Design and Rationale of the HAPIN Study: A Multicountry Randomized Controlled Trial to Assess the Effect of Liquefied Petroleum Gas Stove and Continuous Fuel Distribution
2020; National Institute of Environmental Health Sciences; Volume: 128; Issue: 4 Linguagem: Inglês
10.1289/ehp6407
ISSN1552-9924
AutoresThomas Clasen, William Checkley, Jennifer L. Peel, Kalpana Balakrishnan, John P. McCracken, Ghislaine Rosa, Lisa M. Thompson, Dana Boyd Barr, Maggie L. Clark, Michael Johnson, Lance A. Waller, Lindsay M. Jaacks, Kyle Steenland, J. Jaime Miranda, Howard H. Chang, Dong‐Yun Kim, Eric D. McCollum, Víctor G. Dávila‐Román, Aris T. Papageorghiou, Joshua P. Rosenthal,
Tópico(s)Energy, Environment, and Transportation Policies
ResumoVol. 128, No. 4 ResearchOpen AccessDesign and Rationale of the HAPIN Study: A Multicountry Randomized Controlled Trial to Assess the Effect of Liquefied Petroleum Gas Stove and Continuous Fuel Distributionis accompanied byAir Pollutant Exposure and Stove Use Assessment Methods for the Household Air Pollution Intervention Network (HAPIN) Trialis companion ofDesign and Rationale of the Biomarker Center of the Household Air Pollution Intervention Network (HAPIN) Trial Thomas Clasen, William Checkley, Jennifer L. Peel, Kalpana Balakrishnan, John P. McCracken, Ghislaine Rosa, Lisa M. Thompson, Dana Boyd Barr, Maggie L. Clark, Michael A. Johnson, Lance A. Waller, Lindsay M. Jaacks, Kyle Steenland, J. Jaime Miranda, Howard H. Chang, Dong-Yun Kim, Eric D. McCollum, Victor G. Davila-Roman, Aris Papageorghiou, Joshua P. Rosenthal, and HAPIN Investigators Thomas Clasen Address correspondence to Thomas Clasen, Department of Environmental Health, Rollins School of Public Health, Emory University, 1518 Clifton Rd. NE, Atlanta, GA 30322 USA. Telephone: (404) 727-3480. Email: E-mail Address: [email protected] Department of Environmental Health, Rollins School of Public Health, Emory University, Atlanta, Georgia, USA Search for more papers by this author , William Checkley Division of Pulmonary and Critical Care, School of Medicine, Johns Hopkins University, Baltimore, Maryland, USA Search for more papers by this author , Jennifer L. Peel Department of Environmental and Radiological Health Sciences, Colorado State University, Fort Collins, Colorado, USA Search for more papers by this author , Kalpana Balakrishnan Department of Environmental Health Engineering, ICMR Center for Advanced Research on Air Quality, Climate and Health, Sri Ramachandra Institute for Higher Education and Research (Deemed University), Chennai, Tamil Nadu, India Search for more papers by this author , John P. McCracken Center for Health Studies, Universidad del Valle de Guatemala, Guatemala City, Guatemala Search for more papers by this author , Ghislaine Rosa Department of Disease Control, Faculty of Infections and Tropical Diseases, London School of Hygiene and Tropical Medicine, London, UK Search for more papers by this author , Lisa M. Thompson Nell Hodgson Woodruff School of Nursing, Emory University, Atlanta, Georgia, USA Search for more papers by this author , Dana Boyd Barr Department of Environmental Health, Rollins School of Public Health, Emory University, Atlanta, Georgia, USA Search for more papers by this author , Maggie L. Clark Department of Environmental and Radiological Health Sciences, Colorado State University, Fort Collins, Colorado, USA Search for more papers by this author , Michael A. Johnson Berkeley Air Monitoring Group, Berkeley, California, USA Search for more papers by this author , Lance A. Waller Department of Biostatistics and Bioinformatics, Rollins School of Public Health, Emory University, Atlanta, Georgia, USA Search for more papers by this author , Lindsay M. Jaacks Department of Global Health and Population, Harvard T.H. Chan School of Public Health, Boston, Massachusetts, USA Search for more papers by this author , Kyle Steenland Department of Environmental Health, Rollins School of Public Health, Emory University, Atlanta, Georgia, USA Search for more papers by this author , J. Jaime Miranda CRONICAS Center of Excellence in Chronic Diseases, Universidad Peruana Cayetano Heredia, Lima, Peru Search for more papers by this author , Howard H. Chang Department of Biostatistics and Bioinformatics, Rollins School of Public Health, Emory University, Atlanta, Georgia, USA Search for