An in-vitro evaluation of resonant frequency analysis to measure fixed bridge stability

Conventional fixed bridge prostheses may fail due to one or more loose retainers, which may be difficult to diagnose. An objective and reproducible investigation to identify, at an early stage, loosening of a retainer could be of significant benefit. The aims of the current series of investigations were to record retrospectively the clinical performance of different types of conventional fixed prostheses used to replace missing teeth and to determine whether Resonance Frequency Analysis (RFA) was capable of measuring bridge stability, in-vitro. One hundred and twenty two patients with 168 bridges were referred to two consultants at the Department of Restorative Dentistry at Liverpool Dental Hospital between Jan 2004 - Dec 2008 with fixed prosthesis problems. Fixed-fixed designs were the most common (77.9%), with cantilever bridges constituting 19.0% of the total. The most frequent cause of failure (39.0%) was associated with a post and core abutment. Apical pathology was found in 20.2%, dental caries in 14.8% and loss of retention in 11.9%. Fixed-fixed bridges were therefore chosen for further study. In-vitro pilot studies were subsequently undertaken to determine the feasibility of using resonance frequency analysis (RFA) on all-metal fixed-fixed bridges affixed to different models (wholly dental stone, or incorporating a simulated periodontal ligament) and to determine a reliable method to record this using an Osstell Mentor apparatus. The use of a buccal approach to record RF A values was validated. Based on the results from the pilot studies and a subsequent power analysis to set sample size, 100 models with standardised acrylic tooth abutment analogues and simulated periodontal ligaments were fabricated. All-metal fixed-fixed bridges were constructed from the first molar to the first premolar using standardised methods on models based with dental stone to mimic 100% (n=SO) or SO% (n=SO) bone support. In each case, two equal groups of 2S specimens had either both retainers cemented, (control group) or the premolar left uncemented (test group) to mimic clinical failure, cemented by a second operator to allow blind analysis. A magnetic component (Smartpeg) was subsequently cemented to the bridge using low-shrink composite resin and the Osstell Mentor used to measure bridge stability expressed as Bridge Stability Quotients (BSQ). The BSQ recorded at the premolar site in both 100% and SO% bone support models demonstrated a highly statistical significant difference (p60 suggesting that the fixed bridge was stable (cemented to both abutments) whereas a BSQ >59 indicated a risk of the bridge being uncemented to the premolar. All 100% bone support models were subsequently tested to failure in tension using a Universal Testing Machine with a SOO (N) load cell and cross-head speed of 10mm/min. 37% of specimens from the control group debonded at loads between 82 to 120N with the other 63% failing through extraction of the analogue/fracture of the model. 89.2% of the test group specimens failed by extraction of the tooth analogue from one or both ends at loads below SON. Statistical analysis using Kruskal Wallis tests demonstrated that the destructive testing could detect a highly statistically significant difference between the test and control group (P