Types of studies

We will include primary studies that use qualitative study designs such as ethnography, phenomenology, case studies, grounded theory studies and qualitative process evaluations. We will include studies that use both qualitative methods for data collection (e.g. focus group discussions, individual interviews, observation, diaries, document analysis, and open‐ended survey questions) and qualitative methods for data analysis (e.g. thematic analysis, framework analysis, grounded theory). We will exclude studies that collect data using qualitative methods but do not analyse these data using qualitative analysis methods (e.g. open‐ended survey questions where the response data are analysed using descriptive statistics only). We will include mixed methods studies where it is possible to extract the data that were collected and analysed using qualitative methods. We will include studies regardless of whether or not they are linked to an intervention. We will not exclude any studies based on our assessment of methodological limitations, but will utilise this information to assess our confidence in the synthesis findings.

Topic of interest

The topic of interest in this synthesis is the factors which influence the acceptance of routine childhood vaccination from the perspective of parents and informal caregivers.

Types of participants

The review will include studies that report on views, experiences, decision‐making or factors influencing acceptance regarding routine childhood vaccination, as given by parents or informal caregivers. By ‘informal caregiver’ we mean anyone who is directly involved in caring for the child or making the decision to vaccinate the child, or who has the responsibility to take the child for vaccination (Ames 2017). We will include studies if they report on the views, experiences, decision‐making or acceptance of routine childhood vaccination, as given by prospective parents (e.g. pregnant women, their partners, or both). We will exclude papers if they only report what healthcare providers, policy‐makers, programme administrators and managers or other immunisation stakeholders’ say about the views, experiences or decision‐making of parents and informal caregivers or the factors influencing vaccination acceptance.

Types of interventions

We will include studies about parents’ or informal caregivers’ views, experiences, decision‐making or acceptance regarding routine childhood vaccination irrespective of the vaccination setting or mode of delivery. Vaccination settings and modes of delivery may include, for example, vaccination provided at healthcare facilities or fixed outreach sites, or by mobile health teams in communities (Machingaidze 2013b).

By ‘routine’ childhood vaccination we mean WHO‐recommended routine vaccines for children under six years of age as part of the Expanded Programme on Immunization (EPI). This will include the following vaccines or combinations of vaccines (WHO 2018b).

1. Bacillus Calmette‐Guérin (BCG)

2. Hepatitis B

3. Polio

   1. Oral poliovirus vaccine (OPV)

   2. Inactivated polio vaccine (IPV)

4. Diphtheria, tetanus, and pertussis‐containing

5. Haemophilus influenzae type b (Hib)

6. Pneumococcal conjugate vaccines (PCV)

7. Rotavirus

8. Measles

9. Mumps

10. Rubella

11. Measles, mumps and rubella (MMR)

12. Japanese encephalitis

13. Yellow fever

14. Tick‐borne encephalitis

15. Typhoid

16. Cholera

17. Meningococcal

18. Hepatitis A

19. Seasonal influenza

We will not include the following vaccines because children under six years of age do not routinely receive them as part of the EPI.

1. Human papillomavirus (HPV)

2. Rabies

3. Haemagglutinin type 1 and neuraminidase type 1 (H1N1), and other epidemic vaccinations.

We will exclude studies on hypothetical vaccines, future vaccines or vaccination trials. We will also exclude studies if it is not possible to separate out the data on views of routine childhood vaccination from views of vaccination in other age groups (e.g. adolescents, adults) or views on vaccination in general.

Setting

We will include studies from any setting globally where childhood vaccination is provided. These settings could include healthcare facilities, fixed outreach sites and mobile health teams in communities.

Electronic searches

We will search PDQ‐Evidence (pdq‐evidence.org) for related reviews in order to identify eligible studies for inclusion. We will also search the following electronic databases to identify eligible studies.

