Search strategy

A comprehensive search strategy has been developed to capture all literature relevant to adolescent self-consent procedures for vaccination programmes by a reviewer (HF) experienced in undertaking systematic reviews in the proposed research field and discussed with members of the research team. The original search strategy developed for the Embase database has been adapted for each included database (see below) and comprises a combination of text words and the following medical subject headings (MeSH) indexing terms: ‘child’, ‘adolescent’, ‘active immunization’, ‘immunization’, ‘immunization programs’, ‘mass immunization’, ‘revaccination’, ‘vaccination’, ‘diptheria vaccine’, ‘diptheria tetanus vaccine’, ‘diptheria pertussis tetanus’, ‘haemphilus influenzae type b vaccine’, ‘hepatitis b vaccine’, ‘meningcoccus vaccine’, ‘rubella vaccine’, ‘wart virus vaccine’, ‘papillomavirus vaccines’, ‘decision making’, ‘informed consent’, ‘parental consent’, ‘treatment refusal’ (box 1). Study design filters or restrictions by setting will not be applied as the study aims to be inclusive in relation to study design and settings eligible for inclusion.

Databases

To ensure all the relevant literature is captured, we will search the following 10 databases from inception to January 2018 and reupdated ahead of study completion to inform the wider research study as it progresses: Child Development & Adolescent Studies via EBSCOhost, Cochrane Central Register of Controlled Trials via The Cochrane Library, Cochrane Reviews via The Cochrane Library, Cumulative Index to Nursing and Allied Health Literature via EBSCOhost, Embase via Ovid, Health Technology Assessment Database, Medline via Ovid, PsycINFO via Ovid, Social Care Online via Social Care Institute for Excellence and Web of Science Core Collection: Social Sciences Citation Index and Conference Proceedings Citation Index-Science. All abstracts will be saved using Endnote X8.

Inclusion and exclusion criteria

Quantitative studies will be eligible if vaccine uptake following implementation of self-consent procedures is reported for young people aged between 10 and 18 years.11 Qualitative studies reporting the views and experiences of key stakeholders in relation adolescent self-consent procedures will also be included. Studies related to consent procedures solely targeting parents of adolescents or early childhood and adult vaccination programmes will not be eligible for inclusion. Relevant stakeholders will vary with context but are likely to include young people, parents or primary caregivers, healthcare professionals, policy makers, community leaders and teachers.

We will include a range of study designs. To determine whether self-consent procedures can increase uptake of vaccination programmes, primary studies reporting parallel group randomised controlled trials, quasi-randomised trials, non-randomised controlled trials, controlled before and after studies, historically controlled studies and retrospective or prospective cohort studies that include a control group will be eligible. Qualitative studies which use interviews, focus groups, observations or open-ended questions allowing free-text responses in questionnaires will be included to explore views and behaviours related to young people’s self-consent for vaccination.

Conference abstracts, reviews, editorials, opinion pieces, dissertations, letters and books will only be included if they present original data. There will be no language or country of origin restriction imposed and any relevant full text paper that is not written in English will be translated.

Study selection

Two reviewers will independently assess the titles and abstracts against the predefined eligibility criteria. Full-text publications of all potentially relevant articles will be retrieved and examined for relevance. Any disagreements arising will be resolved by discussion. The reference lists and bibliographies from relevant studies and systematic reviews will be hand-searched for additional primary studies not retrieved by the electronic search.

We will use the reference management software EndNote X8 to remove duplicates and sort exclusions and inclusions. The search strategy and study selection process will be documented using a PRISMA flow diagram.12

Data extraction

At least two reviewers will independently extract data from selected studies using structured and standardised data extraction forms used in our previous qualitative and quantitative systematic reviews. In instances where multiple publications relate to the same study, these will be reported together. The following domains will be retrieved: study characteristics (authors, publication year, country, aim, study time period, study design, location, type of setting, data collection period, data collection method, sampling strategy, analysis), participant characteristics (participant age, sample size, vaccination status of participants, socioeconomic indicators, race/ethnicity, gender and religion) and study results (uptake of vaccine, psychological outcomes, healthcare service use, incidence of vaccine preventable disease, views and behaviours related to self-consent procedures, authors’ reported conflicts of interest and study funding sources). We will also record data relating to the possible harms resulting from self-consent procedures (eg, conflict with parents, healthcare professional anxiety). Where possible, authors will be contacted for missing or incomplete data. Disagreements will be resolved through discussion.

