Table 3.

Characteristics of studies that focused on physical pipe integrity.^a

Study	Location	Source water	Treatment plant	Distribution system	Design	Exposure	Comparison exposure	Outcome assessment	Effect estimate (95% CI)
D’Argenio et al. 1995	Italy	Not reported	Not reported	Presumed to be continuously operated; some chlorine residual; total coliforms, fecal coliforms, and fecal streptococci detected in affected pipe segment	Cohort	Pipeline with fecal contamination from cross-connections	Pipeline with no fecal contamination	Self-report	RR = 2.67 (1.16, 6.11)b
Mohanty et al. 2002	India	Surface water	Conventional treatment with chlorination	Intermittently operated; inadequate residual; total and fecal coliforms detected	Ecological	Unit increase in percentage of cast-iron pipes in service zone	NA	Self-report	Regression coefficient, –0.42 (p = 0.10)
Yassin et al. 2006	Palestine	Groundwater	Chlorination	Intermittently operated; inadequate residual; fecal contamination detected more often than at the source	Cross-sectional	Network > 1 year old	Network ≤ 1 year old	Self-report	RR = 1.51 (0.80, 2.83)b
Abu Amr and Yassin 2008	Palestine	Groundwater	Chlorination	Intermittently operated; inadequate residual; fecal contamination detected more often than at the source	Cross-sectional	Network > 1 year old	Network ≤ 1 year old	Self-report	RR = 1.03 (0.68, 1.56)b
Nygård et al. 2004	Sweden	Surface and groundwater	Chlorination (for surface water only)	Presumed to be continuously operated; low-level residual	Ecological	10-m increase in pipe length per person in municipality	NA	Surveillance records	IDR = 1.12 (1.08, 1.16)
Tinker et al. 2009	USA	Not reported	Not reported	Presumed to be continuously operated; adequate residual	Ecological	ZIP code with long hydraulic residence time	ZIP code with intermediate hydraulic residence time	Emergency department records	OR = 1.06 (1.04, 1.08)c
Abbreviations: IDR, incidence density ratio; NA, not applicable; OR, odds ratio; RR, relative risk. aResults not pooled. bCalculated from data reported in the study. cPooled from estimates for two utilities reported in the study.