Achieving equitable uptake of handwashing and sanitation by addressing both supply and demand-based constraints: findings from a randomized controlled trial in rural Bangladesh.

BACKGROUND: Supply driven programs that are not closely connected to community demand and demand-driven programs that fail to ensure supply both risk worsening inequity. Understanding patterns of uptake of behaviors among the poorest under ideal experimental conditions, such as those of an efficacy trial, can help identify strategies that could be strengthened in routine programmatic conditions for more equitable uptake. WASH Benefits Bangladesh was a randomized controlled efficacy trial that provided free-of cost WASH hardware along with behavior change promotion. The current paper aimed to determine the impact of the removal of supply and demand constraints on the uptake of handwashing and sanitation behaviors across wealth and education levels. METHODS: The current analysis selected 4 indicators from the WASH Benefits trial- presence of water and soap in household handwashing stations, observed mother's hand cleanliness, observed visible feces on latrine slab or floor and reported last child defecation in potty or toilet. A baseline assessment was conducted immediately after enrolment and endline assessment was conducted approximately 2 years later. We compared change in uptake of these indicators including wealth quintiles (Q) between intervention and control groups from baseline to endline. RESULTS: For hand cleanliness, the poorest mothers improved more [Q1 difference in difference, DID: 16% (7, 25%)] than the wealthiest mothers [Q5 DID: 7% (- 4, 17%)]. The poorest households had largest improvements for observed presence of water and soap in handwashing station [Q1 DID: 82% (75, 90%)] compared to the wealthiest households [Q5 DID: 39% (30, 50%)]. Similarly, poorer household demonstrated greater reductions in visible feces on latrine slab or floor [Q1DID, - 25% (- 35, - 15) Q2: - 34% (- 44, - 23%)] than the wealthiest household [Q5 DID: - 1% (- 11, 8%). For reported last child defecation in potty or toilet, the poorest mothers showed greater improvement [Q1-4 DID: 50-54% (44, 60%)] than the wealthier mothers [Q5 DID: 39% (31, 46%). CONCLUSION: By simultaneously addressing supply and demand-constraints among the poorest, we observed substantial overall improvements in equity. Within scaled-up programs, a separate targeted strategy that relaxes constraints for the poorest can improve the equity of a program. TRIAL REGISTRATION: WASH Benefits Bangladesh: ClinicalTrials.gov , identifier: NCT01590095 . Date of registration: April 30, 2012 'Retrospectively registered'.

Quantitative assessment of fecal contamination in multiple environmental sample types in urban communities in Dhaka, Bangladesh using SaniPath microbial approach.

Rapid urbanization has led to a growing sanitation crisis in urban areas of Bangladesh and potential exposure to fecal contamination in the urban environment due to inadequate sanitation and poor fecal sludge management. Limited data are available on environmental fecal contamination associated with different exposure pathways in urban Dhaka. We conducted a cross-sectional study to explore the magnitude of fecal contamination in the environment in low-income, high-income, and transient/floating neighborhoods in urban Dhaka. Ten samples were collected from each of 10 environmental compartments in 10 different neighborhoods (4 low-income, 4 high-income and 2 transient/floating neighborhoods). These 1,000 samples were analyzed with the IDEXX-Quanti-Tray technique to determine most-probable-number (MPN) of E. coli. Samples of open drains (6.91 log10 MPN/100 mL), surface water (5.28 log10 MPN/100 mL), floodwater (4.60 log10 MPN/100 mL), produce (3.19 log10 MPN/serving), soil (2.29 log10 MPN/gram), and street food (1.79 log10 MPN/gram) had the highest mean log10 E. coli contamination compared to other samples. The contamination concentrations did not differ between low-income and high-income neighborhoods for shared latrine swabs, open drains, municipal water, produce, and street foodsamples. E. coli contamination levels were significantly higher (p <0.05) in low-income neighborhoods compared to high-income for soil (0.91 log10 MPN/gram, 95% CI, 0.39, 1.43), bathing water (0.98 log10 MPN/100 mL, 95% CI, 0.41, 1.54), non-municipal water (0.64 log10 MPN/100 mL, 95% CI, 0.24, 1.04), surface water (1.92 log10 MPN/100 mL, 95% CI, 1.44, 2.40), and floodwater (0.48 log10 MPN/100 mL, 95% CI, 0.03, 0.92) samples. E. coli contamination were significantly higher (p<0.05) in low-income neighborhoods compared to transient/floating neighborhoods for drain water, bathing water, non-municipal water and surface water. Future studies should examine behavior that brings people into contact with the environment and assess the extent of exposure to fecal contamination in the environment through multiple pathways and associated risks.

Bias-reduced and separation-proof GEE with small or sparse longitudinal binary data.

Generalized estimating equation (GEE) is a popular approach for analyzing correlated binary data. However, the problems of separation in GEE are still unknown. The separation created by a covariate often occurs in small correlated binary data and even in large data with rare outcome and/or high intra-cluster correlation and a number of influential covariates. This paper investigated the consequences of separation in GEE and addressed them by introducing a penalized GEE, termed as PGEE. The PGEE is obtained by adding Firth-type penalty term, which was originally proposed for generalized linear model score equation, to standard GEE and shown to achieve convergence and provide finite estimate of the regression coefficient in the presence of separation, which are not often possible in GEE. Further, a small-sample bias correction to the sandwich covariance estimator of the PGEE estimator is suggested. Simulations also showed that the GEE failed to achieve convergence and/or provided infinitely large estimate of the regression coefficient in the presence of complete or quasi-complete separation, whereas the PGEE showed significant improvement by achieving convergence and providing finite estimate. Even in the presence of near-to-separation, the PGEE also showed superior properties over the GEE. Furthermore, the bias-corrected sandwich estimator for the PGEE estimator showed substantial improvement over the standard sandwich estimator by reducing bias in estimating type I error rate. An illustration using real data also supported the findings of simulation. The PGEE with bias-corrected sandwich covariance estimator is recommended to use for small-to-moderate size sample (N ≤ 50) and even can be used for large sample if there is any evidence of separation or near-to-separation.