more papers by this author , Dong-Yun Kim Office of Biostatistics Research, National Heart Lung and Blood Institute, National Institutes of Health, Bethesda, Maryland, USA Search for more papers by this author , Eric D. McCollum Eudowood Division of Respiratory Sciences, Department of Pediatrics, School of Medicine, Johns Hopkins University, Baltimore, Maryland, USA Search for more papers by this author , Victor G. Davila-Roman Cardiovascular Imaging and Clinical Research Core Laboratory, Cardiovascular Division, Washington University School of Medicine, St. Louis, Missouri, USA Search for more papers by this author , Aris Papageorghiou Nuffield Department of Women’s and Reproductive Health, University of Oxford, Oxford, UK Search for more papers by this author , Joshua P. Rosenthal Division of Epidemiology and Population Studies, Fogarty International Center, National Institutes of Health, Bethesda, Maryland, USA Search for more papers by this author , and HAPIN Investigators Search for more papers by this author Published:29 April 2020CID: 047008https://doi.org/10.1289/EHP6407Cited by:5AboutSectionsPDF ToolsDownload CitationsTrack Citations ShareShare onFacebookTwitterLinked InRedditEmail AbstractBackground:Globally, nearly 3 billion people rely on solid fuels for cooking and heating, the vast majority residing in low- and middle-income countries (LMICs). The resulting household air pollution (HAP) is a leading environmental risk factor, accounting for an estimated 1.6 million premature deaths annually. Previous interventions of cleaner stoves have often failed to reduce exposure to levels that produce meaningful health improvements. There have been no multicountry field trials with liquefied petroleum gas (LPG) stoves, likely the cleanest scalable intervention.Objective:This paper describes the design and methods of an ongoing randomized controlled trial (RCT) of LPG stove and fuel distribution in 3,200 households in 4 LMICs (India, Guatemala, Peru, and Rwanda).Methods:We are enrolling 800 pregnant women at each of the 4 international research centers from households using biomass fuels. We are randomly assigning households to receive LPG stoves, an 18-month supply of free LPG, and behavioral reinforcements to the control arm. The mother is being followed along with her child until the child is 1 year old. Older adult women (40 to <80 years of age) living in the same households are also enrolled and followed during the same period. Primary health outcomes are low birth weight, severe pneumonia incidence, stunting in the child, and high blood pressure (BP) in the older adult woman. Secondary health outcomes are also being assessed. We are assessing stove and fuel use, conducting repeated personal and kitchen exposure assessments of fine particulate matter with aerodynamic diameter ≤2.5μm (PM2.5), carbon monoxide (CO), and black carbon (BC), and collecting dried blood spots (DBS) and urinary samples for biomarker analysis. Enrollment and data collection began in May 2018 and will continue through August 2021. The trial is registered with ClinicalTrials.gov (NCT02944682).Conclusions:This study will provide evidence to inform national and global policies on scaling up LPG stove use among vulnerable populations. https://doi.org/10.1289/EHP6407IntroductionBackgroundGlobally, nearly 3 billion people rely on solid fuels (wood, dung, coal, charcoal, or agricultural crop waste) for cooking and heating (Bonjour et al. 2013). These fuels are often burned in inefficient and poorly ventilated combustion devices (e.g., open fires, traditional stoves). The resulting household air pollution (HAP) accounts for an estimated 1.6 million premature deaths per year and 59.5 million disability-adjusted life-years (GBD 2018). Despite progress in recent years, this largely preventable exposure remains a leading risk factor for morbidity and mortality worldwide. Poor populations in low- and middle-income countries (LMICs) bear most of this burden (GBD 2018).Several studies have documented association between HAP and multiple diseases or health conditions including chronic lung disease, lung cancer, cancers of the aerodigestive tract, cervical cancer in adults, pediatric acute lower respiratory infections (ALRI), or pneumonia, low birth weight, stillbirth, preterm birth, childhood stunting (short length for