Table 5 Summaries of reviews of peer support interventions

Effect of antenatal peer support on breastfeeding initiation: a systematic review (Ingram et al., 2010) [51]

11 studies

Breast Feeding

Antenatal PS to promote initiating breastfeeding. PS “offered by women who had themselves breastfed, who were usually from the same socioeconomic background and locality as the women they were supporting and who had received appropriate training.” (p. 1740)

For all women regardless of interest in breastfeeding: 4 RCTs – no significant pooled effect; 1/4 showed significance (25%). 3/3 nonRCTs (100%) showed effect. For women considering breastfeeding: 3 RCTs – pooled effect significant (p = 0.04); 2/3 showed significance (67%). 1 nonRCT reported significant effect.

Outcome effectiveness of the lay health advisor model among Latinos in the United States: an examination by role. (Ayala et al., 2010) [108]

17 studies

Varied

Evaluated two roles: Educator Only – “usually involved several home visits and/or group classes” and Educator plus Bridge to other services – “generally consisted of one or two individual contacts in a participant’s home or at the clinic” (p. 827).

Among Educator Only, 5 of 6 (83%) showed effect for health behaviors and 3 of 6 (50%) showed effects in health status

Among Educator plus Bridge, 10 of 11 (91%) reported effects in health behaviors, utilization, and/or clinical status.

A review of the literature on peer support in mental health services. (Repper & Carter, 2011) [109]

40 studies

Mental Health

Varied roles in mutual support, consumer-run services, and providing support as part of broader services, varying from a “reciprocal relationship to a less symmetrical relationship of ‘giver’ and ‘receiver’ of care” (395).

5 RCTs examining effects on Sx, utilization, social functioning, etc.: 2 showed effects of PS (40%)

9 other quantitative studies of effects on utilization, social functioning, etc.: 8 (89%) showed effect of PS

40%–89% (64.5%)

Can community health workers improve adherence to highly active antiretroviral therapy in the USA? A review of the literature. (Kenya et al. 2011) [110]

16 studies

HIV/AIDS

Delivery of culturally appropriate health education, assistance with accessing services, provision of direct services (e.g., medication administration), medication reminders, accompaniment to apptmts. PS personalized based on individual needs and socio-environmental determinants (p. 526)

Many employed Directly Observed Therapy (DOT)

In 13/16 (81%) studies, “CHW model contributed to measurable HIV viral load suppression and/or improved CD4 cell count.” (p. 527).

7/16 (44%) reported significant findings.

12/13 successful interventions used DOT.

2/3 studies that did not find evidence for PS compared PS to “alternative HAART adherence interventions”

“Interventions that lasted at least 24 weeks, provided frequent contact …, and focused on medication management were associated with improved” adherence (Abstract).

44%–81% (62.5%)

An integrative review of community health workers in type 2 diabetes. (Hunt et al., 2011) [97]

16 studies

Diabetes

Support, counseling, education, case management, advocacy, program facilitation, coordinating and conducting educational programs and courses, linking patients and professionals, leading peer support meetings

RCT or controlled designs: 4/5 (80%) reported significant between group tests of program effects

Within-Group, Pre-Post designs: 8/8 (100%) reported effects

80% - 100% (90%)

Breastfeeding peer counseling: From efficacy through scale-up (Chapman et al., 2010) [111]

26 studies

Breastfeeding

Studies classified: low-intensity – only prenatal education, or if postpartum contact primarily by telephone: high-intensity – ≥3 contacts, both prenatal and postpartum support, most contacts in person.

Initiation of Breast Feeding: 3/4 high intensity, 0/3 low intensity

Duration of Breast Feeding: 5/9 high intensity, 1/5 low intensity

Exclusivity of Breast Feeding: 2/5 studies designed to promote breast feeding but not necessarily exclusive breast feeding

7/7 designed to promote exclusivity

Significant reductions in diarrhea in 4 of 5 studies.

0% - 75% (37.5%)

20% - 56% (38%)

40%–100% (70%)

80%

Effectiveness of community health workers in Brazil: A systematic review. (Giugliani et al. 2011) [112]

23 studies

Maternal and Child Health

In Brazil, 240,000 Community Health Agents (staff as part of health system’s primary care teams) serve 118 million citizens. Additionally, Community Health Workers work as volunteers such as in church-based programs.

For categories addressed by at least 4 studies, numbers and %s of studies finding positive results: frequency of weighing children - 4/4, 100%; attend prenatal care – 4/6, 67%; immunizations – 4/5, 80%; breastfeeding – 4/5, 80%; use of oral rehydration for diarrhea – 4/7, 57%; knowledge of oral rehydration – 4/6, 67%; stunting – 0/4, 0%.

