Methods
Patient and public involvement
The systematic review will be conducted in collaboration with a study-specific Parent Advisory Group (PAG). Group members will include parents who have relevant and varied lived experience of having a baby in neonatal care and those who have been involved in providing peer support. Four face to face meetings will take place over the course of the study. In addition, parents will have opportunities to be involved in stages of the project as their circumstances allow.
The PAG will contribute to the protocol development by involvement in finalising the research questions, ensuring outcomes are meaningful to parents as well as professionals, refining the inclusion/exclusion criteria and helping with search terms. There will be opportunities for the group to undertake tasks (with full support and training) within the review process to the extent that they are willing or able to commit their time, e.g. screening papers for inclusion, reading and commenting on papers selected for inclusion, interpreting findings and contributing to the qualitative synthesis.
The PAG will be involved in writing and reviewing the academic papers and in co-producing dissemination products such as plain language summaries.
Objectives
Main objective
To bring together studies which have explored the experience of P2P support from the perspective of the persons giving and receiving P2P support, or those involved in implementing P2P support in the context of the provision of neonatal care.
Secondary objectives
1.
To determine the key characteristics, resources, components and processes that should be included in P2P support for parents of babies cared for in neonatal care settings, whilst on the unit and after discharge home.
2.
Where adequate data are available, to determine the effectiveness and cost-effectiveness of P2P support interventions on the health and well-being of parents and their babies.
3.
To explore what the findings mean for clinical practice and service provision.
Eligibility criteria
We will carry out a systematic review, which brings together all existing research on parent-to-parent support. This will be reported in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement [
31]. The protocol has been produced in accordance with the PRISMA-P reporting guidelines for systematic review protocols [
32] [see completed PRISMA-P checklist in Additional file
2]. The protocol is registered on the International Prospective Register of Systematic Reviews (PROSPERO), registration number CRD42018090569. A plain language protocol summary has been co-produced with the PAG and is available in Additional file
3.
Studies will be included if they describe the effects of P2P support and report data on one or more outcome measures. We will not restrict by study design but will stratify the reporting of results prioritising the most robust evidence. Qualitative studies that do not include a comparison group will not be excluded but will be used to inform the experiences reported by those offering and receiving P2P support where applicable.
Study reports
Editorials, opinions and letters will be excluded. Authors of studies published only as abstracts will be contacted and asked to provide further detail. If no further detail is available, the study will be excluded. Studies may compare support with no support or compare one type of support with another. The inclusion/exclusion criteria will be tested through piloting by two reviewers to establish agreement prior to commencing the study selection process. We will consider the relevance of study setting to the UK health care system. However, we believe that elements of P2P support provision in other health care settings may be transferable to the UK health care system. Only qualitative papers written in English will be included to avoid loss or distortion by translation from studies written in another language. Non-English language quantitative papers will be translated.
Participants
The population of interest is parents of babies cared for in neonatal units accessing the support service at any time (during their time in hospital or back home in the community). We will focus on P2P support interventions for those with experience of neonatal care (where those providing the support may be volunteers or paid parent workers).
Interventions
Studies will be included if they describe the effects of P2P support and report data on one or more outcome measures. We will exclude studies relating to interventions offered by professionals or interventions which offer instruction or training to parents rather than support, studies of peer support specifically for families affected by bereavement or for those whose babies are receiving palliative care and studies which do not adequately describe the intervention. It may not be possible to describe the ‘content’ of the intervention as conversations are individualised and private and do not necessarily follow a formal predefined structure. Adequate description of the ‘context’ of the intervention will therefore be sufficient including description of both parties, level of training of the person offering support, extent of support being offered in terms of time and availability and a description of the purpose of the intervention, for example emotional support, information and practical support. Efforts will be made to contact the authors of relevant studies for further clarification if the description is not sufficient before the study is excluded.
Studies may compare support with no support or compare one type of support with another.
