Abstract
Background
There is increasing interest in the most responsible method for the introduction of new technologies in joint replacement, given the catastrophic consequences of widespread usage of poorly-performing devices. Two factors that make evaluation of new devices particularly difficult is the presence of the learning curve, and the desire of manufacturers to gain early market share for new technologies to recoup initial investment. Both of these factors are expected to lead to inferior early results, however, documentation is lacking. This study examines the effect of different methods of commercial introduction of new devices on early survivorship.
Materials and Methods
We modeled a database of 6000 operations performed using a new implant over a 5 year period. We assumed an average revision rate of 3.4% based on survivorship for hip resurfacing. Four different scenarios were modeled corresponding to the manner of introduction of this device to surgical practice. The “Standard” scenario assumed that 165 surgeons gradually adopted the device over a 5 year period based on the initial favorable experience of a small pilot group. Alternative scenarios were modeled, including limited release of the device (65 surgeons/64 cases each), increased distribution (310 surgeons 20 cases), and rapid early promotion (250 surgeons 24cases). Computer routines were utilized to predict the expected failure rate of each procedure using a standard survivorship curve based on surgeon experience. The sensitivity of the simulation to capture of all cases was also examined by repeating the Standard Scenario will censorship of the first 3 cases, and then the first 5 cases performed by each surgeon.
Results
The method of introduction of the implant had a significant effect on its survivorship. Limited distribution to a smaller number of higher volume surgeons reduced the revision rate by 38% from 3.54% to 2.57% at 5 years. Conversely, use of the device by twice the number of surgeons over a 5-year period reduced survivorship by only 3%. The greatest effect was observed with rapid early promotion of the new implant which led to a 23% increase in revisions at 5 years. Systematic removal of the first few cases performed by each surgeon dramatically changed the predicted number of revisions and thus the cumulative revision rate. When each surgeon's first cases were excluded from consideration, the number of revisions dropped by 11.4% for 3 cases, and 13.6% for five.
Conclusions
-
1.
This approach, using a “synthetic” database, predicts that the rate of commercial introduction of new devices can dramatically affect the survivorship reported by joint registries.
-
2.
Because of the effect of the learning curve on implant survivorship, staged introduction of new devices is recommended to reduce adverse effects of inevitable early failures.
