Sample size

The sample size calculation is based on the test of the research hypothesis that the mean difference in SCIM-SC in the MyndMove intervention group is better than CT control group. The primary measure of effect is the difference in function measured using SCIM-SC at 14 weeks. The criterion for significance (alpha) has been set at 0.05. The test is two tailed, which implies that a mean difference in either direction will be interpreted. The sample was calculated using the power procedure in SAS V.9.2. With the proposed sample size of 30 in each of the two groups (ie, assuming a 1:1 allocation ratio) (ie, total sample size of 60), the study will have power of at least 80% to yield a statistically significant result using t-test (assuming an intention-to-treat principle for the analysis) of the difference between mean SCIM-SC scores at 14 weeks adjusting for baseline SCIM-SC scores at alpha=0.05. It is important to note that using the assumption of a t-test is more conservative in that an analysis of variance (ANCOVA) will lead to better power. This computation assumes that SCIM-SC scores are normally distributed, the mean difference is 3 points and the common within-group SD is 4.05. These estimates are modified estimates from the pilot study,²⁴ which account for the type of intervention planned for in this study. The assumed minimal clinically important difference is considered to correspond to a substantially meaningful improvement on the SCIM-SC, approximately 3 points,^{25 26} and also represents a moderate effect of the intervention.