The effects of cavity preparation and lamination on bond strength and fracture of tooth-colored restorations in primary molars.

PURPOSE This in vitro study compared bond strength and fracture modes of tooth-colored restorations in 2 types of cavity preparations in human primary molars. METHODS Standardized Class II cavities (40 dovetail and 40 box-only preparations) in extracted human primary molars were restored with packable composite resin (PC), resin-modified glass ionomer cement (RMGIC), resin-modified glass ionomer/packable composite resin laminate (RMGIC/PC), or resin-modified glass ionomer/packable composite resin laminate with an experimental bonding agent, K-14 (RMGIC/K-14/PC). The ultimate load at fracture was measured on marginal ridges, and fractured surfaces were examined microscopically. RESULTS The mean (+/-SD) ultimate load at fracture (ULF, in Newtons) of PC and RMGIC/K-14/PC in box-only preparations (400+/-98; 386+/-82) did not differ significantly from that found in dovetail preparations (377+/-80; 317+/-92), and the mean ULF of RMGIC and RMGIC/PC in box-only preparations (307+/-44; 325+/-72) did not differ significantly from that in dovetail preparations (352+/-71; 353+/-70). No interactions were seen between materials and preparations (P=.09). Fracture modes for restorations in dovetail (predominantly mixed) and box-only preparations (predominantly mixed and adhesive) differed significantly (P=.003), but not between restorative procedures (P=.052). CONCLUSIONS Tooth-colored restorations placed in vitro in box-only preparations did not differ in fracture resistance from those placed in dovetail preparations. On fracture loading, resin-modified glass ionomer restorations placed in box-only preparations were more likely to show adhesive failure than those placed in dovetail preparations.