Peer support for postpartum depression: volunteers' perceptions, recruitment strategies and training from a randomized controlled trial

Abstract

A randomized controlled trial evaluated the effect of telephone-based peer support (mother-to-mother) on preventing postpartum depression among high-risk mothers. This paper reports volunteers' perceptions, which showed that peer support is an effective preventative intervention. Two-hundred and five (205) volunteers were recruited and trained to provide peer support to 349 mothers randomized to the intervention group. Volunteers' perceptions were measured at 12 weeks using the Peer Volunteer Experience Questionnaire, completed by 69% (121) of the 175 volunteers who provided support to at least one mother. Large majorities felt that the training session had prepared them for their role (94.2%), that volunteering did not interfere with their lives (81.8%) and that providing support helped them grow as individuals (87.8%). Over 90% stated that they would become a peer volunteer again, given the opportunity. Recruitment and retention of effective volunteers is essential to the success of any peer-support intervention. Results from this study can assist clinicians and program planners to provide effective training, sufficient on-going support and evaluation and appropriate matching of volunteers to mothers who desire peer support and are at high risk of postpartum depression.

INTRODUCTION

In the context of healthcare, peer support is the provision of support from someone who has experienced the same health problem and has similar characteristics as the proposed recipient (Dennis, 2003a). Types of support include emotional support: e.g. expression of caring, encouragement, attentive listening; informational support: e.g. advice, suggestions, factual input; and appraisal support: e.g. communication of optimism, assistance to endure frustration and encouragement to persist in problem resolution (Borkman, 1976). Because individuals assimilate new knowledge and appraisals more effectively when presented by peers, it makes sense to incorporate peer volunteers in the management of many disorders (Hildingh et al., 2003; Macvean et al., 2008; Thomas et al., 2008; Verhaeghe et al., 2008; Hughes et al., 2009; Brownson and Heisler, 2009). Indeed, there is evidence that peer support can enhance positive health outcomes and reduce morbidity in several ways: facilitating psychological adjustment, aiding recovery from traumatic experiences and even extending life for individuals with serious chronic disease (Berkman, 1985; Pearlin, 1985; Thoits, 1986; Cohen et al., 2000).

Of course, the success of peer-support interventions depends upon the participation and commitment of competent volunteers. Recent national studies have found that 29% of Americans (Bureau of Labor Statistics, 2005), 37% of Canadians (Statistics Canada, 2004) and 31% of Australians (Australian Government DoFaCS, 2005) volunteered their time to charities and voluntary organizations. In Canada, the top three reasons for volunteering were: to make a contribution to the community (92%), to use one's skills and experiences (77%) and being affected by the cause supported by the organization (60%). While only a small portion of this formal volunteer activity in Canada is currently directed toward hospitals (2%) and other healthcare organizations (6%), fully 83% of Canadians aged 15 or older reported providing one-to-one, unpaid help directly to others not in their household at least once during the previous year—without involving any formal organization or group. Fifty percent (50%) said this was health-related or personal care ‘such as emotional support, counselling, providing advice, visiting the elderly, or unpaid babysitting’. More than 1 in 10 (12%) reported that they provide such care to others outside their household on a daily or almost daily basis (Hall et al., 2006). Despite the availability of this considerable human resource, the experiences of volunteers are rarely subject to rigorous evaluation or analysis (Macvean et al., 2008; Ley et al., 2010; Stant et al., 2011). The purpose of this paper is to report volunteers' perceptions of their peer support experience in a program aimed to prevent postpartum depression in high-risk mothers. Understanding volunteers' needs and preferences is an important step towards harnessing this key resource in the management of postpartum depression.

