Benjamin F. Arnold

Household Environmental Conditions Are Associated with Enteropathy and Impaired Growth in Rural Bangladesh

We assessed the relationship of fecal environmental contamination and environmental enteropathy. We compared markers of environmental enteropathy, parasite burden, and growth in 119 Bangladeshi children (≤ 48 months of age) across rural Bangladesh living in different levels of household environmental cleanliness defined by objective indicators of water quality and sanitary and hand-washing infrastructure. Adjusted for potential confounding characteristics, children from clean households had 0.54 SDs (95% confidence interval [CI] = 0.06, 1.01) higher height-for-age z scores (HAZs), 0.32 SDs (95% CI = −0.72, 0.08) lower lactulose:mannitol (L:M) ratios in urine, and 0.24 SDs (95% CI = −0.63, 0.16) lower immunoglobulin G endotoxin core antibody (IgG EndoCAb) titers than children from contaminated households. After adjusting for age and sex, a 1-unit increase in the ln L:M was associated with a 0.33 SDs decrease in HAZ (95% CI = −0.62, −0.05). These results are consistent with the hypothesis that environmental contamination causes growth faltering mediated through environmental enteropathy.

Cluster-randomised controlled trials of individual and combined water, sanitation, hygiene and nutritional interventions in rural Bangladesh and Kenya: the WASH Benefits study design and rationale.

Introduction Enteric infections are common during the first years of life in low-income countries and contribute to growth faltering with long-term impairment of health and development. Water quality, sanitation, handwashing and nutritional interventions can independently reduce enteric infections and growth faltering. There is little evidence that directly compares the effects of these individual and combined interventions on diarrhoea and growth when delivered to infants and young children. The objective of the WASH Benefits study is to help fill this knowledge gap. Methods and analysis WASH Benefits includes two cluster-randomised trials to assess improvements in water quality, sanitation, handwashing and child nutrition—alone and in combination—to rural households with pregnant women in Kenya and Bangladesh. Geographically matched clusters (groups of household compounds in Bangladesh and villages in Kenya) will be randomised to one of six intervention arms or control. Intervention arms include water quality, sanitation, handwashing, nutrition, combined water+sanitation+handwashing (WSH) and WSH+nutrition. The studies will enrol newborn children (N=5760 in Bangladesh and N=8000 in Kenya) and measure outcomes at 12 and 24 months after intervention delivery. Primary outcomes include child length-for-age Z-scores and caregiver-reported diarrhoea. Secondary outcomes include stunting prevalence, markers of environmental enteropathy and child development scores (verbal, motor and personal/social). We will estimate unadjusted and adjusted intention-to-treat effects using semiparametric estimators and permutation tests. Ethics and dissemination Study protocols have been reviewed and approved by human subjects review boards at the University of California, Berkeley, Stanford University, the International Centre for Diarrheal Disease Research, Bangladesh, the Kenya Medical Research Institute, and Innovations for Poverty Action. Independent data safety monitoring boards in each country oversee the trials. This study is funded by a grant from the Bill & Melinda Gates Foundation to the University of California, Berkeley. Registration Trial registration identifiers (http://www.clinicaltrials.gov): NCT01590095 (Bangladesh), NCT01704105 (Kenya).

Solar Drinking Water Disinfection (SODIS) to Reduce Childhood Diarrhoea in Rural Bolivia: A Cluster-Randomized, Controlled Trial

Daniel Maeusezahl and colleagues conducted a cluster-randomized controlled trial in rural Bolivia of solar drinking water disinfection, and find only moderate compliance with the intervention and no evidence of reduction in diarrhea among children.

Evaluation of a pre-existing, 3-year household water treatment and handwashing intervention in rural Guatemala

BACKGROUND The promotion of household water treatment and handwashing with soap has led to large reductions in child diarrhoea in randomized efficacy trials. Currently, we know little about the health effectiveness of behaviour-based water and hygiene interventions after the conclusion of intervention activities. METHODS We present an extension of previously published design (propensity score matching) and analysis (targeted maximum likelihood estimation) methods to evaluate the behavioural and health impacts of a pre-existing but non-randomized intervention (a 3-year, combined household water treatment and handwashing campaign in rural Guatemala). Six months after the intervention, we conducted a cross-sectional cohort study in 30 villages (15 intervention and 15 control) that included 600 households, and 929 children <5 years of age. RESULTS The study design created a sample of intervention and control villages that were comparable across more than 30 potentially confounding characteristics. The intervention led to modest gains in confirmed water treatment behaviour [risk difference = 0.05, 95% confidence interval (CI) 0.02-0.09]. We found, however, no difference between the intervention and control villages in self-reported handwashing behaviour, spot-check hygiene conditions, or the prevalence of child diarrhoea, clinical acute lower respiratory infections or child growth. CONCLUSIONS To our knowledge this is the first post-intervention follow-up study of a combined household water treatment and handwashing behaviour change intervention, and the first post-intervention follow-up of either intervention type to include child health measurement. The lack of child health impacts is consistent with unsustained behaviour adoption. Our findings highlight the difficulty of implementing behaviour-based household water treatment and handwashing outside of intensive efficacy trials.

