The Original Publication

This correction is in regard to our publication (Källquist and Salzmann-Erikson, 2019) in the Journal of Child and Family Studies, 28, 2056–2068. We undertook the interpretive meta-synthesis due to the observation that, whereas severe mental illness is primarily a challenge to the individual who experiences it, there is additional burden placed on relatives. Further observing that this phenomenon has received some limited attention in the extant literature, we chose to undertake a qualitative meta-synthesis of childrens’ experiences with parents who have serious mental illness. This method was chosen due to the limited literature on children’s experiences and the desire to illuminate the existing accounts.

As noted in the original publication, the review the existing research was framed by three research questions: “(1) What experiences do adults who grew up with a parent with serious mental illness have from childhood? (2) How is adult life affected by the experience of growing up with a parent with serious mental illness? (3) How do adults, who grew up with a parent with serious mental illness, reflect on the support that they received during their childhood and as adults?” (Källquist and Salzmann-Erikson, 2019, pp. 2057–2058).

To answer these questions, we searched for existing research in relevant databases to collect first person narratives. Figure 1 depicts the Prisma flow that we undertook to identify, search, and sort the research that was located (see later correction). After selecting literature, we conducted a qualitative interpretive meta-synthesis. As noted in the original publication, based on the evidence reviewed, we concluded that having the experience of having a parent with severe mental illness can result in the experience of having relational issues later in life and need for support that can persist into adult life.

Fig. 1
figure 1

PRISMA 2009 Flow Diagram

Full size image

Critique

We were notified by the Co-Editors in Chief of JCFS that concerns had been raised regarding the methods employed, and further, that those methods may have biased the conclusions of the review. The main critique was that we had included in the review three published articles that emanated from the same study and study data (Murphy et al., 2015, 2016, 2017). Our inclusion and treatment of these as three separate studies is not consistent with guidance provided in the Cochrane Handbook for Systematic Reviews and Interventions (Li et al., 2021). Observing that we did not report and/or manage this in keeping with commonly accepted guidelines, the authors of the Letter pointed out the risk of distorted findings due to the disproportionate weight of the Murphy et al. articles. In particular, we understood the concern that this may bias the findings of a meta-synthesis. As authors, we undertook a careful appraisal of our study and its key findings.

Concurrently, two anonymous reviewers were invited by the Co-Editors in Chief to provide expert opinions and make recommendations regarding corrective action. This internal review was conducted with the hypothesis that the findings in the review may have been distorted from our error. The editorial team consulted two external experts in the field of systematic review who reviewed the complaint, our review article, the three Murphy articles, and related materials. The Editors provided the outcome of the review to the authors. The external reviewers concluded that the inclusion of the three articles was incorrect and that a correction from the authors were needed, hence, the Editors requested this correction. However, the external reviewers further concluded that the errors made did not pose a methodologic challenge to the validity of our paper as it was published, since the focus of our original paper was on extraction of themes rather than quantitative findings. Indeed, this is consistent with contemporary views of various forms of qualitative analysis as focusing on “descriptive, rather than statistical, power” (Nye et al., 2016, p. 60).

Moreover, the external reviewers each stated an opinion that the findings of the review did not overstep the contributions made by the literature as a whole and that the findings and implications are largely the same. Besides the fact that we incorrectly treated three articles from one study as three studies and would have collated them as directed by Li et al. (2021), we think that it is necessary to point out the inaccuracies and provide this correction.

In hindsight, we should have treated the three Murphy articles as one study emanating from a single dataset and not treated them as three separate studies. Preferably, we should have determined studies as the unit of interest and explicitly collated the three Murphy-publication as one study from the beginning. We have erroneously but not intentionally risked the possibility of misleading the JCFS readership, which we regret. We think that it is important to reveal and to correct our mistakes in the best way we can which is also our responsibilities as nurse researchers and in accordance with good research practice and research integrity.

The specific corrections to be noted are:

1. The abstract (p. 2056) should state “Fourteen articles met the inclusion criteria…”.

2. In the Method section, in the subheading Literature Search (p. 2058), we did not find 4302 studies, but 4302 articles. In the same paragraph it should be added that out of the 14 articles, three of these were from one study and dataset (Murphy et al. 2015, 2016).

3. Also, on page 2058 (subheading Quality Assessment), we refer to one study was scored as low and excluded and four studies were assessed as medium – both these “studies” should be “articles” or “publications.”

4. On page 2059 (Fig. 1), we wrote “Studies included in qualitative synthesis.” It would be more accurate to state “Articles included in qualitative synthesis.”

5. On page 2067, the second line in the heading “Methodological Limitations”: “fourteen studies” should be “fourteen articles”.