Page 72 - Clinical relevance of current materials for cranial implants
P. 72
70
Chapter 3
Figure 1: Example of cranial defect circumference measurement; in this case 42.6 cm.
Statistical analysis
Multivariate stepwise binary logistical regression analyses were used to identify independent predictive factors for failure of autologous bone flaps. Possible predictive factors were derived from the literature. Non-significant factors were manually and sequentially removed until only significant parameters remained. Odds ratios (OR) and their 95% confidence intervals (95% CI) were determined for significant predictive parameters. Univariable analyses were conducted to detect any differences in patients with and without infection or resorption of the bone flap. Differences in continuous variables were expressed as mean differences (MD) with their 95% CIs, differences in dichotomous variables were presented as risk differences (RD) with their 95%Cis. A Number Needed to Treat (NNT) or Number Needed to Harm (NNH) was calculated is case of a significant RD. Statistical analysis was performed with IBM SPSS Statistics 24.0 (Armonk, NY, USA).
