Table 3 Effects of intervention during follow-up on household-level drinking water quality outcomes.

From: Effects of adding household water filters to Rwanda’s Community-Based Environmental Health Promotion Programme: a cluster-randomized controlled trial in Rwamagana district

Model

Drinking water quality

Intervention

Control

PR (95% CI)

p

1a

≥2 CFU/100 mL (any detectable E. coli contamination)

69.9% (649/929)

87.0% (730/839)

0.80 (0.74, 0.87)

<0.001

2a

≥10 CFU/100 mL (Moderate and higher E. coli contamination)

49.3% (458/929)

74.7% (627/839)

0.66 (0.58, 0.75)

<0.001

3a

≥100 CFU /100 mL (Very high E. coli contamination)

22.4% (208/929)

39.8% (334/839)

0.56 (0.46, 0.68)

<0.001

4b

≥2 CFU/100 mL (any detectable E. coli contamination)

69.8% (644/923)

87.2% (728/835)

0.80 (0.74, 0.87)

<0.001

5b

≥10 CFU/100 mL (Moderate and higher E. coli contamination)

49.2% (454/923)

75.0% (626/835)

0.65 (0.57, 0.74)

<0.001

6b

≥100 CFU /100 mL (Very high E. coli contamination)

22.4% (207/923)

39.9% (333/835)

0.56 (0.46, 0.68)

<0.001

  1. n denotes the total number of household water samples analyzed in follow-up rounds.
  2. aPrevalence ratio (PR), 95% Confidence Interval (95% CI) and p value derived from log-binomial generalized estimating equations with robust standard errors to account for clustering within village. Model only conditions group assignment and drinking water quality outcome.
  3. bPR, 95% CI and p value derived from log-binomial generalized estimating equations with robust standard errors to account for clustering within village. Model further adjusts for government-defined socioeconomic status.