High Fidelity: Delivery and use of an LPG stove intervention during pregnancy in the HAPIN trial

Abstract

BACKGROUND AND AIM: Most previous cookstove trials have reported low intervention adoption and/or high continued traditional stove use, limiting interpretation of results. The Household Air Pollution Intervention Network (HAPIN) trial randomized 3195 pregnant women in resource-poor settings of four countries to a liquefied petroleum gas (LPG) stove and fuel intervention or to control. We assessed implementation fidelity (the extent to which the intervention was delivered as intended) and adherence (the extent to which it was adopted by participants) at the first trial endpoint: the end of the gestational period. METHODS: We tracked intervention fidelity through questionnaires at baseline ( 20 weeks gestation) and at two post-intervention visits (at 24-28 and 32-36 weeks of pregnancy), data collected by study staff (e.g. visual observations at fuel deliveries), and temperature-logging stove use monitors (SUMs) downloaded twice monthly. RESULTS:Over a median [Q1, Q3] follow-up period of 150 [131, 168] days per household, median time to repair for stoves requiring maintenance was 0 [0, 0] days (same day repair), and 87.2% of LPG refills were delivered within one week of request. 96.1% of intervention households reported LPG stove use at both follow-up visits. In the 70.0% of intervention households that retained traditional stoves after receiving their LPG stove, the median rate of traditional stove use was 0.0 [0.0, 1.6] days per 100 days of observation (using SUMs). In 70.6% of these households, traditional stoves were used on fewer than one in every 100 days of observation (Guatemala: 83.3%; India: 85.4%; Peru: 59.6%; Rwanda: 70.2%). CONCLUSIONS:The HAPIN intervention was largely delivered as intended, and participant adherence to the LPG stove intervention during pregnancy was high. Any observed differences in pollutant exposures or health outcomes between control and intervention groups are unlikely to be due to implementation failures during the gestational period. KEYWORDS: Air Pollution, Birth Outcomes, Community-engaged research, International Collaboration, Methodological study design, Particulate Matter