BACKGROUND: Cookstove intervention programs have been increasing over the past two (2) decades in Low and Middle Income Countries (LMICs) across the globe. However, there remains uncertainty regarding the effects of these interventions on household air pollution concentrations, personal exposure concentrations and health outcomes.
OBJECTIVES: The primary objective was to determine if household air pollution (HAP) interventions were associated with improved indoor air quality (IAQ) in households in LMICs. Given the potential impact of HAP interventions on health, a secondary objective was to evaluate the effectiveness of HAP interventions to improve health in populations receiving these interventions.
DATA SOURCES: OVID Medline, Ovid Embase, SCOPUS and PubMED were searched from their inception until December 2015 with no restrictions on study design. The WHO Global database of household air pollution measurements and Members' archives were also reviewed together with the reference lists of identified reviews and relevant articles.
STUDY ELIGIBILITY CRITERIA, PARTICIPANTS AND INTERVENTION: We considered randomized controlled trials, or non-randomized control trials, or before-and-after studies; original studies; studies conducted in a LMIC (based on the United Nations Human Development Report released in March 2013 (World Bank, 2013); interventions that were explicitly aimed at improving IAQ and/or health from solid fuel use; studies published in a peer-reviewed journal or student theses or reports; studies that reported on outcomes which was indicative of IAQ or/and health. There was no restriction on the type of comparator (e.g. household receiving plancha vs. household using traditional cookstove) used in the intervention study.
STUDY APPRAISAL AND SYNTHESIS METHODS: Five review authors independently used pre-designed data collection forms to extract information from the original studies and assessed risk of bias using the Effective Public Health Practice Project (EPHPP). We computed standardized weighted mean difference (SMD) using random-effects models. Heterogeneity was computed using the Q and I2-statistics. We examined the influence of various characteristics on the study-specific effect estimates by stratifying the analysis by population type, study design, intervention type, and duration of exposure monitoring. The trim and fill method was used to assess the potential impact of missing studies.
RESULTS: Fifty-five studies met our a priori inclusion criteria and were included in the systematic review. Fifteen studies provided 43 effect estimates for our meta-analysis. The largest improvement in HAP was observed for average particulate matter (PM) (SMD=1.57) concentrations in household kitchens (1.03), followed by daily personal average concentrations of PM (1.18), and carbon monoxide (CO) concentrations in kitchens. With respect to personal PM, significant improvement was observed in studies of children (1.26) and studies monitoring PM for ≥24h (1.32). This observation was also noted in terms of studies of kitchen concentrations of CO. A significant improvement was also observed for kitchen levels of PM in both adult populations (1.56) and in RCT/cohort designs (1.59) involving replacing cookstoves without chimneys. Our findings on health outcomes were inconclusive.
LIMITATIONS, CONCLUSIONS AND IMPLICATIONS OF KEY FINDINGS: We observed high statistical between study variability in the study-specific estimate. Thus, care should be taken in concluding that HAP interventions - as currently designed and implemented - support reductions in the average kitchen and personal levels of PM and CO. Further, there is limited evidence that current stand-alone HAP interventions yield any health benefits. Post-intervention levels of pollutants were generally still greatly in excess of the relevant WHO guideline and thus a need to promote cleaner fuels in LMICs to reduce HAP levels below the WHO guidelines.
SYSTEMATIC REVIEW REGISTRATION NUMBER: The review has been registered with PROSPERO (registration number CRD42014009768).