Postpartum mood disorders represent the most frequent cause of maternal morbidity after childbirth. These disorders range in severity from the early transient ‘baby blues’ to postpartum psychosis and include postpartum depression, a condition comprised of the often disabling symptoms of dysphoria, emotional lability, insomnia, confusion, anxiety, guilt and suicidal ideation (Beck, 1992; Evins and Theofrastous, 1997; Altshuler et al., 2001). A recent meta-analysis of 30 studies reported prevalence rates for depression ranging between 6.5 and 12.0% in the first year postpartum (Gaynes et al., 2005). The incidence of postpartum depression has been found to be greatest in the first 12 weeks, while duration is often dependent on symptoms severity. However, residual depressive symptoms are common and about 50% of mothers with early-onset postpartum depression will remain clinically depressed at 6 months postpartum (Kumar and Robson, 1984; Whiffen and Gotlib, 1993; Cooper and Murray, 1997). Approximately 25% of untreated postpartum depression continues past the first year.

Postpartum depression has negative health consequences for the mother, child and family. Women who have suffered from postpartum depression are twice as likely to experience future episodes of depression over a 5-year period (Cooper and Murray, 1995). Postpartum depression can also impair maternal–infant interactions (Murray et al., 1996) leading to attachment insecurity (Murray, 1992; Hipwell et al., 2000), emotional developmental delay (Cogill et al., 1986; Hipwell et al., 2000) and social interaction difficulties (Cummings and Davies, 1994; Murray et al., 1999). Yet, postpartum depression often remains undetected, as many mothers are reluctant to disclose symptoms of depression and seek treatment—even when they are in frequent contact with health professionals (Robertson et al., 2004; Dennis and Chung-Lee, 2006).

Detailed analyses of social support variables in predictive studies clearly suggest that the following social deficiencies significantly increase the risk of postpartum depression:

1. not having someone to talk openly with who has shared a similar problem (Brugha et al., 1998),

2. lacking an intimate confidant or friend to converse with (Paykel et al., 1980; O'Hara et al., 1983; Brugha et al., 1998; Romito et al., 1999),

3. not receiving support without having to ask for it (Brugha et al., 1998), and

4. feeling socially isolated (Mills et al., 1995).

Conversely, companionship and belonging to a group of similar others has a protective effect (Cutrona, 1989). In interviews with depressed mothers (n= 60) participating in a population-based study, women were asked for their own explanations as to why they experienced postpartum depression; a ‘lack of support’ and ‘feeling isolated’ were the most common responses. When asked what advice they would give to new mothers currently suffering from postpartum depression, the most common suggestion was ‘find someone to talk to’ (Small et al., 1997). In a previous longitudinal study, scores on the Edinburgh Postnatal Depression Scale (EPDS) were significantly correlated to maternal perceptions of support from other women with children (Dennis and Letourneau, 2007). This prior research suggests that the provision of support from an experienced mother may be a simple intervention with the potential to prevent postpartum depression. To address this issue, a multi-site randomized controlled trial was conducted to evaluate the effect of peer support on the prevention of postpartum depression (Dennis et al., 2009).

METHODS

Trial overview and results

Following ethical approval from the University of Toronto, between November 2004 and September 2006, participants were recruited from seven public health regions across Ontario, Canada where authorities agreed to host the trial. These areas encompass a range of dense and sparse urban neighbourhoods, both old and new suburban housing developments, and rural residences or small towns. They include nearly half the population of the province and include between 35 000 and 45 000 births per year (Statistics Canada, 2004). As part of standard postpartum care in Ontario, each mother receives a telephone call from a public health nurse 24–48 h after hospital discharge. During this call, potential participants were briefly introduced to the study and screened for postpartum depression. New mothers who scored >9 on the EPDS and provided verbal consent were contacted by the trial co-ordinator to explain the study in detail and further assess eligibility and interest. If they agreed to participate, the trial coordinator obtained baseline data and enrolled them in the trial. Eligible participants were all new mothers within 2 weeks postpartum who were at least 18 years of age, were able to speak English, had a live birth and were discharged home from the hospital with their newborn. Exclusion criteria included an infant not discharged home with the mother and current use of antidepressant or antipsychotic medication.