Outcomes
Role of outcomes
Studies that report (quantitatively or qualitatively) at least one of the following outcomes will be included:
-
Experiences of people offering and receiving P2P support, including, but not solely, self-confidence, self-efficacy, knowledge, moving onto or back into employment, sense of belonging/friendship/being part of something
-
Measures of parent-child interactions; parent health, e.g. stress, depression, confidence, isolation, satisfaction and empowerment; child health and development, e.g. cognitive development, behaviour and language; family function including relationships and transitioning home; service use, e.g. readmissions and contact with other services; and resource use where reported
-
Contextual factors and modifiers to implementation of P2P interventions
-
Adverse outcomes
-
Effects on practice
Outcome domains of interest
Primary outcomes will include the experiences of people offering, receiving and implementing P2P support, for example outcomes relating to self-confidence, guilt, self-efficacy, knowledge, moving onto or back into employment, sense of belonging/friendship/being part of something, reassurance of shared experiences, and socialising with others who have similar experiences, agency and self-help, feeling of the baby belonging to the parent (or not), navigating the system, and managing differences between units.
Secondary outcomes will include measures of:
-
Parent-child interactions, ownership (e.g. “feeling like it’s your baby”)
-
Parent health e.g. stress, depression, guilt, confidence, isolation, satisfaction, empowerment;
-
Child health and development e.g. cognitive development, behaviour, language;
-
Family function including partner relationships, sibling relationships, mental health, transitioning between units and moving home;
-
Service use e.g. readmissions, contact with other services;
-
Resource use;
-
Potentially negative impacts of P2P support such as increased anxiety, learning details from others that were not expected/wanted or frightening;
-
Bad experiences of being a peer supporter such as the impact of a negative event (e.g. the death of a child).
Outcome measures of interest
It is likely that different measures and time points will be reported across studies. It is important that the variability is captured in the context of this review; therefore, specific measures/time points are not highlighted here. Where possible, these will be analysed together but where the measures and/or time points are greatly varied these results will be described and discussed narratively.
Minimally important difference
Due to the nature of this review, it is not thought that pre-specified minimally important differences are appropriate, though the relevance and potential impact of any differences in effects will be discussed.
Search methods
Search sources
The search strategy will cover both electronic bibliographic database searching and supplementary search methods to capture both published and unpublished (grey) literature. The database search will be designed by an information specialist (AB) using a combination of controlled vocabulary (e.g. MeSH) and free text terms. The PAG will be consulted to ensure unusual terminology is included. No language, date or study type limits will be applied. The bibliographic databases to be searched will include MEDLINE, EMBASE, PsycINFO, CINAHL, Social Policy and Practice (SPP), The Cochrane Library (CDSR, CENTRAL and DARE), Applied Social Sciences Index and Abstracts (ASSIA), British Nursing Index (BNI), Social Science Citation Index (SSCI), Health Management Information Consortium (HMIC), PQDT, Explore and Midwives Information and Resource Service.
Supplementary search
Supplementary search methods will include citation chasing using SCOPUS and Web of Science, hand searching of relevant journals, web searching, searching of University repositories and searching organisations websites: The National Childbirth Trust, BLISS, The Royal College of Midwives, Neonatal Nurses Association, the National Association of Neonatal Nurses, Bonnie Babies, Tommy’s, the Royal College of Obstetricians and Gynaecologists, the Royal College of Paediatrics and Child Health, the British Association of Perinatal Medicine, the Neonatal Nurses Association and the Neonatal Society. We will also search for reports from the voluntary sector, and local initiatives from primary care trusts (PCTs) that offer P2P support, and university repositories that specialise in Masters of Nursing courses, and will consult stakeholders for suggestions of additional websites.
Search strategy
The main search will include population terms: (neonatal OR premature OR preterm) and (baby OR babies OR infant) and intervention terms: peer support OR support network OR (mother OR father OR parent OR maternal OR paternal) AND (support OR network). Database syntax (e.g. adjacency searching) and additional synonyms will be used where relevant [an example search strategy is shown in Additional file
4].
