be explored and better understood only through qualitative research. Like quantitative studies, multiple qualitative studies are conducted on same or related research questions across the settings, addressing the qualitative aspects. In this background, synthesising findings from multiple qualitative studies, by pooling their reports is a necessity. Meta-synthesis or qualitative evidence synthesis (QES) is the counterpart of the meta-analysis wherein the studies addressing qualitative research questions are systematically reviewed and synthesised [6–8]. QES is defined as “an approach for synthesising the findings from multiple primary qualitative studies”.[9] It is also defined as “an umbrella term for the methodologies associated with the systematic review of qualitative research evidence, conducted either as a stand-alone review or as a part of a review of complex interventions, systems or of guideline development” [10,11].

Applications of meta-synthesis or QES

Meta-synthesis or QES has found applications in multiple domains. It has surpassed the initial phases where it was used to pool and synthesise only the findings of primary qualitative studies. It has been used in establishing the “relative importance of outcomes, the acceptability, fidelity and reach of interventions, their feasibility in different settings and potential consequences on equity across populations.”[11] In the EBM framework, meta-analysis provides the pooled estimate, which is the high-quality evidence desired for taking out recommendations. Similarly, meta-synthesis can majorly help in understanding the perspectives and preference of the patients and communities (patient values dimension) for whom the interventions are intended [11].

Once the evidence is generated and synthesised, the logical conclusion can take place only after they are transformed into a clinically meaningful guidelines or recommendations. Evidence to Decision (EtD) frameworks include multiple aspects among which considerations to the stakeholder preference and other specific qualitative factors are involved [12–14]. A context specific, qualitative evidence synthesis on these factors,[11] can complement the quantitative evidence synthesis (meta-analysis) recommendations. Guideline development groups (GDGs) and the policy makers are advocated to consider the inclusion of qualitative findings along with quantitative evidence in the guidelines involving complex interventions. This can assist the key stakeholders to finalise recommendations (decision criteria),[9] which can ensure that the two of the dimensions of the EBM (high quality evidence and patient values) are incorporated in the guidelines. Additionally, QES can enable the mixed-methods review of complex healthcare interventions. This is achieved by following certain integration mechanisms (Segregated design, sequential synthesis) with the quantitative synthesis [13,15] which can again be a crucial informer in guideline development. QES can also aid in the steps of broader policy making process such as defining the problem, identifying programs that could impact the health problem and identifying potential determinants for implementation of the recommendations at various levels and among various stakeholders [16].

Multiple criteria within the EtD framework such as how people value the outcomes, gender, health equity and human rights impacts, acceptability and feasibility have been populated by means of the QES findings [17]. Guidelines on intrapartum-care episiotomy have reported the QES findings assisting to understand the acceptability of the procedure by the women [17–19]. Similarly, QES by Ames et al informed the guidelines on “Communication interventions to inform and educate caregivers on routine childhood vaccination in the African Region – Face-to-face interventions and community-aimed interventions”, specifically in terms of the gender, health equity and human rights aspect [20]. Apart from the decision making on a specific healthcare research question, QES also enables the decision on the scope of the research question itself. World Health Organisation’s (WHO) antenatal and intrapartum care guidelines was informed by the qualitative systematic reviews undertaken to identify the relevant outcomes for the guidelines [9,21,22]. WHO guidelines on digital interventions for health system strengthening, and antenatal care for a positive pregnancy experience are examples where QES findings were integral in informing the recommendations [19,23]. Implementation of recommendations has also been guided by the QES findings. In the guideline where epidural analgesia was recommended for pregnant women, [24] QES informed the implementors regarding the setting specific factors affecting the feasibility and acceptability of the epidural analgesia [22].

Thus, QES has a wide range of applications, particularly in fields where understanding complex human experiences, behaviours, and contexts is essential. It informs evidence-based practice by providing a comprehensive understanding of qualitative findings, guiding healthcare practitioners

in tailoring interventions to patient needs and preferences [25]. QES is instrumental in policymaking by helping decision-makers incorporate diverse perspectives and values into policies [13]. It aids in developing theoretical frameworks and conceptual models by integrating findings from multiple qualitative studies [26]. In clinical settings, it supports the design of patient-centered care approaches and the development of tools that address specific community or individual needs [27]. It plays a critical role in education by synthesizing insights into learning experiences and teaching methods, enhancing curricula and instructional design [28]. QES is valuable for understanding disparities and barriers in access to healthcare, enabling the creation of equitable health interventions [29]. Additionally, it supports systematic reviews by providing context and depth to quantitative findings, making the results more actionable [30]. Overall, QES serves as a vital tool for bridging the gap between research, practice, and policy.

