Kalyango 2012a.

Study characteristics
Methods	Design: cluster‐randomized controlled trial Unit of randomization: groups of villages (parishes)
Participants	Inclusion criteria: children aged 6–59 months in study villages who received treatment from CHWs for any illness; identified from CHW registers, traced to their homes and enrolled in study. All enrolled children were included in the analysis for treatment outcomes. Only children with pneumonia symptoms were included in the analysis for prompt and appropriate antibiotics for pneumonia symptoms Exclusion criteria: none reported
Interventions	Intervention Recruiting and training lay health workers (CHWs) to provide iCCM for malaria and pneumonia (ARI) among children aged 4–59 months Recruiting and training other types of health workers to provide IMNCI Implementing simplified IMCI‐adapted clinical guidelines for iCCM providers Implementing referral of children under 4 months of age and children with severe disease to health facilities Providing iCCM providers with drugs and equipment Training supervisors of lay health workers (iCCM for intervention and CCM for control) Providing supervision to lay health workers (iCCM for intervention and CCM for control); frequency monthly (content and approach not reported) Comparison Usual facility services + CCM for malaria
Outcomes	Coverage of appropriate treatment: Coverage of appropriate treatment (antibiotics) for pneumonia Coverage of appropriate treatment (antibiotics) for pneumonia by an iCCM provider Coverage of appropriate treatment (antibiotics) for pneumonia within 24 hours Coverage of careseeking to an 'appropriate provider'of treatment services Careseeking for children with suspected pneumonia to an iCCM provider Careseeking for children with fever to an iCCM provider Coverage of careseeking to an appropriate provider of treatment services for any illness Coverage of careseeking to an iCCM provider as first source of treatment for any illness
Notes	Objective: to determine the effect of integrated malaria and pneumonia management, compared to malaria only management by CHWs, on receiving prompt and appropriate antibiotics for pneumonia symptoms. Location: Eastern Uganda, Iganga Municipality. Funding source: SIDA and UNICEF/UNDP/World Bank/WHO Special Program for Research and Training in Tropical Diseases.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "Randomization was done by a statistician that was independent of the study using stratified block randomization. Iganga‐Mayuge HDSS has 65 villages which make up 26 parishes that were divided into eight urban and 18 rural clusters (parishes). The clusters from the rural area were further grouped into three strata based on the population size of children less than five years: i) 190–320, ii) 321– 390, and iii) 391 and above, resulting in six clusters in each of these strata. The clusters from the urban area were grouped into two strata based on population sizes of iv) 280–430, and v) 431 and above. Random numbers were generated in blocks of six for the rural clusters and in blocks of four for the urban clusters."
Allocation concealment (selection bias)	Low risk	Quote: "Randomization was done by a statistician that was independent of the study using stratified block randomization."
Blinding of participants and personnel (performance bias) All outcomes	High risk	No blinding of participants and personnel. Lay health workers would have known if they received additional training and this may have biased their performance. Allocation was by village and parents may have known that the health workers at their primary health centre had received additional training and this may have biased their care seeking behaviour or responses to questionnaires, or both.
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Data collectors were not blinded; however, they were independent of the intervention. It is not clear whether being independent would have mitigated the risk of detection bias due to not being blinded.
Incomplete outcome data (attrition bias) All outcomes	Low risk	Quote: "All children enrolled on day 1 were assessed on day 4."
Selective reporting (reporting bias)	High risk	Mortality was the primary outcome measure of the registered trial (ISRCTN52966230), but this outcome has never been published.
Baseline outcomes similar	Unclear risk	Baseline outcomes (careseeking and quality of care) were not assessed. The history of children with illness at baseline was similar between arms, with the exception of the % of children with fast breathing per respiration count by field assistants on day 1 – which was higher in the intervention arm compared to the control arm. This may have had an effect on outcomes for careseeking and quality of care. Imbalances in the number of children treated per arm could have resulted in a loss of power, possibly dampening any effect of the intervention.
Baseline characteristics similar	Low risk	Baseline characteristics were similar except for higher % rural population in control clusters.
Contamination	Unclear risk	There were no buffer zones between the intervention clusters and control clusters and caregivers from the control clusters may have accessed care in the intervention clusters, possibly dampening any positive effect of the intervention.
Other bias	Unclear risk	No other apparent source of bias.