A total of 388 mandibular third molars were removed (a98 fron the ringt side and 190 from the left side of the mount). All information available was collected for these 388 molars, but for some of the parameters information was missing (see Tables I, II and III).
Risk indicators for extended operation time
Operation time was 8 minutes or less for 25% of the patients, 10 minutes orless for 90% of the operations. Operation time was 31 minutes or longer for 5% of the operations. The was 31 minutes or longer for 5% of the operations. The logistic regression analyses revealed risk indicators for an extended operation time using the preoperative variables (Table IV). Patients in the youngest age group (18.0-23.0 years) were found to have a lower risk of an extended operation timeolder patients: that is, patients 23.1-24.5 year (OR 1.94: P = .043), patients aged 24.6-26.5 years (OR 2.25; P = .014), and patients older than 26.5 years (OR 2.60; P = .007) had a 2 to 2.5 times higher risk for an operation time above 10 minutes.
When looking at the radiographic findings, 3 variables were found to increase the risk for longer operation time. If the molar in question washorizontally positioned, it was more likely that the operation would take an extended period of time than if vertically positioned (OR 2.33; P = .014), if the molar was found to have 2 roots there was a risk of a longer operation time than if judged to have I root complex (OR 2.01; P = .014), and if the molar was found to be in close proximity to the mandibular canal, it was at higher risk of delonger operating time than if judged to be not close to the canal (OR 2.09; P = .003).
Risk indicators for severe pain (VAS) 4 hours postoperatively
Some patients (8.7%) did not mark on the VAS since they had no pain 4 hours postoperatively, 25% de patients marked pain as 50 or less on the VAS, 50% marked 56 or less, and 90% marked 70 or less; 5% marked 75 or higher on the VAS. Two logisticregression models were made, one with the preoperative information (Table V) and one with the operative information as the independent variables (Table VI). In the preoperative model, only root morphology was a significant indicator for severe pain perception since molars with curved roots were more likely to cause a high VAS score compared with molars with straight root (OR 1.98; P = .024).
Twooperative findings related to the molar were indicators for a high VAS score. If the removed molar had been semi-impacted, the patients scored higher on the VAS than if the molar had been fully impacted (OR 2.14 P= .025=. if the mandibular nerve had been visible after the removal of the molar, it was more likely that the patient scored high on the VAS than if the nerve had not been visible (OR 3.19; P =.030=.
Risk indicators for postoperative pain
In 59 (15.2%) cases, the patient experienced more severe pain than she or he had expected during the postoperative week. Postoperative pain was recorded as a subjective statement from the patient at the removal of the sutures 1 week after surgery. As before, 2 models were made, 1 with the preoperative information (Table VII and 1 with the operativeinformation as the independent variables (Table VIII). Males were found to have a lower risk for reporting more severe postoperative pain compared with females (OR 0.31; / OR 0.35; P = .002 in the models. If a patient had a molar with radiographically recorded curved roots, there was a lower risk that this patient would report pain than a patient whose molar root(s) were straight (OR 0.37, P =.02).
One operative variable was found to be an indicator for reported postoperative pain. When the mandibular nerve was visible after removal, there was a higher risk for the patient to experience severe pain than if the nerve had not been visible (OR 3.01; P = .032).
Risk indicators for postoperative general infection
In 11 (2.8%) cases postoperative general infection was recorded (Table IX...