These terms will be discussed with the project team and PAG before confirming the full search strategy to ensure all relevant terms are included to the best sensitivity and specificity of the search to ensure appropriate use of research resources. This main strategy will then be adapted for other databases.
We will use iterative searches to explore where necessary. For instance if named interventions are identified as being particularly relevant, further citation chasing, web-based or database searches may be carried out and the audit trail will reflect this. Search results will be interrogated to ensure that key papers are identified.
Search restrictions
No restrictions will be used in the search for literature.
Data collection and analysis
Inclusion decisions
The abstracts and titles of references retrieved by the electronic searches will be screened independently by two reviewers (out of HH, RA, KB and AB) using the pre-specified inclusion/exclusion criteria. The full text of potentially relevant studies will be obtained. Using the same methods, full text articles will be assessed for inclusion by two reviewers (HH, RA, KB and AB). Any quantitative papers in non-English languages will be translated.
Discrepancies will be resolved by discussion, with a third reviewer (JTC) if necessary. We will exclude duplicate papers. A PRISMA-style flowchart will be produced detailing the study selection process with the reason for exclusion of each full-text paper reported.
Data collection process
Quantitative data will be extracted by one reviewer (HH) into a piloted, standardised data extraction form. This will be checked by another reviewer (RA). Discrepancies will be resolved by discussion with the involvement of a third reviewer (JTC) if necessary. We will use the template for intervention description and replication (TIDieR) checklist [
19] to inform the systematic collection of this information.
For qualitative studies, we will extract details of the study aim, the sample, and the type and nature of the intervention/programme. We will also collect data on the theoretical approach, the methods used to collect the data, the analytic processes and the quotes, themes and concepts pertinent to our research questions. This process will be conducted by two reviewers independently (HH plus RA). Any discrepancies will be resolved through discussion. A structured summary will be produced for each paper and the extracted data will be tabulated allowing comparison between studies.
Requests for data
Where papers provide insufficient details of the intervention, e.g. what is delivered and by whom, we will contact authors to obtain unpublished details.
Data items
The following data will be tabulated: details of the intervention type and content; study date and publication date; the setting and the provider; sample characteristics of the included population; and the type of outcomes measured.
Missing data
Missing data will be identified and recorded within the review. Where means and standard deviations are not available, we will collect effect estimates, confidence intervals, test statistics, P values and individual participant data where reported. Authors will also be contacted for clarification of any missing data.
We will use appropriate quality assessment tools depending on the design of the included studies using the guidance produced by the NHS Centre for Reviews and Dissemination [
33] and the Cochrane Collaboration.
Risk of bias assessment process
The risk of bias of individual studies will be assessed using a suitable tool by one reviewer (HH), checked by a second reviewer (RA), and any disagreement will be resolved by discussion involving a third reviewer (JTC) if necessary. We will include all reported effect measures used to describe effect sizes in included studies and meta-analyses (e.g. risk ratio or odds ratio, mean difference or standardised mean difference). We will make a judgement on the most appropriate measures for synthesis depending on availability and appropriateness of available data, and clearly document our decision-making processes.
Quantitative synthesis
We will use methods of systematic review and evidence synthesis as outlined by the Cochrane Handbook [
34]. Quantitative data will be tabulated and discussed narratively in the first instance. If sufficient quantitative data are located, appropriate meta-analysis will be used to synthesise the effectiveness of similar interventions using a random effects model. It is not expected that enough RCTs will be found to conduct specific analyses of heterogeneity or other sub-group analyses (but potential sub-group analyses could involve, e.g. parent-to-parent peer support versus other forms of peer support).
