Predictive models of pain following root canal treatment: a prospective clinical study

Abstract

Aim: To determine the probability of the incidence, intensity, duration and triggering of post-endodontic pain, considering factors related to the patient (age, gender, medical evaluation) and to the affected tooth (group, location, number of canals, pulp vitality, preoperative pain, periapical radiolucencies, previous emergency access, presence of occlusal contacts with antagonist).

Methodology: A total of 500 one-visit root canal treatments (RCTs) were performed on patients referred to an endodontist. Shaping of root canals was performed manually with Gates-Glidden drills and K-Flexofiles, and apical patency was maintained with a size 10 file. A 5% NaOCl solution was used for irrigation, and canals were filled with lateral compaction and AH-Plus sealer. Independent factors were recorded during the treatment, and characteristics of post-endodontic pain (incidence, intensity, type and duration) were later surveyed through questionnaires. Of the 500 questionnaires, 374 were properly returned and split in two groups for two different statistical purposes: 316 cases were used to adjust the logistic regression models to predict each characteristic of post-endodontic pain using predictive factors, and the remaining 58 cases were used to test the validity of each model.

Results: The predictive models showed that the incidence of post-endodontic pain was significantly lower when the treated tooth was not a molar (P = 0.003), demonstrated periapical radiolucencies (P = 0.003), had no history of previous pain (P = 0.006) or emergency endodontic treatment (P = 0.045) and had no occlusal contact (P < 0.0001). The probability of experiencing moderate or severe pain was higher with increasing age (P = 0.09) and in mandibular teeth (P = 0.045). The probability of pain lasting more than 2 days was increased with age (P = 0.1) and decreased in males (P = 0.007) and when a radiolucent lesion was present on radiographs (P = 0.1).

Conclusions: Predictive formulae for the incidence, the intensity and the duration of post-endodontic pain were generated and validated taking account of the interrelation of multiple concomitant clinical factors. A predictive model for triggering post-endodontic pain could not be established.

Keywords: logistic regression; pain duration; pain intensity; post-endodontic pain; postoperative pain; predictive models.