Evaluating Impact of Improved Floors on Health (SABABU)

Evaluating the Impact of an Improved Household Flooring Intervention on Enteric and Parasitic Infections in Rural Settings in the Counties of Bungoma and Kwale, Kenya

Kenya Medical Research Institute, International Centre of Insect Physiology and Ecology (ICIPE), Jomo Kenyatta University of Agriculture and Technology

The goal of this intervention study is to learn about the impact of household flooring on health in rural Kenya, and test whether providing an improved (cement stabilised, washable) floor improves the health of children and their care providers. The main questions the study aims to answer are: What is the effect of providing a sealed, washable floor on the prevalence of infections that cause diarrhoea, intestinal worms and sand flea infections? To what extent does the intervention reduce contamination of floors with pathogens within the home? What is its effect of the intervention on the wellbeing of caregivers and children? Over the course of a year, do the new floors remain undamaged, with no cracks? Do participants living with the new floors, and the masons that helped to install the floors, like them and feel they are practical and affordable? The study will involve a trial, where half of the recruited households will be randomly chosen to receive the new floor in addition to some support on how to care for the floor and keep it clean. The other half of households will not receive anything at first, but at the end of the research project will also receive a new floor. Before the new floors are installed, the investigators will make several assessments in all study households. These will include a survey to measure household characteristics; a stool survey, to measure how many people are infected with diarrhoea-causing microorganisms and parasitic worms; a jigger flea examination among children; wellbeing assessments among children and caregivers; and soil sampling to identify microorganisms on the floor of the household. When households receive the new floor, participants will have to move out of their house for up to 7 days during installation. Participants will also be asked to attend some group meetings to discuss ways of taking care of the floor and keeping it clean. Assessments will be repeated 12 months after the floor has been delivered, and additional interviews will be held with a small number of randomly selected participants. Throughout the 12 months following delivery of the intervention, investigators will make unannounced visits to households to check the condition of the floor. Participants will also be offered treatment for parasitic worm infections after assessments have been completed at the start and end of the project.

