Summary
Both articles explore the often forgotten parental experiences of having and caring for a child with an eating disorder, including the impact on mental health and the whole family.
McCormack and McCann (2015) aimed to explore the subjective experience of 10 Irish parents of adolescents with anorexia nervosa through a series of semi-structured qualitative interviews. Findings showed that the parental experience can be summed up in four main themes: negative impact on the domestic family life, psychological impact on parental mental health, experience with primary care services and the relentless hope for a better future for the child and the family. Based on these findings, it was concluded that parents need more support in caring for adolescents with an eating disorder, which can be achieved through involvement in treatment programs for the child to improve family life, reduce the psychosocial impact on parents and aid recovery, but also by providing more educational training for non-specialist mental health nurses.
Svensson, Levi and Suarez’s (2013) text focused on the experience of 10 Swedish parents and their caregiver roles to children with a wider range of eating disorders. Using a phenomenological-hermeneutical qualitative approach, the parents that had already participated in family-based treatment were administered questionnaires and semi-structured interviews. This study yielded three main themes, disruption of social life in terms of isolation and interpersonal conflict, psycho-emotional influence on the parents and diverse parental coping mechanisms. The integration of parents into treatment approaches to reduce carer distress, improving accessibility to further information about eating disorders and facilitation of supportive communication were seen as essential implications for clinical practice.
Evaluation
Theoretical foundation and study rationale
In their short literature review, both McCormack et al. (2015) and Svensson et al. (2013) highlight the well-established finding that caring for a child with any eating disorder can be straining for a parent’s mental and physical well-being (Cohn, 2005). This is especially due to the family’s important role in the child’s recovery that demands strength, while simultaneously impacting all facets of the classic family environment (Beale, McMaster & Hillege, 2005).
However, while McCormack et al. (2015) then briefly review established interventions to reduce carer distress, namely family-inclusive therapies and coping skills training, Svensson et al. (2013) focus on ways in which family members’ reactions can aggravate the disorder. Nonetheless, while family experiences with children with eating disorders were the focus in both studies, only Svensson et al. (2013) mentioned that family-based treatments are the most frequently recommended therapy approaches for children with eating disorders (Lock, 2011; Whitney, Murray et al., 2005), a grave omission in the review by McCormack et al. (2015).
It is also worth noting that Svensson et al.’s (2013) literary review fails to explore the reasons for carer distress in eating disorders, only mentioning that it strains family relationships severely. McCormack et al. (2015) take this exploration further and include mentions of stigma and lack of information as underlying sources of carer distress. Haigh and Treasure (2003) suggested that many carers experience distress due to holding incorrect views of how the illness arises, mainly due to stigma surrounding the disorder which results in feelings of blame and guilt. While McCormack et al. (2015) touch upon this important facet of the parental experience, there is no mention of the role of stigma in Svensson et al.’s (2013) literary review. However, both studies present a relatively short theoretical foundation, due to the nature of qualitative research structures.
In terms of study rationales, both articles highlight that there has not been extensive research into the subjective experience of parents, stating this as their rationale. However, despite exploring nearly the same phenomenon albeit in different nationalities, McCormack et al. (2015) failed to include Svenson et al. (2013) in their literary review, even though almost identical findings were obtained. Therefore, stating a lack of research as rationale for a study is highly insufficient: for example, key qualitative studies by Cottee-Lane, Pistrang and Bryant-Waugh (2004), McMaster, Beale and Hillege (2004) and Zabala, MacDonald and Treasure (2009), that specifically focus on carer distress in eating disorders, have been omitted from at least one of both papers, suggesting that rather than a lack of previous research, there seems to be a gap in the literary review.
Clearly, the high prevalence of the disorder elaborated by McCormack et al. (2015) and the high co-morbidity and potential mortality highlighted by Svensson et al. (2013) provide better rationales to explore the family’s experience of eating disorders and the development of interventions to reduce carer burden is essential, but are not a clearly specified research motivation in either text.
Methodology
Both studies under review used a qualitative semi-structured interview approach to explore parental experiences of caring for children with eating disorders. This method was more appropriate than quantitative methods, as it enables in-depth exploration of the phenomenon in a “living” environment, rather than aiming for generalizability, which was the rationale. Both studies also addressed the biggest limitation of their research, namely the limited sample size and subsequent non-generalizability of findings.
However, the samples are not perfect: Svensson et al. (2013) used both parents for three participants when exploring the experience of being a caregiver to a child with an eating disorder. However, it is likely that both parents’ accounts are not excessively different, since they were exposed to the same living situation. McCormack et al. (2015), on the other hand, managed to recruit ten parents of ten different patients, therefore obtaining a slightly more faceted experience overview. While Svensson et al.’s (2013) sample is less representative, it has been subjected to more selection criteria: age of the patients, severity of the eating disorder, participation in family-based therapy and duration of the illness. McCormack et al.’s (2015) method of recruitment is limited to “rural and urban communities” and while inclusion criteria have been mentioned, they have not been further elaborated in the paper.
While one strength of both studies is that the sample consisted of both mothers and fathers, in McCormack and McCann’s (2015) study, only a small number of fathers chose to participate. Qualitative evidence suggests that mothers caring for daughters with an eating disorder experience higher levels of depression and anxiety, accompanied by higher emotional over-involvement than fathers (Modes & LeGrange, 1993). These findings suggest that for mothers, the demands of caring for a child with eating disorders may take a greater toll. Therefore, to correctly understand the discrepancy and reasons for differential levels of strains, recruiting an equal number of mothers and fathers would have provided a better picture.