Seven hundred one new mothers identified as being at high risk for postpartum depression using the EPDS were recruited and randomized to either a control group (standard postpartum care, n= 352) or intervention group (standard care plus peer support, n= 349). In this trial, peer support consisted of individualized, telephone-based, mother-to-mother support initiated within 48–72 h of randomization and provided by a volunteer recruited from the community. Providing such support by telephone has been advocated for a number of health concerns, but this trial is among the first efforts to determine its effectiveness and risks (Dale et al., 2008). Participant follow-up rates were greater than 85% for both intervention and control groups. At 12 weeks postpartum, significantly fewer women randomized to the intervention group (13.5%) had an EPDS score >12 compared with those in the control group (24.8%; χ²= 12.2, p= 0.0005, OR = 2.1, 95% CI = 1.38–3.20). These results indicate that support provided by peer volunteers can be an effective strategy for preventing postpartum depression and its impact on mothers and their infants (Dennis et al., 2009).

Recruiting and training of peer volunteers

A paid volunteer coordinator was hired to organize the recruitment of peer volunteers, obtain informed consent, conduct peer volunteer training sessions, match postpartum women with an appropriate peer volunteer, assist in the organization of peer volunteer meetings and discussions on an electronic bulletin board, monitor the intervention implementation and provide support to peer volunteers as required. Between summer 2004 and spring 2006, flyers were distributed and newspaper ads were placed in the participating regions to enlist peer volunteers. Over 300 women from the community volunteered were able to speak and understand English, and had a self-reported history of and recovery from postpartum depression. Two hundred five of them attended the mandatory 4-h training session, which focused on developing the skills required to provide effective telephone-based peer support and making referrals to health professionals, if necessary.

A ‘Mothers Helping Mothers with Postpartum Depression’ manual was developed, pilot-tested and distributed to trainees (Dennis et al., 2009). This 121-page training manual outlined the professional services available for referral and covered relevant topics including: (i) introduction to peer support; (ii) potential benefits of peer support; (iii) relationship development; (iv) techniques for effective telephone support; (v) general postpartum depression information; and (vi) the helping process. Role-playing and strategizing were important components of the training session. All peer volunteers were provided clear guidelines regarding thoughts of self-harm and indicating when to refer mothers to professional health services or crisis care. The training workshop also provided an important opportunity to select applicants who were best suited for peer support with postpartum mothers. Those whose communication skills were deemed inadequate, who demonstrated difficulties participating in discussions about postpartum depression or who showed evidence of unresolved depression were excluded from the peer support program. Of those who attended the training, 175 (86.8%) were accepted as peer volunteers and were matched with at least one new mother in the trial.

At the end of the training session, peer volunteers received activity logs to document all support they provided up to 12 weeks postpartum and postage-paid, addressed envelopes for returning them. Among women allocated to the intervention group, the volunteer coordinator matched participants and peer volunteers based on residency and ethnicity, if the mother desired. Peer volunteers initiated telephone contact within 2–3 days of their enrolment and were instructed to make a minimum of four contacts, after which they should interact as deemed necessary (Dennis et al., 2009). Peer volunteers were required to maintain strict confidentiality regarding counselled mothers. Over the course of the trial period, peer volunteers were also encouraged to incorporate regular support from each other. To facilitate this, they were given the telephone numbers of other peer volunteers in their training session and provided with an electronic listserv. A newsletter was also distributed regularly to give an update on trial progress and to offer encouragement and support. The volunteer coordinator further assisted the peer volunteers via telephone in addressing participant difficulties and developing problem-solving strategies on an as-needed basis.

Volunteer roles and peer support interaction

On average, peer volunteers supported two new mothers during the trial (mean = 2.3, SD = 1.5). The shortest length of time supporting a mother was approximately 5 weeks (mean = 5.5, SD = 4.6), while the longest was 11 weeks (mean = 10.9, SD = 6.5). When the length of support time was short, peer volunteers usually attributed its discontinuation to maternal preferences. Either the mothers were doing well and decided they did not require the support, or the peer volunteers could not connect with the mothers. For example, one peer volunteer wrote, ‘Both mothers stopped taking my calls after a certain point. Both were doing very well and may not have felt they needed support.’ Approximately half (51.2%) of the peer volunteers reported having difficulty contacting a mother at some point.