age), tuberculosis, and impaired cognitive development (Bruce et al. 2015; Smith et al. 2014; Gordon et al. 2014; Quansah et al. 2017; Thakur et al. 2018). However, except for pediatric ALRI, current methods used for the global HAP-attributable disease burden calculations are focused primarily on cardiovascular disease (CVD) and chronic respiratory diseases in adults using integrated exposure–response functions drawn substantially from estimates of effects of other sources of air pollution (Burnett et al. 2014). Additionally, adverse birth outcomes such as low birth weight and preterm birth, along with childhood development and growth, are not included in the global burden of disease estimates. Therefore, the current estimate for the burden of disease related to HAP is uncertain and likely underestimated.The state of the science illustrates several compelling reasons to undertake a multicountry randomized controlled trial (RCT) for reducing HAP using a clean fuel intervention. We will address several knowledge gaps in this study. First, liquefied petroleum gas (LPG) is currently the most widely available clean fuel in LMICs (IEA 2017), but to date, no LPG trials have been conducted that demonstrate significantly improved health outcomes among children and adults. Currently available cleaner-burning biomass combustion stoves are unlikely to achieve or sustain health-relevant exposure reductions (Clark et al. 2013; Bruce et al. 2015; Anenberg et al. 2013; Sambandam et al. 2015). This trial will provide needed evidence from a randomized LPG stove intervention to support policy formulation on national levels. Second, focusing on a combination of child (birth weight, pneumonia, and stunting) and adult cardiovascular [blood pressure (BP)] primary outcomes is strategic for public health goals in LMICs, as these conditions contribute the most to HAP-associated health burdens in these settings (GBD 2018). Although HAP has been identified as a risk factor for these outcomes, intervention efforts directed at birth outcomes, child health, and noncommunicable disease risk have been included in a few RCTs aimed at reducing HAP. Furthermore, we will evaluate biomarkers and other indicators that are known to predict noncommunicable disease occurrence and/or severity across the lifespan. Third, establishing exposure–response relationships across a range of personal exposures to HAP is needed to close critical gaps in our current understanding of exposure to disease relationships (Steenland et al. 2018). Exposure–response relationships are also critical for transferability of trial results across settings and for benchmarking future intervention efforts. Finally, most stove intervention studies have failed to adequately investigate and address the behaviors necessary to overcome concurrent use of polluting stoves (a practice known as stacking) to ensure consistent and sustained use of cleaner stoves and displacement of polluting ones (Rosenthal et al. 2017).This paper summarizes the rationale, study design, and methods of the Household Air Pollution Intervention Network (HAPIN) trial, a recently launched RCT that seeks to provide the evidence necessary for policy makers to determine what health benefits that can be achieved by implementing a scalable (in many areas) intervention aimed directly at reducing HAP in LMICs. The trial represents the first multicountry RCT to assess the effect of a stove intervention LPG on exposure to HAP and on a broad range of maternal, child, and adult health outcomes.Study AimsThe study has three specific aims. The first is to determine the effect of a randomized LPG stove and fuel intervention on health in four diverse biomass-using LMIC populations across the world using a common protocol. We hypothesize that compared to pregnant women (18 to <35 years of age) in control households (n=1,600), those who receive LPG stoves and fuel (n=1,600) will have offspring with increased birth weight, reduced severe pneumonia incidence, and improved growth [less stunting, defined as length-for-age z-score less than 2 standard deviations (SD) below the median z-score based on World Health Organization (WHO) child growth standards] up to 12 months of age. We also hypothesize that compared to control households, older adult women (40 to <80 years of age) living in households that receive LPG