0% - 100% (67%)

Lay health workers providing primary care for maternal and child health. (Lewin et al., 2010) [113]

82 studies

Maternal and Child Health

Cochrane Collaborative review of lay health workers, “paid or voluntary…who: performed functions related to healthcare delivery, was trained in some way in the context of the intervention, but had received no formal professional or paraprofessional certificate or tertiary education degree.” (p. 7)

Numbers of RCTs (%, RR for effect when significant) reporting significant effects: immunizations – 3/6, 50%, 1.23; mortality under 5 years – 0/3; neonatal mortality – 0/4; reported childhood illness – 0/7; care seeking – 1/3, 33%; initiated breastfeeding – 6/12, 50%, 1.36; any breastfeeding – 5/12, 41.7%, 1.24; exclusive breastfeeding – 7/10, 70%, 2.78; cure for TB – 1/4, 25%, 1.22; cure for new TB – 1/2, 50%; cure and completed trtmt for TB – 1/3, 33%; completed Isoniazid trtmt for TB prev – 0/3

0% - 70% (33%)

The effect of peer support programs on depression. (Pfeiffer et al., 2011) [114]

7 studies

Depression

Regular contact with at least one other person with depression. Groups could be professionally led, however needed to … be described as peer support (or mutual support or self-help) or to be organized so participants determined majority of topics, content of discussion. Included varied formats, e.g., group, pairs, telephone.

CBT conditions were group delivered.

Pooled standardized mean difference, PS vs UC = − 0.59, p = 0.002. 5/7 studies showed significant differences favoring PS.

Pooled standardized mean difference, PS vs group CBT = 0.10, NS. 0/8 studies showed significant differences favoring PS. NB: showed PS equal to group CBT.

0% - 71% (35.5%)

Evaluating outcomes of CHW programs. (Viswanathan et al., 2010) [2]

53 studies

Varied

Performs health-related tasks beyond peer counseling or peer support alone to create bridge between community members, especially hard-to-reach populations, and health care system. Health training associated with the intervention shorter than professional worker, not part of a tertiary education certificate. Recognized or identified as member of the community in which works (p. 793).

Outcomes in specific areas:

Knowledge – Moderate evidence in 2/3 areas (67%)

Health Behavior – Moderate evidence in 3 of 21 areas (14%)

Health Care Utilization – Moderate evidence in 4 of 12 areas (33%)

Costs & Cost-Effectiveness – Moderate evidence in 1 of 4 (25%)

14% - 67% (29%)

Peer support telephone calls for improving health. (Dale et al., 2008) [115]

7 studies

Varied

Telephone calls (of any duration) in which the peer has similar or relevant health experience (p. 4).

Increases in mammography, maintained mammography, healthy diet in post-MI patients, continuation of breastfeeding, reduced Sx of post-partum depression.Numbers of studies (%) finding significant effects:

Physical Health Outcomes: 0/3 (0%)

Psychological Health Outcomes: 2/5 (40%)

Self Efficacy: 0/2 (0%)

Mental Health: 2/2 (100%)

Quality of Life: 0/1 (0%)

Satisfaction with Intervention: 1/2 (50%)

Health Behaviors: 4/5 (80%)

0% - 100% (40%)

Systematic review of peer-support programs for people with cancer (Hoey et al., 2008) [83]

43 studies

Cancer

Peers provided support to people with cancer; peer had been diagnosed and/or treated for cancer; primary purpose of the program was to provide support to cancer patients (p. 316).

Mostly qualitative findings: “high level of satisfaction” and/or indicators of acceptance. Of 8 RCTs, 3 (37.5%) reported effects. Of 4 nonRCT studies w/ quantified findings, 2 (50%) reported effects.

38% - 50% (44%)

Effectiveness of community health workers programs for hypertension. (Brownstein et al., 2007) [116]

14 studies

Hypertension

Health education re: behavioral risks, changes in lifestyle, adherence, barrier reduction, facilitate services (e.g., insurance), instrumental support (e.g., transportation for care), measuring and monitoring blood pressure, social and emotional support, and mediation with health care and social services.

Numbers of studies (%) finding effects:

Behavioral changes: 9/10 (90%)

Adherence: 5/5 (100%)

Blood pressure: 9/10 (90%)

90%–100% (90%)

Systematic Review of U.S.-Based Randomized Controlled Trials Using Community Health Workers (Gibbons & Tyus, 2007) [3]

12 studies

Varied

“community members who serve as connectors between health care consumers and providers to promote health among [those traditionally lacking] … adequate access to care” (p. 371).

Included home visits, educational sessions, distribution of health education materials, personalized counseling.

Overall, 10/12 (83%) RCTs “demonstrated … efficacy in enhancing outcomes” (abstract).

Numbers of RCTs (%) reporting effects in categories indicated:

Mammography: 3/3 (100%)

Cervical cancer screening: 3/5 (60%)

Other areas (enrollment in research, early intervention for developmental disabilities, healthy diet, blood pressure, maternal and child health): 4/6 (67%)

60%–100% (75%)

Community health worker programs for diabetes management. (Norris et al., 2006) [117]

18 studies

Diabetes

Any healthcare worker who: (i) carried out functions related to healthcare delivery; (ii) trained in some way in the context of the interventions; (iii) no formal professional or paraprofessional training in healthcare; and (iv) had relationship with the community served” (p. 545).

