Data sources: The databases Medline/PubMed, EMBASE, the Cochrane Library, Chinese Biomedical Literature Database and the China National Knowledge Infrastructure were searched. Additional hand searches were conducted in the journals.
Study selection: Randomised controlled trials, prospective cohort studies and retrospective studies with follow-up of 36 months or longer were included.
Data extraction and synthesis: Data were extracted independently in duplicate. The annual core and veneer fracture rates of various tooth types were estimated and compared using Poisson regression.
Results: 37 studies were included; two RCTs, 25 prospective cohorts and 10 retrospective studies. Based on the calculated results, all-ceramic crowns had an acceptable overall five-year fracture rate of 4.4% irrespective of the materials used. Five-year fracture rates were significantly higher for molar crowns (8.1%) compared to premolar crowns (3.0%), and the difference between anterior (3.0%) and posterior crowns (5.4%) also achieved significance. Core fracture rates had a five-year incidence of 2.5%, and a significantly higher core fracture rate was found in the posterior region (3.9%). The overall five-year incidence of veneer fracture was 3.0%, and no clear difference was found between restored tooth types, with incidences of 2.0%, 2.5%, 1.0%, and 3.0% for incisor, canine, premolar, and molar crowns, respectively.
Conclusions: Within the limitations of this study's protocol, the current evidence suggests that dental ceramic materials demonstrated acceptable five-year core and veneer fracture incidences when used for tooth-supported single crowns in both anterior and posterior segments. Higher fracture tendency for posterior crowns was the trend for all-ceramic crowns, while molar crowns showed a significantly higher fracture rate than premolar crowns. Randomised controlled trials with large sample sizes be undertaken to obtain more definitive results.