Table 3.
Statistical analysis of change from baseline in tactile threshold (g) over time in the mITT population
| Change from baseline | Comparison with negative control | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Visit | Study product | n | Baseline tactile threshold (mean ± SD) | Adjusted mean (SE)a | 95% CIa | p-valuea | Adjusted mean difference (SE)a | % Adjusted mean difference | 95% CIa | p-valueb |
| Week 4 | Test | 64 | 11.5 ± 2.33 | 9.22 (1.577) | 6.11, 12.33 | < 0.0001 | 1.80 (2.204) | 24.3 | –2.54, 6.16 | 0.6978 |
| Positive control | 64 | 12.1 ± 2.79 | 4.59 (1.577) | 1.48, 7.70 | 0.0040 | –2.82 (2.203) | –38.1 | –7.17, 1.52 | 0.1682 | |
| Negative control | 67 | 11.8 ± 2.71 | 7.42 (1.538) | 4.38, 10.45 | < 0.0001 | |||||
| Week 8 | Test | 65 | 16.63 (2.393) | 11.91, 21.35 | < 0.0001 | 4.81 (3.356) | 40.7 | –1.81, 11.43 | 0.8130 | |
| Positive control | 64 | 11.05 (2.412) | 6.30, 15.81 | < 0.0001 | –0.76 (3.370) | –6.4 | –7.41, 5.89 | 0.3937 | ||
| Negative control | 67 | 11.82 (2.353) | 7.18, 16.46 | < 0.0001 | ||||||
Test Sensodyne Sensitivity and Gum®, Negative control Crest Cavity Protection®, Positive control Sensodyne Repair and Protect®
ANCOVA analysis of covariance, CI confidence interval, mITT modified intent-to-treat, n number of observations, SD standard deviation, SE standard error
aAnalysis was performed using ANCOVA model with study product and baseline Schiff stratification as factors and baseline tactile threshold as a covariate
bP-value from van Elteren test. Positive % adjusted mean difference favours test/positive control