Using rapid indicators for Enterococcus to assess the risk of illness after exposure to urban runoff contaminated marine water

BACKGROUND Traditional fecal indicator bacteria (FIB) measurement is too slow (>18 h) for timely swimmer warnings. OBJECTIVES Assess relationship of rapid indicator methods (qPCR) to illness at a marine beach impacted by urban runoff. METHODS We measured baseline and two-week health in 9525 individuals visiting Doheny Beach 2007-08. Illness rates were compared (swimmers vs. non-swimmers). FIB measured by traditional (Enterococcus spp. by EPA Method 1600 or Enterolert™, fecal coliforms, total coliforms) and three rapid qPCR assays for Enterococcus spp. (Taqman, Scorpion-1, Scorpion-2) were compared to health. Primary bacterial source was a creek flowing untreated into ocean; the creek did not reach the ocean when a sand berm formed. This provided a natural experiment for examining FIB-health relationships under varying conditions. RESULTS We observed significant increases in diarrhea (OR 1.90, 95% CI 1.29-2.80 for swallowing water) and other outcomes in swimmers compared to non-swimmers. Exposure (body immersion, head immersion, swallowed water) was associated with increasing risk of gastrointestinal illness (GI). Daily GI incidence patterns were different: swimmers (2-day peak) and non-swimmers (no peak). With berm-open, we observed associations between GI and traditional and rapid methods for Enterococcus; fewer associations occurred when berm status was not considered. CONCLUSIONS We found increased risk of GI at this urban runoff beach. When FIB source flowed freely (berm-open), several traditional and rapid indicators were related to illness. When FIB source was weak (berm-closed) fewer illness associations were seen. These different relationships under different conditions at a single beach demonstrate the difficulties using these indicators to predict health risk.

Epidemiological methods in diarrhoea studies—an update

Background Diarrhoea remains a leading cause of morbidity and mortality but is difficult to measure in epidemiological studies. Challenges include the diagnosis based on self-reported symptoms, the logistical burden of intensive surveillance and the variability of diarrhoea in space, time and person. Methods We review current practices in sampling procedures to measure diarrhoea, and provide guidance for diarrhoea measurement across a range of study goals. Using 14 available data sets, we estimated typical design effects for clustering at household and village/neighbourhood level, and measured the impact of adjusting for baseline variables on the precision of intervention effect estimates. Results Incidence is the preferred outcome measure in aetiological studies, health services research and vaccine trials. Repeated prevalence measurements (longitudinal prevalence) are appropriate in high-mortality settings where malnutrition is common, although many repeat measures are rarely useful. Period prevalence is an inadequate outcome if an intervention affects illness duration. Adjusting point estimates for age or diarrhoea at baseline in randomized trials has little effect on the precision of estimates. Design effects in trials randomized at household level are usually <2 (range 1.0–3.2). Design effects for larger clusters (e.g. villages or neighbourhoods) vary greatly among different settings and study designs (range 0.1–25.8). Conclusions Using appropriate sampling strategies and outcome measures can improve the efficiency, validity and comparability of diarrhoea studies. Allocating large clusters in cluster randomized trials is compromized by unpredictable design effects and should be carried out only if the research question requires it.

Optimal Recall Period for Caregiver-reported Illness in Risk Factor and Intervention Studies: A Multicountry Study

Many community-based studies of acute child illness rely on cases reported by caregivers. In prior investigations, researchers noted a reporting bias when longer illness recall periods were used. The use of recall periods longer than 2-3 days has been discouraged to minimize this reporting bias. In the present study, we sought to determine the optimal recall period for illness measurement when accounting for both bias and variance. Using data from 12,191 children less than 24 months of age collected in 2008-2009 from Himachal Pradesh in India, Madhya Pradesh in India, Indonesia, Peru, and Senegal, we calculated bias, variance, and mean squared error for estimates of the prevalence ratio between groups defined by anemia, stunting, and underweight status to identify optimal recall periods for caregiver-reported diarrhea, cough, and fever. There was little bias in the prevalence ratio when a 7-day recall period was used (<10% in 35 of 45 scenarios), and the mean squared error was usually minimized with recall periods of 6 or more days. Shortening the recall period from 7 days to 2 days required sample-size increases of 52%-92% for diarrhea, 47%-61% for cough, and 102%-206% for fever. In contrast to the current practice of using 2-day recall periods, this work suggests that studies should measure caregiver-reported illness with a 7-day recall period.