A number of peer volunteers (17.4%) indicated developing a friendship with a new mother they supported and several (9.1%) met their matched mother face-to-face. Twelve peer volunteers (9.9%) interacted with a health professional on behalf of a new mother and 33 (27.3%) referred a mother to a professional health service. When explaining why they had referred mothers to a health professional, 12.4% reported seeing increasing symptoms of depression or difficulty coping. In a minority of cases, the referral was for breastfeeding assistance or help with other health issues.

Evaluating the peer support program by peer volunteers

At the completion of the trial, or when a peer volunteer discontinued providing support, a Peer Volunteer Experience Questionnaire was mailed to them with a postage-paid return envelope (Dennis et al., 2009). This questionnaire had been used successfully in previous peer-support projects (Dennis et al., 2002; Dennis, 2003b) and includes closed- and open-ended questions related to program training and expectations, interpersonal effect, recruitment and retention, interactional characteristics and volunteer roles.

Data analysis

The data from the closed-ended questionnaire items are presented using descriptive statistics. For the open-ended items, content analysis was used to develop response categories (Krippendorf, 1980). All responses were coded by the researcher. The codes were then analysed, grouped thematically and frequencies were calculated.

RESULTS

Of the peer volunteers who supported at least one mother in the trial, 121 (69.1%) returned the volunteer experience questionnaire. Their socioeconomic and ethno-cultural characteristics are presented in Table 1. Most (78.6%) were between the ages of 25 and 45, with the average age being 37.6 years old. The majority (69.4%) identified themselves as White, or Caucasian Canadians, while only 11.6% reported being of other racial groups—although a large number (19%) did not identify their race. More than half (58.7%) said they are of Anglo-Canadian cultural background (British, Scottish, Irish) and only 4.1% indicated being Francophone or French Canadians, but many (18.2%) did not report their ethno-linguistic heritage and 19% reported another ethnicity. These included other European groups (such as Dutch, Finnish and Italian), Latin-American and Caribbean nationalities (Venezuelan, Jamaican), African countries (Ghana, Tanzania), Middle-Eastern (Turkish, Lebanese) or South Asian origins (Indian, Punjabi), East/Southeast Asian ethnicity (Chinese, Vietnamese) and even minority religious background (Jewish, Sikh). This strongly reflects the growing multi-ethnic immigrant mix of Ontarians.

Table 1:

Characteristics of peer volunteers (n= 121)

Age group
25–29	8 (6.6%)
30–34	38 (31.4%)
35–39	32 (26.5%)
40–44	17 (14.1%)
45–49	11 (9.1%)
50+	8 (6.6%)
Age unknown	7 (5.8%)
Racial group
White/European/Caucasian	84 (69.4%)
Black (African/Caribbean)	3 (2.5%)
Middle Eastern/South Asian	6 (5.0%)
Asian/Southeast Asian	4 (3.3%)
Multi-racial	1 (0.8%)
Unknown race	23 (19.0%)
Ethno-Linguistic group
Anglophone	71 (58.7%)
Francophone	5 (4.1%)
Other	23 (19.0%)
Unknown	22 (18.2%)
Education level
Elem./some high school	9 (7.4%)
Completed high school	37 (30.6%)
Some college/trade school	47 (38.8%)
Completed college/trade school	26 (21.5%)
Unknown	2 (1.7%)
Marital status
Married/common law	102 (84.3%)
Single (never married)	2 (1.7%)
Separated/divorced	14 (11.6%)
Unknown	3 (2.5%)
Number of children
1	44 (36.4%)
2	52 (43.0%)
3	16 (13.2%)
4–5	7 (5.8%)
Unknown	2 (1.7%)
Employment status
Self-employed	10 (8.3%)
Full-time	30 (24.8%)
Part-time/casual	23 (19.0%)
On maternity leave	9 (7.4%)
Unemployed	10 (8.3%)
Stay-at-home mom	36 (29.8%)
Unknown	3 (2.5%)
Annual household income
<CAD 20 000	5 (4.1%)
CAD 20 000–39 999	16 (13.2%)
CAD 40 000–59 999	17 (14.1%)
CAD 60 000–79 999	26 (21.5%)
CAD 80 000+	49 (40.5%)
Unknown	8 (6.6%)

