Nursing and Health Interventions. Souraya Sidani
reviews include systematic reviews, meta‐analyses, and umbrella reviews that focus on the relationships among determinants and the health problems. The reviews apply a systematic process for searching the literature; identifying and selecting studies that meet the pre‐specified criteria; appraising the study quality; and extracting data on the existence, direction, and magnitude of the associations between the determinants and the problem. However, they differ in the method for synthesizing the evidence. Systematic reviews use vote count or content analysis to integrate evidence on the associations. Meta‐analyses use statistical techniques to estimate the direction and magnitude of the association reported in each study; to determine the extent of variability in the estimates across studies. When the variability is minimal, the estimates are synthesized to indicate the average level of associations between the problem and its determinants. Where there is high variability, statistical tests are used to identify conceptual and/or methodological factors that account for the observed variability in the estimates. Recently, umbrella reviews (also called reviews of reviews, meta‐reviews or meta‐evaluation/meta‐epidemiology; Gough et al., 2012) have been conducted to compare, contrast, reconcile, and/or explain differences in the results of systematic reviews and meta‐analyses (Paré et al., 2015). Findings of systematic reviews, meta‐analyses, and umbrella reviews indicate: the extent to which the associations between the determinants and the health problem exist within particular client populations and contexts, or are replicated across populations and settings; the expected direction and magnitude (or size) of the associations; and the factors that moderate the associations.
Qualitative Reviews
Meta‐syntheses entail methods for integrating findings of qualitative studies (Edwards & Kaimal, 2016) that investigated the health problem. Examples of methods include meta‐ethnography and thematic synthesis that consist of searching the literature, identifying and selecting relevant studies, extracting and coding the findings of each study, then comparing and contrasting the codes to generate themes (Sandelowski & Barraso, 2007; Snilstveit et al., 2012). The themes can provide a list of determinants and explanation of the inter‐relationships among the determinants and the problem, embedded within or across contexts.
Mixed Reviews
Mixed reviews are increasingly performed to understand the health problem. They consist of integrating evidence synthesized from quantitative and qualitative studies. Mixed reviews are justified with the increasing complexity of the inter‐relationships among determinants (occurring at different levels) and health problems (experienced in different domains), and the acknowledgement of the unique and complementary contribution of quantitative and qualitative research methods (Edwards & Kaimal, 2016; Fleming, 2010; Hong et al., 2017). Mixed method reviews (e.g. critical interpretive synthesis) follow the same steps for searching the literature; identifying, selecting, and appraising quantitative and qualitative studies; and extracting data on the experience of the health problem and its determinants from each study. In some mixed‐method reviews, the synthesis is conducted separately for the quantitative (e.g. through vote counting, estimating effect size) and the qualitative (e.g. coding, generating themes) findings; then, the synthesized quantitative and qualitative evidence is compared and contrasted to identify convergence and divergence. In other mixed‐method reviews, the quantitative and qualitative data are integrated (i.e. brought together) during the analysis and synthesis stages of the review.
The integration of the quantitative and qualitative evidence can be done by aggregation or configuration. Aggregation is used when quantitative and qualitative findings address the same association between a determinant and the health problem, and the purpose is to examine convergence (i.e. whether the findings confirm each other). The aggregation may be accomplished at the data or study level. Data‐based aggregation involves the transformation of one type of data into another (e.g. quanticizing qualitative codes/themes) and analyzing the transformed data using the same analytic method (e.g. descriptive statistics). Study‐based aggregation consists of juxtaposing the findings of quantitative and qualitative studies, using grids, tables or matrices, where the interface between the two types of findings occurs in the cells; the latter evidence is compared and contrasted to identify convergence (Fleming, 2010; Hong et al., 2017). Aggregation by configuration entails the arrangement of diverse findings into a coherent form or model. The quantitative and qualitative findings are carefully examined and analyzed to determine if they complement, extend, explain each other (e.g. why and how a determinant contribute to the health problem), or reflect variability in the problem experience and in the association of the determinant with the problem in different contexts (Sandelowski et al., 2012).
All types of literature review are useful for understanding the health problem. The mixed‐method reviews are most promising in generating a comprehensive list of factors contributing to the problem, delineating direct and indirect relationships that are context‐dependent, and providing explanation of why and how the determinants contribute to the health problem.
Methods
The process for conducting literature reviews is systematic. It has been extensively described in different sources (e.g. Heyvaert et al., 2016) as comprised of several steps. The steps are briefly reviewed with a particular emphasis on understanding the health problem.
1 Clarifying the health problem: This first step involves delineating the nature of the health problem and specifying the context (i.e. characteristics of the client population and context) of interest. This step is important for specifying the key words to be used in searching the literature and the criteria for selecting literature sources. The goal is to enhance relevance of the literature to the problem and context of interest.
2 Specifying the key words: The key words are the terms used to refer to the health problem. They include the specific words (e.g. insomnia) and its synonyms (e.g. sleep problem, disorder, difficulty) frequently appearing in the scientific literature and mentioned in lay conversation or documents. Consultation with librarians is very helpful in finalizing the list of key words. The key words are used independently or in combination with other key words representing the target client population (e.g. older adults), context (e.g. primary care), and determinants of the problem (e.g. stress) if the determinants are known (e.g. derived from pertinent middle range theories); alternatively, the terms “factors,” “determinants,” “causes,” “predictors,” or similar ones are used in the initial search; and additional terms reflecting the specific factors identified through the review of the initial literature sources are used to refocus the search.
3 Conducting the search: The search for literature relevant to the experience and determinants of the health problem is carried out using multiple bibliographic databases pertaining to health literature in general or to specific health‐related disciplines (e.g. CINAHL, MEDLINE, PSYCHINFO), for comprehensiveness. The search may yield a large number of studies whose abstracts should be reviewed to determine their relevance. Setting reasonable limits to the search (e.g. language of publication and time period) enhances the relevance of the literature to the most recent conceptualization of the problem; further, the limits may render the review more focused and manageable.
4 Specifying the selection criteria: The criteria ensure the relevance of the literature sources to be selected. As mentioned previously, theoretical or conceptual and empirical sources can be included in the review if they address any aspect of the health problem (e.g. presentation of its conceptual definition, description of its indicators, and examination of its determinants). Quantitative and qualitative studies utilizing a range of research designs are sought, regardless of their quality. Quality is defined in terms of appropriately preventing or addressing biases that threaten validity of findings. Quality of the studies can be assessed with available tools and taken into account during the data analysis and synthesis steps.
5 Identifying and selecting sources: Having a list of the preset selection criteria and clearly specified definitions and indicators of these criteria facilitates the review of the sources. The selection is done in two stages. First, the abstract of the identified source is read and its content evaluated for relevance (i.e.