A 20-year historical prospective cohort study of root canal treatments. A Multilevel analysis

Aim: To evaluate the number of healthy and functional root filled teeth of patients included in a recall programme for at least 20 years. Methodology: Teeth were root filled by a single specialist following manual canal instrumentation, lateral/vertical compaction of gutta-percha and restored with glass–ionomer cements and bonding system/composite resin. In a large percentage of teeth, a metal-ceramic crown was placed during follow-up. Patients included in the recall programme (n = 130) were blindly assessed both clinically and radiographically (every 2 years) to evaluate clinical symptoms and periapical status (PAI). The following variables were analysed: age, tooth location, tooth type, initial diagnosis, PAI, root filling length and coronal restoration type. Chi-square test and multilevel analysis were performed to detect variables associated with treatment functionality and disease/lesions (P < 0.05). A cumulative teeth survival curve was constructed by means of Kaplan–Meier using extractions as the end-point. Results: At the 20-year recall, 72 patients (31 M, 41 F; mean age 57.7 ± 8.29 years; 196 teeth) completed the follow-up. Thirty-six patients were excluded for medical complications or died before the end of the study. Drop-outs consisted of 22 patients (17%) who did not complete the follow-up. Single metal-ceramic crowns were positioned after 4–6 months in 40% of teeth. Composite restorations were replaced with single metal-ceramic crowns during the follow-up in 53% of teeth after 8–19 years. Of 196 teeth, 155 were classified as Survived (79%), 128 of which (65%) were Healthy (PAI ≤ 2). Thirty-nine teeth (20%) were extracted for nonendodontic reasons. Twenty-nine teeth (15%) were classified as: re-exacerbation (11 teeth; 5.6%) or persistent asymptomatic lesions (18 teeth; 9%). Only two re-exacerbated teeth were extracted. Multilevel analysis confirmed the clinical relevance of tooth type (P = 0.001) on Survived and healthy teeth (P = 0.007). Tooth location (P = 0.0045) and initial diagnosis (P = 0.019) significantly affected only Healthy teeth. Conclusions: Root filled teeth were more frequently extracted for non-endodontic reasons rather than for endodontic disease. The majority of teeth with adequate root fillings, adequate restorations and included in a recall programme remained functional and healthy for more than 20 years.