Designing a Metasynthesis Study in Pediatric Oncology Nursing Research

Metasynthesis Research Designs

Qualitative metasynthesis is a general category for a collection of qualitatively driven approaches to synthesizing the findings of research reports. There have been at least 12 variations of metasynthesis identified (Barnett-Page & Thomas, 2009). We suggest that the following metasynthesis types identified by Polit and Beck (2012), with the addition of the Joanna Briggs Institute (JBI) meta-aggregative approach, as the most applicable to nursing research:

The myriad metasynthesis methodologies can be divided into 2 main categories, methodological approaches that are mainly interpretive and methodological approaches that are mainly aggregative (Paterson, 2012).

Mainly interpretive approaches to metasynthesis include meta-ethnography, grounded formal theory, meta-study, and critical interpretive synthesis (Barnett-Page & Thomas, 2009). The researcher applies an interpretive method to develop a theory or gain a greater conceptual understanding than what is found in individual research reports. For example, in meta-ethnography a process of reciprocal translation is used to create an overarching theory by recognizing and configuring related concepts that occur across reports (Noblit & Hare, 1988). An interpretive approach to metasynthesis is most suitable when integrating the findings from interpretive studies such as those conducted by phenomenology, grounded theory, or ethnography. This approach to metasynthesis is considered data-far as the extracted findings and the final synthesis product have both undergone significant interpretive transformation (Sandelowski, 2012). An interpretive metasynthesis, while useful in developing conceptual understanding, does not always give information that can be immediately applied to clinical practice. The purpose of the metasynthesis project is a primary consideration when choosing a metasynthesis research design. Furthermore, interpretative metasynthesis projects cannot always be considered systematic reviews of qualitative evidence as they often use a theoretical approach to sampling by selecting research reports to support and explain an emerging theory. Researchers use a systematic approach for the database search in aggregative projects to include findings from all published research reports on the topic of interest.

Aggregative approaches to metasynthesis include meta-aggregation, metasummary, and thematic synthesis. The final product of an aggregative meta-synthesis is considered data-near as there is a limited interpretive transformation of the study findings. An aggregative approach is employed to systematically integrate the findings from research reports that present summaries or topical surveys of findings that have undergone little or no interpretive transformation (Sandelowski, 2012). For example, in the meta-aggregation approach developed by the JBI, similar findings that are found across research reports are assembled and used to produce statements that directly inform health care decision making and clinical practice (Hannes & Lockwood, 2011). Even though some level of interpretation is used in aggregative metasynthesis projects, the degree of interpretation used separates interpretive and aggregative metasynthesis projects.

Developing a Metasynthesis Study

The following section describes the steps that are used to develop and carry out a metasynthesis research project. Although these steps are presented in a linear fashion, it is sometimes necessary to move back and forth among steps as the project unfolds.

Consider the Appropriateness of Metasynthesis for the Research Problem

The skill mix of the proposed research team is an important consideration when planning a metasynthesis project. Each team member should be familiar with qualitative research methods, and ideally at least one team member should be an experienced qualitative researcher (Jensen & Allen, 1996). It is equally important to make sure that no other metasynthesis has been conducted on the topic as this negates the need for metasynthesis (Finlayson & Dixon, 2008).

Determine the Rationale for Conducting Metasynthesis Research

Researchers undertaking a metasynthesis project must first name and define the phenomenon they wish to study, giving consideration to clinical priorities, gaps in knowledge, and personal interest (Hannes & Lockwood, 2011).

Once the phenomenon of interest is identified, the research question is identified. The type of research question is dependent on the metasynthesis method chosen. An interpretive metasynthesis is framed by a broad purpose like the inductive purpose statements found in primary qualitative research studies (Britten et al., 2002), whereas aggregative projects are framed by a more specific research question that forms the basis for a fixed protocol (Pearson, 2004). The PICo (population, phenomenon of interest, context) mneumonic (Joanna Briggs Institute, 2014) and the SPIDER (sample, phenomenon of interest, design, evaluation, research type) mneumonic (Cooke, Smith, & Booth, 2012) are 2 devices that can be used to aid in developing a metasynthesis research question/purpose statement. The breadth of the research question needs to provide for sufficient coverage of the phenomenon in question without sacrificing the creation of meaningful results (Finfgeld, 2003).

