Caroliene Meijndert

48 Chapter 3 Statistical methods Statistical analysis was performed to determine if there was an association between the obtained measurements. A statistical software package was used (IBM Corp. Released 2015. IBM SPSS Statistics for Windows, Version 23.0. Armonk, NY: IBM Corp). Normality of the data was checked with the Shapiro Wilk test before determining the statistical test. Papilla index, plaque index and bleeding index, being ordinal variables, were presented as medians and interquartile range. Pocket probing depth, being not normally distributed, was presented as median and interquartile range. Bone level change and aesthetic evaluation, being normally distributed, were presented as means and standard deviations. Inter-group comparison (augmented versus not augmented) of non-parametric data (pocket depth, bone level change, PES, WES and VAS score of patient satisfaction) was calculated using the Mann-Whitney U test. Inter-group comparison of ordinal data (Papilla Index) was analysed using the Chi-square test. The null-hypotheses was that there was no difference in bone level change between the two subgroups. The between-time interval comparison for non-parametric continuous and ordinal data was calculated using the Wilcoxon signed-rank test. Results Patients A total of 50 patients (23 male, 27 female, mean age at time of inclusion was 37 years old) were available for evaluation after 5 years (T 60 ) of which 23 had been subjected to pre-implant reconstructive surgery. Ten patients had dropped out of the study because they either moved to another country (n=5) or moved without leaving an address (n=5). The assumption was made that not attending the evaluation was independent of the clinical or the radiographic condition as well as that it was independent of the patients’ satisfaction. Statistical analysis was done on the evaluated 50 patients. Implant and crown survival No implants were lost during the 5-year follow-up (implant survival 100%). At the time of this follow-up 24 patients showed signs of peri-implant mucositis and one patient was diagnosed with peri-implantitis. That patient was referred for treatment when this was diagnosed at the T 60 follow up. Crown survival was 98% (one crown fractured within the first year in function). In addition to this was the observation of porcelain chipping of