Table 5.
Multilevel multivariate logistic regression analyses of the association between interventions and receipt of ORS and zinc
| GUJARAT | ||||||||
|---|---|---|---|---|---|---|---|---|
|
Outcome: ORS use |
Outcome: ORS and zinc use |
|||||||
|
Model 1* aOR (95% CI) |
Model 2† aOR (95% CI) |
Model 3‡ aOR (95% CI) |
P-value for Model 3 |
Model 1* aOR (95% CI) |
Model 2† aOR (95% CI) |
Model 3‡ aOR (95% CI) |
P-value for Model 3 |
|
|
Main effects |
||||||||
| Survey round |
||||||||
| Early phase |
Ref |
Ref |
||||||
| Late phase |
0.33 |
0.31 |
0.17 |
<0.01 |
1.86 |
1.26 |
0.31 |
0.08 |
| (0.21, 0.53) |
(0.19, 0.49) |
(0.08, 0.34) |
(0.85, 4.08) |
(0.56, 2.83) |
(0.08, 1.15) |
|||
| District intervention category |
||||||||
| Light touch district |
Ref |
Ref |
||||||
| Focal district |
2.48 |
1.88 |
0.93 |
0.85 |
11.06 |
6.86 |
1.89 |
0.31 |
| (1.65, 3.72) |
(1.21, 2.90) |
(0.45, 1.93) |
(5.28, 23.19) |
(3.15, 14.95) |
(0.55, 6.49) |
|||
| Mass media exposure |
||||||||
| Not exposed to mass media campaign |
Ref |
Ref |
||||||
| Exposed to mass media campaign |
1.75 |
1.73 |
1.19 |
0.58 |
1.41 |
1.39 |
1.92 |
0.26 |
| (1.32, 2.32) |
(1.31, 2.29) |
(0.67, 2.07) |
(0.96, 2.06) |
(0.95, 2.04) |
(0.62, 5.93) |
|||
|
Interactions |
||||||||
| Survey round * Focal district |
N/A |
4.01 |
3.42 |
<0.01 |
N/A |
12.93 |
15.02 |
<0.01 |
| (1.68, 9.68) |
(1.39, 8.33) |
(2.80, 60.34) |
(2.97, 75.19) |
|||||
| Survey round * Focal district * Mass media |
N/A |
N/A |
1.66 |
0.12 |
N/A |
N/A |
0.69 |
0.56 |
| (0.86, 3.16) |
(0.21, 2.32) |
|||||||
|
Model fit - AIC (lower is better) |
2226.63 |
2218.78 |
2218.48 |
1539.06 |
1530.35 |
1532.00 |
||
|
UTTAR PRADESH | ||||||||
|
Outcome: ORS use |
Outcome: ORS and zinc use |
|||||||
|
Model 1* |
Model 2† |
Model 3‡ |
p-value |
Model 1* |
Model 2† |
Model 3‡ |
p-value |
|
|
aOR |
aOR |
aOR |
for |
aOR |
aOR |
aOR |
for |
|
|
(95% CI) |
(95% CI) |
(95% CI) |
Model 3 |
(95% CI) |
(95% CI) |
(95% CI) |
Model 3 |
|
|
Main effects |
||||||||
| Survey round |
||||||||
| Early phase |
Ref |
Ref |
||||||
| Late phase |
1.36 |
1.55 |
1.02 |
0.91 |
1.59 |
1.64 |
1.61 |
0.20 |
| (1.00-1.86) |
(1.12-2.17) |
(0.70-1.51) |
(0.89-2.84) |
(0.90-2.98) |
(0.77-3.38) |
|||
| District intervention category |
||||||||
| Light touch district |
Ref |
Ref |
||||||
| Focal district |
1.65 |
1.08 |
0.66 |
0.17 |
1.44 |
1.24 |
1.72 |
0.92 |
| (1.40-1.96) |
(0.79-1.49) |
(0.36-1.20) |
(1.13-1.84) |
(0.70-2.22) |
(0.34-3.25) |
|||
| Mass media exposure |
||||||||
| Not exposed to mass media campaign |
Ref |
Ref |
||||||
| Exposed to mass media campaign |
1.56 |
1.56 |
1.34 |
<0.01 |
1.94 |
1.95 |
1.54 |
<0.01 |
| (1.34-1.82) |
(1.34-1.82) |
(1.09-1.63) |
(1.53-2.47) |
(1.53-2.47) |
(1.12-2.11) |
|||
|
Interactions |
||||||||
| Survey round * Focal district |
N/A |
2.74 |
2.29 |
0.01 |
N/A |
1.37 |
1.02 |
0.97 |
| (1.45-5.16) |
(1.19-4.39) |
(0.44-4.35) |
(0.31-3.35) |
|||||
| Survey round * Focal district * Mass media |
N/A |
N/A |
1.38 |
0.02 |
N/A |
N/A |
1.57 |
0.04 |
| (1.04-1.84) |
(1.01-2.46) |
|||||||
| Model fit - AIC (lower is better) | 5778.20 | 5770.18 | 5767.13 | 3037.65 | 3039.34 | 3037.20 | ||
ORS – oral rehydration salts, aOR – adjusted odds ratio, AIC – Akaike’s information criterion
*Model 1 is a multivariate logistic regression model including all covariates.
†Model 2 is a multivariate logistic regression model including all covariates and a 2-way interaction between survey round and focal district interventions.
‡Model 3 is a multivariate logistic regression model including all covariates, a 2-way interaction between survey round and detailing, and a 3-way interaction between survey round, focal district interventions, and mass media exposure.
§All models are adjusted for child gender, child age, source of care, respondent education level, urban/rural, and household wealth status.