Abstract
Selection Criteria
Electronic searching, with English language restriction, was performed in MEDLINE and Cochrane Library up to November 2011. References listed within the retrieved articles were searched for further publications. The inclusion criteria were randomized controlled trials of fully edentulous patients older than 18 years and radiographic data with a mean follow-up period of at least 12 months. Non-English language articles and articles with insufficient information on prognosis were excluded. Titles and abstracts of the identified articles were screened independently by the 2 authors, with further screening by 3 independent reviewers on the basis of inclusion criteria. Disagreements were resolved by discussion. The following data were extracted by the authors: postload implant loss, peri-implant bone loss, overdenture longevity, peri-implant clinical parameters, overdenture retention and stability, patient satisfaction, surgical aftercare, and overdenture complications. Outcome variables included implant and bone loss that were analyzed by meta-analysis. All the individual outcomes were pooled by using the metan command written for the Stata software package. Risk difference estimates were used to summarize dichotomous outcomes and mean difference for continuous data, and all measures of effect were presented with 95% confidence intervals (CIs).
Key Study Factor
Implant number, complete edentulousness, and type of overdenture attachments mechanism (ball or bar attachments).
Main Outcome Measure
Dental implant failure and marginal bone loss.
Main Results
No articles were found that provided information regarding the maxilla. Eleven studies on the mandible were included for the final comparative analysis. Meta-analysis showed no statistically significant difference in implant loss with the following 2- and 4-implant configurations: 2 implants with ball attachments vs 4 implants with a bar; 2 vs 4 implants joined by bars; 1 implant (ball) vs 2 implants (ball); and 2 implants splinted vs unsplinted. The amount of bone loss varied from 0.1 to 1.75 mm. The mean difference in bone loss between 2 splinted and 4 splinted implants was -0.26 mm (95% CI: -0.57 to 0.05), but this had a high estimated statistical heterogeneity (I-2 = 92.6%; P < .001). The mean difference between splinted and unsplinted implants was 0.00 mm (95% CI: -20.26 to 0.26), with low statistical heterogeneity.
Conclusion
The authors of this review concluded that, in mandibular implant retained or supported overdentures, implant failure, bone loss, patient satisfaction, and number of complications do not significantly differ by the number of implants supporting the overdenture. Furthermore, splinting 2 implants does not seem to offer additional value.