Risk of bias and quality assessment

For eligible primary studies, quality assessment will be undertaken to illustrate potential sources of bias. As we anticipate the majority of eligible studies will be observational, studies will not automatically be excluded on the basis of ‘low’ quality assessment if they are considered to contribute relevant information. We propose using: the Cochrane Collaboration’s handbook for the assessment of risks of bias for systematic review of randomised controlled studies and quasi-randomised intervention studies;13 Risk Of Bias in Non-Randomised Studies of Interventions (ROBINS-I);14 the NIH Quality Assessment Tool for Observational Cohort and Cross-Sectional Studies15 and the Critical Appraisal Skills Programme criteria adapted for qualitative studies for evaluating qualitative research.16 Quality assessment of primary studies will be undertaken independently by two reviewers and recorded in an excel spreadsheet. An overall assessment of ‘high’, ‘medium’, ‘low’ or ‘unclear’ will be assigned and reported.

Data synthesis: quantitative studies

We anticipate that the primary quantitative studies will be reported narratively as preliminary searches specifically related to HPV vaccination programmes indicated a lack of published studies and the likelihood of heterogeneity in relation to study design and reported outcomes. However, if sufficiently similar studies are captured, we will consider combining individual study results through meta-analyses. To assess the heterogeneity between studies, we will use the Q-statistic and the I² -statistics.17 Evidence of heterogeneity will be classified as weak, moderate and strong for corresponding I² of 25%, 50% and 75%, respectively. If heterogeneity between studies is classified as weak, analyses will comprise adjusted ORs where available, with unadjusted ORs used if not reported. Analyses will be undertaken using the meta-analysis function18 available in Stata V.15. We do not anticipate sufficient data being available to undertake subgroup analyses. However, if sufficient data were reported, we propose two subgroup analyses to compare impact of self-consent procedures by: (1) setting (healthcare vs school) and (2) age of participants (less than 14 years old vs 14 years and greater).

Data synthesis: qualitative studies

The socioecological model19 considers that behaviour is shaped by a complex interaction between factors operating at public policy, community, organisational, interpersonal and intrapersonal levels. In a previous qualitative synthesis, we have shown that young women’s access to the HPV vaccine is shaped by decisions at different levels of the socioecological model.5 During the analysis, we will use the socioecological model to provide a framework for understanding how barriers and facilitators operating at different levels of the model can provide access to, or prevent, young people self-consenting in the context of vaccination programmes.

To analyse the qualitative data, the methodology for thematic synthesis reported by Thomas and Harden,20 assisted by the Framework method of qualitative data management,21 will be used. These methods are suited to studies with a priori aims and objectives. The overall purpose of the synthesis will be to ‘pool’ the results from individual primary studies by initially separating the findings, coding and interpreting the text and then combining them through the identification of key themes across the studies as well as similarities and differences within those themes.22 Thematic synthesis will be led by one reviewer reporting to the wider team about interpretation of the data as analysis progresses.

Familiarisation with the dataset will begin with reading the full papers. Pertinent sections of the text reported in each primary study will represent the basic units for analysis. Primary charts of the text will be constructed around key issues using the Framework Matrix within QSR NVivo11 software. For example, initial charts are likely to focus on ‘barriers’ and ‘facilitators’ to adolescent self-consent. The primary charts will be retained and revisited as required. Streamlined versions will be produced as the process of coding, summarising and synthesising the data progresses. In subsequent charts, key terms and phrases will be retained while repetition within studies and extraneous text are removed. During this process, overarching themes will be identified and differences or similarities explored within these emerging themes.

Data synthesis: interrogation

The final stage of the analysis will aim, first, to test whether the recommendations developed from the qualitative studies have been addressed in evaluative studies retrieved for the review and, second, to examine whether interventions that match the recommendations result in higher uptake in vaccination.9

Patient and public involvement

The Bristol Young People’s Advisory Group comprises young people aged 10–17 years who are interested in healthcare and research. They meet regularly to help researchers with their projects and have been consulted about the design of the wider study and participant materials. They will also be invited to an event at the end of the study to consider findings and recommendations with the young people, parents, immunisation nurses and school staff involved in the study.