Article Text

Download PDFPDF

The synthesis of qualitative research and evidence-based nursing
  1. Kate Flemming, RGN, MSc
  1. University of York
    York, UK

Statistics from Altmetric.com

Readers of Evidence-Based Nursing will be familiar with “review articles.” They are among the most common types of research abstracted for the journal, and although systematic reviews of randomised controlled trials (RCTs) predominantly feature in the Treatment section, reviews that address questions of assessment, causation, quality improvement, prognosis, and quality assessment have also been abstracted. Reviews of qualitative research have been abstracted less frequently—7 times since the journal began in 1998.

Qualitative research is an important component of the evidence base for nursing. Almost one quarter of all abstracts published in Evidence-Based Nursing (n = 205) have been primary qualitative research articles, and they have examined several areas of nursing practice. Like quantitative researchers, qualitative researchers have become interested in synthesising qualitative research in order to develop new cumulative knowledge.1 This notebook will explore how syntheses of qualitative research may contribute to the practice of evidence-based nursing.

Systematic reviews and meta-analyses of RCTs are at the top of the hierarchy for providing evidence of the effectiveness of interventions. This is because all individual trials have limitations, such as low statistical power, researcher or expert bias, and contextual variability.2 Similarly, single qualitative studies can be heavily influenced by specific issues of context and generalisability. By synthesising qualitative research, these difficulties can potentially be overcome.

THE DEVELOPMENT OF THE SCIENCE OF THE SYNTHESIS OF QUALITATIVE RESEARCH

Several methodological approaches have been developed to synthesise qualitative research over the past 2 decades. Some have been adapted from methods used to analyse primary qualitative research, whereas others have been specifically developed for this purpose.3

Qualitative synthesis is being promoted by a number of organisations based on the recognition that providing health care involves complex, multifactorial decisions that require a range of evidence. Qualitative synthesis complements the “rationalist” model of synthesis provided by traditional systematic review methods.4 One such organisation is the Cochrane Qualitative Methods group (http://www.joannabriggs.edu.au/cqrmg/about.html), which was founded in 1998 with an aim to develop methods for including and incorporating synthesised findings from qualitative research into systematic reviews on the effectiveness of healthcare interventions. Looking more broadly at social care, the Campbell Collaboration (http://www.campbellcollaboration.org/MG/index.asp) was established in 1999 to prepare, maintain, and disseminate systematic reviews of studies of social interventions, which, by their nature, require the synthesis of complex forms of evidence. The Joanna Briggs Institute (http://www.joannabriggs.edu.au/about/home.php) in Australia has developed an approach to the synthesis of qualitative research (with supporting computer software) that enables the synthesis of qualitative research and its integration with quantitative reviews.5

Since the late 1990s, the efforts of specific research groups have resulted in an increase in publications reporting methods and findings of qualitative syntheses.6,7,8,9,10,11,12 Many of these syntheses addressed topics relevant to nursing and health care, such as the lay meanings of medicines, experiences of diabetes care, and factors influencing healthy eating among children.

Several different approaches can be used to synthesise qualitative research, each producing different types of review. Noblit and Hare developed the methodology of meta-ethnography specifically to synthesise qualitative research.13 They distinguish between integrative and interpretative reviews. Integrative reviews aggregate quantitative data and use statistical analysis to generate new findings, whereas interpretative reviews create understanding from multiple cases, accounts, narratives, or studies derived from qualitative research. Concepts identified in these studies are pulled together to create “higher-order” theoretical structures.

Dixon-Woods et al3 focus on different ways of synthesising evidence. Rather than focusing on different forms of evidence, they propose that integrative syntheses summarise data through pooling, meta-analysis, or description. Interpretative synthesis, however, is best viewed as the development of concepts and theories that integrate these concepts, grounded in data from primary studies. Interpretative syntheses can incorporate both quantitative and qualitative evidence. Both types of syntheses exist on a continuum in which integrative reviews interpret data and interpretative reviews aggregate data. What drives the method of synthesis is the nature of the research question asked.3 Although the interpretative end of the synthesis continuum is mainly populated with methodologies synthesising qualitative research, in reality, many syntheses include elements of both interpretation and integration. Indeed, combining interpretation and integration is an underlying principle of good syntheses.

