Rural Kenyan Household Stored Water Quality Open Access

Dismer, Amber Marie (2012)

Permanent URL: https://etd.library.emory.edu/concern/etds/dz010q65n?locale=en
Published

Abstract


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.

Table of Contents

Table of Contents
1. INTRODUCTION....................................................... 1
1.1 PATHWAYS OF SPREADING DISEASE.......................... 2
1.2 FACTORS THAT IMPACT THE SPREAD OF DISEASE........ 2
1.3 MEASURING WATER QUALITY.................................... 3
1.4 RESEARCH GAP...................................................... 5
1.5 INTRA-CLUSTER CORRELATION COEFFICIENT............... 6
1.6 OBJECTIVES.......................................................... 6
1.7 FOCAL STUDY: THE SPRING IMPROVEMENT PROJECT

HOUSEHOLD SURVEY AND BI-WEEKLY MONITORING............ 8
1.8 KENYAN CONTEXT.................................................. 8
1.9 LITERATURE REVIEW............................................... 9
1.9.1 Health Impacts attributed to poor water,

sanitation, and hygiene................................................ 9
1.9.2 Measuring water quality....................................... 10
1.9.3 Indicators of fecal pathogens................................ 10
1.9.4 Millennium Development Goals' safe water indicator... 13
1.9.5 Presenting statistics............................................ 15
1.9.6 JMP's New Guidelines........................................... 16
1.9.7 Source variation................................................. 16
1.9.8 Seasonality........................................................ 17
1.9.9 Human Behavior.................................................. 18
1.9.10 Relevant Studies............................................... 20
Table 1: Studies included in the Literature Review............. 21
1.9.11 Conclusion........................................................ 23
2. METHODS............................................................. 25
2.1 SPRING IMPROVEMENT PROJECT................................ 25
Figure 1: Spring Improvement Project study region............ 26
2.2 SPRING SELECTION................................................. 26
2.3 HOUSEHOLD SELECTION FOR SIP-H STUDY.................. 27
2.4 SIP-H SURVEY ROUNDS............................................ 27
Table 2: SIP Spring & Household Selection........................ 28
2.5 SIP-H INTERVENTIONS............................................. 28
2.5.1 Spring protection.................................................. 28
2.5.2 WaterGuard......................................................... 29
Figure 2: Spring Improvement Project Design..................... 29
2.6 CRITERIA FOR SAMPLES USED IN THIS ANALYSIS.......... 29
2.6.1 Bi-weekly Monitoring Sample (BWM sample)............... 30
2.6.2 Sub-sample......................................................... 31
Table 3: Viable samples from unprotected springs............... 31
& non-chlorinating unprotected spring household users........ 31
2.7 DATA.................................................................... 31
2.7.1 Survey data........................................................ 31
2.7.2 Lab data............................................................ 33
2.8 ANALYTICAL METHODS: ANALYSIS PLAN..................... 34
2.8.1Summary Statistics............................................... 34
Table 4: Analysis Plan.................................................. 34
2.8.2 Logarithmic MPN of E. coli..................................... 36
2.8.3 WHO drinking water risk levels................................ 37
2.8.4 Geometric means of E. coli..................................... 38
2.8.5 Presence absence tests........................................ 39
2.8.5 Intra-cluster correlation coefficients........................ 40
2.8.6 Power and sample size.......................................... 42
2.9 LIMITATIONS......................................................... 43
3. RESULTS................................................................ 44
3.1 BI-WEEKLY MONITORING ROUND RESULTS OVERVIEW..... 44
Figure 3: BWM log MPN of E. coli of springs and households.. 44
Table 5: BWM Round..................................................... 46
Table 6: Sample size for the BWM round by month of the

year........................................................................... 46
3.1.1 Seasonality measured by Monthly Variability.............. 47
Figure 4: BWM log MPN of E. coli..................................... 48
Figure 5: WHO risk level by month.................................... 49
Figure 6: Geometric mean MPN of E. coli by month............... 50
Figure 7: Presence absence E. coli test by month 51
3.1.2 Seasonality measured by Bi-weekly Monitoring Round

Variability.................................................................... 52
Figure 8: Log MPN of E. coli by BWM round........................ 52
Figure 9: Geometric mean MPN of E. coli by BWM round........ 53
Figure 10: WHO risk level by BWM rounds.......................... 54
Figure 11: WHO risk level by BWM rounds.......................... 56
3.2 SUBSAMPLE OVERVIEW............................................. 56
Table 7: Sub-sample Statistics........................................ 58
Table 8: Sample size for the sub-sample by month of the

year........................................................................... 58
Figure 12: Sub-sample logarithmic MPN of E. coli by month... 60
Figure 13: Sub-sample geometric means of MPN of E. coli

by month.................................................................... 61
Figure 14: Sub-sample WHO risk levels of households'

stored drinking water by month....................................... 62
Figure 15: Subsample WHO risk levels of springs water by

month........................................................................ 64
Figure 16: Sub-sample presence absence test of

households by month..................................................... 66
Figure 17: Subsample presence absence test of springs

by month..................................................................... 67
3.3 INTRA-CLUSTER CORRELATION COEFFICIENTS (ICC)....... 68
Table 9: Intra-cluster Correlation Coefficients..................... 69
3.4 POWER AND SAMPLE SIZE CALCULATIONS..................... 70
Figure 18: Effective Sample Sizes..................................... 71
4. DISCUSSION............................................................ 73
4.1 HOUSEHOLDS' STORED WATER QUALITY....................... 73
4.2 SPRINGS' WATER QUALITY......................................... 75
4.3 LIMITATIONS........................................................... 77
5. CONCLUSION........................................................... 78
5.1 STRENGTHS............................................................. 78
5.2 CONCLUSIONS FOR PUBLIC HEALTH PRACTITIONERS

USING THE JMP RECOMMENDATIONS................................. 78
5.3 CONCLUSIONS FOR PUBLIC HEALTH RESEARCHERS...........80
REFERENCES................................................................ 82
APPENDIX..................................................................... 88

About this Master's Thesis

Rights statement
  • Permission granted by the author to include this thesis or dissertation in this repository. All rights reserved by the author. Please contact the author for information regarding the reproduction and use of this thesis or dissertation.
School
Department
Degree
Submission
Language
  • English
Research field
Keyword
Committee Chair / Thesis Advisor
Committee Members
Partnering Agencies
Last modified

Primary PDF

Supplemental Files