Search and Data Retrieval

The hallmark of a well-executed search strategy is the inclusion of enough precision to find the relevant research reports without including large numbers of irrelevant reports in the search results (Sandelowski, 2012). The search strategy adopted will depend on the approach to metasynthesis being taken. A theoretical sampling approach is sometimes used in interpretive metasynthesis projects, such as meta-ethnography, and is in line with the theoretical sampling strategy used in primary qualitative research projects (Jones, 2004). Aggregative projects use a systematic search strategy identical to what is used in a systematic review of quantitative research. Another issue in searching is the fact that there is no standard definition of what constitutes qualitative research within or across databases. Barroso et al. (2003) give detailed instructions for locating qualitative research studies in bibliographic databases. The following is the general procedure for obtaining research reports for a metasynthesis project.

All relevant bibliographic databases need to be searched and suitable explanations given for the choice of databases. The search strategy for a metasynthesis project must identify the keywords used and be documented with enough detail so the search could be reproduced by another researcher. All metasynthesis projects in pediatric oncology nursing should, at a minimum, include searches of the CINHAL (Cumulative Index to Nursing and Allied Health Literature), MEDLINE (Medical Literature Analysis and Retrieval System Online), and Cancer.gov (U.S. National Cancer Institute). Other databases to consider, especially for interpretive projects, are those that cover the social science literature such as PsycINFO, Social Sciences Citation Index, and Sociological Abstracts. Most of these databases are accessed through commercial platforms, so consultation may be necessary with your institution’s librarian to determine availability and access. In order to maximize search results, hand searches of key journals from the target subject are recommended. Additionally, footnote chasing, which is the review of the references listed in the research reports located by the electronic database search, is recommended (Sandelowski & Barroso, 2007). Finally, a decision needs to be made on whether to include unpublished literature such as graduate theses and dissertations (available in the ProQuest Dissertations and Theses Database). The inclusion of unpublished works is recommended as there is evidence that their exclusion can lead to an overrepresentation bias, particularly because unpublished works tend to contain more negative findings (Conn, Valentine, Cooper, & Rantz, 2003).

Review and Quality Appraisal

The review and appraisal process has 2 phases: initial screening followed by strict screening (Saini & Shlonsky, 2012). Initial screening is accomplished by reading the titles and abstracts of the research reports found by the search to decide which reports meet the inclusion criteria of the project. Each report is screened by 2 people, and the reasons for excluding a report needs to be documented. In order to ensure objectivity, a third reviewer should be used in cases of disagreement between the primary reviewers. Saini and Shlonsky (2012) recommend using the following questions to guide the screening process:

Once initial screening is finished, the selected reports are retrieved. It is recommended that an electronic citation management software program (e.g., EndNote, Mendeley) be used to organize the reports. The use of citation management software is especially important if a large number of reports are to be included in the project (Havill et al., 2014).

Strict screening involves thoroughly reading the reports selected during the initial screening phase in order to further determine content and applicability. Two reviewers are likewise required to perform strict screening, with the addition of a third reviewer to resolve any conflicts (Saini & Shlonsky, 2012). Moreover, the decision whether or not to exclude reports based on the quality needs to be made. There are a number of quality appraisal tools available for use in metasynthesis projects. However, there is no consensus on the role of quality assessment in selecting reports for inclusion in a metasynthesis research project (Hannes, Lockwood, & Pearson, 2010; Sandelowski, 2012). The Joanna Briggs Institute (2014) method utilizes JBI-QUARI, a critical appraisal of qualitative evidence checklist, to rate research reports for the purposes of excluding reports considered to be of inferior quality. However, Noblit and Hare (1988), who originated meta-ethnography, recommend that all reports be included with more consideration being given to findings from reports considered higher quality. Sandelowski and Barroso (2007) recommend against eliminating individual research reports for reasons of quality due to the subjective nature of what constitutes quality in qualitative research. The decision-making process on the role of quality appraisal in any metasynthesis project needs to be transparent, and the reasoning behind this decision needs to be defined and congruent with the specific goals of the project.