Numbers of studies (%) reporting effects in categories indicated:

Knowledge of diabetes/self care: 5/7 (71%); Blood glucose (Hemoglobin A1c): 4/11 (36%); Lipids: 2/5 (40%); Blood pressure: 2/4 (50%); Emergency and/or hospital care: 3/4 (75%)

36%–75% (50%)

Social support interventions for diabetes. (van Dam et al., 2005) [118]

6 studies

Diabetes

Variety including group medical visits, peer group, peer internet, inclusion of spouse, family of friends in intervention.

Numbers of studies (%) reporting effects in specific categories:

Knowledge of diabetes: 2/3 (67%); Self management: 3/3 (100%); Psychosocial factors, QOL: 4/5 (80%); Clinical, biomedical (e.g., HbA1c): 2/5 (40%)

40%–100% (73.5)

Use of community health workers in research with ethnic minority women. (Andrews et al., 2004) [119].

24 studies

Varied

Varied roles: educator – 18 studies, outreacher – 14 studies, case manager – 4 studies, data collector, e.g., Pap tests, breast exams in remote villages – 1 study.

Qualitative, descriptive, quasi-experimental findings: effective in increasing access to health services, knowledge and behavior change among ethnic minority women (abstract).

Numbers of studies (%) reporting effects in specific categories:

Knowledge: 4/6 (67%)

Behavior: 7/9 (78%)

Access: 14/14 (100%)

67%–100% (78%)

Health related virtual communities and electronic support groups: Systematic review of the effects of online peer to peer interactions. (Eysenbach et al., 2004) [21]

38 studies

Varied

“virtual community” defined as individuals with similar health related interests and predominantly nonprofessional backgrounds who interact and communicate publicly through a computer communication network (p. 1167).

From 38 studies, identified 6 RCTs:

4 reported effects (67%)

2/3 evaluating impacts on depression reported effects (67%)

1/2 evaluating impacts on diabetes reported effect (50%)

50%–67% (68.5%)

Outcome evaluations of CHW programs. (Swider, 2002) [1].

20 studies

Varied

Identified papers using terms “community health worker,” “community health advocate,” “promotora de salud,” “community health promoter,” “lay health worker,” and “community outreach worker.”

Numbers (%) reporting effects of PS in categories indicated:

Health status: 3/4 (75%)

Behavior change: 5/6 (83%)

Cost: 1/2 (50%)

50%–83% (75%)

Peer support programs for cancer. (Campbell et al., 2004) [47].

18 studies

Cancer

One-to-one, group, telephone and internet support programs, some with professional facilitation.

Across varied designs, “consistent informational, emotional and instrumental effects were identified” (abstract). However, 3 RCTs evaluating peer-led support groups found mixed and/or negative results on QOL; see text Discussion of Possible Harm of Unmoderated PS.

0/3 (0%)

Indigenous healthcare worker involvement for indigenous adults and children with asthma (Chang et al. 2010) [48]

1 paper

Asthma

Review of Indigenous Health Workers “Indidgenous” as “group of people who have inhabited a country for thousands of years, which often contrast with those of other groups of people who reside in the same country for a few hundred years” (p. 3), e.g., Australian Aboriginal, First Nations, Native Hawaiian.

Found only one study with children with asthma meeting Cochrane Collaboration criteria. Significant difference on asthma knowledge favored group with Indigenous Health Worker, but “although not statistically significant, all the outcomes favoured the group that had IHW involvement in the asthma education program” (Abstract)

Not Applicable

Review of CHW evaluations. (Nemcek & Sabatier, 2003) [49]

10 studies

Varied

Outreach, culturally sensitive care, health education/counseling, advocacy, home visits, health promotion/lifestyle change, transportation/homemaking

Identified 18 studies through 10 papers.

11 of 18 assessed outcomes, each of which reported ≥1 effect of PS.

100%

Peer support intervention trials for individuals with heart disease: A systematic review (Parry & Watt-Watson, 2010) [5]

6 studies

Heart disease

“peer mentors,” “lay health workers,” and “peer informants” delivered one-to-one sessions, telephone calls, combination of one-to-one and telephone calls, or self-help/support groups.

Some evidence for effects but authors indicated methodological problems preclude generalizations. Three of 6 studies reported some effect for peer support.

50% (50%)

Community health workers and environmental interventions for children with asthma. (Postma et al. 2009) [50]

7 studies

Asthma

PS worked in homes with families to promote behaviors that would reduce environmental triggers for asthma (e.g., controlling exposure to cockroach, dust mite, cigarette smoke).

From abstract: “Overall, the studies consistently identified positive outcomes associated with CHW-delivered interventions, including decreased asthma symptoms, daytime activity limitations, and emergency and urgent care use” (p. 564)

From results provided, not possible to characterize significance of PS vs control comparisons

Not Applicable