Review question/objective

Peer-based interventions have the potential to enhance quality of life and functioning; however their role specifically within the older population has not been fully investigated. The objective of this review therefore is to locate, appraise and synthesize evidence on the effectiveness of peer-based interventions on changes in health behaviors, specifically for the older population.

The specific question to be answered is: "what is the effectiveness of peer-based interventions on health promoting behaviors in older adults, when compared to interventions that are not peer-based?"

Background

The proportion of population over 65 years in Australia is predicted to rise from 14% in 2012 to between 18.3% and 19.4% in 2031,1 while globally the prediction is that more than two billion people will be over the age of 60 by 2050.1 Due to declining birth rates, the United Nations reports that China will have the most rapidly aging population.1 With an aging population comes the need for innovative, cost-effective and clinically effective interventions that will promote physical, social and psychological functioning and independence in later years. Peer support and peer mentor programs have been suggested by the World Health Organization (WHO) as strategies that may assist older people to age in a positive and active manner.2

Generally the term "peer" is used to refer to someone who has been trained to deliver an intervention or to support a person of a similar background or who have experienced similar events. However, programs incorporating peer support are often ill defined and complexities within the term "peer" can be misunderstood.3,4 According to Dennis,4 diversity exists within peer roles pertaining to health promotion activities and although the primary focus may be to provide encouragement and empathic understanding, other peer roles may include providing education and/or emotional support. Therefore, the term "peer-based" is used within this review to include programs defined as peer-led as well as those including peer mentors or peer educators. These terms are often used interchangeably but there may be some differences between how the peer is incorporated into the intervention and the actual role of the peer. Definitions and differences between the terms are outlined below.

The term "lay-led" is also sometimes used interchangeably with peer-led.5 A "lay" leader/support person may differ from a peer support person in that the lay person may not have had the same experiences as the people they are assisting. A key concept in peer interventions is that the peer is considered an equal among their group according to the characteristic being targeted by the intervention4 For example; a peer supporter for an alcohol rehabilitation program could be someone who has successfully managed their own alcohol dependency. Due to the ambiguities with definition of terms, "lay" supports will also be considered for this review but will only be included if the lay support person is defined in the research as someone who has similar characteristics to a peer supporter, more specifically, that he/she shares the same characteristics as the participants in the group being addressed. Close analysis of studies that refer to lay-led programs will be required to identify if they fall under the definition of peer. For example, a study by Coull et al.6 reported an intervention utilizing "lay health mentors". Further explanation of the characteristics and training of the mentors identified that they were volunteers with a similar health condition to the people they were helping in the intervention; this description would therefore also fit the definition of a "peer mentor".

Peer-based interventions have been shown to be useful to effect change in health behaviors7 and interventions including a peer component have demonstrated some short-term benefits in self-management programs for patients with chronic disease.5 A Cochrane systematic review on telephone peer support was undertaken in 2008 and identified some evidence of effectiveness for improving rates of mammography screening, depressive symptoms and some self-management aspects of heart disease; however results overall were ambivalent and more well designed studies are needed.8 A theoretical understanding of the process of change expected within the context being examined is a major step in designing effective healthcare interventions and explaining the change in behaviors or the causal chain between components of interventions and outcomes.9 Peer-based self-management interventions based on Bandura's self-efficacy theory have been found to promote psychosocial wellbeing and physical functioning among diabetes patients9 and stroke survivors.11

Since 2002, as part of health policy reform, the United Kingdom invested considerable time and money into an Expert Patient Program, which is based on a peer intervention. The program consists of a six-week training course for people suffering from a chronic health condition. Facilitators of the course include people with a similar chronic condition as well as some health professional trainers. Initial evaluation of the program during its pilot phase indicated that the leaders of the program had the most influence over a program's success; however, recruitment issues were the biggest barrier to the program.12 Further critique of the program has provided mixed opinions of its effectiveness.13

