This paper presents the study protocol of a mixed methods systematic review on early bereavement interventions following the loss of a parent, adhering to the Preferred Reporting Items for Systematic Review and Meta-Analysis Protocols guidelines (PRISMA-P) [1]. This protocol paper seeks to strengthen the quality, reliability, and transparency all the way through the completion of the systematic review.

The death of a parent has been documented as a major stressful and disturbing experience for children [2,3]. Parentally bereaved children are more prone to functional impairment and other multiple negative outcomes, including psychological and behavioral problems [4,5]. Although former clinical assumptions stressed children’s lack of ability to grieve [6], it is now well accepted that bereaved children experience a grieving process [7] and some may develop extended psychiatric conditions [8-11]. Moreover, the death of a parent is also associated with an increased mortality risk in children [12,13]. As a result, a diversity of theoretical frameworks and psychosocial interventions have been proposed.

Grief intervention usually consists of quite diverse intervention approaches such as, for example, self-help, family interventions, support groups, counseling, and therapy. Such interventions are delivered by a variety of (para)professionals (eg, psychologists, social workers, nurses, pastoral staff) in varied formats (eg, individual or group, Internet, telephone, home visiting) [14]. Similarly to what was shown in the literature for bereaved adults [15], quantitative reviews on the effectiveness of bereavement interventions for children showed small to moderate effect sizes [16,17]. Nonetheless, interventions targeting high-risk children [16,17] and interventions starting closer to the loss were proved to be more effective [16]. On average, the length of time to initiate intervention after the death was a year and a half, but several studies presented a time interval of 5 years postloss.

An emphasis has been placed in starting interventions early in the mourning process [18,19]. When interventions take place later in the mourning trajectory, children may not be focused on their loss any longer and may have changed in a maladaptive way [16]. Therefore, early psychological interventions for bereaved children and their parents have been highlighted as a tool to decrease acute distress levels [20] and prevent future psychopathology [11,21], namely posttraumatic and complicated grief reactions [22]. In addition, bereaved parents require early qualified help not just for themselves but for their children in particular [23,24]. Likewise, bereaved children ask for early notification and involvement [25,26]. Nevertheless, contradicting results concerning grief therapy for adults still prevail. While some documented that interventions closer in time to the death are more effective [18], others did not find a significant relation between the effect of time since loss on outcome [15]. Additionally, a meta-analysis examining the short- and long-term effects of grief interventions for adults did not support the usefulness of preventive approaches [19]. The authors pointed out a series of methodological limitations among preventive studies such as, for instance, the lack of proper grief measures in screening at-risk groups. However, they did not elaborate on the fact that some studies did not report the mean time since death at study entry, whereas other studies reported a broad time range of 1 month to 2 years postloss. Such diverse timeframes have been noticed as one of the accountable variables for lowering the main effects of grief interventions [16,27].

According to the International Classification of Diseases, 11^th Revision (ICD-11) and the Diagnostic and Statistical Manual of Mental Disorders, 5^th Edition (DSM-5), it looks like the long time intervals of the contemporary quantitative reviews run counter to the viewpoint emphasizing early intervention. The proposal for ICD-11 revision acknowledges grief reactions as a possible form of psychiatric disorder whenever severe grief complications persist beyond 6 months postloss [28]. The current DSM-5 revision suggests a minimal watchful period of 6 months postloss for children and 12 months for adults [29]. A previous review suggested that grief counseling would be more successful if provided within 6 to 18 months following the death. However, this same review also considered that different types of support may be required early and/or later in the bereavement process [27]. Despite the fact that the focus on early intervention has been raised by both clinicians and researchers, there is a profound lack of unanimity. Its conceptualization, timeframes, and effectiveness have not been thoroughly explored. The dearth of high-quality and methodologically equivalent studies [30-32] seem to have led to the general conclusion that the grief field remains in its infancy [33].

Building on the existing quantitative reviews for bereaved children, evidence from qualitative, quantitative, and mixed methods studies will be collected. Based on the ICD-11 and DSM-5 diagnosing systems and the users’ perspectives studies, the mixed methods systematic review will focus on intervention studies initiated within 6 months after parental death in regard to the prevention of grief complications. The bereavement interventions will be dependent on the child's age (≤18 years of age), a setting that is compatible with the early intervention standpoint of promoting an optimal developmental trajectory.

The overall purpose of the mixed methods systematic review is to provide a rigorous synthesis of early intervention after parental death in childhood. The following two questions will be addressed:

As shown in Textbox 1, the Population, Interventions, Comparator, Outcomes, and Study Designs (PICOS) structured approach [34] is used to frame inclusion and exclusion criteria for studies. In view of the overall family context, each study will contemplate parentally bereaved children (≤18 years of age) and/or their parents with no limits on further sociodemographic indicators such as gender, ethnicity, or socioeconomic status. Any causes of parental death will be included, whether due to illness, accidents, or other causes. Any type of bereavement psychosocial intervention (eg, crisis intervention, support groups, counseling) starting up to 6 months postloss will be included.