METHODS OF SYNTHESISING QUALITATIVE RESEARCH

Many of the methods associated with the linear process of conventional systematic reviews (eg, question formulation, searching, quality appraisal, and methods of pooling data) are still being developed or evaluated for their application to synthesis of qualitative research. Qualitative synthesis will probably never evolve into the relatively standardised process of producing a systematic review of clinical trials or diagnostic studies. More likely, the vision of qualitative synthesis will include a reviewer drawing on a “methodological palette,” in which the review processes are defined by the question and nature of the research to be included in the review.

This does not mean that syntheses of qualitative research are not being done. One of the most frequently used methods is meta-ethnography.13 Meta-ethnography uses techniques familiar to primary qualitative researchers to synthesise qualitative research reports and is not limited, as its name might suggest, to ethnographic studies. It is an interpretative approach, with the findings of studies being “translated” into one another. Meta-ethnography was the method used in 3 of the 7 syntheses of qualitative research that were abstracted in Evidence-Based Nursing. The clinical questions of these syntheses were

  • How have healthcare relationships in chronic illness changed over the past 15 years?14

  • What are patients’ experiences of recognising symptoms of cancer and seeking help?15

  • How do people with diabetes mellitus describe their lived experience?16

Meta-ethnography is a methodology capable of answering clinical questions and is one of the few areas in qualitative synthesis with an active programme of funded methodological work.1 Other methods for synthesising qualitative research include meta-synthesis, meta-summary, meta-study, meta-narrative mapping, cross-case analysis, narrative summary, thematic analysis, content analysis, case survey and the qualitative comparative analysis method.1,3 The volume of methods available indicates why underdevelopment and underevaluation are so common.

HOW CAN SYNTHESISING QUALITATIVE RESEARCH CONTRIBUTE TO EVIDENCE-BASED NURSING?

The development of the synthesis of qualitative research has created a “new” form of research evidence, which impacts on the evidence available for clinical decision making. The number of completed qualitative syntheses is small, but growing. Just as it has taken time for the repository of systematic reviews of effectiveness to build up to a useful mass whereby fruitful searching occurs, it will take time for the number of syntheses of qualitative studies to grow to a practical number so as to enhance clinical decision making. However, syntheses of qualitative research that have already been done address many topics of interest to nurses, including

  • what constitutes quality end-of-life care17

  • adapting to and managing diabetes8

  • patients’ help-seeking experiences and delay in cancer presentation18

  • the experience of physical restraint19

  • development of a user focused stroke service in primary care20

The synthesis of qualitative research can contribute to the evidence base for nursing in ways that are relevant to practitioners and researchers (table).

Contribution of qualitative research to practice and research

Development of theory

One of the primary arguments for synthesising qualitative research is that relevant theories for clinical practice and research can be made stronger.21,22 An example is found in the area of end-of-life care. From a synthesis of primary qualitative explorations of patient and family perspectives and narratives, a shared conceptual framework of the constituents of quality end-of-life care was developed.17 The review determined that quality of life and quality of dying are intertwined with quality of care from the perspectives of patients and their families. Before the synthesis, these 3 concepts had been identified and measured separately. Practice applications arising from these findings included development of clear outcome measures to evaluate quality end-of-life care and prioritisation of strategies for improving quality of care at this time.

