Hartinger 2010 PER.

Methods	Cluster‐RCT Allocation sequence: unclear Allocation concealment: unclear Blinding: participants and assessors Inclusion of participants in the analysis: unclear Length of follow‐up: 12 months Cluster‐adjustment method: unclear
Participants	Number: 534 households (267 intervention, 267 control) with 534 children (267 intervention, 267 control) Inclusion criteria: at least one child aged 6 to 35 months living in the home using wood or solid fuel as main energy source for cooking not being connected to public sewage tenants planning to stay in their home for the next 12 months Exclusion criteria: the child had any congenital abnormalities or suffered from a chronic debilitating illness families that had two or more households in different geographical areas with migration within sites that lasted more than 6 months during the year (mainly for migratory agriculture practices) Age: 6 to 35 months
Interventions	Intervention (see Table 5 for detailed description): Hygiene education with small and large group meetings 51 community clusters received integrated home‐hygiene intervention package Control: Psychomotor‐stimulation package
Outcomes	Diarrheal episodes Not used in this review: Prevalence of cough and fever Duration of days spent ill Average number of days for health care seeking Child growth outcomes (stunting, wasting and underweight)
Notes	Location: San Marcos province, Cajamarca region, Peru Duration of trial: March 2008 to January 2010 (23 months)
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Randomized "...using covariate‐based constrained randomization as proposed by Moulton (2004)". Researchers went to extra lengths to ensure integrity of the randomizations.
Allocation concealment (selection bias)	Unclear risk	Not described.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	"As a strategy to reduce non‐blinding bias, a child psychomotor development intervention was implemented in the control arm as an equivalent to the IHIP in the intervention arm".
Blinding of outcome assessment (detection bias) All outcomes	Low risk	"...and data collection was done by an independent team of field workers, which was not part of the initial education and re‐enforcement of the interventions during the follow‐up period". We consider this an attempt to blinding outcome assessors.
Incomplete outcome data (attrition bias) All outcomes	Low risk	Researchers presented a detailed account of the randomization and follow‐up in a PRISMA flow diagram.
Selective reporting (reporting bias)	Low risk	None observed.
Other bias	Low risk	None observed.