Furthermore, both sets of authors fail to disclose their own standpoints in relation to the topic of exploration. Qualitative research is never completely objective and a statement of interests from the authors is therefore essential, but has not been included in these studies (Greenhalgh & Taylor, 1997). Observer bias is an important shortcoming of qualitative research if left unaddressed. In terms of ethical considerations, both articles provided sufficient ethical approval from the parents, but only McCormack et al. (2015) sought approval from the patients too.
The materials employed by both articles included semi-structured interviews, but only Svensson et al. (2013) provided an interviewer guide to facilitate understanding and interpretation of the results. McCormack et al. (2015) failed to do so, therefore making it difficult for readers to interpret the results in light of the interview questions. Potentially, this interviewer guide may have been included in the appendices, which were not accessible to the general public.
Finally, in terms of the data analysis, both studies transcribed the interviews verbatim and subsequently searched for themes. Svensson et al. (2013) improved reliability by having three researchers analyse the interviews and come up with their own findings of patterns and differences. Inter-rater reliability was therefore assured as far as possible in qualitative research. However, this inter-rater reliability is lacking in McCormack et al. (2015): a computer package (NVivo 9) was used for data analysis and only the finished coded transcripts and frameworks were reviewed by one other person. Efforts were made to ensure rigour, validity and reliability of the research, but no software can analyse qualitative data to the extent that humans can and should be used as support only (Zamawe, 2015; Armstrong et al., 1997).
Interpretation/discussion of findings
Svensson et al. (2013) and McCormack et al. (2013) identified similar main themes during their interviews: the impact of the disorder on family life, parental emotional well-being and the mechanisms of coping, especially by holding out hope for improvements in the future. Both articles also mention the issue of stigma that leads to social isolation, strained relationships and potentially financial issues. However, while McCormack et al. (2015) further explored the parental experience with primary care service usage, Svensson et al. (2013) identified burden-reducing coping mechanisms. Both highlight the need for information to facilitate the process of recovery. All of these points have repeatedly been found in earlier studies, such as by Cottee-Lane et al. (2004), Goddard et al. (2011) and the Department of Health’s (2014) report on eating disorders. Hence, the findings are supported by prior research.
However, interpretations were not always supported by the data that were produced: McCormack et al. (2015) concluded that parents spending more time at home suffered more than those who did not, even though such findings were at no point supported by interviewee statements. While similar findings have been made in earlier studies (Dimitropoulos et al., 2008), there is no justification for this interpretation in this study. Similarly flawed interpretations were not found in Svensson et al.’s (2013) research, whose most salient points were not only identified in the data, but also supported by prior research by Whitney et al. (2005), the National Institute for Clinical Excellence (2004) and Raenker et al. (2013).
In terms of implications of the findings for clinical practice, both articles have identified the issue of guilt, stigma and their impact on parental coping. Both conclude that it is necessary to improve parental support by including them in the treatment and improving access to information to provide better knowledge about the situation. Sepúlveda et al. (2012) found that including parents in the therapy not only reduced their social isolation and feelings of guilt, but also improved recovery rates. Nonetheless, both studies fail to identify the need for more information for the general public, before the illness’ onset: in McCormack et al. (2015), one participant cites failing to identify the early signs of eating disorders and not intervening earlier as the origins of their guilt. Thus, concluding that providing more information for parents after the onset of the disorder is a necessity is incomplete and does not address the severe lack of general information.
Similarly, some of the suggestions for clinical practice are unrealistic. Svensson et al. (2013) mention the enormous impact of financial difficulties on carer burden and a possible solution for reimbursement through the free health care system. However, given that cases of anorexia tend to last on average 8 years (Theander et al., 1985), expecting public government funding healthcare to be sufficient is highly short-sighted and unreasonable. While there may not be any better solution available at the moment, presenting an impossible suggestion is not appropriate. Similarly, McCormack et al.’s (2015) suggestion to train non-specialist staff to provide better parental support is not only vague, but also shows a naïve ignorance of worldwide staff shortages, especially in nursing departments: a meta-analysis by Beckmann et al. (1998) suggested that 81% of nursing staff surveyed declared inability to cope with current patient loads. Therefore, other support options, such as peer groups, communities and even social services should have been explored further.
Finally, based on the large number of limitations identified here but not addressed by the authors, it is safe to say that both articles’ discussions of limitations were incomplete. In terms of suggestions for future research, however, both correctly mentioned the need to investigate effectiveness of family-inclusive treatments in improving recovery rates for children with eating disorders. More recent studies, such as Lock et al. (2016), found that hospitalisation days in adolescents with anorexia nervosa tended to reduce when family-based treatment was introduced at the point of diagnosis.
Conclusion
To conclude, it is safe to say that Svensson et al. (2013) and McCormack et al. (2015) have made valuable contributions to the limited research field of parental experiences in caring for children with eating disorders. Both identify family-based treatments as an important step towards recovery for the child but also the need for reduction of carer distress, due to guilt and family conflict. Another strength of both articles is the use of appropriate qualitative research methods, correctly addressing their biggest limitation and inclusion of both mothers’ and fathers’ experiences. While Svensson et al. (2013) especially excels in the rigid selection criteria applied to the sample, a solid interpretation of the results, providing detailed materials and assuring inter-rater reliability, McCormack et al. (2015) shows consideration for informed consent by both parents and patients. However, both studies also show some important omissions in their literary review, vague research rationales, the failure to disclose the authors’ personal interests in the topic and unrealistic suggestions for implications for clinical practice. Therefore, it is important to consider these articles in light of the findings in this critical review. Nonetheless, their value should not be underestimated, especially in their suggestions for future research and potential development of distress-reducing interventions.