Forty-six studies (approximately 5,800 participants) were included in the review. There were 24 RCTs (including one cluster RCT), 9 CCTs, 4 controlled observational studies and 9 uncontrolled observational studies. The sample size ranged from 13 to 433 participants.
Of the 24 RCTs, one was judged to be of good methodological quality, six satisfactory, ten poor and seven very poor. Of the remaining studies, one was of satisfactory quality, six were poor and fifteen were very poor. Reasons for poor quality included small sample size, high drop-out rates, lack of intention-to- treat analysis, possible contamination of treatment groups, and lack of control for confounding.
Results across all indications.
Of the 26 comparisons with a no-treatment control, 19 reported at least one statistically significant difference. Of the 22 comparisons with baseline, 14 reported at least one statistically significant difference. When considering only higher quality studies (those rated good or satisfactory), all comparisons with a no-treatment control gave at least one statistically significant effect of the self-care support network. There were no clear patterns to suggest optimal features of self-care support networks when the results were explored by indication, study quality, study design, duration of follow-up, peer or professional leadership, level of professional input, group process, or whether researchers set up the self-care group.
Findings by indication.
Weight loss/obesity: 13 studies (7 RCTs, 1 CCT and 5 observational studies) were identified. Overall, the studies provided evidence that peer-led weight loss programmes can result in statistically significant but clinically modest weight loss, although most studies were of a poor quality.
Carers: 9 studies (7 RCTs, 1 CCT and 1 observational study) were identified. The better quality RCTs were of carers for people with mental health conditions in Hong Kong. They found improvements in patients' psychosocial functioning, rehospitalisation, family functioning, family burden, carer distress and quality of life associated with self-care support networks.
Diabetes: 4 studies (2 RCTs and 2 CCTs) were identified. The studies suggested that self-care support networks can result in long-term improvements in several outcomes, including glucose tolerance and quality of life, although they were of poor methodological quality and may not be reliable.
Rheumatic disease/arthritis: 3 studies (2 RCTs and 1 CCT) were identified. One study suggested a beneficial effect of a self- care support network on self-efficacy and health care costs for osteoarthritis. Two studies found no clear benefits for rheumatic diseases.
HIV/AIDS: 3 studies (1 RCT and 2 observational studies) were identified. The RCT evaluated a very short programme which was not as effective as a professionally led intervention. There were methodological problems with the other 2 studies.
Eating disorders: 2 studies (1 RCT and 1 observational study) were identified. The RCT suggested that self-help groups may be of benefit to bulimics over time, but professionally led interventions can be more effective for some outcomes. The results from the observational study were considered to be unreliable.
Depression: 2 RCTs were identified. One RCT found that a self- care support network may be beneficial for women with postnatal depression, particularly if it involved physical activity. The other RCT found significant benefits of a number of group interventions, including one self-care support network.
Chronic pain: 2 studies (1 RCT and 1 observational study) were identified. The RCT found no clear indication of a beneficial effect of a self-care support network. The other study was very small.
Other indications (8 studies): 8 studies were identified. One study each was included for the following indications: cardiac recovery (1 observational study), epilepsy (1 observational study), healthy living (1 CCT), injury (1 CCT), general mental health (1 RCT), psoriasis (1 CCT), systemic lupus erythematosus (1 observational study) and tuberculosis (1 CCT). The only study considered to be of sufficient methodological quality found that self-help clubs can improve adherence for people with tuberculosis in Northern Ethiopia.