Steps in qualitative evidence synthesis

The methodology of QES and meta-analysis are similar till the stage of identification of the eligible studies, following all major steps of systematic reviews [1]. Overall steps of the QES are outlined in Figure 1.

Figure 1: Steps in undertaking the qualitative evidence synthesis (QES)

Formation of QES team

Forming a team comprising of the subject experts (health concepts studied in the meta-synthesis) and methodological experts (experts in qualitative research and the QES) shall be the first step while starting the QES. The review team should be in constant touch with all relevant stakeholders affected or associated with the impact of the findings of the QES.

Identification of research question and the framework

This shall be followed by identification and refining of the research question, which is an iterative process. Preliminary, multiple rounds of systematic searches in one of the relevant databases

(preferably MEDLINE through PubMed) may be done to refine and finalise the research question. Continuous discussion between the subject and methodology experts in the team should take place during the process of finalising the research question. Once the research question is finalised, appropriate framework must be identified, and the inclusion and exclusion criteria for each component in the research question must be elaborated. It also aids in developing the search strategy for the database searches. Frameworks used to define primary qualitative research question can be used for the QES as well (Table 1).

Table 1: Common frameworks which can be employed for research questions in QES[8,31]

Protocol registration

A detailed protocol describing the research question and the subsequent steps (which are discussed below) needs to be prepared before starting the review. The protocol shall be prospectively registered with one of the appropriate registries such as Prospero, INPLASY, Protocols.io, among others [32–35].

Source of the studies

The source of the potentially eligible studies needs to be explicitly mentioned. Common sources include databases, websites, policies, and other grey literature. Database search will be the primary source and the major databases searched during QES are MEDLINE, Scopus, EMBASE, Cochrane Central Register of Controlled Trials (CENTRAL), PsycInfo, CINAHL, Web of Science, ProQuest dissertations and theses, sociological abstracts and EBSCOHost [36–44]. Each database requires specific search strategy to be built. Tailored search strategies for retrieving qualitative studies can be made in CINAHL, [45] EMABSE, [46] MEDLINE, [47] and PsycINFO [48]. Clinical trial registries may also be searched, according to the research question. Frameworks such as STARLITE and PRISMA-S can be used for reporting literature searches in the QES [49,50]. It has been shown that 93.1% of the eligible qualitative studies can be retrieved using four databases (PubMed, CINAHL, Scopus, ProQuest Dissertations and Theses Global) with the maximum unique retrieval rates for the following databases: Scopus (83.6%), PubMed (72.3%) and EMABSE (71.7%) [51]. However, 5.6% of the studies could not be located in any of the commonly used nine databases [51]. Thus, other sources and methods can be explored to identify further qualitative studies. Google Scholar may be utilised as secondary search source. Preprint servers such as medRxiv, arXiv, bioRxiv, BioRN, ChemRxiv, and SSRN can also be explored for unpublished literature [52–56]. Citation searching of eligible studies, hand searching in the relevant online sites and speciality journals, can also be undertaken [57,58].

Screening of the studies

The search in multiple sources yield potentially eligible studies, which needs to be deduplicated to yield a set of unique studies. Tools such as Mendeley, EndNote, Rayyan, COVIDENCE and NESTED Knowledge assists in the process of deduplication [1]. Following this, screening of the studies shall be done in same pattern (Two pass, Dual screening) as it is done in the systematic review of quantitative studies: title abstract screening followed by full text review of the articles (Two pass), which is undertaken by two independent reviewers (dual screening). The screening shall be done based on the elaborate inclusion and exclusion criteria made for each component in the research question framework. The “7 S” approach elaborated by the Booth and colleagues can assist the QES authors in undertaking the literature search and defining the eligibility (in terms of inclusion and exclusion criteria) [44,59].

Data extraction

The final list of eligible primary studies after screening will undergo data extraction. The manuscript meta-data, baseline characteristics and qualitative findings of the articles will be extracted. The qualitative data can be in the form of verbatims, themes, sub-themes derived by the authors, novel theories and other interpretations expressed by the authors based on the above data. These data could be presented as narrative or summaries, depicted as tables, infographics or models [13]. A key consideration during QES data extraction is that it is not a sequential or linear process, but an iterative process [59]. More often it could be back and forth between the review stages such as data extraction, analysis and synthesis [59]. Continuous meetings of the authors, at times along with the other stakeholders relevant to the project, needs to be undertaken during the data extraction [8]. It is also recommended to report the verbatims from the primary studies included in the QES [60]. Qualitative data management software such as NVIVO, Atlas-Ti can also be utilised for data extraction and analysis in QES, especially when large volume of data is present [59,61,62].

