Table 5.

Selected effects of programmes on quality of care related outcomes*

Programme	Access, quality of care, health outcomes	Quality of evidence†
Backpack	Optimize access and service efficiency; increased community trust thanks to better communication; reduced error rates thanks to improved tools and higher guidance. Better treatment thanks to enhanced guidance and real–time support (source: project statement); reduced stock outs.	+
Concern Rwanda‡ – Home Based Management Malaria Programme	Increase access and use of prompt treatment for presumed malaria (20% [CI 13% to 23%] to 43% [CI 35 to 51%]); increase access to zinc for diarrhoea (5% [CI 2% to 8%] to 22% [CI 15% to 30%]); more practice of giving increased liquids for diarrhoea (36 [CI 30% to 42%] to 57% [48% to 66%]); increase vitamin A coverage (66% [CI 61% to 71%] to 86% [78% to 94%]); increase practice of hand–washing with soap on key occasions (2% [CI 1% to 4%] to 19% [CI 11% to 26%])§. Notable improvements in treatment–seeking between 2005 and 2010 (greater in KabehoMwana districts). Treatment seeking from any provider for all three conditions combined increased from 16% to 46% in the KM districts vs 26% to 40% in non– KabehoMwana districts. Other indicators shown differences: soap availability, vitamin A supplementation, diarrhoea management, respiratory disease management.#	+++
	In most Health Centres assessed, reported malaria cases decreased during the peak malaria season in the year after implementation of HBM, compared to the year before.	++
CORE – Cameroon	Changes from baseline in the percentage of sick children correctly assessed and managed for danger signs (10.5% to 33.9%) and specific diseases (for example diarrhoea: from 23% to 66.7%).Coverage of certain interventions (eg, vaccination). Mothers’ knowledge.	++
CORE – Malawi	Estimated 1114 lives were saved over the life of the project, 474 from malaria (applying the lives saved calculator to data). Estimated cost per life saved US$ 1200 (based on the project’s total budget).	++
	Mothers continued breastfeeding children even when pregnant; children and pregnant women were more likely to eat eggs, food high in protein and essential micronutrients. Care–seeking for childhood illness increased from 71% to 84%; childhood vaccinations increased from 69% to 96%; vitamin A dosing increased from 54% to 82%; exclusive breastfeeding jumped from 40% to 82%.	++
	Residents far less likely to use traditional healers; people stopped using bed nets for fishing; a significant number of traditional healers abandoned their practice and joined the program as volunteers, isolating and undermining the credibility of those who remained working as traditional healers.	+
IRC Sierra Leone	Care–seeking changes (2010 to 2013): overall (82.0% [CI 76.7% to 88.2%] to 72.4% [CI 62.6% to 80.5%]) and malaria /fever (57.4% [CI 49.7% to 64.9%] to 83.8% [CI 77.9% to 88.4%]); time delays reduced for diarrhoea but increased for pneumonia. First sources of health care in 2010 and 2013 for sick children (CHW52.0% to 52.9%), for children who died (governmental health facility CBD 52.9% to 24.9%; CBD37.7% to 30.8%). Treatments/child/y given by CBD (2010 vs 2013): malaria (0.56 to 1.37), diarrhoea (0.52 to 0.88), pneumonia (0.46 to 0.31). Appropriate treatment (2010 to 2013): malaria (54.7% to 80.4%), diarrhoea (33.1% to 53.7%), pneumonia (0.0% to 67.8%). Prevalence: malaria (46% to 36%), diarrhoea (5% to 7%), pneumonia (1% to 6%). Mortality 2 to 59 mo: statistically non–significant reduction, from 2010 to 2013.	+++
Living Goods Uganda	Better access to diagnostics. The results are consistent with a simple experience model where biomedical misconceptions decrease consumers’ ability to infer quality.	+
MC Uganda	Communication outcomes: sick child job aid is a trusted guide for both CHW and caregivers and appears to contribute to quality of care; interpersonal skills are the key drivers of caregivers’ satisfaction, impacting positively on the CHWs’ clinical skills.	+
MOH Madagascar	CHW referred to health facilities: 71.6% (69.9% to 73.3%) of children with severe illness or other indications; chose the appropriate life–saving treatment when it was needed only 53% (43.3% to 63.1%); chose RTDs when indicated only 55% of the time; assess contraindications for oral contraceptive use only 41% of encounters.	+++
MOH Malawi	Communities are using the sick child services.	+
PSI Cameroon	Reduction of mortality in one district but not in another one (from 96.8 to 86.7/1000 life birth); increased access to the poorest (52% among the poorest vs 35% among the less poor); for ACTs: 45% vs 33%. Improved quality of care.	+++
PSI Malawi CIDA	Slight reduction in stock–outs and slight increase in health seeking behaviour for diarrhoea, fast breathing and fever.	+
PSI Mali	Treatment target (80%) was exceeded (average 81% and 86% at the end of the period). 713 474 DALYs (8399 deaths averted).	++
	Mild to moderate improvements in appropriate treatments, positive care–giver feed–back.	+
PSI Madagascar	Trust of community members; although some are sceptical. Statements on supply management and sales.	+
PSI Myanmar	Access to RDT. Increase of ORS use. Cost–effectiveness of ORS distribution (US$ 431/DALY).	++
PSI Uganda	No evidence of changes in coverage; changes in case management comparable to national levels; may be stronger gains in children from less poor households.	+
Save Malawi	CHW were the main source of care in intervention areas (at baseline the source was the public sector); shifting care from public to CHW care; checking breathing with timer not systematic; non–statistical significant increase of appropriate treatments. Improved on equity in access.	+++
Save Mozambique	Health seeking for fever: higher in intervention areas for the formal sector (intervention clusters 83.2% (CI 76.3 to 90.0); comparison areas 66.3% (95% CI 57.8 to 74.9)); CWH were the main source; pneumonia: lower in intervention areas in the formal sector; diarrhoea: CHW main source in intervention areas; first–line antimalarials: CHW preferred in intervention areas and formal sector in comparison areas. Improve of early treatment for malaria and diarrhoea (significant) and pneumonia (hardly significant). 2/3 RDT and less used timer for breathing. Provision of correct drugs (80%). Increased knowledge by mothers in intervention areas.	+++
IRC Ethiopia	Some practices not followed (check for danger signs, correct assessments).	+
Sizika	100% completeness in electronic data upload; 98.44% of promptness of treatment; 57% of treated children; 70% of relay/CHW were supervised; 75.5% to 98.44% of early malaria treatments (depending on month).	++

