Table 4.

Summary of quantitative outcomes from review papers

Author	Summary of key communicable disease control outcomes from community engagement interventions in low and lower middle income countries
	HIV
Cornish et al. [7]	High risk populations (including FSW and MSM) - Of three studies measuring HIV prevalence following community mobilisation interventions, one case control study showed that greater programme intensity was significantly associated with a lower HIV prevalence in three of six Indian states tested. In a cohort study of the same programme, CMI were associated with a significant reduction in the prevalence of HIV, whereas a further cohort study found no significant reduction in HIV. -Of three studies measuring the impact of CMI on other STIs, one showed a significant reduction in syphilis, and chlamydia and/or gonorrhoea in FSW, while another showed a significantly lower likelihood of HSV-2 and syphilis in FSW and MSM. A further study showed a significant increase in the prevalence of HSV-2, alongside significant decreases in syphilis, trichomonas, chlamydial infection and gonorrhoea. -Significant increase in condom use in four CMI studies with FSW, (although increases were non-significant under certain circumstances in two of these studies). Mixed evidence on condom use in MSM following CMI. -One study showed a significant increase in social support but not political participation following a community mobilisation intervention. -Significant association between being a member of a help group and experiencing higher perceived collective efficacy and support in one study. Youth population -Four studies showed non-significant effect of CE intervention on HIV incidence or prevalence. - Effects on the incidence/prevalence of other STIs were mainly non-significant; one study showed a significant reduction in gonorrhoea and syphilis incidence following the CE intervention. Another study showed a significant decrease in HSV-2 incidence. -One study showed a significant increase in the rate of condom use with casual partners, however four studies showed no significant changes in condom use with regular partners. -HIV testing was significantly increased in one intervention using community based voluntary testing compared to standard care.
Skevington et al. [20]	-Condom use significantly increased in two of five studies following the ‘Stepping Stones’ (SS) CE intervention. No significant changes were seen in the other studies. -Of the two studies that reported the effect on multiple sexual partners, one showed a significant reduction following the intervention. -Two of five studies showed a significant decrease in alcohol use before sex following the CE intervention. One study showed that communities participating in SS used significantly less alcohol than non SS villages. -One of five studies reported a significant increase in individual knowledge following the SS intervention, another study showed a significant increase in knowledge at a community level compared to non-participating villages. -Of two studies measuring changes in gender equity, one study showed significant improvements in some attitudes following the intervention. -Two of five studies reported improvements in attitudes towards those living with HIV and AIDs following the SS intervention, one of which reported statistical significance.
Kerrigan et al., 2013 [8]	-Two of three studies measuring HIV infection showed an odds ratio that was significantly protective in favour of the community empowerment intervention at a follow up of 2.5 years. - In meta-analysis of three studies, community empowerment was associated with decreased odds of gonorrhoea but not chlamydia. - Condom use was measured in six studies. Five studies showed that community empowerment was associated with significantly higher odds of condom use with clients, however there was statistical heterogeneity in this result. -Three studies measured consistent condom use with regular non-paying partners and no significant associations with CE were found.
Kerrigan et al., 2015 [9]	All relevant results are from community empowerment studies conducted in India; -Results from nine intervention sites were combined in meta-analysis and showed a significantly reduced prevalence of HIV in sex workers following the community empowerment intervention (heterogeneity was high). -Meta-analysis of results from four intervention sites showed a significant reduction in the odds of syphilis. -Of ten intervention sites measuring the impact of community empowerment interventions on gonorrhoea prevalence, five showed significantly reduced odds of gonorrhoea. -Similarly, of ten intervention sites measuring the impact on chlamydia risk, four showed significant reductions in the odds of chlamydia, (five showed non-significant reductions and one showed a significant increase in the odds of chlamydia). -Condom use was measured in one RCT and showed a significant improvement over time in intervention participants compared to controls. -Meta analysis of results from cross sectional studies over six intervention sites showed significantly increased condom use with regular clients (heterogeneity was high). -A further seven and five intervention sites reported significant increases in condom use with all clients and condom use with new clients respectively.
Nachega et al. [21]	-Of seven RCTS and two cohort studies measuring the impact of community-based delivery of antiretroviral therapy (ART), one RCT showed a significant decrease in all-cause mortality in the intervention group compared to control group. The remaining studies showed no significant differences between groups. - Virologic suppression at 12 and/or 24 months after ART initiation was measured in six RCTS and two cohort studies- no significant differences between the intervention and control groups were found. -Two of five RCTS that measured optimal ART adherence levels showed a significant increase following the community-based initiative, while three showed a non-significant reduction in adherence levels. -Six RCTs and two cohort studies measured retention in care, and no statistically significant differences were found between those receiving the community-based initiative and those in the control group.
Medley et al. [22]	-Of five studies measuring the impact of peer education on STI infection, one study showed a significant decrease in STI infection, one showed a significant increase in STI infection and three studies showed non-significant reductions in STI risk. - Of ten studies reporting the impact of peer education on condom use, five showed significant increases in the likelihood of use, four showed non-significant increases in the and one showed a non-significant decrease in the likelihood of condom use. -Of ten studies, seven showed a significantly positive impact on HIV knowledge associated with peer education interventions.