Techniques such as sub-group analysis may also be planned according to the context and available data in the primary studies to further improve the findings of the QES. Under the characteristics of the included studies table, it is essential to extract information on the i) context and participants, such as the settings, study population characters, specific participants characters, description of the intervention or concept studied; ii) study design and methods applied in the primary study for conducting, sampling, collect, code, analyse data and the models used to interpret the findings.[8] Tools such as TIDieR and ICAT_SR may assist in identifying important information during the data extraction [8,63,64]. QES authors may build a data extraction sheet, either by modifying the existing standard sheet or by their own as per the requirement. They may use a priori framework to obtain data, or they may use software to code and extract the data which build themes inductively [25,65,66].

Risk of bias assessment

The risk of bias assessment of the individual studies included in the meta-synthesis needs to be done using appropriate tools such as “Critical Appraisal Skills Program (CASP)” Qualitative Research Checklist, QARI, COREQ, Mays and Pope [67–70]. Both data extraction and quality assessment of the included studies are recommended to be done by two reviewers independently.

Any of the following conflict resolution methods can be adopted by the reviewers to sort out the contradictions which might arise out of the steps wherever two independent reviewers were involved (screening, data extraction, coding, analysis, quality assessment): i) consensus meeting between the two reviewers, ii) adjudication by the third reviewer, iii) consensus meeting by including the third reviewer.

Artificial Intelligence-Machine Learning (AI-ML) enabled semi-automated tools such as NESTED Knowledge, EsySLR have become available to ease and quicken the literature search, screening, risk of bias assessment and data extraction in the systematic reviews [71,72]. These can be leveraged while undertaking the QES.

Synthesis of the findings

The QES method/approach adopted, its principle and the outcomes need to be prespecified in the methodology. Broadly the QES approaches are divided into integrative and interpretive [11]. Integrative is more of deductive approach which takes the concepts and themes explicitly mentioned in the pre-existing primary research. Interpretive is more of inductive approach in which the themes and concepts are developed based on the data and information from the primary research [11]. The specific method which needs to be chosen depends on the type of research question, and the framework and the context in which factors are explored in the QES [11]. Commonly used methods in QES are enumerated in Table 2 [16,25,65,73]. Thematic synthesis can be the choice of QES method when the data is quite ‘thin’ to create descriptive themes as well as where the data are ‘thicker’ to formulate descriptive themes into a more “in-depth analytic themes” [11]. When the QES involves complex interventions, framework synthesis can be the better choice of QES method, since it can accommodate any complexity within its framework. Best fit synthesis can be applied in QES when there is broad agreement regarding the characteristics of the interventions and their effects [11,13]. Meta-ethnography requires presence of ‘thick/rich’

data in the primary studies included in the QES review. Being an interpretive synthesis, meta-ethnography can generate higher order constructs, and can assist in the guideline development process [11]. Overall, the choice of the QES methods depends on the level of conceptual and contextual details in the included primary qualitative research, with time, resources and data available as the limiting factors [11]. RETREAT framework, which is composed of seven criteria (Review question – Epistemology – Time/Timescale – Resources – Expertise – Audience and purpose – Type of Data), has been formulated to assist the evidence synthesis researchers for choosing the appropriate QES approach [74].

Table 2: Common QES methods[16,25,65,73]

Certainty in evidence

After the synthesis of the findings from the included studies, it is essential to ascertain the certainty of the evidence. Tools such as GRADE-CERQual are available to assess the certainty in evidence from the meta-synthesis [75,76]. GRADE-CERQual evaluates the synthesis findings through four domains: Methodological limitations, Relevance, Coherence and Adequacy.

Reporting guidelines for QES

Enhancing transparency in reporting the synthesis of qualitative research (ENTREQ) guidelines can be adopted while preparing and reporting the QES manuscript [60]. It is a 21 item checklist which provides section-wise reporting guidelines for QES manuscript. Reporting guidelines can improve transparency of QES. In addition to the above items, certainty in the evidence and the tool used to assess the certainty can also be ensured while reporting the QES. Other reporting tools such as RAMESES (realist syntheses and meta-narrative reviews) and eMERGe (where methods used to translate meaning from one study to another) may also be used [16].

Ethical considerations

Systematic reviews (which also includes QES) generally do not have any ethical issues involved, since they are based on the data of the primary studies which were already published in journals. Hence, the review or approval from the ethics committee or institutional review board is usually not required for undertaking QES. However, the authors shall adhere to the general ethical practices in undertaking and publishing systematic reviews, and register their protocol with a prospective registry [77,78].