1. MEDLINE (Ovid)

2. Embase (Ovid)

3. CINAHL (EBSCO)

4. Anthropology Plus (EBSCO)

5. Web of Science Core Collection (Clarivate Analytics)

6. PsycINFO (Ovid)

Using guidelines developed by the Cochrane Qualitative & Implementation Methods Group for searching for qualitative evidence (Harris 2018), we will develop search strategies for each database. We will not apply any limits on language or geographic location. We will search all databases from 1974 to the date of search. This date range is intended to capture parents' and informal caregivers' views and experiences of routine childhood vaccination since the introduction of the WHO's EPI. We will include a methodological filter for qualitative studies. See Appendix 1 for the MEDLINE search strategy, which we will adapt for other databases. We will provide appendices detailing all strategies used.

Selection of studies

We will collate records identified from different sources into one database and remove duplicates. Two review authors (SC, ES) will use Covidence (https://www.covidence.org) to independently assess the titles and abstracts of the identified records to evaluate eligibility. We will retrieve the full text of all the papers identified as potentially relevant by one or both review authors. These papers will then be assessed independently by the same two review authors (SC, ES). Disagreements will be resolved via discussion or, when required, by seeking a third review author’s opinion. Where appropriate, we will contact the study authors for further information. We will include a table listing the studies excluded from our synthesis at the full‐text stage, and the main reasons for exclusion. Where the same study (using the same sample and methods) has been presented in different reports, we will collate these reports so that each study (rather than each report) is the unit of interest in the review. We will include a PRISMA flow diagram to show our search results and the process of screening and selecting studies for inclusion.

Translation of languages other than English

Articles will only be selected if they are published in languages spoken by the review authors. We have chosen this approach due to the challenges associated with translating papers reporting qualitative research, the resources required for this, and our use of a meta‐ethnographic approach which requires a full understanding of the meaning of concepts. This means that articles in French, English and German will be included. For papers published in French or German, we will attempt to translate the core meaning of the relevant themes or concepts into English. However, we acknowledge that some phrases and concepts are difficult to translate from one language into another, and that what constitutes the most appropriate translation is a judgement.

Sampling of studies

Large numbers of studies can impair the quality of the analysis in qualitative evidence syntheses (Ames 2017). Moreover, syntheses of qualitative studies aim for greater variation in concepts and depth of understanding, as opposed to an exhaustive overview of every study (Hannes 2013). Therefore, once all eligible studies have been identified, we will consider whether the amount of data in the included studies is too large and could impact the quality of the synthesis. If this is the case, we will sample from the included studies. Here we will explore the appropriateness of different purposeful sampling strategies (Suri 2011). One option might be to deploy a maximum variation (heterogeneity) sampling approach. This will comprise identifying potential areas of variation (e.g. geographic setting, type of health settings, type of vaccine, country income level, and study method), creating a sampling frame based on these dimensions and then mapping all eligible studies onto the frame (Suri 2011). A second option might be to use a theoretical sampling approach, whereby specific constructs and associated operational definitions will be predetermined, and studies will be selected based on their revelation of important theoretical insights about the predetermined constructs (Suri 2011). We will make decisions about whether to deploy sampling, and if so which strategy to use, once all eligible studies have been identified and we are more familiar with the available literature.

Data extraction

We will perform data extraction using a data extraction form designed specifically for this review. The form will be used to extract information on: first study author, date of publication, country of study, context (urban, rural; high‐income country, low‐ and middle‐income country), participant groups (first‐time parent/informal caregiver, older parent/informal caregiver, etc.), number of participants, vaccine, study design, objectives, guiding theoretical or conceptual framework, and data collection and analysis methods.

Data management and synthesis

We will use a meta‐ethnographic approach, following the steps outlined originally by Noblit and Hare (Noblit 1988), and the eMERGe meta‐ethnography reporting guidance (France 2019). Meta‐ethnography is an interpretive, rather than aggregative, qualitative synthesis approach which is well‐suited to producing new concepts, theories and theoretical models (France 2014; Noyes 2018). As a central aim of this synthesis is to derive new conceptual understandings of the factors influencing parents’ and informal caregivers’ acceptance of routine childhood vaccination, we decided that a meta‐ethnographic approach was the most appropriate synthesis method. Meta‐ethnography involves induction and interpretation to translate and synthesise conceptual data identified within included studies into a higher‐order interpretation. The analysis is built up study by study, in a manner that both preserves the context of the primary data within individual studies, and facilitates an understanding of how concepts in different studies are related to each other (France 2019).

