Fourteen studies were located, but two were excluded because no data could be extracted. A total of 1241 patients were involved, 95 in 2 prospective studies, 104 in a prospective/retrospective combination study (no details on how many prospective vs retrospective), and 1042 in retrospective studies (131 in pediatric studies).
One study had no cases of meningitis in either group, hence an OR could not be calculated. For the remaining 11 studies, there were no significant differences between the OR's for individual studies (p=.443). The combined OR was 1.15 (95% CI 0.68, 1.94, p=0.678). A sensitivity analysis was conducted by removing a single study, re-conducting the analysis, replacing the study then removing the next study, and so on. No significant changes were found.
Results from combining the nine studies (n=547) which reported CSF leakage rates, none of the individual study OR's differed significantly (p=.278). The combined OR was 1.34 (95% CI 0.75, 2.41, p=.358). The sensitivity analysis again showed no significant changes.
In a second sub-set analysis of the three studies of children showed the combined OR was 1.04 (95% CI 0.07, 14.90, p=1.000).