Once the relevant research reports are selected for inclusion, the search and review process should be summarized with a diagram. The diagram displays the number of reports identified and reasons for the inclusion and exclusion of reports at each stage of the review process (Moher, Liberati, Tetzlaff, & Altman, 2009).

Data Extraction

The data extraction procedure involves both the recording of demographic and methodological information from each report and locating and recording the research findings. “Findings in qualitative research reports are the data-based interpretations researchers advance about the target phenomenon under investigation (e.g., a grounded theory of the process, a phenomenological description of an experience)” (Sandelowski, 2012, p. 26). Metaphors, themes, concepts, and categories are all labels applied to findings in research reports. Due to the range of ways findings are displayed in research reports, the data extraction procedure can be complicated and time-consuming, requiring several readings of each research report (Noblit & Hare, 1988; Sandelowski & Barroso, 2002). A finding is considered suitable for a metasynthesis when there is evidence substantiating it presented in the report. If a finding cannot be substantiated, it should not be included in the project (Hannes & Lockwood, 2011; Pearson, 2004).

Data Synthesis

The biggest difference between aggregative and interpretive metasynthesis projects are seen in the data synthesis phase. However, regardless of the metasynthesis method used, the goal of data synthesis is not to summarize the extracted research findings. The findings are organized and analyzed to create a new explanation of a process or to create lines of action for decision making (Hannes & Lockwood, 2011; Polit & Beck, 2012). Data synthesis procedures used in meta-ethnography will now be reviewed, followed by a review of meta-aggregative data synthesis.

There are three distinct steps to data synthesis in meta-ethnography: determining how studies are related, translating studies into one another, and synthesizing translations. In order to determine how studies are related, the findings that were extracted from each study are organized into a grid demonstrating how each study is related (Noblit & Hare, 1988). “Studies can be related in three ways: reciprocal (directly comparable), refutational (in opposition to each other), and in a line of argument other than either reciprocal or refutational” (Polit & Beck, 2012, p. 670). Step 2, translating studies into one another, is accomplished by comparing the identified concepts in order to determine if any of the separately identified concepts are the same (Britten & Pope, 2011). Finally, the remaining concepts are analyzed to produce a theory that explains the overall process present across reports in a similar fashion to analyzing primary qualitative research (Atkins et al., 2008).

Data synthesis in meta-aggregation is also a 3-step process. In the first step, each extracted finding is assigned a level of credibility, either unequivocal, credible, or unsupported. Findings considered unsupported are those that are not substantiated with evidence in the research report. The researcher must decide whether or not to contact the author of the report containing the unsubstantiated finding to ask for other evidence or to discard the finding in question. The remaining findings are then placed into categories with at least 2 findings per category. Categories are then analyzed, in the same manner as in primary qualitative research, into at least 1 synthesized finding (Joanna Briggs Institute, 2014).

Validate the Data

Measures to ensure validity are just as essential in qualitative metasynthesis as they are in other research projects. Descriptive validity is maintained in a metasynthesis project by ensuring all relevant research reports are located and the findings are accurately extracted and reported. Interpretive validity is supported by transparent reporting that acknowledges that the findings in each research report are the constructed views of the reports’ authors and not of the original study participants. The theoretical validity of a metasynthesis project includes an assessment of the credibility of the methods used to arrive at the metasynthesis product and the metasynthesis product itself. Pragmatic validity is assessed by the degree of generalizability and transferability of the synthesis product (Sandelowski & Barroso, 2007).