stoves and fuel will have reduced BP. In addition to these primary outcomes, the study will assess multiple secondary outcomes on mothers, infants, and older adult women.The second aim is to evaluate the exposure–response associations for HAP and health outcomes in four diverse LMIC populations. Using repeated 24-h personal and indirect measurements of exposure to fine particulate matter with aerodynamic diameter ≤2.5μm (PM2.5), carbon monoxide (CO), and black carbon (BC), we will characterize the exposure–response associations for all primary and secondary outcomes (assessing potential nonlinearity) while adjusting for confounders. While evidence for an overall effect of the intervention will be available from the first aim, analysis of exposure–response is critical for quantitative risk assessment and policy determinations of acceptable levels of HAP regardless of the cooking technologies and/or fuels in use.The third aim is to evaluate the extent to which biomarkers of exposure and health effects, including targeted and exploratory (e.g., metabolomics) analyses, are associated with intervention status or exposure. We hypothesize that participants residing in households that receive LPG stoves and fuel or have lower levels of exposure to HAP will have lower levels of carcinogenic polycyclic aromatic hydrocarbons and volatile organic compounds, such as urinary 1-OH-pyrene, 2-naphthol, 9-phenanthrene, as well as of chronic disease indicators, such as inflammatory, endothelial, inflammatory, oxidative stress, and glycemic control/diabetes biomarkers [e.g., C-reactive protein (CRP), endothelin-1, E-selectin, interleukin 6, and hemoglobin A1c (HbA1c)] when compared to participants in control households.By comparing an LPG stove and continuous free fuel intervention with standard cooking practices (typically traditional solid biomass in these settings), the HAPIN trial will provide an estimate of a potentially achievable level of HAP reduction and the associated impact on select maternal, infant, and adult health outcomes. Establishing exposure–response relationships in LMIC settings will allow for estimates of the range of improvement that can be expected across real-world conditions where clean stoves and fuels are often combined with traditional biomass stoves.MethodsStudy OverviewThe study is an RCT of an LPG stove and continuous fuel distribution intervention and promotion of its exclusive use among 3,200 households in four LMICs (India, Guatemala, Peru, and Rwanda). Following an 18-month period of planning, piloting, and formative research, the study began recruiting participants in May 2018 and is expected to complete enrollment in February 2020; follow-up data collection will continue through August 2021. In each country, eligible pregnant women are recruited and their households randomly assigned to intervention and control groups on a 1:1 ratio, and they are followed for ∼18 months until their newborn child is 1 year old. Intervention households receive a free LPG stove and free unlimited supply of LPG for the 18-month follow-up period. Control group households do not receive an LPG stove and fuel during the study period, and it is anticipated that they will continue cooking with solid biomass fuels during the trial. After enrollment, assessments will be made on a regular schedule over the course of the pregnancy (baseline, 24–28 wk gestation, 32–36 wk gestation), at 3 months of age, 6 months of age, and 12 months of age for the child, and at the same time points for the older adult woman in the household (Table 1). Control group compensation is summarized below and described elsewhere (Quinn et al. 2019).Table 1 Schedule of exposure and outcome assessment.Table 1 has eight columns, child age (study time point); less than 20 weeks gestation (baseline); 24 to 28 weeks gestation (1 to 3 months post-randomization); 32 to 36 weeks gestation per birth (3 to 5 months post-randomization); approximately 3 months old (approximately 9 months post-randomization); approximately 6 months old (approximately 12 months post-randomization); approximately 9 months old (approximately 15 months post-randomization); and approximately 12 months old (approximately 18 months post-randomization).Child age (study time point)<20wk gestation (baseline)24–28 wk gestation (1–3 months postrandomization)32–36 wk