Table 2.
Prevalence of diarrhoea by water treatment intervention group among various age groups, western Kenya
|
Person level observations
|
Compound level observations*
|
|||||||
|---|---|---|---|---|---|---|---|---|
|
Person weeks of observation
|
Weeks with diarrhoea
|
Compounds under observation
|
Weeks with diarrhoea/100 weeks
|
Absolute difference in diarrhoea prevalence v control (%)
|
Absolute difference in diarrhoea prevalence v sodium hypochlorite (%)
|
|||
| Age group(intervention) | Crude | Adjusted (95% CI)† | Crude | Adjusted (95% CI)† | ||||
| Age <2 years | ||||||||
| Flocculant-disinfectant | 3381 | 251 | 171 | 7.65 | -21 | -25 (-40 to -5) | -2 | -9 (-29 to 16) |
| Sodium hypochlorite | 3151 | 258 | 169 | 7.81 | -19 | -17 (-34 to 4) | Ref | Ref |
| Control | 3467 | 342 | 176 | 9.64 | Ref | Ref | — | — |
| All ages | ||||||||
| Flocculant-disinfectant | 34 775 | 699 | 201 | 2.22 | -17 | -19 (-34 to -2) | 8 | 9 (-12 to 34) |
| Sodium hypochlorite | 36 438 | 675 | 203 | 2.06 | -23 | -26 (-39 to -9) | Ref | Ref |
| Control | 37 262 | 929 | 201 | 2.67 | Ref | Ref | — | — |
Ref=reference category.
*
Data aggregated and compared at the level of randomisation (compound level).
†
Compared by generalised linear model with log link binomial distribution and adjustment for overdispersion.