Fransen, R. (Ruud) (2014) De voorspelbaarheid van antrumperforaties. thesis, Dentistry.
![[img]](https://umcg.studenttheses.ub.rug.nl/style/images/fileicons/text.png) |
Text
FransenR.pdf Restricted to Registered users only Download (570kB) |
Abstract
During an extraction of an upper (pre) molar there is a chance of creating an oroantral communication. The aim of this study was to clarify the use of a panoramic radiograph in predicting the occurrence of oroantral communication in case of extraction. This research was performed by measuring 7 characteristics on an OPG, representing the relationship of the (pre)molar with the maxillary sinus. In retrospect patients who have experienced an oroantral communication due to an extraction were analyzed including their OPG. Subsequently a matched control group was formed which has been analyzed too. The differences between both groups concerning the 7 characteristics were analyzed statistically. It was aimed to give the characteristics a predicting power as an independent variable on the outcome of the dependent variable which is whether an oroantral communication will occur following extraction. This predicting power is given as oddsratio’s that were formed by univariate en multiple regression analysis. After a correlation analysis the characteristics were dichotomized. Hereafter a second correlation analysis was performed. Almost all the characteristics appeared to correlate with each other. This correlation reduced after the dichotomization. Before this transformation was performed the characteristics were analyzed in a univariate logistic regression analysis. Not every characteristic appeared to be significant. After the dichotomization every characteristic, except one, appeared to be significant. By performing a forward selection procedure 3 significant characteristics, A_aantmm_wortels_in_antrum”, “B_aantmm_restwortel” en “C_bodemhoogte”, were incorporated in the multiple model with oddsratio’s of 0,149, 0,154 and 3,271. In conclusion it can be stated that the dichotomized characteristics that make up the multiple model stand for big oddsratio’s. However the correlation of these characteristics should be kept in mind.
| Item Type: | Thesis (Thesis) |
|---|---|
| Supervisor name: | Stegenga, Prof. dr. B. |
| Faculty: | Medical Sciences |
| Date Deposited: | 25 Jun 2020 11:06 |
| Last Modified: | 25 Jun 2020 11:06 |
| URI: | https://umcg.studenttheses.ub.rug.nl/id/eprint/2552 |
Actions (login required)
 |
View Item |