gestation/birth (3–5 months postrandomization)∼3 months old (∼9 months postrandomization)∼6 months old (∼12 months postrandomization)∼9 months old (∼15 months postrandomization)∼12 months old (∼18 months postrandomization)PWCOPWCOPWCONMCONMCONMCONMCOPersonal exposure 24-h PM2.5, CO, BCXXXXXXXXXXXXXXX Urinary biomarkersXXXXXXXXXXXXPrimary outcomes Birth weightXa Severe pneumoniab StuntingX Blood pressureXXXXXXSecondary outcomes Maternal blood pressureXXXXXX Fetal growthXXX Child linear growth (continuous)XaXXXX Preterm birthXa Child developmentXXXX WHO severe pneumoniaa BARTXX CIMTXX SGRQXX SF-36XX Expenditures/time useXXX Chronic disease biomarkersXXXXXX Metabolomics/microRNAcXX Hemoglobin (anemia)XXXXXaCovariates SociodemographicsXXX Clinical historyXXXXXXXXXXXXXXX Weight/height/BMIXXXXXXXXX Diet/food securityXXXX IYCFXXXXBiospecimens collected UrineXXXXXXXXXXXX Dried blood spotsXXXXXXaXXXXXXXNote: BART, brachial artery reactivity testing; BC, black carbon; BMI, body mass index; C, child; CIMT, carotid intima-media thickness; CO, carbon monoxide; IYCF, infant and young child feeding practices; NM, new mother; O, older adult woman in household; PM2.5, fine particulate matter with aerodynamic diameter ≤2.5μm; PW, pregnant woman; SF-36, Short Form 36 survey; SGRQ, St. George Respiratory Questionnaire.aMeasured at birth.bRecorded whenever children present to HAPIN health facilities with respiratory symptoms.cMetabolomics/microRNA biomarker discovery in 100/site for the older adult woman and child.Study Settings and Formative ResearchThe study is being conducted across four LMIC settings in which large portions of the population use solid biomass as the primary fuel type. To increase generalizability, the settings were purposefully selected to represent a diversity of characteristics expected to influence intervention effects, including altitude, population density, cooking practices, baseline pollution levels, and sources of pollution other than cooking (Table 2). Other factors that may influence intervention effects, such as fuel types, dwelling characteristics, and socioeconomic conditions, are being measured and recorded. Within each country, candidate sites were selected after evaluation in formative research over 12 months. The formative research consisted of four phases: a) initial scoping to identify potentially suitable sites and developing contextually grounded behavior change strategies to promote intervention adoption, b) pilot intervention to determine the HAP exposure contrast that might be expected from the intervention, c) pilot assessment to test trial procedures and methods, and d) respiratory rate/pulse oximetry assessment to define context-specific tachypnea and oxyhemoglobin saturation thresholds in the study sites.Table 2 Summary of key characteristics of international research centers by country based on sampling, government information, or published studies.Table 2 has five columns, country, India, Guatemala, Peru, and Rwanda.CountryIndiaGuatemalaPeruRwandaState/province(s)District(s)Tamil Nadu, selected blocks:Kalrayan Hills (Villupuram District); Kezhvelur, Keelayur, Vedaranyam, and Thalaignayiru (Nagapattinam District)Jalapa municipality:Regions: Santa Maria Xalapan, Ladinos PardosDepartment of Puno:Provinces of Puno, San Roman, Azángaro, Huancané, El Collao, and ChicuitoEastern Province:Kayonza District:Selected sectors: Kabare, Kabarondo, MuramaAltitude (m above sea level)10–165871–2,6773,8251,300–1,700Population density (per km2)892–1,09823317.6274Cooking practices (stoves, fuel, location)Traditional plastered clay/mud stoves fueled with biomass, with 90% cooking indoors. Use of chimneys is negligible.Chimney stoves and open fires; 97% wood use; cooking mainly indoorsRural households use traditional biomass (typically dung)-burning stoves daily for cookingTraditional three-stone fires (62.6%) or Rondereza (34.6%) fueled with wood (89.9%) or charcoal (8.1%), cooking indoors (71.9%)Baseline 24-h PM2.5 concentration (μg/m3)Mean:Kitchen area: 210Personal: 155Median (Q1–Q3):Personal: 115 (80–265)Kitchen: 349 (244–524)Bedroom: 25 (17–120) Patio: 30 (16–50)Median indoor: 130Mean 24 h:Personal: 329Kitchen: 437Outdoor: 44Other sources of air pollutionIncense, mosquito coils, etc. (55% of households), trash burning (38% of households)Trash burning and seasonal burning crop