New insights from primary research

The synthesis of qualitative research may improve the use of primary research findings, particularly those that relate to rare or infrequent events that are often under-represented in single studies.23 By systematically examining primary studies, similarities that were not individually recognised as major themes can be integrated and re-examined, and new insights can be gained. A review on adapting to and managing diabetes combined a series of qualitative research studies that had focused on specific concepts of diabetes management, rather than the totality of the experience.8 Synthesising the findings highlighted the connections between key concepts in diabetes management. Of particular note were the divergent views of healthcare professionals and people with diabetes regarding the focus of management. Healthcare professionals favoured managing and controlling diabetes through compliance with prescribed regimens. People with diabetes, however, sought to balance self care and striving for a normal life.

Because one of the key advantages of qualitative research generally is the ability to access sensitive, rare, and “difficult” phenomena, synthesising qualitative research allows maximum value to be gained from primary studies that have overcome problems in accessing and researching difficult settings.22 Qualitative syntheses then are a useful step to consider before undertaking further research in these types of study areas.

Enhanced generalisability of primary studies

The nature and extent of the generalisability of qualitative research is heavily contested.24,25 Qualitative research methods do not seek to produce generalisable knowledge in the same way as statistical predictions. Single qualitative studies can produce indepth, insightful, and rich descriptions of phenomena, which can be difficult to generalise beyond the original sample. Such studies can end up isolated and used little in practice. Synthesising qualitative research addresses this limitation by drawing single studies together and providing new cumulative knowledge with broader applicability.1

In a qualitative review of 32 studies of patients’ help-seeking experiences and delays in cancer presentation, Smith et al18 developed 2 higher order categories that had direct implications for nursing practice: the influence of gender on help seeking and the influence of sanctioning on help seeking. These findings were shown to hold true across patients with different types of cancers, generalisability that would not have been apparent if only the individual studies had been considered. The synthesis also revealed the difficulties faced by patients with cancer when attempting to communicate what might be perceived as trivial symptoms to healthcare professionals.

Providing context for systematic reviews of effectiveness

When combined with systematic reviews of effectiveness, qualitative syntheses can provide a valuable contextual dimension.26 Qualitative and quantitative research methods may be synthesised separately, with one method used to inform the other (combined synthesis), or they may be integrated throughout the process of synthesis itself.

Combining research syntheses is becoming more common in response to the challenge of evaluating complex interventions in health care. Such evaluations are crucial, not only in terms of applications at practice levels, but also for informing policy.27 This is similar to the that way mixed methods research (the combination of qualitative and quantitative research methods) has been advocated for the evaluation of complex interventions in primary research.28,29 As awareness of the limitations of “mono-method” reviewing grows, calls for greater methodological heterogeneity in the studies included in reviews are increasing. Combining or integrating syntheses of qualitative research alongside empirical evidence may be efficient and increase the richness, relevance, and usefulness of evidence for decision makers.2 Although underevaluated,3 methods for undertaking “multi-method” reviews are one of the most promising areas of evidence synthesis.2

An illustration of a combined synthesis of qualitative and quantitative research is a review by Thomas et al, which examined barriers to, and facilitators of, increased fruit and vegetable intake in children.30 The review included a meta-analysis of intervention studies designed to increase intake and a synthesis of qualitative research into children’s views on eating fruit and vegetables. The 2 separate reviews were then combined using a matrix that matched the interventions found to be effective with children’s views of those interventions. In doing so, Thomas et al found that the promotion of fruit and vegetable intake needs to be separated or packaged in different ways within the same intervention as children see fruit and vegetables as different types of food. There needs to be less emphasis on health messages—fruit and vegetables need to be branded as “tasty” rather than “healthy”. Promotion of fruit and vegetables in educational materials needs to be accompanied by access to fruit and vegetables. Combining qualitative and quantitative syntheses in this way illustrates not only which interventions work and why, but also the impact (or lack of impact) on those they are designed to target.