Causal inference methods to study nonrandomized, preexisting development interventions

Empirical measurement of interventions to address significant global health and development problems is necessary to ensure that resources are applied appropriately. Such intervention programs are often deployed at the group or community level. The gold standard design to measure the effectiveness of community-level interventions is the community-randomized trial, but the conditions of these trials often make it difficult to assess their external validity and sustainability. The sheer number of community interventions, relative to randomized studies, speaks to a need for rigorous observational methods to measure their impact. In this article, we use the potential outcomes model for causal inference to motivate a matched cohort design to study the impact and sustainability of nonrandomized, preexisting interventions. We illustrate the method using a sanitation mobilization, water supply, and hygiene intervention in rural India. In a matched sample of 25 villages, we enrolled 1,284 children <5 y old and measured outcomes over 12 mo. Although we found a 33 percentage point difference in new toilet construction [95% confidence interval (CI) = 28%, 39%], we found no impacts on height-for-age Z scores (adjusted difference = 0.01, 95% CI = −0.15, 0.19) or diarrhea (adjusted longitudinal prevalence difference = 0.003, 95% CI = −0.001, 0.008) among children <5 y old. This study demonstrates that matched cohort designs can estimate impacts from nonrandomized, preexisting interventions that are used widely in development efforts. Interpreting the impacts as causal, however, requires stronger assumptions than prospective, randomized studies.

Simulation methods to estimate design power: an overview for applied research

BackgroundEstimating the required sample size and statistical power for a study is an integral part of study design. For standard designs, power equations provide an efficient solution to the problem, but they are unavailable for many complex study designs that arise in practice. For such complex study designs, computer simulation is a useful alternative for estimating study power. Although this approach is well known among statisticians, in our experience many epidemiologists and social scientists are unfamiliar with the technique. This article aims to address this knowledge gap.MethodsWe review an approach to estimate study power for individual- or cluster-randomized designs using computer simulation. This flexible approach arises naturally from the model used to derive conventional power equations, but extends those methods to accommodate arbitrarily complex designs. The method is universally applicable to a broad range of designs and outcomes, and we present the material in a way that is approachable for quantitative, applied researchers. We illustrate the method using two examples (one simple, one complex) based on sanitation and nutritional interventions to improve child growth.ResultsWe first show how simulation reproduces conventional power estimates for simple randomized designs over a broad range of sample scenarios to familiarize the reader with the approach. We then demonstrate how to extend the simulation approach to more complex designs. Finally, we discuss extensions to the examples in the article, and provide computer code to efficiently run the example simulations in both R and Stata.ConclusionsSimulation methods offer a flexible option to estimate statistical power for standard and non-traditional study designs and parameters of interest. The approach we have described is universally applicable for evaluating study designs used in epidemiologic and social science research.

Swimmer illness associated with marine water exposure and water quality indicators: Impact of widely used assumptions

Background: Studies of health risks associated with recreational water exposure require investigators to make choices about water quality indicator averaging techniques, exposure definitions, follow-up periods, and model specifications; however, investigators seldom describe the impact of these choices on reported results. Our objectives are to report illness risk from swimming at a marine beach affected by nonpoint sources of urban runoff, measure associations between fecal indicator bacteria levels and subsequent illness among swimmers, and investigate the sensitivity of results to a range of exposure and outcome definitions. Methods: In 2009, we enrolled 5674 people in a prospective cohort at Malibu Beach, a coastal marine beach in California, and measured daily health symptoms 10–19 days later. Concurrent water quality samples were analyzed for indicator bacteria using culture and molecular methods. We compared illness risk between nonswimmers and swimmers, and among swimmers exposed to various levels of fecal indicator bacteria. Results: Diarrhea was more common among swimmers than nonswimmers (adjusted odds ratio = 1.88 [95% confidence interval = 1.09–3.24]) within 3 days of the beach visit. Water quality was generally good (fecal indicator bacteria levels exceeded water quality guidelines for only 7% of study samples). Fecal indicator bacteria levels were not consistently associated with swimmer illness. Sensitivity analyses demonstrated that overall inference was not substantially affected by the choice of exposure and outcome definitions. Conclusions: This study suggests that the 3 days following a beach visit may be the most relevant period for health outcome measurement in recreational water studies. Under the water quality conditions observed in this study, fecal indicator bacteria levels were not associated with swimmer illness.