Measure characteristic | Baseline | Fluoxetinea Endpoint | Change | Baseline | Placebob Endpoint | Change | Difference in Changec (95% CI) | pd | F |
---|
CDRS-R | 53.0 ± 2.32 | 34.60 ± 3.22 | -18.40 ± 2.94 | 53.94 ± 2.46 | 31.31 ± 3.42 | -22.63 ± 3.12 | -4.23 (-12.95–4.49) | .33 | 0.98 |
CGI-S | 4.28 ± 0.15 | 2.88 ± 0.28 | -1.39 ± 0.28 | 4.37 ± 0.16 | 2.87 ± 0.30 | -1.50 ± 0.30 | -0.11 (-0.95–0.73) | .71 | 0.07 |
CGI-I | -- | 2.61 ± 0.32 | -- | -- | 2.44 ± 0.34 | -- | -0.17 (-1.12–0.77) | .79 | 0.14 |
BDI | 17.20 ± 3.10 | 7.62 ± 1.96 | -9.58 ± 3.29 | 13.00 ± 3.33 | 8.12 ± 2.11 | -4.88 ± 3.54 | 4.70 (-5.23–14.63) | .34 | 0.95 |
BHS | 7.33 ± 1.38 | 4.07 ± 1.10 | -3.27 ± 1.38 | 7.69 ± 1.49 | 5.46 ± 1.18 | -2.23 ± 1.48 | 1.04 (-3.12–5.19) | .61 | 0.26 |
CGAS | 53.06 ± 2.06 | 69.63 ± 3.62 | 16.56 ± 3.50 | 51.21 ± 2.20 | 65.93 ± 3.87 | 14.71 ± 3.75 | 1.85 (-8.67–12.37) | .72 | 0.13 |
- Values represent mean ± SE from fixed effects parameter estimates. CDRS-R = Children's Depression Rating Scale-Revised; CGI-S = Clinical Global Impressions – Severity scale; CGI-I = Clinical Global Impressions – Improvement Scale; BDI = Beck Depression Inventory; BHS=Beck Hopelessness Scale; CGAS = Children's Global Assessment of Functioning
- aFluoxetine: last observation carried forward: for BDI, n = 15, for BHS, n = 15, for CGAS, n = 16.
- bPlacebo: last observation carried forward: for BDI, n = 13, for BHS, n = 13, for CGAS, n = 14.
- cDifference in Change shows the difference in average change score of fluoxetine-treated patients in comparison to placebo-treated patients. The 95% confidence intervals for the differences are shown in parentheses below. If the entire 95% confidence interval is greater than zero, this indicates a 95% or greater probability that the mean change associated with fluoxetine treatment is greater than the mean change associated with placebo.
- dGeneralized linear model (GLM) procedures were used for analysis of continuous baseline and endpoint values; Value for difference in mean change between treatment groups, using Type III difference of least square means.