Two different types of commercially available ultrasonic devices, set at their lowest intensity for endodontic purpose, were used to instrument the internal portion of five different implant types, either with or without cooling. Intermittent anti-clockwise strokes were made, assuring that the tip was constantly in contact with the inner implant wall, as much as possible mimicking the motion that would have been used in clinical practice.

The ultrasonic devices used were the Satelec Suprasson T Max (Acteon Group, Merignac, France) and the EMS miniMaster (EMS, Electro Medical Systems SA, Nyon, Switzerland) with non-diamant, non-cutting tips ET 20, Satelec, and Instrument A (EMS). The device allowed for internal cooling of the tip with the cooling liquid at 31°C during instrumentation.

The implants used were from different brands; all 8-mm long but with various diameters and designs: Astra 3.5 mm (Dentsply Implants, Mölndal, Sweden), bone level Regular CrossFit (RC) 4.1 mm, bone level Narrow CrossFit (NC) 3.3 mm, Straumann tissue level regular body regular neck 3.3 mm (Straumann, Basel, Switzerland), and Straumann tissue level wide body regular neck 4.8 mm (Straumann AG, Basel, Switzerland). A single implant per group was used. They were embedded in epoxy resin, with a thermocouple (TC-08, Pico Technology, St. Neots, Great Britain) glued to the outer implant surface, at a level corresponding with the anticipated marginal bone level in uncompromised conditions. The change in temperature was registered for 30 s, followed by a 30-s cooling down period, at 5-s intervals (Figure 1).

The primary outcome variable was defined as the difference in temperature between the start of instrumentation and after 30 s when comparing the different implants and the maximum rise in temperature (deltaTmax) where results were averaged per experimental condition (type of device, with or without coolant). All tests were performed three times, and the results were averaged per condition. Differences between several experimental conditions were analyzed by means of independent t-tests and univariate analysis of variance, after verification of normal distribution by human eyeballing and the Kolmogorov-Smirnov test. Where appropriate, post hoc analysis was performed using the Student-Newman-Keuls multiple comparison test. The value for α was set at 0.05 to distinguish statistical significancy.