At our field site in southern Nepal, acute respiratory illness (ARI) has been a leading cause of mortality among young children. Besides immunization there is little evidence for effective primary preventive approaches for ARI on a population basis. Low birth weight is highly prevalent in this population as well affecting approximately 30% of live born infants. Low birth weight is a key determinant of neonatal mortality and has also been resistant to cost-effective interventions in resource poor settings. Given the lack of appropriate interventions for poor, rural areas in developing countries and the strong observational association between open burning of biomass fuel sources and ARI in young children and low birth weight, we have designed a community-based randomized trial to determine if reductions in household indoor smoke exposure can reduce the incidence and duration of acute lower respiratory infections in children <36 months of age and low birthweight among newborn infants. Household indoor smoke reduction will be accomplished by replacing the current cook stove in the household with a locally appropriate, inexpensive model that is more efficient and vented to the exterior. In addition, we will assess the impact on respiratory function and symptoms among adults in the household. The project has 2 phases. Phase 1 is a cluster-randomized, community-based, step-wedge trial of cookstove replacement in a rural population of southern Nepal. Households will be randomized to receive replacement of their cook stove with an appropriately designed, efficient stove that is vented to the exterior at different time periods during the course of the study. An initial period of surveillance for ARI and low birth weight will establish a baseline rate for all clusters. This will be followed by the randomized, serial replacements of cook stoves over a 12 month period. Surveillance will continue throughout this period and for an additional 6 -18 months depending on when the stove was replaced. Phase 2 is a individually randomized trial in a subset of households that will receive either the improved biomass stove from phase 1 or a LPG stove and gas. Follow-up for phase 2 will be for 12 months with the same outcomes as phase 1. Measurement of indoor air particulate concentration will be conducted in all households before and after stove replacement. The analysis for both phases will focus on estimating the impact on incidence of ARI in children and low birth weight among live births as a result of stove replacement. Approximately 4200 children 1-35 months of age will be required to observe a minimum 10% reduction in risk of ARI with 90% power in phase 1. Given the expected number of live births to occur in these clusters, we can detect a 50 gram difference in birthweight with over 90% power and a type I error of 5%. Phase 2 will have lower power (total of 1800 households).

See brief summary above

Phase 1

Inclusion Criteria:

• All households with traditional open burning cookstoves in the study area.

Exclusion Criteria:

• Houses with walls of thatch or bamboo

Phase 2

Inclusion criteria:

• households in the phase 1 study area that either participated in phase 1 or are newly eligible household since phase 1 began. Eligible households are those with either a resident married pregnant woman or at least one child less than 24 months of age.