Challenges

Certainty assessment in the QES is still an evolving concept [75]. One of the major challenges is that there are fewer innovations and less awareness about the synthesis of qualitative data. The relatively subjective aspects of the qualitative study findings and the heterogeneous population poses practical challenges while trying to interpret the synthesised findings. This becomes more challenging when the team undertaking the QES does not include researchers from all the countries and settings from where the primary qualitative studies were included in QES. Maintaining a balance between the synthesised findings from multiple primary studies potentially from different contexts, without compromising their original meaning and complexity, is another challenge [79]. Lack of a more specific critical appraisal tool for the meta-synthesis studies is also a challenge while assessing the methodological rigor of the meta-synthesis. Relatively large number of studies for screening, language bias and database bias are other factors that poses challenges in undertaking the QES [79]. Machine learning (ML) can enable optimal and effective use of resources in evidence synthesis [80]. Semi-automated artificial intelligence-machine learning (AI-ML) based tools such

as NESTED Knowledge can enable the QES reviewers to overcome the challenges in quantum of the studies for screening. Creating a broad base of expert panel for the QES review, by including authors from all the settings/countries of the primary articles, can ensure better interpretation of the synthesis.

Opportunities

Increased incorporation of QES in the guideline development process has also opened the scope for undertaking QES as an essential activity for enabling and better informing the GDGs. Commissioning QES on the patient/stakeholder preferences, attitudes towards various modalities of treatment/interventions/concepts evaluated and other complexities in healthcare interventions and process can be done as a part of the development of the evidence-based guidelines. Introduction of living systematic reviews (LSRs) has led to the concept of living guidelines (LG),[81,82] wherein multiple LGs are already ongoing [82,83]. Undertaking LSRs of QES towards informing the LG is being explored, albeit with challenges,[82] which can be another opportunity to work in this domain.

QES are very much context specific and hence requires separate grouping of the studies for different context. Hence, there is an existing geographic specific scope and gap to work in QES of studies, especially in low- and middle-income countries. Further work can be undertaken in the development of critical appraisal tools specific for the QES, and to expand the reporting guidelines of QES by incorporating the certainty assessment components and AI-ML enabled QES methods. Improvement in the assessment tools for reporting the certainty in the evidence of QES is another domain that can be worked upon. There is also scope for conducting umbrella reviews of QES of related and similar research questions. A rapid search in MEDLINE (through PubMed) on 11.12.2024 revealed only six results for the umbrella reviews of QES, and the earliest one being conducted as recent as 2020 only.

Recommendations

QES shall be incorporated into the guideline development process, from the stage when the research question is refined and the relevant outcomes are decided, to the final guideline finalisation stage. It is essential to inform the GDGs on the availability and utility of the QES in various stages of guideline development. A QES methodology expert may be included as a part of the GDG, who can assist the other GDG members and the funders in recognising the qualitative aspects of the health condition and weave the QES findings into EtD [17]. Further methodological research into the application and strategies to identify, present and incorporate the QES findings into the guideline development is also warranted.

Conclusion

Overall, QES is an evolving concept within the broader evidence synthesis specialty. It can assist in the guideline development of complex healthcare interventions, by informing the scope of the guidelines, outcomes to be studied, and stakeholders of various context specific determinants. Albeit the existing gaps and challenges, it is recommended to include QES research questions and methods during the disease/health state specific guideline development. This can enable a more robust practice of EBM. Further research needs to be undertaken to improve the techniques and tools applied in undertaking, evaluating, reporting and certainty assessment of QES. Effective methods to incorporate the QES findings into the guideline development for informing the policies also needs to be framed through future research.

Abbreviations

EBM: Evidence based medicine

QES: Qualitative evidence synthesis

EBPH: Evidence based Public Health

AI-ML: Artificial Intelligence-Machine Learning

Supporting information: None

Ethical Considerations: Not applicable

Acknowledgments: The authors acknowledge the guidance and support provided by the Technical Resource Centre (Centre for Evidence for Guidelines) at the Department of Community Medicine, All India Institute of Medical Sciences, Nagpur, India for undertaking the review.

Funding: This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

Author contribution statement: All authors (APG, PD, AD) contributed equally and attest they meet the ICMJE criteria for authorship and gave final approval for submission.

Data availability statement: Data included in article/supp. material/referenced in article.

Additional information: No additional information is available for this paper.

Declaration of competing interest: Dr Aravind P Gandhi is the managing editor of the journal and he did not handle the peer review process and/or the final decision of publication of the manuscript.

Clinical Trial: Not applicable

Consent for publication: Note applicable

References