Objectives

The main objective of this systematic review is to identify and synthesise the existing literature that explores the barriers and facilitators experienced by children and young people themselves and their carers and healthcare professionals when managing symptoms in ICYP at end-of-life.

Eligibility criteria

The criteria outlined below will be used for study selection. Following the recommendations of the Cochrane Qualitative and Implementation Methods Group Guidance,15 we have used STARLITE (Sampling Strategy, Type of study, Approaches, Range of years, Limits, Inclusions and exclusions, Terms used, Electronic sources17) to report our search methods.

Searches

Data management

All records and data will be saved to Endnote X8.20 This software will be used to identify potential duplicates. The researchers will check this and remove all confirmed duplicated articles.

Selection process

Articles will be screened by title and abstract by one reviewer (KG) with a second reviewer (SH) assessing 10% of the articles, randomly selected. At this stage, articles will be judged as either (1) ‘not relevant’ or (2) ‘potentially relevant’. Both reviewers will read and screen all remaining potentially relevant articles. The reviewers will independently apply the criteria at all stages of the selection process. Intercoder agreement will be evaluated using Cohen’s kappa coefficient. A minimum kappa value of 0.75 will be taken to represent high agreement.21

The full text of all remaining potentially relevant articles will then be obtained. If the relevance of a study cannot be ascertained from the abstract, then the full article will be obtained. The full articles will be read by two reviewers independently (KG and SH) to make the final decision about whether they should be chosen for inclusion in the review. A third reviewer (CL) will resolve any uncertainties. Additional information will be sought from authors if necessary at the stage of full-text assessment.

Data collection process and items

The following information will be extracted into a piloted data collection form for all included studies: study aims; patient population (infant/child/adolescent); participant population (patient/caregiver/healthcare professional); inclusion and exclusion criteria; sample size; recruitment; design; intervention and comparator group (where applicable); date and duration of data collection; setting; country; data collection; analysis methods; data describing the participants’ views/experiences of barriers and facilitators to symptom management. For qualitative data, the authors’ interpretations (presented through themes and categories) will represent these data.22 KG will extract this information and SH will check it, with any disagreements resolved through discussion with CL.

Quality assessment (including risk of bias)

A quality appraisal of included studies will be conducted independently by two reviewers (KG and SH). Disagreements will be resolved by discussion between KG and SH, with CL if required.

Three checklists will be used depending on each study’s design. These were chosen since they are all validated and have been used in published systematic reviews within healthcare research. For each study type, Cohen’s kappa coefficient will be used to measure inter-rater agreement between the two reviewers. A minimum kappa value of 0.75 will be taken to represent high agreement with disagreements resolved via discussion with CL.

As recommended by the Cochrane Qualitative and Implementation Methods Group,15 we will not calculate total quality scores across domains since domains of quality are not equal. Instead, KG, SH and CL will determine how each study’s methodological limitations affect confidence in the findings via discussion. We will not exclude studies based on poor quality but will record and highlight methodological issues.

Qualitative studies will be quality appraised using the Critical Appraisal Skills Programme for Qualitative Studies (CASP).23 CASP assesses clarity of research aims, research design, recruitment methods, data collection, relationships between participants and researchers, ethical issues, analyses, description of findings and valuableness of the research. It is comprised of nine closed questions (eg, ‘Was there a clear statement of the aims of the research?’ Yes/Can’t tell/No) and one open-ended question (‘How valuable is the research?’). For each question, there is the option to add comments to explain the reasoning for each rating. Currently, CASP is the most frequently used qualitative research synthesis tool in the Cochrane Library and WHO guideline research24 and has been used in similar systematic reviews assessing barriers and facilitators within healthcare research .25 26 However, because the CASP tool does not address aspects of the research validity and can favour papers that are less insightful as long as they comply with ‘expectations of research practice’,27 in addition, the evaluative criteria of credibility, transferability, dependability and confirmability28 will be applied. Included studies will be assessed as to whether they apply the techniques suggested for ensuring study quality according to Guba and Lincoln’s28 criteria, that is, prolonged engagement, persistent observation, peer review, triangulation, negative case analysis, referential adequacy and member checking to ensure credibility; thick description for transferability; inquiry audit for dependability; confirmability audit, audit trail, triangulation and reflexivity to ensure confirmability. Studies will be rated as ‘high quality’ if they meet at least three of the four criteria, ‘medium quality’ if they meet two of the criteria and ‘low quality’ if they meet one or none.