A systematic review by Webel et al.7 in 2010 appraised peer-based health behavior interventions for adults and identified three models for peer interventions, namely group-based education whereby the peer was the facilitator (peer educator) for the group, individual peer support or a combination of the two models. Overall findings from the review were equivocal due to heterogeneity of outcomes resulting from low study quality. The authors also highlighted a lack of standardized time units to measure delivery of interventions and suggested further research using rigorous methods. As the review included studies on participants over 18 years of age, results cannot be extrapolated to populations over 60. While results from individual studies may suggest potential benefits, a systematic review specifically for older adults will identify if peer interventions are actually effective for improving health behaviors in this population. Single studies which incorporate a peer component to a health promotion intervention have been undertaken for this population in relation to improving physical function,14 social participation15 as well as chronic disease management.16,17

In their policy framework on active ageing, the WHO calls for an international collaborative approach to provide positive interventions that provide opportunities for older adults to participate. Being able to function physically, socially and psychologically in the older years underpins the concept of active aging, the goal of which, is to change provision of care from responding solely to a person's needs, to also considering their rights.2 In 2012, the Victorian Government Health Department compiled a literature review to assess existing evidence on national and international healthy aging strategies.18 The literature review was based on the definitions used by the WHO Active Ageing Policy Framework2 and summarized the evidence found, into the following areas specific to healthy aging: physical activity, falls prevention, healthy eating, management of health, social participation, intergenerational programs, ageism, and environments that can improve health. Peer based interventions were found within some of these categories but with varying levels of evidence.

This review will synthesize the best available research evidence on peer-based interventions in older people. Due to increasing health care demands, the complexity of care needs and limited healthcare resources, a strong evidence base is needed to support implementation of effective and innovative health programs for promoting well-being in older populations.

Definitions

Peer: A person of equal standing.4 For health interventions a peer is someone who has experienced or is experiencing the same specific condition or behavior as others within the group.

Peer support: In its simplest explanation, peer support is defined as the process of giving and receiving help or encouragement by someone of similar or equal character4 or defining characteristics.

Lay personnel: Someone who has received informal training in order to deliver an intervention but is not a formally trained health professional.19

Peer mentor: Sometimes referred to as peer coaches, a peer mentor is someone who has successfully managed the behavior or condition as the person they are mentoring. They may have received some training but voluntarily provide personal help, empathy and skills to assist their mentee to achieve set goals.20

Inclusion criteria

Types of participants

The World Health Organization suggests that while a generally accepted definition of "older" may refers to age 60 and above, using a specific chronological age to define older adults may be inappropriate as differing contexts will have different determinants of aging.2 Therefore this review will consider studies that include persons over age 60 as well as any participant within a population group defined in the study as "older adult". The setting will be community based interventions, primary care or residential aged care facilities. Interventions undertaken in acute care settings will be included if the population of the intervention is targeted at older persons.

Types of intervention(s)/phenomena of interest

This review will consider studies that evaluate interventions comprising any peer-based strategy that is delivered to effect a change in health promoting behaviors. The intervention may include peer mentors, peer educators or peer support. Peer strategies as part of chronic disease management are not the primary focus of this review. The intervention may relate to promoting management of a specified health behavior, e.g. smoking, or general health promotion activity, e.g. physical activity. "Lay-led" interventions will be included if the layperson is equivalent to a peer, i.e. that they have or have had the same experience or condition as the people included in the intervention group. If there is ambiguity in the terminology used, primary authors will be contacted for clarification.

The comparator will be interventions that are not peer-based.

Types of outcomes

Studies that measure outcomes of any behavioral change related to promoting health will be included. Outcomes may pertain to a specific disease state or general change in health.

The outcome measures include:

Quality of life as assessed by general or disease-specific quality of life measures (e.g. SF-36).

Self-efficacy as measured by general or disease-specific self-efficacy measures (e.g. General Self-Efficacy Questionnaire)

Symptom management as assessed using self-reported instruments (e.g. Chronic Respiratory Disease Questionnaire)

Self-care as measured using self-care management scales (e.g. Hypertension Self-care Profile)

Social participation as assessed by questionnaires of community reintegration or participation (e.g. London Handicap Scale).

Physical function as measured by standardized functional outcome measures (e.g. Barthel Index).

Health Behavior as measured by standardized or disease specific outcome measures (e.g. Miller's Health behavior Scale).

Types of studies

This review will consider for inclusion any experimental study design including randomized controlled trials, non-randomized controlled trials, quasi-experimental or before and after studies.

