Hedges’ g describes the precise differences between baseline and follow-up measures and should be interpreted just like Cohens d, but is additionally correcting for smaller sample sizes. Therefore, it is appropriate to analyze the single reductions or improvements by t-tests for depended samples and calculate the effect size Hedges’ g from these statistics. Our data strongly agreed with the predetermined statistical hypothesis. Third, the effect size to evaluate the meaningfulness of deviation-eta square-is rather global as well as less specific to interpret.
Second, in case of a significant ANOVA result, further analysis needs to be conducted to determine the exact nature of the difference. Because the effects described here are supposed to reduce negative outcome measures consequently over time (not randomly) and the primary results show very strong differences in dependent measures, an overall test of deviation seems inadequate for this analysis. First, an ANOVA tests for any possible difference in outcomes without specific assumptions on when outcomes should deviate from the overall mean or not. However, three reasons justify more adequate statistics. Many consider repeated measures analysis of variance (ANOVA) a standard analytical approach for the given experimental design. Appropriate lead position was determined through intraoperative device programming to confirm paresthesia overlap with the painful regions. In some rare cases, we performed preoperative intraforaminal test injections with local anesthetics. In most cases, the decision about the level to be stimulated is made intraoperatively based on the results of the test stimulation. Leads were advanced in an anterograde fashion and then directed into the intervertebral foramen near the DRG using curved stylets under fluoroscopic guidance. Leads were placed via an epidural approach, with access gained using standard loss-of-resistance technique. Thereafter, local anesthesia using a 1:1 mixture of 0.75% ropivacaine with 1% prilocaine to the operation field was performed. The operation field was prepared using standard practice of our clinic. Anxious patients received midazolam (1-3 mg) in combination with sufentanil (5-10 μg) intravenously as a premedication.
All procedures were performed under monitored anesthesia care.