Unit of analysis issues
As we will not restrict by study design, we will take a pragmatic approach to addressing clustering, matching, double counting or other design features of the included studies using the guidance produced by the NHS Centre for Reviews and Dissemination [
33] and the Cochrane Collaboration.
Studies with more than two groups
It is considered unlikely that studies will be identified with multiple arms. If studies with more than one study arm are identified and included within our review, we will analyse multiple intervention groups in an appropriate way that avoids arbitrary omission of relevant groups and double-counting of participants using the guidance produced by the NHS Centre for Reviews and Dissemination [
33] and the Cochrane Collaboration.
Qualitative synthesis
We will extract details of the methods of each study along with quotes, themes and concepts pertinent to our research questions. Structured summaries will be produced for each paper. This is produced so that the same information, recorded in the same order, is given for each study. To facilitate this step, a number of additional strategies may be used, such as grouping the studies according to elements such as study design, the type of intervention under study, its context, populations and types of outcome measures and so on. Tabulation can also assist synthesis and may be needed to produce the initial descriptions. The findings of each qualitative paper will be coded and grouped together into broad themes through discussion. The PAG will be invited to a qualitative synthesis reference group meeting to discuss the qualitative synthesis to establish relevance and transferability to a wider audience. This meeting may be followed up with emails and telephone conferences to explore themes and ideas further.
Narrative synthesis
We will use formal methods of narrative synthesis outlined by Rodgers and colleagues to bring together the findings of the qualitative and quantitative reviews [
35]. Narrative synthesis is a method of synthesising quantitative and qualitative findings about effectiveness
and implementation and relies on text to “tell the story” of such findings. A number of possible tools and techniques are offered as mechanisms for manipulating, analysing and comparing the results of studies included in a review. We will use the descriptive paragraphs produced during the synthesis of the qualitative information and numerical information presented in a common rubric (where possible). Various techniques can be used to explore within-study and between-study relationships. Visual tools include ideas webbing (spider diagrams) and concept mapping (using diagrams and flow charts to represent relationships between and within studies that are being explored by the reviewer), and textual case descriptions, and triangulation (to consider how methodological and theoretical approaches may have impacted on the outcomes). The exact choices of methods will be developed through the synthesis process and reasons for our choices recorded and reported.
Risk of reporting bias across studies
In line with current advice on the dangers of misinterpretation [
33] in the event that we identify statistical heterogeneity and small studies, we will not produce funnel plots to explore potential reporting bias across studies. This would be due to their inappropriateness where statistical heterogeneity is present, where studies are under-powered or there are small numbers of studies being assessed.
Addressing risk of bias
It is likely that we will encounter bias due to the nature of the studies within the inclusion criteria. We will assess the impact of risk of bias using appropriate methods according to the guidance produced by the NHS Centre for Reviews and Dissemination [
33] and the Cochrane Collaboration.
Subgroup analyses
It is unlikely that data will be available for sub group analyses, and we have not pre-planned any subgroup analyses.
Methods for qualitative research evidence
Findings from each qualitative study will be synthesised using the methods described above. We hope to develop an explanatory synthesis to explain the effectiveness of P2P support interventions for the health and well-being of parents and their premature babies, and explain what these findings mean for clinical practice and service provision.
Summary of findings table
After review by our PAG, the outcomes to be reported in the summary of findings table will be those relating to the primary outcomes of personal experiences of people offering and receiving P2P support (including, but not solely, self-confidence, self-efficacy, knowledge, moving onto or back into employment, sense of belonging/friendship/being part of something); measures of parent-child interactions; parent health, e.g. stress, depression, confidence, isolation, satisfaction, empowerment; child health and development (e.g. cognitive development, behaviour, language); family function including relationships and transitioning home; service use (e.g. readmissions, contact with other services); and resource use where reported.
The most appropriate effect measures reported for each outcome will be selected depending upon included study designs.
Acknowledgements
We would like to acknowledge members of the Project Advisory Group for their contributions to the project to date.
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (
http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (
http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.