Systematic Review: Family Resilience After Pediatric Cancer Diagnosis

Marieke Van Schoors, MSc; Line Caes, PHD; Lesley L. Verhofstadt, PHD; Liesbet Goubert, PHD; Melissa A. Alderfer, PHD


J Pediatr Psychol. 2015;40(9):856-868. 

In This Article


As outlined by Eiser, Hill, and Vance (2000) and the Cochrane Collaboration (Deeks, Higgins, Altman, & Green, 2011), we complied with a strict scientific methodology to ensure comprehensiveness, minimal bias, and reliability of our systematic review. To this end, the following consecutive steps were taken: (a) formulation of the scope of the review and research question, (b) thorough literature search, (c) detailed data extraction, and (d) integration of the major findings and implications. Meta-analysis was not conducted because of heterogeneity across studies in terms of sample characteristics (e.g., different stages of treatment) and outcomes measured, as well as our desire to integrate qualitative findings to ensure a comprehensive review.

Literature Search and Inclusion Criteria

Web of Science, Pubmed, PsycInfo, Cochrane, and Embase were searched using keywords selected in collaboration with a library information specialist and three researchers familiar with the field (details available on request). Studies selected for inclusion examined aspects of functioning among families of children with cancer including: (a) a subjective (qualitative) or objective appraisal of changes since diagnosis (longitudinal data); or (b) standardized scores, clinically meaningful categorization of families (e.g., based on cut-scores), or a comparison with norms or control groups. Included studies investigated families of children diagnosed with any type of cancer before the age of 18 years. Studies published in languages other than English and nonempirical articles (i.e., reviews, case reports, books, book reviews, commentaries, practice guidelines, conference abstracts, and dissertations) were excluded. Articles concerning families of children who died from cancer were also excluded, as theirs is a distinct family experience. Reference lists of the selected papers were also reviewed to ensure inclusion of all relevant papers.

Study Selection

The database search was undertaken in July 2014, identifying 5,496 unique papers. The first and second author independently screened all titles for decisions regarding exclusion with 89% agreement. Disagreements were discussed and resolved. The 1,592 remaining abstracts were then screened for exclusion, with an agreement rate of 83%. Again, disagreements were discussed and resolved. Finally, the first author screened the full texts of the remaining 427 studies for final decisions regarding inclusion. The second author screened 25% with 93% agreement. Disagreements were discussed and, if necessary, a third reviewer was consulted; 72 studies were selected. Thirteen articles were added based on reference chaining, resulting in a final set of 85 papers (see Figure 1).

Figure 1.

Flow chart of article selection.

Data Extraction

The first and second author extracted data from the studies in a systematic and standardized way, summarizing study characteristics and general findings on abstraction sheets (available on request). Year of publication, journal, and database were recorded along with methodological characteristics such as type of design (cross-sectional or longitudinal), measures used, and sample achieved (e.g., sample size and demographics). In addition, the theoretical framework, unit of measurement used, and characteristics of the statistical analyses (interdependence of data reported by multiple respondents) were evaluated. Findings of the studies were extracted by summarizing the results in a few lines. The last author reviewed the information extracted against original publications to ensure accuracy. Authors were contacted for papers and information, as needed, but unpublished data were not requested.