follow-up studies using tools which are appropriate to capture meaningful changes to QoL with age are required. It is difficult to draw conclusions regarding the effect of age on DMD based upon evidence in the literature. Future efforts to study QoL in DMD or develop a DMD-specific QoL instrument should anticipate and accommodate the complexities of eliciting responses from young children, adolescents and young adults whose ability to communication will reduce over time.

Due to the inclusion of studies from around the world there were some geographical differences observed in the DMD research populations. Unsurprisingly the type and amount of formal care offered to patients and their families varies between countries [56]. One study documents that adults with DMD had higher HRQoL scores in Western Europe than those in countries in Eastern European countries [28]. One study conducted in India found that the caregivers participating in the study were mostly male, owing perhaps in part, to cultural differences in research participation [33].

Limiting study eligibility to the last 5 years means that relevant studies published prior to 2010 may have been missed. However, as QoL themes that are known to be relevant are most likely to continue to receive attention it is less likely that important themes have not be captured by the data conducted in the last 5 years. Moreover, it could be argued that themes that have not been investigated in the last 5 years are less likely to be relevant to patients today. Similarly, the restriction to academic papers published in the English language could equally result in missed relevant themes and therefore a more inclusive approach to future comprehensive reviews is recommended. Moreover, it is also unlikely that the omission of studies will affect the output of this review as the focus is on thematic analysis. The aim is therefore to encompass the most relevant themes rather than including all published evidence. The consensus of themes discussed in the primary studies adds confidence to the probability that data saturation was reached in this review and the majority of relevant themes were captured.

This research highlights that factors relevant to QoL in DMD span all domains of the included HRQoL instruments and beyond. Many of the nuances specific to DMD, such as the impact on the carer and wider family, are not currently captured by standard HRQoL instruments. The breadth and diversity of themes and emerging trends in this review highlight the need to consult with patients and their families about what aspects of QoL need to be covered by the QoL instruments used. For example, themes relating specifically to accessibility and adaptation to key stages in the condition, communication, access to appropriate healthcare services, family resources, carer burden, and impact on the wider family require consideration to obtain a more holistic picture of QoL.

Health professionals and clinicians who work in DMD may be aware of the complex picture of QoL for people living with DMD but must work with researchers to raise the issue of the difficulties in capturing QoL in DMD as in other diseases occurring in children which are severe and incurable, and therefore impacting substantially on patients and carer’s QoL. Future research in this area should therefore build on the themes identified from this review and consult with people with DMD and their families. A qualitative research approach could focus on specific areas, from the variety of themes identified, that require attention and identify which should be given more weight to capture QoL in this population. This work either could inform the use of generic HRQoL instruments or could lead to the development of a specific HRQoL instrument for DMD. The aim would be to ensure that primary research conducted in the field of DMD would eventually use specific standardised, user-validated methods to study QoL.

The importance of perspective (patient versus carer) when collecting QoL data needs to be highlighted in future research. Owing to the specific nature of DMD where communication of patients’ QoL is/becomes difficult, the value of using accessible instruments that reflect user needs should be considered. Moreover, the complexities at play when interpreting differences between self and/or parent proxy report should be given due consideration.

Evidence from the literature demonstrates that DMD impacts upon an individual’s QoL, however the measurement of this has been using either existing generic instruments, condition-specific measures for given aspects of QoL (such as anxiety or depression), or using clinical indicators as a proxy. There is little robust evidence exploring the impact of DMD from the patient perspective using qualitative methods. Recent approaches to instrument development in other areas have utilised qualitative data techniques to ensure high face and content validity to the items contained within a questionnaire [57‐59].