Efficacy and cost-effectiveness of a community-based smoke-free-home intervention with or without indoor-air-quality feedback in Bangladesh (MCLASS II): a three-arm, cluster-randomised, controlled trial

BACKGROUND: Exposure to second-hand smoke from tobacco is a major contributor to global morbidity and mortality. We aimed to evaluate the efficacy and cost-effectiveness of a community-based smoke-free-home intervention, with or without indoor-air-quality feedback, in reducing second-hand-smoke exposure in homes in Bangladesh. METHODS: We did a three-arm, cluster-randomised, controlled trial in Dhaka, Bangladesh, and randomly assigned (1:1:1) mosques and consenting households from their congregations to a smoke-free-home intervention plus indoor-air-quality feedback, smoke-free-home intervention only, or usual services. Households were eligible if they had at least one resident attending one of the participating mosques, at least one adult resident (age 18 years or older) who smoked cigarettes or other forms of smoked tobacco (eg, bidi, waterpipe) regularly (on at least 25 days per month), and at least one non-smoking resident of any age. The smoke-free-home intervention consisted of weekly health messages delivered within an Islamic discourse by religious leaders at mosques over 12 weeks. Indoor-air-quality feedback comprised providing households with feedback on their indoor air quality measured over 24 h. Households in the usual services group received no intervention. Masking of participants and mosque leaders was not possible. The primary outcome was the 24-h mean household airborne fine particulate matter (<2·5 microns in diameter [PM(2·5)]) concentration (a marker of second-hand smoke) at 12 months after randomisation. Cost-effectiveness was estimated using incremental cost-effectiveness ratios (ICERs). This trial is registered with ISRCTN, 49975452. FINDINGS: Between April 11 and Aug 2, 2018, we enrolled 1801 households from 45 mosques. 640 households (35·5%) were assigned to the smoke-free-home intervention plus indoor-air-quality feedback group, 560 (31·1%) to the smoke-free-home intervention only group, and 601 (33·4%) to the usual services group. At 12 months, the adjusted mean difference in household mean 24-h PM(2·5) concentration was −1·0 μg/m(3) (95% CI −12·8 to 10·9, p=0·88) for the smoke-free-home intervention plus indoor-air-quality feedback group versus the usual services group, 5·0 μg/m(3) (–7·9 to 18·0, p=0·45) for the smoke-free-home intervention only group versus the usual services group, and −6·0 μg/m(3) (–18·3 to 6·3, p=0·34) for the smoke-free-home intervention plus indoor-air-quality feedback group versus the smoke-free-home intervention only group. The ICER for the smoke-free-home intervention plus indoor-air-quality feedback versus usual services was US$653 per quality-adjusted life-year (QALY) gained, which was more than the upper limit of the Bangladesh willingness-to-pay threshold of $427 per QALY. INTERPRETATION: The smoke-free-home intervention, with or without indoor-air-quality feedback, was neither effective nor cost-effective in reducing household second-hand-smoke exposure compared with usual services. These interventions are therefore not recommended for Bangladesh. FUNDING: Medical Research Council UK. TRANSLATION: For the Bengali translation of the abstract see Supplementary Materials section.