br Methods br Study design settings and sample br
Study design, settings, and sample
A qualitative inductive content analysis was conducted (Sandelowski, 2000). Childhood cancer has a tremendous impact on a child's family, especially the parents (Benedetti, Garanhani, & Sales, 2014). Usually, the parents of a patient determine the coping of the en-tire family. Purposive sampling was employed to select parents as inter-viewees. With the help of the nurse, we obtained access to the electronic medical records of the patients. The sampling included PFK-158 with children hospitalized for varying periods (including readmission) to capture the maximum variation in the coping experience. One of the parents from each family was invited as a proxy if she or he had a child aged ≤14 years hospitalized for the treatment of cancer in an on-cology ward. Parents of children hospitalized for b24 h, or those who ex-perienced an emergency (e.g., massive hemorrhage) were excluded from the present study.
Data collection and analysis
Data were collected using semi-structured interviews conducted from November 2017 to June 2018. The settings for this study were the pediatric oncology departments of four Grade III Level A hospitals in mainland China. The interviews were conducted by a researcher. The researcher recruited the participants with the assistance of a nurse from each pediatric oncology department. The researcher ex-plained the objectives and significance of the present study to the par-ents. The parents were requested to sign the statement of informed consent prior to their participation in the face-to-face interviews. The investigator responsible for conducting the interviews was experienced
in conducting qualitative research. The interviewer had previously per-formed a cross-sectional study involving this population, and had established rapport with the parents of the patients. Recruitment and data collection regarding strategies of family coping proceeded until data saturation. Of note, the interviews were conducted in a designated room in each of the participating institutions. The researcher asked the participants regarding their preferences for the interview place. Most parents wished to continuously monitor their children; thus, the inter-views took place in the wards at the children's bedsides.
Initially, the interviews consisted of general questions, followed by open-ended questions related to family coping. An interview guide based on the Double ABC-X model was used during each interview. The Double ABC-X model describes how a family responds to a stressful event (Kong, 2010; Lee, Parker, DuBose, Gwinn, & Logan, 2006). Coping is the central concept in this model. It is used to describe families' efforts to achieve a new level of balance after a stressful event (Kong, 2010). The interview guide explored parental perceptions pertaining to:
a) the impact of the child's hospitalization for cancer treatment upon the family; b) the coping strategies families used to deal with their child's hospitalization for cancer treatment; and c) the effectiveness of their coping strategies. Interview questions used in the study are pre-sented in Table 1. Owing to the nature of the semi-structured inter-views, the questions were rephrased based on responses or understanding of the question. At the end of each interview, the re-searcher restated or summarized the provided responses, and asked
the participant to confirm the accuracy of the record. The interviews were conducted in Mandarin. The duration of each interview was 45–90 min. All participants agreed to have their interviews audio-recorded. Information related to the hospitalized children was mainly obtained from their medical records. The audio recordings were transcribed verbatim within 24–48 h after the interview, and the transcripts were analyzed through qualita-tive content analysis (Graneheim & Lundman, 2004; Hsieh & Shannon, 2005). The transcripts were initially open-coded word-by-word and line-by-line. Two experienced qualitative researchers examined the data in detail to identify “meaning units” and encode them. Upon com-pletion of coding, a 3-day meeting was arranged to discuss the codes, re-solve any disagreements, and reach a consensus. The codes were sorted into subcategories based on their interrelation. Depending on their rela-tionships, the number of subcategories was reduced into categories. Fol-lowing the formation of the categories, the members of the research team convened to reach a consensus. An audit trail was applied to pro-vide evidence that a process of data analysis, reduction, and synthesis was performed, leading to the reported findings.