residue (February–April)Low ambient air pollutionMinimal unless near road or in more urban/dense area.Prevalence of smoking in the homeWomen rarely smoke. Other smokers reported in 24% of households but smoking inside house reported in less than 1% of households.Women rarely smoke (<1% ever smoked). Secondhand smoke in 20% of homes, but limited to 1–2 cigarettes/daySelf-reported daily smoking is 0.2%.Province level:Never: 76.4%Daily: 21.9%Note: Data are from Guatemala: Fujisada H et al. 2012; Government of the Republic of Guatemala 2019; Johnson M et al. 2018. Peru: Pollard SL et al. 2014; Jaganath D et al. 2015; Hollada J et al. 2017. India: Balakrishnan et al. 2013, 2018. Rwanda: National Institute of Statistics of Rwanda, 2010, 2012; unpublished data from Kirby et al. 2016; HAPIN formative research results (unpublished). PM2.5, fine particulate matter with aerodynamic diameter ≤2.5.Eligibility Criteria, Screening, and RecruitmentStudy teams led by experienced local investigators work in collaboration with clinics and community health workers in each country to identify candidate pregnant women. To be eligible to participate in the study, a pregnant woman must meet the following inclusion criteria: confirmed pregnancy (human chorionic gonadotropin–positive blood or urine test); 18 to <35 years of age (confirmed by government-issued ID, whenever possible), cooks primarily with biomass stoves, lives in the study area, 9 to <20wk gestation with a viable singleton pregnancy with normal fetal heart rate confirmed by ultrasound, continued pregnancy at the time of randomization (via self-report), and agrees to participate with informed consent. Eligible pregnant women are excluded if they currently smoke cigarettes or other tobacco products, plan to move permanently outside the study area in the next 12 months, use a clean fuel stove predominantly, or are likely to use LPG or another clean fuel predominantly in the near future. Ultrasound measurements are conducted by trained personnel (who are also additionally certified centrally) in a clinic or home setting to determine eligibility and assess fetal growth using a portable ultrasound [Edge (Edge Ultrasound System), Sonosite/Fujifilm Edge (FUJIFILM SonoSite Inc.)].Across the trial locations, up to 800 older adult women 40 to <80 years of age (confirmed by government-issued ID whenever possible) who reside in the same households as an enrolled pregnant woman are being recruited (one per household), provided they do not fall within the following exclusion criteria: currently smoking cigarettes or other tobacco products, pregnant (via self-report), or planning to permanently move out of current household in the next 12 months.Baseline Surveys and Assessments; RandomizationFollowing recruitment and informed consent, a baseline visit is made to the household by a trained fieldworker to conduct surveys and other assessments. This baseline visit includes a survey that covers a range of topics like cooking behaviors, household composition, socioeconomic and demographic information, housing characteristics, and pregnancy-related information. Pregnant women are also surveyed about their medical and gynecological history, including medication use. Separate questionnaires assess physical activity, dietary diversity, household food insecurity, and household expenditures.The baseline visit includes assessments of health, biomarkers, and exposure for both the pregnant woman and older adult woman. For pregnant women, a trained fieldworker or nurse measures resting BP (model HEM-907XL; Omron®) in triplicate and maternal weight (seca 876/874 scales; Seca) and height (seca 213 stadiometer; Seca) in duplicate. Among older adult women, respiratory symptoms are assessed using the Saint George’s Respiratory Questionnaire (SGRQ) ( http://www.healthstatus.sgul.ac.uk/sgrq), and quality-of-life measures are assessed using the Short Form 36 (SF-36) ( https://www.rand.org/health-care/surveys_tools/mos/36-item-short-form.html). Further, carotid intima-media thickness (CIMT) and, in Peru, only due to its exploratory nature, brachial artery reactivity testing (BART) are measured among older adult women using the ultrasound devices described above. Common carotid artery (CCA) ultrasound will be performed as a marker of atherosclerosis using a high-resolution linear transducer to