The authors aimed to determine whether the Chiari Severity Index (CSI), and other clinical variables, can be used as a predictor of postoperative outcomes for Chiari type I malformation (CM-I) using the modified Chicago Chiari Outcome Scale (mCCOS) as the postoperative measure.
The cohort included patients 18 years of age and younger who were treated for CM-I between 2010 and 2015 who had at least 12 months of clinical and radiographic follow-up. CSI grades were assigned using preoperative clinical and neuroimaging data. Clinical, radiographic, and operative data were obtained from medical records. Kruskal-Wallis tests and Spearman correlations were conducted to assess for differences among CSI grades. Linear and ordinal regressions were conducted to evaluate predictors of the mCCOS and its components. Statistical significance was set a priori at p < 0.05.
A total of 65 patients were included in the final cohort. The average age at the time of surgery and the mean mCCOS score were 9.8 ± 4.9 years and 10.4 ± 1.4, respectively. There were no significant differences in the mean mCCOS scores or CSI grades. Pre- and postoperative syrinx sizes were similar across the total patient cohort with median sizes of 7.4 and 3.7 mm, respectively. After controlling for age at the time of surgery, whether duraplasty and/or arachnoid dissection was performed, CSI preoperative score did not predict postoperative mCCOS score. No clinical variable could predict total mCCOS score. When the mCCOS was broken down into 3 subcomponents (pain, non-pain, and complications), only one relationship was identified. Those patients who presented with no headache had a statistically significant decrease in their pain (neck pain, shoulder pain, or dysesthesia in the upper extremities) as measured by the pain component of the mCCOS (χ2 [2, n = 20] = 6.43, p = 0.04). All other preclinical predictors, including CSI grades, were nonsignificant in demonstrating correlations to the mCCOS subcomponents.
CSI grade was not found to be a marker of surgical outcome as measured by the mCCOS in this study. There were no correlations between the clinical variables and covariates investigated with the mCCOS. The lack of variation in mCCOS scores across this cohort may suggest that the mCCOS is not adequate for detecting differences in postsurgical outcomes. Further investigation is warranted to make this determination.