This study is a two-arm household cluster randomised controlled trial (RCT) evaluating the impact of an improved household flooring intervention on enteric and parasitic infections among participating households in two contrasting settings in western and coastal Kenya. The flooring intervention will involve retrofitting a cement-stabilised earth floor that is sealed, washable and durable and that covers the total interior floor space of a household dwelling. A key pathway through which an improved floor is expected to reduce exposure to enteric and parasitic infections is facilitating a more hygienic domestic environment. As such the proposed intervention would also include a behaviour change component aiming to promote sustained adoption of appropriate domestic hygiene behaviours. This proposed trial would be the first of its kind to comprehensively assess the effects of combining improved flooring technologies with tailored behaviour change programming on a wide range of parasitic and enteric outcomes, providing an important step towards the establishment of transformative, community-driven, integrated approaches to WASH-related disease control. Exploring these relationships across contrasting contexts helps ensure findings will be of relevance to settings outside Kenya where similar housing, WASH infrastructure and disease risk are found. Results from this trial will help guide global and national environmental health priorities, at a time when the WHO is re-evaluating global targets for NTD control and elimination beyond 2020. Beyond evaluating the intervention's effects on health outcomes there is a need to understand how practical the intervention is, and to assess its feasibility and acceptability among target communities, as these factors will affect how relevant the findings are to control programmes. Results from the formative research indicate that there may be high levels of heterogeneity in how household members interact with the floor and adapt their behaviours such as cooking, animal husbandry and sleeping, all of which may play an important role in mitigating the success of the intervention. As such a dedicated process evaluation will take place alongside the RCT to explore implementation fidelity, intervention acceptability, and how the intervention is integrated into households' daily routines. THE INTERVENTION * The first two activities will be conducted in the intervention arm at month 0, and in the control arm at the end of the study after all data collection is complete: A low-cost, cement-stablised floor shall be installed in each room of the dwelling (including kitchen area) to meet the following requirements: (i) non-absorbent, durable and smooth; (ii) possess good wear resistance; (iii) acceptable appearance; (iv) be affordable. The proposed structure will consist of a base layer made from compacted cement-stabilized murram and a final sand and cement mortar finish. All materials required to build the floor will be provided by the study. Floors will be installed by trained masons supervised by the investigators, with the support of additional laborers. Household members will not be expected to contribute to labour or costs of laying floors, but they will need to vacate their dwellings for up to 7 days whist floors are laid and cured. The logistics around this will be discussed in detail with community leaders, and trial participants, during initial community engagement activities. Group meetings facilitated by the investigators will be held periodically post intervention to allow households to provide peer support on routines or challenges relating to living with the new floor and to give space to allow mutually accepted norms and standards around floor cleaning and maintenance to be established among intervention households. These group meetings will be complimented by individual household meetings which will take place at 4 weeks and 8 weeks post intervention, which will serve to help households develop and adhere to plans around floor hygiene, personal storage, livestock housing, and cooking arrangements. Annual mass treatment for STH infections (400 mg albendazole) and treatment of tungiasis in those affected by heavy infections (at 0 and 12 months) according to county DoH recommendations will be provided in both study arms. METHODS * this study will take place in two study sites; one within Kwale county (Dzombo ward) and the other in Bungoma county (South Bukusu ward and Kabula ward). In each cluster (i.e., household) all residents will be sampled immediately before and twelve months post-installation of floors. Faecal samples will be collected from the sampled population and will be assessed via multiplex PCR for enteric infections (in those aged under 5 years) and via Kato Katz for STH infections (for those aged >1 year old). Additional clinical examinations will be performed for tungiasis on all children aged under 15 years immediately prior to installation of floors, and then at 12-months post installation. In addition to these primary outcomes, quality of life measures in enrolled children and their caregivers will be recorded immediately before and twelve months post-installation of floors, and environmental sampling will be conducted on floors and surfaces of all enrolled households 12 months post-installation of floors. Alongside the trial, a process evaluation will be undertaken to investigate intervention fidelity, acceptability, durability and practicality. After the endline assessments, all control households will be offered an improved floor. Sample size and power calculations are based on the primary outcomes (prevalence of enteric infections in children under 5 years of age; prevalence of at least STH infection in all household members over 1 year old; prevalence of tungiasis infection in children under fifteen years of age) and have been informed by existing data from Kenyan populations. These include data from the national school-based deworming programme, community-based tungiasis surveys, and the Global Enteric Multicenter Study (GEMS) study, which was a large case-control study of moderate to severe diarrhoea in children younger than 5 study that included Nyanza Province, Kenya. Enrolled individuals will be clustered within households and calculations are thus based on the principles of cluster randomised trials, assuming an ICC of 0.1 based on small cluster size. Effect sizes for tungiasis and STH are based on expert opinion of the smallest meaningful public health effect. Effect sizes for enteric pathogens are based on earlier WASH efficiency studies. Tungiasis prevalence will be evaluated per-site while data on STH and enteric infections will be pooled across-sites. The primary outcome for which the largest sample size is required is STH prevalence, measured in enrolled children and their caregivers. Expect STH prevalence is 15% in the control arm and 10% in the intervention arm. Assuming five enrolled participants per household and a 15% loss to follow up, 220 households per arm in total across two sites would provide 80% power to observe this difference at 0.05 significance. This sample size is also sufficient to detect at 80% power and 0.05 significance: (i) the expected difference in enteric infection risk in children <5 years old - assuming one <5 year old child per household, and an expected prevalence post-intervention of 70% in the control arm and 56% in the intervention arm; and the expected difference in tungiasis prevalence in children <15 years at a site level - assuming two children <15 per household and an expected prevalence post-intervention of 30% in the control arm and 15% in the intervention arm. Based on these estimates, the target sample size is 220 clusters (households) per arm - thus ensuring 220 children (aged <5 years) and 440 children (aged <15 years) per arm in both sites. * ANALYSIS * Analysis of the primary and secondary outcomes for this research question will be carried out on groups as randomised (intention-to-treat). Results will be presented as appropriate effects sizes with a measure of precision (95% CIs), using generalised estimating equations to account for clustering by household. Incidence of caregiver-reported gastrointestinal illness will be analysed using interrupted time series methods. Pre-specified faecal indicator bacteria and specific pathogens of interest will be quantified in each study arm. Generalised linear models with robust standard errors will be used to estimate differences in overall pathogen prevalence in the dwelling environment at endline. Data from the caregiver questionnaire will be used to quantify scores on the EQ5D wellbeing index (caregivers) and the EQ-5D-Y (children only). Generalised linear models with robust standard errors will be used to estimate differences in wellbeing scores between study arms at endline. Qualitative data. Data from in-depth interviews with caregivers will be used to explore different pathways through which the intervention has changed caregiver and child daily routines and if these changes have wrought any impact on wellbeing. Pre-identified themes to explore include caregiver self-efficacy, social status, pride, and availability of free time. Following transcription and translation, data will be coded and analysed thematically using a case-memo approach. Results will be triangulated with data from the quantitative caregiver wellbeing questionnaire.