The large majority of peer volunteers (84.3%) was married or in a stable, common-law relationship, while 11.6% were separated or divorced. Only a few (1.7%) were single and never married. A slight majority of peer volunteers (60.3%) had some post-secondary education and just over half (52.1%) were employed at least part-time. However, more than one-third (37.2%) were currently focused on childcare, being on maternity leave or as stay-at-home moms. Peer volunteers had two children, on average, although nearly 20% had three or more. As is common with those doing other forms of charity work, nearly two-thirds of peer volunteers were from relatively high-earning households, given that almost half had only one income earner: 62% had an annual income of more than CAD 60 000/year.

More than anything else, though, what makes these volunteers peersto the new mothers in this trial is the fact that they all had experience with postpartum depression (Table 2). Peer volunteers had experienced postpartum depression an average of 7.4 years prior to participation in the program, which had lasted more than a year (13 months) on average. Almost two-thirds (64.5%) of peer volunteers reported having needed professional assistance for their postpartum depression and about half (53.7%) took antidepressant medication to overcome it.

Table 2:

Peer volunteers' experience with postpartum depression (n= 121)

Mean age	37.6 years old (SD = 7.6 years)
Mean number of children	1.9 (SD = 0.93)
PPD experienced mean # years ago	7.4 years ago (SD = 7.8 years)
PPD started mean # weeks postpartum	3.5 weeks (SD = 4.0 weeks)
PPD lasted mean # months	13.0 months (SD = 12.6 months)
Received professional help for PPD	78 (64.5%)
Took medications for PPD	65 (53.7%)

Program training and expectations

The overwhelming majority of peer volunteers (n= 114, 94.2%) felt that the training session had prepared them for their volunteer role. Only seven felt the training was inadequate due to insufficient length or depth. However, one quarter (n= 30) of the peer volunteers either felt that they would have liked additional information in some specific area—such as more detail about factors that contribute to postpartum depression and medications used to treat it—or felt that they would have benefited from more practice, role-playing scenarios. One wrote: ‘The session was too short. I felt as though my ability to assess a new mother was limited and it was so vital.’ Two peer volunteers wanted examples of (open-ended) questions to use with new mothers to get them to ‘open-up’. Another wanted clues about ‘how to pick up on subtle hints that a new mother may be experiencing PPD’ and one wrote that she would have liked to have ‘a triage chart with phrases to listen for [that may indicate signs of PPD]’.

Asked what they liked best about the training, 19.8% of peer volunteers said they most enjoyed meeting and connecting with other women at the session who had also experienced postpartum depression, i.e. ‘being with other women who had “been there”’. The friendly and caring atmosphere (18.2%) and the skill, knowledge and enthusiasm of the trainers (14.0%) were also frequently mentioned as the best part of the session. Other favourite aspects included the informative nature of the session (13.2%), the thorough resource manual (9.1%), the role-playing and case examples (8.3%) and the small group size (5.8%). In describing what they would like done differently, 12.4% expressed a desire for more role-playing during the training session or more opportunity to discuss potential scenarios that might occur with a new mother. Many (9.9%) would have liked some follow-up sessions with the trainer. In fact, when specifically asked, the majority (80.2%) of peer volunteers thought there should be on-going educational sessions throughout the trial period to refresh their skills and provide them continuous support. The average suggestion was for two to four sessions per year.

Forty-three percent (43%) of peer volunteers responded to the open-ended question ‘What expectations should be placed on a peer volunteer in order that she properly supports a new mother?’ Most responses related to three themes:

1. that she be dedicated, making regular contact with the new mothers (e.g. ‘the commitment to see the match through to the end’);

2. that she be empathetic, caring and non-judgmental (e.g. ‘not to judge, but be there for moral support … to show there is a light at the end of the tunnel and they are not the only mother feeling this way’); and

3. that she provide appropriate referrals to health professionals, if required.

