Table 3. Efficacy outcomes: analyses of covariance.
| ANCOVAa for t1 score | ANCOVAa for t2 score | ||||||
|---|---|---|---|---|---|---|---|
| n | Regression co-efficient | Partial eta squared b | n | Regression co-efficient | Partial eta squared | ||
| ( 95% CI ) | ( 95% CI ) c | ( 95% CI ) | ( 95% CI ) b | ||||
| DCS | PDA | 23 | -3.5 (-12.6 to 5.6) | 0.014 (0.000 to 0.143) | — | ||
| No PDA | 23 | ||||||
| EPDS | PDA | 23 | 1.4 (-1.6 to 4.5) | 0.020 (0.000 to 0.158) | 21 | -1.3 (-4.1 to 1.5) | 0.021 (0.000 to 0.163) |
| No PDA | 23 | 23 | |||||
| STAI (state) | PDA | 23 | 5.1 (1.0 to 11.3) | 0.062 (0.000 to 0.230) | 20 | -3.8 (-11.4 to 3.8) | 0.025 (0.000 to 0.174) |
| No PDA | 23 | 23 | |||||
| Knowledge questionnaire | PDA | 23 | 0.13 (-0.86 to 1.1) | 0.002 (0.000 to 0.090) | – | ||
| No PDA | 23 |
aANCOVA model of effect of allocation, adjusted for baseline score of the relevant outcome measure. bStandardised effect sizes were estimated using partial eta squared (by convention, values of 0.01, 0.09, and 0.25 indicate small, medium, and large effect sizes respectively). cThe lower limit of the confidence interval was estimated as 0.0 for all analyses.
ANCOVA = analysis of covariance. CI = confidence intervals. DCS = decisional conflict scale. EPDS = Edinburgh postnatal depression scale. PDA = patient decision aid. STAI = state-trait anxiety inventory.