The Quality Assessment Tool for Quantitative Studies (QATQS) will be used to assess all clinical studies with or without randomisation and control groups, including quasi-experimental and before-and-after studies.29 The QATQS is comprised of 22 closed questions and an overall rating of strong, moderate or weak in eight sections: selection bias; study design; confounders; blinding; data collection; withdrawals and dropouts; intervention integrity; analysis. It has been shown to be a valid tool for assessing quality, comparing studies and addressing threats to validity of findings.30

The Mixed Methods Appraisal Tool (V.11) will be used to assess the quality of any mixed methods studies.31 This tool consists of five closed questions assessing the research question, research design, integration of qualitative and quantitative methods, integration of qualitative and quantitative data and consideration of methodological limitations in mixed methods studies. As reported by the Cochrane Qualitative and Implementation Methods Group, this tool has been used widely in systematic reviews and has the advantage of being able to assess interdependent qualitative and quantitative elements of mixed-methods research.24

Outcomes and prioritisation

The main outcomes sought are carers’, healthcare professionals’ and patients’ (CYP) views on the barriers and facilitators to effective symptom management in ICYP at end-of-life.

Data synthesis

Although it is unlikely that the majority of included studies will be quantitative, if this is the case, then random-effects meta-analysis will be conducted to synthesise group means and standard deviation from individual studies using Comprehensive Meta-Analysis version 3.32

For meta-analysis to be conducted, data must be available from two or more eligible studies reporting similar barriers or facilitators. The studies must report the number of participants reporting that barrier/facilitator and the total number of valid participant responses for that survey item. A random-effects model will be used for all analyses since; unlike a fixed-effects model, this can be used when statistical heterogeneity (I²) is present in the results of the included studies.33 Where evidence of statistically significant heterogeneity is present, sensitivity analyses will be conducted where possible to verify the robustness of the study conclusions, assessing the impact of methodological quality, study design, sample size and the potential effects of missing data. We will use funnel plots to detect potential reporting biases and small-study effects where data are available from 10 or more studies.34

If the included studies are all qualitative or a combination of quantitative and qualitative, there are several approaches that could be taken for data synthesis. Some of the most commonly used methods to synthesise qualitative health research include thematic analysis,35 grounded theory36 and meta-ethnography.37 38 However, there is no consensus on the best approach, which will depend on the type and number of included studies39 and the form and nature of the research question.38 As such, we will make a final decision on the most appropriate method after selecting and quality assessing the included articles, as recommended by the Cochrane Qualitative and Implementation Methods Group.39

We will first analyse and synthesise data related to the experience of patients, care providers and healthcare professionals separately before deciding whether it is appropriate to aggregate data between these groups. These data will likely include themes, concepts and categories of information. If data are relatively ‘thin’, then we will consider using thematic synthesis to undertake line-by-line coding and development of descriptive and analytic themes. If the included articles include sufficient ‘thick’ data (eg, details about the context and background of the studies and participants40), we will consider a more interpretative approach such as meta-ethnography.37 This method goes beyond aggregating data to generate new interpretations of the findings.

As recommended by the Cochrane Qualitative and Implementation Methods Group,24 the GRADE‐CERQual (Confidence in the Evidence from Reviews of Qualitative research41) will be used to summarise our confidence in synthesised qualitative findings (eg, in the themes that we identify). The CERQual is made up of four key components, that is, methodological limitations of included studies, coherence of the review finding, adequacy of the data contributing to a review finding and relevance of the included studies to the review question. After assessing each of the four components, overall confidence will be graded as high, moderate, low or very low. The barriers and facilitators to symptom management will be divided into overarching themes for each group (patients, healthcare professionals, carers), and presented in a matrix along with our CERQual assessment of confidence in the evidence of each theme and an explanation of this assessment.

The GRADE guidelines42 will be used to appraise the quality of any quantitative findings. The GRADE guidelines include four elements for which quantitative findings will be rated against: risk of bias (‘Study limitations’), inconsistency, indirectness, imprecision, publication bias.

Patient and public involvement

Consultation with young people, parents and healthcare professionals has been used to determine their perception of the barriers and facilitators they experience when managing symptoms in ICYP at end-of-life. It is based on their perspectives that this systematic review was deemed to be timely and crucial to conduct to inform further research work. Moreover, patient and public involvement (PPI) is represented in the authorship (MJ) of this manuscript. MJ is the PPI representative at the UK National Institute for Health Research (NIHR) Pain and Palliative Care Clinical Studies Group-Children. MJ is a parent of four children, one of whom died from T-cell acute lymphoblastic leukaemia at the age of 12. She has supported many families of children with cancer including those receiving palliative and end-of-life care.