The review will encompass all types of study designs (quantitative, qualitative, and mixed methods). It will include studies published in English language peer-reviewed scientific journals as well as dissertations. Exclusions will apply to systematic or other forms of literature reviews, letters, commentaries/editorials, and conference abstracts/presentations. Study protocols will only be consulted to provide deficient details of later related articles.

Two reviewers (MP and IJ) will independently undertake the database searches. An initial-stage screening of titles and abstracts will be performed, and the studies will be assessed against the predetermined inclusion and exclusion criteria. Articles that explicitly do not meet the eligibility criteria will be excluded, while potentially eligible studies will be imported into EndNote citation software (Clarivate Analytics) and duplicates will be deleted. The full text of potentially eligible articles and studies for which a decision grounded on title/abstract cannot be made will be saved for investigation. Any disagreements will be reconciled through discussion and achieving consensus. A second-stage screening of the full-text articles will be independently conducted by MP and IJ for further eligibility assessment. Reasons for exclusion will be documented for future inclusion in the PRISMA flow diagram [34]. Following this, a citation search of all eligible studies and any pertinent reviews attained during the first- and second-stage screenings will be performed to search for additional studies. Once again, disagreements will be resolved via discussion. Multiple articles pertaining to the same study will be gathered and placed under the main study.

A comprehensive literature search will be carried out on the largest medical, psychological, and nursing databases: Medical Literature Analysis and Retrieval System Online (MEDLINE) (via Ovid), PsycINFO (via Ovid), Excerpta Medica Database (EMBASE) (via Ovid), and Cumulative Index to Nursing and Allied Health Literature (CINAHL) (via EBSCO). To secure a contemporary overview, possible eligible studies will be searched from January 1980.

The search strategy was developed using the PICOS framework [34]. It was informed by the methods sections of previous reviews on bereavement interventions [32,35] and was devised together with a university librarian. The search terms will combine keywords referring to the intervention and the population. Table 1 shows an example of the MEDLINE search strategy that will also be used for the other databases. There will be no methodological filters so that quantitative, qualitative, and mixed methods studies can be screened.

A data extraction template was developed using Cochrane existing guidelines [36] as well as current proposals identifying central aspects of interventions [37]. It comprises information concerning (1) eligibility (eg, child’s age, time since death, reasons for exclusion); (2) study characteristics (eg, aims, design, causes of death); (3) participant demographics (eg, age, sex, socioeconomic status); (4) intervention features (eg, theoretical basis, setting, contents/components); and (5) outcomes (eg, frameworks, types of measurements). This data extraction form was independently piloted by two reviewers (MP and IJ).

Despite the likelihood of some unreported data, specifically regarding intervention features, this review will not ask original authors to provide additional information. This is in line with the overall aim to specify and qualify what has been disclosed in peer-reviewed publications. Two reviewers (MP and IJ) will independently extract the data and disagreements will be reconciled through discussion.

The interest in mixed methods studies has been increasing over the last 2 decades. However, the emergence of mixed methods systematic reviews is rather novel and there is no consensus regarding its best critical appraisal measures [38]. The Mixed Methods Appraisal Tool (MMAT) [39] consists of a pilot-tested quality assessment instrument that simultaneously assesses the most prevalent types of quantitative, qualitative, and mixed methods studies. The MMAT was chosen because it allows different study designs to be qualified with the same measure. It was specifically developed to be used in mixed methods systematic reviews [38], and it has been used by others in the grief field [35].

Two reviewers (MP and MH) will independently use the MMAT to assess the quality of eligible studies. Both quality scores and general brief descriptions will be provided, and the results will be compared for consistency. Disagreements will be resolved by discussion.

Since this review covers a wide range of research designs and multiple findings, it will make use of an interpretative framework. This approach is well suited to systematic reviews including studies expected to be too heterogeneous to ensure a quantitative overview [40]. The Guidance on the Conduct of Narrative Synthesis in Systematic Reviews, proposed by Popay and colleagues [40], will be followed. It seeks to “tell the story” of the review findings through the iterative implementation of four stages: (1) developing a theory of how the intervention works, why, and for whom; (2) developing a preliminary synthesis of findings of included studies; (3) exploring relationships in the data; and (4) assessing the robustness of the synthesis. In line with these guidelines, a number of different tools and techniques (eg, textual descriptions, tabulation, grouping and clusters, qualitative case descriptions, and reflecting critically on the synthesis) will be used to accommodate the data at each stage. This will be done through an iterative and inductive process, and the emerging results will be discussed with the research team throughout.