Informations et ressources scientifiques
sur le développement des zones arides et semi-arides

Accueil du site → Master → Etats Unis → 2012 → Rural Kenyan Household Stored Water Quality

Emory University (2012)

Rural Kenyan Household Stored Water Quality

Dismer, Amber Marie

Titre : Rural Kenyan Household Stored Water Quality

Auteur : Dismer, Amber Marie

Université de soutenance : Emory University

Grade : Master 2012

Background : The Joint Monitoring Program recommended collecting presence absence tests of households’ stored water quality after 2015 for all future demographic health surveys, multiple cluster indicator surveys, and living standards measurement surveys. The main objective of this thesis is to provide guidance on collecting clustered water quality measurements for practitioners and health scientists to quantify gains in households’ and springs’ water quality. Methods : The Spring Improvement Project Household Survey’s cluster randomized controlled trial design in western Kenya provides a unique opportunity to test the presence absence method against the logarithmic MPN of E. coli, the World Health Organization’s drinking water risk-levels, and the geometric means method. Clustered analyses were performed for two sampling timeframes across months of the year and across bi-weekly rounds for households’ stored water quality. Similar analyses were conducted for the springs where water was collected. Results : Over 15 bi-weekly data collection rounds, households’ stored water quality did not differ across the months of the year or the bi-weekly round. The logarithmic MPN of E. coli method (F= 3.83, p=0.0001) and the presence absence test (Wald x2= 27.88, p=0.03) detected significant variability of springs’ water quality by bi-weekly round. The intra-cluster correlation coefficients (ICCs) for households’ stored water in spring clusters (logarithmic : 0.09 ; presence absence : 0.04) were smaller than the ICCs of clusters of springs’ water across time (logarithmic : 0.46 ; presence absence : 0.10). In multi-year rounds, monthly variability of households’ water quality was detected at the baseline by all four methods. However, only the logarithmic MPN of E. coli method and the WHO risk levels detected baseline differences of springs’ water quality by month. Conclusion : The presence absence test yielded the same results for households’ water quality as the logarithmic MPN of E. coli method ; however, more samples per cluster are required for water quality interventions.


Page publiée le 13 décembre 2013, mise à jour le 30 mars 2019