Diarrhoeal Disease, Hookworm Infection, Ascaris Lumbricoides Infection, Trichuris Infection, Tungiasis

Replacement of rudimentary floors with an improved floor, Support for behaviour change through 'floor clubs'. Annual mass treatment for STH infections (400 mg albendazole) Treatment of tungiasis in those affected by heavy infections (at 0 and 12 months) according to county DoH recommendations

Annual mass treatment for STH infections (400 mg albendazole) Treatment of tungiasis in those affected by heavy infections (at 0 and 12 months) according to county DoH recommendations

Replacement of rudimentary floors with a cement stabilised floor in all rooms in the dwelling, combined with support for behaviour change through 'floor clubs'.

Pevalence of enteric infections (detected through PCR) in children under 5 years old will be assessed using cross-sectional stool surveys including all enrolled children under 5 years before installation of floors and 12 months after receiving the floors. Pathogens to be identified from those observed at baseline using a multipathogen panel on a subset of 100 samples.

Pevalence of tungiasis (detected through clinical examination of hands and feet) in children under 15 years old will be assessed using cross-sectional clinical assessment surveys including all enrolled children under 15 years before installation of floors and 12 months after receiving the floors.

Pevalence of at least one STH infection (hookworm, acsaris and trichuris infections; detected through kato katz) in all household members 12 months and older will be assessed using cross-sectional stool surveys including all enrolled people over 12 months of age before installation of floors and 12 months after receiving the floors.

Prevalence of gastrointestinal illness in children under 5 years - assessed in both study arms based on caregiver reported symptoms

Intensity of tungiasis and severity of acute and chronic tungiasis-associated pathology in children <15 years - assessed using clinical severity scores.

Quality of Life assessed in children ages 8 to 14 years in both arms using the standardised EQ5D-Y tool

Ascaris lumbricoides (roundworm) infection prevalence in all household members >1 year assessed using Kato Katz

T. trichiura (whipworm) infection prevalence in all household members >1 year assessed using Kato Katz

Environmental contamination for human-specific and animal faecal markers - assessed from dust/soil samples from household cooking areas and living rooms using PCR.

Contamination of floors with eggs, larvae, pupae and adults of T. penetrans - assessed through entomology soil surveys

Inclusion Criteria: Household with a child under 5 years of age that meets structural criteria (unimproved earthen flooring throughout, structurally sound), with members willing to temporarily relocate. Exclusion Criteria: Households that are intending to move within the next 12 months, or that have improved flooring in any rooms or are not structurally sound.