2.3 Data analysis and Statistical methods
SPSS® version 25 (IBM, Armonk, NY, USA) was used for statistical analysis. Independent T test was used to compare the means of 2 normally distributed groups (salivary vs neurogenic tumors). A Pearson correlation coefficient was calculated to determine the Correlation of tumor’ s tissue of origin and determinant factors for selecting surgical approach to the post-operative complications. Binary logistic regression analysis was performed to measure the odds of having neurological complications in salivary and neurogenic PPS tumors. A p-value <0.05 was significant. GraphPad Prism 8 (GraphPad Software, San Diego, California, USA) was used to draw the bar charts. The free available software RevMan 5.3 (Cochrane Collaborative, Oxford, England) was used for the meta‐analysis and creation of the forest plot. The risk ratios (RRs) of neurological complications and their 95% confidence intervals (CIs) were calculated for each study. Statistical heterogeneity was assessed using the Cochran Q  statistic (P  value for heterogeneity) and the I 2 statistic (total percentage of variation resulting from heterogeneity). In case of significant heterogeneity (I 2 ≥ 50) the random‐effect model was used, while the fixed‐effect model was used in absence of significant heterogeneity. Herein, we solely applied the fixed‐effect model to obtain RR, HR, 95% CI, and P ‐value.