Most peer volunteers (67.7%) felt that they should have some supervision, although when asked what type of oversight would be helpful, their answers revealed a desire for trainers or the volunteer coordinator to act as a resource, rather than actually monitoring their peer-support activities. ‘How would you supervise phone conversations?’ one wondered incredulously. Most suggested a weekly or monthly ‘check-in’ from the coordinator to help monitor the progress of new mothers. Others indicated that help should be available as needed when the peer volunteers confront a difficult problem. Most of the peer volunteers (73.6%) also felt that there should be an evaluation process to ensure they are providing high-quality support to new mothers. Many (27.3%) specifically stated that the mothers they were supporting should be asked directly, but were unsure whether this should be done formally, via questionnaire, or informally, by a simple phone call from the volunteer coordinator. Only three (2.5%) suggested ‘test calls’ (e.g. ‘Have investigator or “fake” caller do quality control’) or surveillance (e.g. ‘Listen-in on calls periodically’). A majority of peer volunteers (76%) felt they themselves had been adequately supported throughout the program.

Intrapersonal effect

While the majority of peer volunteers (81.8%) said volunteering did not interfere with their lives, eight (6.6%) did report time-commitment issues. Only one peer volunteer revealed that ‘Past anxieties and sadness reoccurred after contact with a new mom.’ A large majority (83.5%) felt that participation in the peer support program helped them grow as individuals. Ways how can be divided equally into four categories. Participation helped them:

1. resolve their own experiences with postpartum depression (e.g. ‘Volunteering helped in my healing process from having postpartum depression’);

2. broaden their knowledge about community resources and the treatment of postpartum depression;

3. realize that they were not alone and improved their sense of social integration (e.g. ‘I knew that anyone could be affected—not just crazy me’); and

4. feel good about themselves (e.g. ‘My first match found our discussions to be very helpful and she told me I saved her life—that felt great’).

On the other hand, almost half (46.2%) of peer volunteers felt that they had been disappointed while providing support to a new mother. Of those that specified why, responses usually related to working with new mothers who did not need support or who did not return their calls (e.g. ‘One mother was not returning phone calls and I found it frustrating, especially after initial contact exposed her feelings of loneliness’). Over all, 39% of peer volunteers felt uncomfortable at least once while supporting a new mother. Specific concerns included being worried about interfering in the mother's life, feeling that the support they provided was not sufficient or not useful, being nervous about the first contact, or being uncomfortable discussing emotional issues. One peer volunteer wrote about her discomfort: ‘I wasn't sure at times how much to continue to nudge. I did not want to be intrusive or detached!’

Recruitment and retention

Most peer volunteers learned about the support program through ads in newspapers (42.2%), posters/flyers (31.4%) or word of mouth (14.1%). When asked why they had become a peer volunteer, about half indicated wanting to help mothers who were experiencing postpartum depression (e.g. ‘I went through postpartum depression and wanted to be there for someone else who may experienced it too’). Many believed that they would have benefited from a similar program at the time they experienced their own depression (e.g. ‘I felt alone when I went through postpartum depression and wished I had someone to talk to that understood’). They felt that sharing their experience would be useful to new mothers (e.g. ‘I felt I could give some support because I had been through postpartum depression twice before’).

Specifying what they liked most about being a peer volunteer, 26.5% of comments involved being able to help other mothers (e.g. ‘I felt thrilled to make a difference to someone in a very personal way’). They also liked the connections they formed with the new mothers and, in some cases, felt that participation in the program helped them heal from their own experience of postpartum depression. One wrote: ‘I actually felt like I was redeeming some of my own lost time due to postpartum depression by making a mark against it with other mothers.’ Asked what they disliked the most about being a peer volunteer, the most common response (10.7%) was not having enough time to devote to the task. Other areas of dislike included difficulty establishing the initial connection, not being called back by the mothers, being unsure of their own skills and helpfulness, feeling that the mothers did not need or want the support, and not being matched to enough new mothers. A large majority (89.3%) of peer volunteers reported that they were satisfied with the program overall and 90.1% stated that they would become a peer volunteer again.