image the distal 1-cm CCA region (just proximal to the bifurcation) (Stein et al. 2008). CIMT will be obtained from end-diastolic B-mode images as the average of the posterior wall segments from both right and left CCAs, measured using an automated system with an edge detection algorithm and manual override capacity (100 separate dimensional measurements are obtained from the 1-cm segment and averaged to obtain mean and maximal CIMT values). BART will be performed with a portable ultrasound to assess endothelial function using a high-resolution linear transducer to image the brachial artery (BA) above the antecubital fossa to measure rest diameter (Corretti et al. 1995). The BA will be occluded for 5 min by a BP cuff inflated to suprasystolic levels in the forearm; after 5 min of occlusion, the BP will be released, and the BA diameter will be imaged every 30 s for 2.5 min after cuff deflation to calculate percent BA dilation after hyperemia.Urine samples (first morning void) and dried blood spots (DBS) via finger prick are obtained from all participating pregnant women and older adult women. Finally, pregnant women and the older adult women are monitored for 24-h personal and household level exposure to HAP (PM2.5, CO, and BC) using the procedures described below. Baseline data on ambient temperature and humidity in the home and primary cooking area are also collected by the data loggers of the PM2.5 samplers, and dimensions of the kitchen are measured by the surveyor.After the baseline surveys and assessments are completed, households are randomly assigned to intervention or control arms, stratified by country. In India and Peru, additional stratified randomization is used to ensure a balance between discrete geographical regions within the study area. In Rwanda and Guatemala, the study areas are deemed homogenous so that such further stratification is not necessary.InterventionThe intervention consists of an LPG stove, a continuous supply of LPG fuel delivered to the homes for 18 months, as well as education and behavioral messages (described in future papers) to promote safe, exclusive use of the LPG stove for cooking. Stoves are procured in each country and vary (footprint, base, burner size, nob position, and griddle) based on local cooking practices; however, all include at least two burners and meet applicable safety requirements. The intervention (stove and fuel) is provided free of charge to all intervention households after baseline measurements are conducted. On each visit to provide additional fuel cylinders in Rwanda, Guatemala, and Peru, stove condition is examined, any necessary repairs performed, and the weight of LPG cylinders measured and recorded in order to help monitor use and anticipate the need for additional refills. In India, per national governmental regulations, the public sector oil marketing company is responsible for stove installation, cylinder refills, and repairs; study staff facilitate those activities for intervention participants. The rate of LPG usage is monitored by calculating average kilograms of LPG used per household member per day (using data on fuel cylinder weights and the number of days between installation and exchange of each cylinder).Control CompensationControl households receive compensation designed to meet three aims. First, it must comply with applicable ethics requirements for treatment of controls. Second, we are compensating control participants for the burden associated with this study, with the view of minimizing losses to follow-up. Third, we offset the economic advantage to intervention households accorded by the provision of free stoves and fuel. While the details vary across the four countries, compensation was designed based on a uniform set of trial-wide principles that address the above aims, with details informed by focus group discussions in the communities selected for the intervention. Controls receive either an LPG stove and a supply of fuel at the end of the trial or preferred alternatives of comparable value during or at the end of the trial. Details concerning the development of the compensation strategy are provided elsewhere (Quinn et al. 2019).Primary OutcomesOur primary health outcomes are birth weight, severe pneumonia in the first 12 months of life, stunting at 12 months of age, and BP in the ol
Referência(s)