Early implant failure. Prognostic capacity of Periotest: retrospective study of a large sample.

OBJECTIVES The objectives of this study were to determine the accuracy of Periotest to monitor primary implant stability at first-stage surgery, to identify by multivariate analysis the variables associated with early implant failure and to compare Periotest with radiographic study in the diagnosis of implant stability at second-stage surgery (during osseointegration period). MATERIAL AND METHODS A 10-year retrospective study was conducted on 1084 Brånemark implants placed in 316 patients. Clinical variables, implant diameter and length, Periotest values (PTVs) and radiological variables were analyzed in bivariate and multivariate studies in order to determine their influence on early implant failure. RESULTS After examination of the sensitivity and specificity values obtained for different PTV cutoff points, a cutoff PTV of -2 was selected (84% sensitivity and 39% specificity). In the bivariate analysis, early failure was significantly related to smoking habits, implant location, bone type, implant features and PTVs (-2 and >or=-2). In the final multiple logistic model, only age (odds ratio (OR)=4.53; 95% confidence interval (CI), 1.34-15.27), smoking habits (OR=2.5; 95% CI, 1.3-4.79), bone type (OR=1.93; 95% CI, 1.01-3.7) and PTV at first surgery (OR=3.01; 95% CI, 1.5-6.02) were independently related to early failure. CONCLUSIONS The Periotest (with -2 cutoff) at first surgery offers high sensitivity in the prognosis of early implant loss and shows a greater capacity to evaluate stability during the osseointegration period compared with radiographic study.