Report the Findings

A metasynthesis project report follows the format of a standard research report and includes the headings, introduction, methods, results, discussion, and conclusion. The results section of a meta-ethnography should include a visual diagram of the theory developed and/or a detailed explanation of the line of argument discovered (Noblit & Hare, 1988). The results section of a meta-aggregation should include lines of action to guide clinical practice based on the synthesized statements (Hannes & Lockwood, 2011). The discussion section should include any limitations of the synthesis and state the implications for pediatric oncology nursing practice.

Illustrations of Metasynthesis Research Design and Key Design Features

In this section, we present 2 published metasynthesis research studies from the oncology nursing literature to illustrate the 2 primary research designs discussed in this article.

Coffey (2006) used meta-ethnography (Noblit & Hare, 1988) to explore parenting in the context of childhood chronic illness. The researcher’s rationale for choosing metasynthesis by meta-ethnography was “to create a comprehensive chronicle of the phenomena of parenting a child with a chronic illness” (Coffey, 2006, p. 51). The population of interest was parents of a child with a chronic illness. The phenomenon of interest was parenting, including family life. The study context was not reported beyond qualitative reports or reports with a qualitative component focused on parenting published between 1989 and 2000. The researcher conducted searches in the following electronic databases: CINAHL, ERIC, Psyclit, Sociological Abstracts, PubMed, and Dissertation Abstracts. The author selected 11 reports based on the purpose of the project; the total number of reports retrieved by the search was not stated. The author extracted metaphors from each study and organized each metaphor into a table under 1 of 7 themes. The themes reported included living worried, staying the struggle, and survival as a family. A quotation to support each theme was included in the metasynthesis report in order to enhance validity. The report concluded with a conceptual statement on the everyday experience of parents of children with a chronic illness and considerations for clinical practice (Coffey, 2006).

Lee et al. (2012) used meta-aggregation (Joanna Briggs Institute, 2014) for the purposes of examining body image in children and adolescents with cancer. Their rationale for choosing metasynthesis by meta-aggregation was the integration of existing qualitative research study findings to increase the understanding of the participants’ experiences to make sure they receive the proper support and services. The population of interest was individuals 6 to 20 years old. The phenomenon of interest was the experiences of children and adolescents with cancer. Moreover, the context was living in acute care, at home, or in the community, in any cultural context. Published and unpublished qualitative research of all designs, written in English or Chinese, was considered. Initial searches were conducted in the PubMed and CINAHL databases in order to find relevant keywords and terms. These initial searches were supported by extensive searches of the aforementioned databases and the following databases: Scirus, Mednar, ProQuest Dissertations and Theses, Chinese electronic theses and dissertation systems, and Chinese electronic periodical services. In addition, the reference lists of the items located were searched (footnote chasing) in order to find relevant studies. Three hundred and thirty-seven items were found and evaluated against the study criteria. Twelve items were found to meet the project criteria and were independently evaluated for quality by 2 researchers using the JBI critical appraisal instrument (Joanna Briggs Institute, 2014). A third reviewer was not needed as disagreements of quality were not found between the 2 researchers. The detailed analysis (strict screening) afforded by the appraisal process led to the elimination of 4 reports that did not meet the project criteria, leaving 8 reports to be synthesized. A total of 41 findings were extracted from the study reports using the JBI’s SUMARI software (Joanna Briggs Institute, 2013). The 41 findings were aggregated into 9 categories and then grouped into the following 4 synthesis statements: being distanced from the body, loss of self-identity, self-protective strategies and support, and getting rid of physical shackles (Lee et al., 2012). Validation was accomplished by regular team meetings and the use of crosschecking between research team members. The published report of the metasynthesis concluded with 9 proposed guidelines for clinical practice with children and adolescents with cancer, including: listening to and respecting their thoughts and needs about body image changes; assessing the emotional support, illness-related information, level of substantial social support required, and encouraging them to think positively; and discussing with them their plans for the future (Lee et al., 2012).

