preview

Outcome Data Will Be Analyzed For Any Subjects That Drop Out During The Course Of Study

Decent Essays
Outcome data will be analyzed using intention to treat analysis to account for any subjects that drop out during the course of study. Outcome measures of pain using NRS-11 and jaw ROM using the TheraBite ROM Scale will be assessed at baseline, immediately after the 2-week intervention, and at 2 week, 4 week, 6 week, and 8 week follow up periods. Although NRS-11 pain measures are technically classified as ordinal measures, most studies using NRS-11 as an outcome measure utilize parametric tests and consider the measures as interval or ratio data rather than ordinal measures.27 Therefore, assuming homogenous variance and a normal distribution, a parametric 2-tailed mixed-ANCOVA will be used as the primary statistical analysis. Prior to…show more content…
The ANCOVA will control for initial pretest differences by equating the two groups based on covariate scores and adjusting the posttest scores of pain and ROM appropriately. Once these posttest scores are determined, the adjusted posttest scores will be analyzed to see if there was a difference between groups. A significance level of α = 0.05 will be chosen to minimize the chance of a Type I error.
If a significant difference is found, Tukey’s test will be run as a post hoc test to see at which period in time the significant difference occurred. If data is not found to be normal or the variance is not homogenous, then a nonparametric 2-tailed Friedman’s ANOVA by ranks and a 2-tailed Kruskal-Wallis ANOVA by ranks will be run instead with the same significance level.
Sample size is estimated a priori based on a previous study analyzing the short-term effects of hypnosis on oral function and pain in subjects with TMD.33 Pain is chosen as the priority primary outcome measure in comparison to the other primary outcome measure, jaw ROM, as many TMD patients struggle with chronic pain, which diminishes their quality of life. Therefore, the posttest NRS-11 mean pain scores and standard deviations for the control (X1=3.9 and s1=1.5) and hypnosis group (X2=2.9 and s2=2.4)33 were utilized to first calculate effect size, d (Equation A.1). Because the standard deviations, s, were different between the two groups,
Get Access