CI –confidence interval, CIDA – Canadian International Development Agency, DALY – disability adjusted life years, DRC – Democratic Republic Congo, HBM – home based management of malaria, IRC – International Red Cross, MC – Malaria Consortium, MOH – Ministry of Health, ORS – oral rehydration salts, PSI – Population Services International, RDT – rapid diagnostic test

*Not all programme reported information on this area.

†Sources of evidence are qualitative data or opinions (+), quantitative methods described in the source documents (++), or findings are presented with some measure of statistical significance (+++).

‡Two similar documents sharing a common author and similar sources were available on the Peer Support Group topic specifically so the outcomes were captured from the most comprehensive document (PSG Review paper). Case study on Peer Support Group drawing from primary and secondary data collected as part of the KabehoMwana project final evaluation, prepared for the iCCM Symposium evidence review; consists of an adaptation of a longer paper describing experiences from the USAID-funded KabehoMwana project in Rwanda. Review paper on Peer Support Group combining participant observations from designers and implementers of the CHW PSG model, with project monitoring and routine monitoring data, findings from primary and secondary data collected as part of the KabehoMwana final evaluation; and other available studies in the grey literature.

§Final project external evaluation (2011) with a knowledge, practice and coverage survey, comparison with 2007 baseline. Interviews were conducted with 120 mothers of children 0 to 23 mo, and 395 mothers of children 0 to 59 months who had been sick in the last two weeks with at least one the following conditions: fever or malaria, diarrhoea, respiratory symptoms. In total, 120 villages were sampled, and 473 households were interviewed. Household selection was made according to an algorithm.

#Demographic and Health Surveys 2005 and 2010.