	MALARIA
Atkinson et al. [25].	Eight studies provided quantitative results; Biological outcomes -One community engagement intervention showed a statistically significant reduction in prevalence of STD symptoms in the intervention compared to control groups. - A community engagement intervention for the treatment of malaria showed a significant reduction in mean incidence of malaria per 10,000 person weeks over 2 years compared to control. Behavioural -A community delivered intervention showed significantly increased coverage for vitamin A supplementation, bed nets and anti-malaria treatment compared to control districts, however no significant difference was found in directly observed therapy (DOT) between the intervention and control areas. - In a study of lymphatic filariasis, no significant difference was found in drug distribution and consumption when this was devolved entirely to communities or delivered routinely by medics. - A study of environmental modification plus community participation showed significantly higher perceived benefits of drain cleaning in the intervention communities compared to the control group (61% vs 30%). Psychosocial - A study of health and feedback committees in communities in Cambodia found engagement of existing community-based structures more effective for community participation than externally introduced structures. - A community directed intervention (CDI) approach using traditional kinship systems for the treatment of onchocerciasis showed significantly better disease knowledge, significantly lower control by leaders and increased treatment coverage compared to a standard CDI approach.
Okwundu et al. [26].	Ten studies were included in this review; -One trial showed a significantly reduced risk of mortality in the home or community-based programme compared to facility-based care. - Two trials to measure parasitaemia showed mixed results- one showing a significantly reduced risk in the intervention group, the other not. - Evidence from one trial showed no significant impact on hospitalisation for children, when mothers had been trained to treat fevers. - Pooled results from two trials showed a significant increase in prompt treatment with anti-malarials in the intervention group, compared to control. -The pooled results of two trials showed that the use of rapid diagnostic testing compared to clinical diagnosis in community-based programmes reduced prescribing of antimalarials however there were no differences in hospitalisation or all cause morbidity.
	CHILD AND MATERNAL HEALTH
Prost et al. [23].	-Meta-analysis of seven RCTS showed exposure to women’s groups was associated with a 23% non-significant reduction in maternal mortality, a 20% significant reduction in neonatal mortality and a 7% non-significant reduction in stillbirth, with significant heterogeneity for maternal and neonatal results (NB these results represent all-cause mortality). -Five of seven studies measured ‘increased handwashing by attendants before home deliveries’: Of these five studies, there was a significant difference between intervention and control groups in three studies. -Four of seven studies reported increased use of clean delivery kits for home births. Of these four studies, three found significant differences between intervention and control groups.
Farnsworth et al. [2].	Communicable disease specific outcomes were measured in five studies: Biological outcomes - Significant decrease in child deaths due to malaria in one study using an eight stage CE intervention. - Reduced prevalence of fever in relation to community-based control of malaria, in one study. Behavioural outcomes -Improved hygiene in birth delivery practices in one study following a range of collaborative approaches and CE techniques. - Two studies showed significant increases in net use for malaria prevention following a community engagement intervention; Another study showed significant increases in water disinfectant use with a study utilising volunteer health promoters to deliver motivational interviewing. -Increased care seeking for malaria was found in one study that followed a health promotion approach with participation, empowerment and contextualisation. Psychosocial outcomes - Two studies showed improvements in knowledge following community engagement interventions, in the areas of malaria knowledge and water disinfectant use. - Social cohesion was increased in two studies, alongside increases in social capital and trust following CE interventions. Collective self-efficacy (community empowerment) increased in three studies.
Gilmore et al. [24]	Biological outcomes -Of five studies measuring the impact of community health worker programme on rates of diarrhoea, four showed significantly reduced rates of diarrhoea in infants or children, two using educational approaches, one through breastfeeding promotion and one through the promotion of Kangaroo care. Another breastfeeding intervention showed no significant difference in the prevalence of infant diarrhoea in the intervention and control group, despite demonstrating significantly higher breastfeeding rates. -One study of CHWs reported a reduction in under 5 year mortality rates of 53%, at 18 months following the intervention (no tests of significance provided). The same study reported that malaria and or fever prevalence was significantly reduced by 5.8% in the intervention group. - A trial using CHWs to promote DPT-3/Hep B vaccination demonstrated that full immunization rates were 32% higher in the intervention group at 4 months. Behavioural and psychosocial outcomes -A further study of CHWs in antimalarial treatment and bed net distribution reported significantly higher rates of bed net use in pregnancy and rates of antimalarial treatment in the intervention group compared to the control. -In a study of CHWs in an urban slum, poor sanitation and hygiene practices were significantly reduced in the intervention group compared to the control. In addition, there was significant improvement in mother’s knowledge, attitude and practice regarding diarrhoea etiology and sanitation and hygiene.
Salimi et al. [28]	Three relevant studies were included in this review of community based participatory research; -One cluster RCT, using a participatory learning and action cycle with women’s groups in Nepal showed a significant reduction in neonatal mortality and in maternal mortality rates in the intervention group compared to the control over 2 years. There were no significant differences in stillbirth rates. -A longitudinal, experimental study using participatory action research (PAR) with high risk heterosexual males in the Philippines showed significant increases in condom use and attitude towards condom use at post-test and 6 months compared to baseline. The reported STI incidence also decreased significantly at post- test and 6 months’ time points. -A further cluster RCT using participatory approaches with community leaders to promote a healthy living environment showed a significant increase in scores relating to ‘healthy living environment competencies’ following the intervention. These competencies were in areas such as sanitation, hygiene and prevention of diseases. No significant changes in these competencies were seen in the control group.
	TB
Musa et al. [27].	- Pooled outcome from five studies shows no significant difference in TB treatment success when TB care was delivered by lay health workers compared to facility-based care. However, stratified analysis of a small number of studies showed that LHW interventions in rural settings significantly increased TB treatment success compared to standard facility-based care with no significant difference in urban studies.