Influence of Two Different Approaches to Reporting Implant Survival Outcomes for Five Different Prosthodontic Applications

Background: Over the years, the definition of implant failure has varied, with some investigators accounting for all implants placed, while others discard failures that occurred before clinical loading. The influence of stresses transmitted to various bone densities, by different prosthetic appliances as well as the method used to determine failures, needs to be clearly understood. This paper reports on the influence of two different methods of determining 36-month survival of implants used to support different dental prostheses.

Methods: More than 2,900 implants with six different designs were placed in 829 patients at 32 study centers and followed for 3 years. The first method of determining survival accounted for all failures from placement through 36 months, while the second method counted only failures from post-loading of the prosthesis to 36 months. Survival curves were used to determine differences in survival outcomes for the two methods.

Results: For the maxillary single-tooth prosthetic application, implant survival from placement to 36 months was 94.7% when all failures were counted and increased to 98.3% with the post-loading method. For upper completely edentulous applications, implant survival was 85.3% with all failures counted and 95.6% with the post-loading method. This 10.3% difference is clinically important. The survival for implants in lower completely edentulous applications increased by 4.4% simply by using the post-loading approach. Implants used for upper posterior, partially edentulous applications involved only hydroxyapatite (HA)- coated implants, and the survival rates were similar (96.4% for all implants and 98.2% for the post-loading method). The difference in reported survival rates for implants in the lower posterior, partially edentulous application was 5.8%. Since failures tended to occur in the earlier phases of treatment, the post-loading approach always resulted in more favorable survival data. With the post-loading approach, valuable information related to implant performance before loading is lost. In all comparisons, HA-coated implant survival was always better than non-HA implants. Clinical investigators should clearly state the method used for determining failures. For all implants included in the study, survival curves illustrated different failure patterns for each method of determining overall survival.

Conclusion: Reporting of implant survival rates based on the post-loading method provides more favorable survival rates; however, accounting for all implants provides a more accurate method of determining survival. Ann Periodontol 2000;5:90-100.