We will begin by identifying conceptual data (e.g. concepts, themes, ideas, metaphors) through reading and rereading the included studies. In this process, participant quotes will be understood as reflecting primary themes or first‐order constructs. Secondary themes or second‐order constructs will be understood as concepts developed by the authors of primary studies (Britten 2002). However, we recognise that all constructs are the result of author selection and interpretation (Schutz 1971). One review author (SC) will extract first‐ and second‐order constructs, as well as relevant data on study characteristics and contexts, from all the studies. Two other authors (BS, NL) will extract these data from a sample of studies. Data will be extracted from the full primary studies and any additional linked online files of papers where available. The three authors (SC, BS, NL) will compare their extractions and resolve any differences through discussion and reference to the original studies.

The findings, as well as other potential aspects of the studies (such as research design, contextual factors), will then be compared to determine how studies are related. A wide variety of methods exist for this process; we will make a decision about the most appropriate method for comparison once we are more familiar with the included studies. One author (SC) will lead the process of comparison, with discussion and input from the rest of the review team.

The next step will involve translating the studies into each other. We will consider doing a ‘reciprocal analysis’ to translate concepts from individual studies into one another, or a ‘refutational analysis’ to explain differences and to explore and explain incongruities, exceptions and inconsistencies. We will make a decision regarding which translational approach to use (or a combination of both) once we have determined whether the studies and concepts within the studies relate reciprocally or refutationally (or both). Three review authors (SC, BS, NL) will conduct the translation, returning to the full text of papers frequently to ensure the emerging translation is meaningful.

The final step will involve a 'lines of argument synthesis’ to develop synthesised translations (or third‐order interpretations) and an overarching conceptual framework. We will compare the translated themes and interpretations, considering if and how they might be linked. Three review authors (SC, BS, NL) will consider the linkages and develop an initial conceptual framework which describes the factors influencing parental acceptance of childhood vaccination. At this stage, we will examine other, existing frameworks on vaccination acceptance (See 'Description of the topic' section, above). We will consider whether our initial framework ‘fits’ with any existing framework(s); whether our initial framework can be used to refine/adapt/amend any existing framework(s); or whether our initial framework suggests a new conceptual framework needs to be developed. These considerations will be discussed between the three authors (SC, BS, NL), and then the rest of the review team, to develop an integrated, refined or new conceptual framework. The conceptual framework will be discussed and re‐worked until a team consensus is reached. This final framework will provide a theoretical understanding of what, and how, different factors come together to influence parental acceptance of childhood vaccination. It will resemble a re‐interpretation of meaning across studies, and represent more than an aggregative descriptive account (Barnett‐Page 2009). This framework will also potentially be used to link our findings with the related Cochrane intervention reviews (Saeterdal 2014; Oyo‐Ita 2016; Jacobson 2018; Kaufman 2018).

Assessment of methodological limitations in the included studies

Our inclusion criteria specify that included studies need to use both qualitative data collection and analysis methods. This constitutes a basic quality threshold as we will exclude studies that have used qualitative methods to collect data but not to analyse these data. In addition, two review authors (SC, BS) will independently assess methodological limitations for each study using an adaptation of the Critical Appraisal Skills Programme (CASP) quality assessment tool for qualitative studies (Atkins 2008; CASP 2013). Disagreements will be resolved through discussion between the two authors or through consultation with a third author (AS or CC). The adapted tool includes the following 8 questions which will be used to assess methodological limitations.

