The frequency of replacement of dental restorations may vary based on a number of variables, including type of material, size of the restoration, and caries risk of the patient.

SUBJECTS: The authors analyzed the dental records of 2780 Navy (cohort 1 = 1078 entered the Navy in 1997) and US Marine Corps recruits (cohort 2 =1053 entered the USMC in 1999-2000; cohort 3 = 649 entered the USMC in 2002-2005). The records were reviewed at 16 US Navy dental treatment facilities at the following time periods: cohort 1, 2001; cohort 2, 2002-2003; and cohort 3, 2005-2006. The mean age of the subjects was 20 years, and 85% were men. Only posterior teeth (not third molars) with amalgam or resin-based composite (including glass ionomer restorations) were evaluated. Teeth that had been restored with more than one material and restorations that did not involve the occlusal surface were excluded. The minimum follow-up time was 2 years with at least 2 periodic exams following the initial exam. KEY EXPOSURE/STUDY FACTOR: The primary factor of interest was the type of restorative material (amalgam versus resin-based composite). Secondary factors included tooth number, number of restored surfaces (single or multiple), and caries risk of the patient. Caries risk status was defined using the Navy Dental Corps Oral Disease Risk Management protocol. MAIN OUTCOME MEASURE: The primary outcome measure of interest was the determination of the relative risk of replacement of an initially intact restoration during the subject's first years of military service. Restorations were classified as clinically acceptable or requiring replacement either as a result of new primary caries, secondary caries, defective restorations, or endodontic therapy. MAIN
RESULTS: At the initial exam, 964 (15.2%) of the amalgam restorations and 199 (17.4%) of the resin-based composites required replacement and were excluded from further analysis. Of the remaining restorations, an additional 14.2% of the amalgam and 16.7% of the composite restorations required replacement during the observation period. The mean follow-up time was 3.0 years (cohort 1, 3.4 years; cohort 2, 3.1 years; cohort 3, 2.3 years). Replacement rates for resin-based composite restorations compared with amalgam were significantly higher owing to all causes (adjusted hazard ratio [HR], 1.28; P < .05) and for replacement owing to restoration failure (adjusted HR, 1.64; P < .01). Multiple surface restorations demonstrated higher rates of replacement than single surface restorations from all causes (adjusted HR, 1.39; P < .01) and for replacement of existing restorations (adjusted HR, 1.82; P < .01). High-caries-risk subjects experienced more than twice the risk of retreatment than did low-caries-risk subjects when considering all replacements (adjusted HR, 2.04; P < .01) and 50% higher risk of replacement of previously restored surfaces (adjusted HR, 1.48; P<.01)
CONCLUSIONS: Approximately 30% of all posterior restorations required replacement either at the initial or subsequent exams during the observation period. The number of resin-based composite restorations requiring replacement was significantly higher than amalgam restorations. The authors concluded that because of the extra cost, time, and potential for increased frequency of replacement, posterior composite restorations should be limited to restorations of appropriate size and placed under meticulous restorative technique with strict adherence to manufacturer's instructions.