Benefits of Using Metasynthesis Designs in Pediatric Oncology Nursing Research

It is apparent that qualitative metasynthesis research is a useful tool for building evidence to inform pediatric oncology nursing practice. It can be used in various forms, much in the way of primary qualitative research, for enhancing and refining our understanding of childhood cancer experience (Woodgate, 2000b). Furthermore, metasynthesis research studies have the potential to go beyond understanding by generating transferable knowledge that can be used to incorporate the patient’s perspective into nursing practice (Bondas & Hall, 2007). Metasynthesis research studies also have the potential to present findings of atypical events that do not stand out in individual studies (Thorne, 1994). Metasynthesis projects maximize the value of primary research studies that are often conducted with participants under uncommon circumstances (Dixon-Woods, Agarwal, Young, Jones, & Sutton, 2004)and can be used to examine a phenomenon before proceeding on to further study (Flemming, 2007). Finally, metasynthesis projects can aid in overcoming the bias associated with small sample sizes found in qualitative studies by generating theory or guidelines based on the experiences of participants from multiple studies (Estabrooks et al., 1994).

Engaging in metasynthesis research also creates opportunities for publication and research dissemination. There are a growing numbers of high-impact journals publishing metasynthesis research in the fields of nursing, pediatrics, and psychology (Hannes & Macaitis, 2012). In addition to publishing in traditional peer-reviewed journals, researchers interested in conducting an aggregative metasynthesis project have the choice to submit their review protocols to the JBI for peer review and publication. If the protocol is published a priori, there is then the opportunity to publish the completed review in the JBI Database of Systematic Reviews and Implementation Reports (Joanna Briggs Institute, 2014). Metasynthesis projects can also serve as a bridge for qualitative researchers to collaborate with quantitative researchers in the systematic review process. For example, the Cochrane collaboration has published guidelines for including qualitative evidence in systematic reviews of interventions in order to interpret and explain quantitative findings (Noyes, Popay, Pearson, Hannes, & Booth, 2008).

Challenges of Using Metasynthesis in Pediatric Oncology Nursing Research

The main goal of a metasynthesis research project is to produce a synthesis that is both transferable and generalizable, and this process is not without challenges (Finfgeld-Connett, 2010; Thorne, 2009). A major challenge in conducting qualitative metasynthesis is establishing credibility of this methodology. There is a tendency for some qualitative researchers to view the synthesis or generalization of qualitative findings as incongruent with this research paradigm. However, it is common practice to see recommendations for clinical practice in qualitative research reports, building a firm foundation for a metasynthesis (Finfgeld-Connett, 2010). Another challenge is the generalizability of the synthesis to contexts other than those reported in the synthesized research reports. This challenge can be addressed by using an integrated knowledge translation approach where clinicians and decision makers are made part of the research synthesis team at the outset of the project (Canadian Institutes of Health Research, 2012). It is the clinicians on the research team who will ultimately decide whether or not the synthesis is generalizable (Finfgeld-Connett, 2010; Lincoln & Guba, 1985).

A major challenge in transferability of metasynthesis research is the limited reporting of the synthesis methods used in published metasynthesis report. This includes incongruence in the reported synthesis method and the actual method used (Dixon-Woods, Booth, & Sutton, 2007). For example, a recent review of published metasynthesis research found that many published meta-ethnographies did not present a line of argument, which is an integral part of using meta-ethnography (Hannes & Macaitis, 2012; Noblit & Hare, 1988).

In order to overcome these challenges, researchers reporting the results of metasynthesis projects must be transparent in their choice of the method and make sure that the method is consistently applied (Hannes & Macaitis, 2012). Furthermore, clinicians need to critically analyze the synthesis methods used in a report before making decisions based on the findings.

Conclusion

Pediatric oncology nurses in both research and practice settings require knowledge of metasynthesis methods to evaluate the range of synthesized research. The primary purpose of this article was to introduce qualitative metasynthesis research to pediatric oncology nurses. Although only 2 of the many metasynthesis research designs were described here, we expect this article will help develop the capacity of pediatric oncology nurses to use and/or conduct metasynthesis research. Well-executed metasynthesis research studies can increase the uptake of qualitative research in pediatric oncology nursing research and foster practice environments that are sensitive to children’s needs and experiences.