Monitoring failure rates of commercial implant brands; substantial equivalence in question?

The aim of this study was to report on the failure rates of two distinct dental implant systems in a clinical practice setting.

Date of implant placement and loss were entered prospectively in a data registry system. Failure rates of two commercially pure titanium implants, one with a porous oxydized surface (POS) and the other with a chemically altered surface (CAS), were assessed using a quality control chart and survival analyses.

A total of 860 POS and 759 CAS implants were placed. A warning of an increased failure rate of the CAS implant was identified by means of the quality control chart. Survival analyses indicated that the CAS implant failure rate was twice that of the POS implant (Hazard Ratio: 2.08; 95% CI: 1.33-3.28, P-value < 0.0012). After adjusting for alternative explanations, the CAS implant remained associated with a 95% increased failure rate (95% CI: 1.14-3.35; P-value = 0.0146). Abandoning the CAS implant and returning to a POS implant was associated with a non-significant 64% drop in the implant failure rate within less than a year (HR: 0.36; 95% CI: 0.12-1.14; P-value < 0.0826).

The difference between a 4% failure rate with the POS implant and an 8% failure with the CAS implant appears inconsistent with the assumption of substantial equivalence.