Identification of new research topics

Qualitative synthesis can generate future research topics. This may be in the form of developing theory for future hypothesis testing in effectiveness research or identifying gaps in research agendas for specific content areas. For example, in synthesising qualitative studies of experiences of physical restraint, Evans and Fitzgerald found that research had only been done in acute care settings.19 Consequently, they identified a need for research on patient experiences of restraint in residential care environments. Similarly, a synthesis of qualitative research on stroke care identified patients that were under-represented in primary care based research.20 Only one study included in the review addressed the issues facing younger patients with stroke, and this provides an impetus for future research to address the discrepancy.

Improved quality and reporting of primary qualitative research

Syntheses of qualitative research can also help to improve the quality and reporting of primary qualitative research. Systematic reviews of RCTs led to the development of the CONSORT statement,31 which provides explicit criteria for reporting of trials. As reviewers of qualitative research begin to grapple with similar issues of quality and reporting, an analogous initiative may be initiated.

FUTURE DEVELOPMENTS IN THE SYNTHESIS OF QUALITATIVE RESEARCH

Before addressing the future developments for qualitative research synthesis, it is useful to acknowledge the debate over whether we should be attempting to synthesise qualitative research at all. Synthesising primary qualitative studies is anathema to those who consider that each piece of qualitative research is a unique representation of multiple realities or truths.1 In order to accept that synthesis of qualitative research is a feasible methodological option, a position of “subtle realism”32 is required. From this position, one acknowledges that research studies attempt to describe and capture (albeit in different ways) an underlying social reality.1 For those concerned with answering questions relevant to clinical practice and policy decision making, adopting this pragmatic stance makes the synthesis of qualitative research methodologically feasible.

Several methodological issues remain under debate among those attempting to synthesise of qualitative research, such as whether different qualitative research methodologies should be combined, whether studies included in reviews should be appraised for quality, and if so, using which methods. Further areas of debate include the need for comprehensive searching to identify all relevant studies as opposed to a sampling of papers for inclusion and methods of synthesising studies. As more qualitative syntheses are completed, so will such methods become replicated and validated. In addition to developing the methods for synthesising research, researchers are continuing to advance techniques for searching33,34 and critically appraising qualitative studies.35

Readers of syntheses of qualitative research may also be looking for a quality appraisal checklist by which to judge the quality of the synthesis, such as those available for other research methods. As yet, such a tool does not exist, and it is unlikely that a single checklist could cover all qualitative syntheses. This reflects the fact that syntheses of qualitative research can be done using various methods. This is appropriate given the range of questions that may be answered using this methodology. Quality criteria that have been suggested as guidance include a clearly stated purpose for the review, an explanation of inclusion criteria, description of the sample of papers, procedures for handling and synthesising data, and rules for interpretation.36

Further work is needed to develop the terminology associated with the synthesis of qualitative research. Readers may be baffled by the range of language used, but perhaps a more important implication of the lack of consensus on terminology is that indexing within electronic databases is haphazard, making searching for qualitative syntheses difficult. Clearly, this makes life more difficult for teachers of evidence-based nursing who instruct students on methods of searching and who wish to identify qualitative syntheses in their searches. Until some consensus is reached on indexing, the contribution of syntheses of qualitative research to evidence based nursing may be limited.

SUMMARY

This notebook has shown that although considerable developmental work is needed to create mutually agreed upon approaches to qualitative research synthesis, the existing methodologies can produce influential findings relevant to nursing practice.

The synthesis of primary qualitative studies creates a cumulative body of evidence that builds and develops theory for practice in ways that individual studies cannot. Syntheses of qualitative research may also complement quantitative research syntheses, providing context for, and giving meaning to, evidence of the effectiveness of interventions.

Syntheses of qualitative research will increasingly become available to nurses as nursing journals become more open to publishing qualitative syntheses and related methodological papers. Such papers will add to the evidence base for nursing practice, informing clinical decision making and policy.

References

Footnotes

  • Thanks to Dr Karl Atkin and Professor Trevor Sheldon for comments on an earlier draft of this paper and Laurie Gunderman for providing me with the data on the number of qualitative articles abstracted in Evidence-Based Nursing.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.