Please activate JavaScript!
Please install Adobe Flash Player, click here for download
ePaper created 2018-02-14, 12:02:22 | version 1.38.0
P e r i i m p l a n t s o f t - t i s s u e a n d b o n e l e v e l s w i t h d i f f e r e n t i m p l a n t n e c k d e s i g n s Figs. 2a & b a Figs. 2c & d c b d Figs. 2a–d Adimensional measurements of marginal bone loss (a) at prosthesis placement and (b) after 3 years of follow-up. Adimensional measurements across the implant diameter with the objective of calibrating the bone level measurements, knowing the true width of the implant: calibration of (c) the prosthesis placement and (d) the 3-year follow-up radiographs. Marginal bone loss was measured with the soft- ware ImageJ (National Institute of Health, Md., U.S.) to process JPG files as obtained from intra- oral radiographs. Two reference points were marked on each implant at the implant– prosthesis interface and joined with a line rep- resenting height 0. Two vertical lines were traced perpendicular to the 0 line up to the first mesial and distal bone–implant contacts (Figs. 2a & b). Differences between these per- pendicular lines in radiographs taken at the dif- ferent time points (T0 and T1) were used to cal- culate bone loss. The highest difference value was chosen between the mesial and the distal values. A line was traced across the implant diameter (Figs. 2c & d) with the objective of cal- ibrating the periapical radiograph measure- ments, knowing the true width of the implant. S t a t i s t i c a l a n a l y s i s The principal predictor variable was the implant neck designs and neck surface treatments (group A and group B). The outcome variables of interest were periimplant tissue health and radiographic bone loss after 3 years of func- tional loading. A descriptive analysis of the parameters was performed. Sample distribution of bone loss was assessed, and due to lack of adjust- ment to normal distribution and dependence of observations, the corresponding nonpara- metric tests were applied: method for longitu- dinal data of Brunner and Langer, providing an analysis of variance statistic. Generalized esti- mating equations models were estimated to analyze the probability of the neck design affecting the various clinical variables through the Wald chi-squared statistic. For the variables BoP and presence of mucositis, a binary logis- tic regression model was estimated. For PPD and width of keratinized mucosa, an ordinal logistic regression model was estimated. The statistical analysis was performed using SPSS (statistical package for Microsoft Windows, Version 15.0, SPSS, Chicago, Ill., U.S.) and R software (Version 2.15.0, R Foundation for Sta- tistical Computing, Vienna, Austria). The Journal of Oral Science & Rehabilitation Volume 3 | Issue 4/2017 19