Search strategy

The search strategy aims to find both published and unpublished studies. A three-step search strategy will be utilized in this review. An initial limited search of MEDLINE and CINAHL PLUS will be undertaken followed by an analysis of the text words contained in the title and abstract, and of the index terms used to describe the article. A second search using all identified keywords and index terms will then be undertaken across all included databases. Thirdly, the reference list of all identified reports and articles will be searched for additional studies. Studies published in English and Chinese language will be considered for inclusion in this review. No date restriction will be applied to this review.

The databases to be searched include:

Articles published in English

The electronic bibliographic databases to be searched will include: MEDLINE, CINAHL PLUS, EMBASE, Cochrane Central Register of Controlled Trials, All EBM Reviews, PsycINFO, EBSCOhost, Health and Medical Complete, Health Sciences, ProQuest Dissertations & Theses, ISI Web of Science, Academic OneFile, Bandolier - Evidence Based Health Care, SCOPUS, and Scirus.com.

Articles published in Chinese

Electronic databases to be searched for primary publications written in Chinese will include: WanFang Data, China Journal Net, Chinese Biomedical Literature Database, Chinese Medical Current Contents, Hong Kong Index to Chinese Periodical Literature, Chinese Electronic Periodical Services, Digital Dissertation Consortium, HyRead, and Taiwan Electronic Periodical Services. The Chinese search terms will be based on the terminologies used in Taiwan and China.

Grey literature

The databases to be searched for grey literature or unpublished studies will include: Dissertation Abstracts International, Digital Dissertations, Index to Theses, MEDNAR, Althealth Watch, Netting the Evidence, Lancashire Care Library and Information Service, Grey Literature Report (via The New York Academy of Medicine), National Library of Medicine Gateway, The Networked Digital Library of Theses and Dissertations, Academic Archive On-line, and Agency for Healthcare Research and Quality.

Initial keywords to be used will be: peer base*, peer support, peer mentor*, peer intervention*, peer*, lay person*, lay led, health promotion, health behavio?r or behavio?r, older adult* or elder*.

Assessment of methodological quality

Papers selected for retrieval will be assessed by two independent reviewers for methodological validity prior to inclusion in the review using standardized critical appraisal instruments from the Joanna Briggs Institute Meta-Analysis of Statistics Assessment and Review Instrument (JBI-MAStARI) (Appendix I). Kappa statistics tests will be performed to assess agreement between reviewers. Any disagreements that arise between the reviewers will be resolved by discussion. If resolution is not obtained, a third reviewer will be asked to assess the papers.

Data extraction

Data will be extracted from papers included in the review using the standardized data extraction tool from JBI-MAStARI (Appendix II). The data extracted will include specific details about the interventions, populations, study methods and outcomes of significance to the review question and specific objectives. A second reviewer will independently check for its accuracy. Discrepancies between the reviewers will be resolved by discussion and if resolution is not obtained, a third reviewer will be consulted. Authors of primary studies will be contacted if there are any clarifications required in regard to missing or unclear data.

Data synthesis

Quantitative data will, where possible, be pooled in statistical meta-analysis using JBI-MAStARI. All results will be subject to double data entry. Analysis of outcomes will be made using intention-to-treat results where possible. For continuous data that are collected using the same scale, the mean difference and 95% confidence interval will be calculated for each included study and used as the summary measure of effect; while for continuous data collected using different scales, the standardized mean differences and their 95% confidence interval will be calculated. For dichotomous data, relative risks, odd ratios and their 95% confidence interval will be calculated and used as a summary measure of effect. The studies will be assessed for clinical heterogeneity by considering the settings, populations, interventions and outcomes. Heterogeneity will be assessed statistically using the standard Chi-square. Where indicated, further assessment using I2 calculations will be conducted. A fixed effects model will be applied for pooling if there is no clinical or statistical heterogeneity, while a random effects model will be used in the absence of clinical heterogeneity but with the presence of statistical heterogeneity. Where statistical pooling is not possible the findings will be presented in narrative form including summary tables and figures to aid in data presentation where appropriate.

Conflicts of interest

There are no known conflicts of interest to disclose.

References