1. Are the setting(s) and context described adequately?

2. Is the sampling strategy described, and is this appropriate?

3. Is the data collection strategy described and justified?

4. Is the data analysis described, and is this appropriate?

5. Are the claims made/findings supported by sufficient evidence?

6. Is there evidence of reflexivity?

7. Does the study demonstrate sensitivity to ethical concerns?

8. Any other concerns?

We will conduct a pilot on three included studies to assess the feasibility of the use of this tool and ensure integrity of the assessment. We will not use the quality assessment approach to exclude studies but rather to judge the relative contribution of each study to the development of explanations and relationships and as part of the assessment of how much confidence we have in each finding (see below). We will report our assessment of methodological limitations for each study in the 'Characteristics of included studies' tables.

Assessment of confidence in the synthesis findings

Two review authors (SC, BS) will use the GRADE‐CERQual (Confidence in the Evidence from Reviews of Qualitative research) approach to summarise our confidence in each finding (Lewin 2018). CERQual assesses confidence in the evidence, based on the following four key components.

1. Methodological limitations of included studies: the extent to which there are concerns about the design or conduct of the primary studies that contributed evidence to an individual review finding.

2. Coherence of the review finding: an assessment of how clear and cogent the fit is between the data from the primary studies and a review finding that synthesises those data. By cogent, we mean well supported or compelling.

3. Adequacy of the data contributing to a review finding: an overall determination of the degree of richness and quantity of data supporting a review finding.

4. Relevance of the included studies to the review question: the extent to which the body of evidence from the primary studies supporting a review finding is applicable to the context (perspective or population, phenomenon of interest, setting) specified in the review question.

After assessing each of the four components, we will make a judgement about the overall confidence in the evidence supporting the review finding. We will judge confidence as high, moderate, low, or very low. The final assessment will be based on consensus among the review authors. All findings start as high confidence and will then be graded down if there are important concerns regarding any of the CERQual components.

'Summary of Qualitative Findings' table

To facilitate understanding and use of the review findings, we will present them in a 'Summary of Qualitative Findings' (SoQF) table. The table will display a structured summary of each review finding and references to the studies contributing data to each finding. It will also provide our assessment of confidence in the evidence as well as an explanation of this assessment, based on the GRADE‐CERQual approach (Lewin 2018). All review findings will be reported in the SoQF table regardless of their associated level of confidence. The final conceptual framework developed will also be presented in a summarised narrative, and visually if appropriate.

Linking the synthesised qualitative findings to a Cochrane intervention review

As part of data synthesis, we plan to explore how the findings from our review relate to the findings of the related Cochrane intervention reviews (Saeterdal 2014; Oyo‐Ita 2016; Jacobson 2018; Kaufman 2018). Our review findings might help to explain and contextualise why some interventions to improve vaccination uptake are effective (or more effective) and some are not, and may help generate hypotheses about potential important differences (e.g. setting, population, or vaccine) for planned subgroup analyses in future effectiveness reviews. Our review may also help identify factors that are considered important to parents and informal caregivers, for consideration in future intervention development and trials. The findings of this review will also complement those from a Cochrane qualitative evidence synthesis on parents' and informal caregivers' views and experiences of communication about routine childhood vaccination (Ames 2017).

We will explore the appropriateness of using two different supplementation approaches, and make a decision based on our synthesis findings. One option will be to deploy a logic model approach (Glenton 2013; Shepherd 2014), to link the review findings with a selection of conclusions and outcomes drawn by the intervention reviews. This will involve using the conceptual framework developed through our review as a starting point to develop a logical flow of theoretical hypotheses through which different factors, based on evidence from the qualitative synthesis, could affect the outcomes explored in the related Cochrane intervention reviews. The objective will be to present theories or assumptions about possible links, rather than to ‘prove’ causal associations. A second option will be to use a matrix model approach (Thomas 2004; Candy 2011), to configure and compare the findings of the reviews. This will involve developing a comparative table to explore how the reviews relate, and how the findings from the qualitative review inform the findings from the intervention reviews and vice‐versa. For example, we might consider whether the interventions studied in the related Cochrane reviews contain or address features that parents and informal caregivers identified as important, or perceived as facilitators to vaccination acceptance. At least two review authors will work together to assess how our findings relate to the results of the intervention reviews, and how best to present them.