12 Australian Dental Journal 2002;47:1.

A comparison of stainless steel hand and rotary nickel-titanium instrumentation using a silicone impression technique JL Chen, * HH Messer* Background: Root canal preparation using rotary nickel-titanium instruments has been reported to be superior to hand instrumentation in terms of root canal shape, centring and frequency of procedural errors. The purpose of this study was to evaluate canal preparations using a sequential silicone impression technique to assess root canal morphology. Methods: Pre-operative canal impressions were obtained from 24 extracted single-rooted premolars. Canals were randomly assigned to be firstly, instrumented with stainless steel K-type files or rotary nickel-titanium instruments (Profile), then further apically enlarged with rotary nickel-titanium instruments (Profile or Lightspeed). Post instrumen-tation impressions were taken and digitally photographed in a bucco-lingual and mesio-distal orientation at low-power magnification. Images were evaluated for procedural defects, changes in canal curvature, canal enlargement and canal rounding at 1, 3 and 7mm from the working length. Results: Silicone impressions could accurately reproduce the detailed root canal morphology of single-canal premolars. Repeated impressions of each canal using a standardized procedure allowed a detailed comparison of instrumentation techniques at various stages. Hand instrumentation incurred more errors than rotary nickel-titanium instrumentation (Profile). Further apical enlargement using Lightspeed instruments incurred fewer errors than with Profile instruments. Differences among techniques were highly significant (p<0.005). All techniques resulted in slight canal straightening, as well as rounder and enlarged canals particularly in the apical third. Conclusions: Rotary nickel-titanium instrumentation, especially Lightspeed, may produce better canal shape by reducing procedural errors. Key words: Rotary nickel-titanium, root canal preparation, silicone impressions, procedural errors. (Accepted for publication October 2000.