CONCLUSION

Many people appreciate the considerable value of peer support and policymakers, administrators and health professionals increasingly look to peer interventions to mobilize informal support in order to provide tangible assistance, improve skills or health outcomes, offer emotional support and enjoyable social relationships and monitor and promote safety. Peers can provide individualized assistance that matches the strength, needs and contexts of women and their families. Informal support interventions use systematic activities to improve the existing quality, level or function of an individual's personal social network or to create new relationships and networks for them with volunteers. Although there are many different types of volunteer programs, the unifying characteristic of volunteer roles is that helping occurs with no expectation from recipients in return for that help.

Overall, the peer volunteers viewed their experience positively. They reported that their training session provided them with the necessary information and anticipatory guidance. This finding is consistent with another peer support program for breastfeeding mothers (Dennis et al., 2002). Furthermore, this trial's limited training maintains the status of ‘peer’ in line with a concept analysis that warned against over-training peer volunteers and turning them into paraprofessionals.

Almost all the peer volunteers indicated they grew as individuals through participation in this trial. The rewards of volunteering evident in this study are consistent with the findings of the Canada Survey of Giving, Volunteering and Participating (Statistics Canada, 2004; Hall et al., 2006). Volunteers reported that their experience provided them with interpersonal skills, communication skills, increased knowledge and organizational and managerial skills. Improved job opportunities can also be a benefit of volunteer experience (Hall et al., 2001). Peer volunteers in this trial also benefited from enriched interpersonal relations.

The recruitment and retention of effective volunteers is essential to the success of any planned peer support intervention (Dennis, 2003a; Macvean et al., 2008). In the ‘Mothers Helping Mothers’ program, we recruited and retained 175 volunteers for the course of a 12-week intervention. Satisfaction with the program was very high. Most peer volunteers derived satisfaction from helping other mothers. We identified several areas where peer volunteers preferred changes to the programs. These included better screening of mothers to ensure that they wanted peer support, on-going training sessions for peer volunteers to discuss difficult cases and provide reasonable monitoring of progress, and evaluation of their efforts by the new mothers to ensure accountability.

Major strengths of this study include the large number of peer volunteers and the detailed evaluation questionnaire. Social support experts have recommended comprehensive analysis of peer support interactions to promote the development of more effective interventions (Cohen et al., 2002; Gotay et al., 2007). In-depth analysis such as this provides insight into the peer volunteer's role and can help delineate which components of the peer-mother relationship are most important for retaining volunteers and improving health outcomes. However, there are some limitations to the generalizability of our results. This evaluation was undertaken in the context of a large randomized controlled trial, which had funding to hire a paid volunteer co-ordinator. It is uncertain whether similarly high levels of satisfaction and retention would be obtained in the context of a community program if paid staff were not available. On the other hand, a community peer-volunteer program that did not require strict eligibility criteria and randomization would have more flexibility to match volunteers with mothers who truly desired peer support. This might alleviate some of the concerns mentioned by volunteers.

While peer volunteers felt that they benefited from the experience, a significant proportion also felt negative feelings such as discomfort and disappointment through the course of their participation. The training session could be used as an opportunity to prepare volunteers for this and to help them cope with these very natural feelings without discouragement. Interestingly, many peer volunteers also reported that participation in the program helped them work through some unresolved feelings and heal from their own experiences of postpartum depression. This process too could be discussed in the training session and peer volunteers could be better equipped to cope with unresolved issues should they arise.

This paper presents volunteers' perspectives of a peer support program from those who had contact with at least one new mother. Future research with volunteers who did not establish supportive relationships could give us insight into why peer relationships sometimes fail to develop. It is promising that most volunteers were driven to participate for the purpose of helping others. Program planners can harness this altruism most effectively by providing sufficient training, on-going support and appropriate matching of volunteers to new mothers.

FUNDING

This work was supported by the Canadian Institutes of Health Research [grant number MCT66874].