Each year, orthopaedic surgeons are presented with numerous new devices, technologies, and implants. However, little is known regarding what proportion of these new devices will become important parts of the orthopaedic armamentarium and what proportion will be on the market for a short time. Orthopaedic surgeons often are faced with the difficult task of evaluating new implants and procedures for their patients, yet have essentially no baseline data on what proportion of the devices will prove clinically successful. If the success rate of new orthopaedic devices is low, then surgeons would be more cautious to adopt new devices.
Little data exist on the epidemiology of medical devices.3,15 Our clinical experience has been that numerous devices are introduced with enthusiasm, only to disappear from the market shortly thereafter. One example is the flexible humeral nail (Synthes, Paoli, PA), which was reported to have good results14 but no longer is on the market in the United States. We sought to investigate the natural history of what happens to new devices after they are introduced to the orthopaedic marketplace.
We hypothesized that a large portion of technologies introduced would not be available 10 years later. We sought to determine: (1) What percentage of devices are available on the market five and 10 years after approval?; (2) Are devices approved 10 years ago less likely to be available than devices approved 5 years ago?; (3) Is there any relationship between company size and availability?; (4) Why are some devices no longer available?; and (5) Were there any major clinical performance problems with the devices?
MATERIALS AND METHODS
We obtained a list of all devices approved by the United States Food and Drug Administration (FDA) 5 and 10 years before the study period from the routine publication of premarket approvals (PMA) and substantially equivalent approvals (510k) found in The Gray Sheet (FDC Reports, Inc, Chevy Chase, MD). We randomly selected 100 devices from the 4000 devices approved that year. The proportion of devices that are orthopaedic is not known. We then contacted the companies and ascertained what happened to the devices after FDA approval and whether the devices were still on the market. We attempted to determine the reasons for withdrawal of a device (eg, company out of business, poor clinical results, device-related problems). We determined the proportion of devices that still were available 5 and 10 years after approval, and analyzed the data for factors (such as company size, type of device, year of approval) that might be related to a greater chance that a device would still be on the market.
The sample size was chosen because a power analysis showed a sample size of 100 devices would provide 90% power to detect a difference of 25% between groups that were or were not on the market at the p < 0.05 level of significance.
A cohort of devices was selected randomly using a random number list generated by the Microsoft® Excel® (Microsoft Inc, Redmond, WA) random number function. An orthopaedic surgeon (TB) reviewed the device list. If the random number said five, the fifth next orthopaedic device was selected. If the random number selected seven, the seventh next orthopaedic device was selected. The process continued until 50 devices had been selected from each of the study years. Companies were classified based on their market capitalization at the time of approval into small capitalization (< $500 million), midcapitalization ($500 million to $1 billion), and large capitalization (> $1 billion).
We then contacted the companies for information regarding the devices using a sequential protocol. First, companies were mailed a written survey. If no response was obtained, up to three phone calls were made. We also performed internet searches for the companies and devices using major search engines (google.com and anywho.com). When possible, an E-mail address of the individual responsible for regulatory affairs at the company was obtained, and the party was sent an E-mail. Additional data were obtained from the FDA's MAUDE database (http://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfMAUDE/search.CFM) of device-associated problems. We were able to obtain data from the companies on 93% of the devices (Table 1). Forty-eight of 55 companies (87%) responded to our requests. Seven companies accounting for seven devices (13%) did not respond to multiple requests.
A device was considered available on the market if it was in production and able to be purchased for use in its original or a modified form. The device could be available from the original company or a company that purchased the technology.
The final cohort consisted of 100 devices. Forty-nine devices were approved by the FDA in 1994 (10 years before the study), and 51 devices were approved by the FDA in 1999 (5 years before the study). The majority of devices (97%) were approved through the 510k process, indicating the devices were substantially equivalent to devices already available on the market. Three percent of devices required a PMA, indicating they were substantially different than existing devices and required a more rigorous approval process.
Of the devices studied, 63% were surgical implants, including total joint prostheses, and 20% were surgical instruments (Table 1). Thirty-eight percent of the technologies had been acquired by another company, and 94% of devices were introduced to the market within 1 year of approval by the FDA.
The data then were analyzed using SPSS 10.0 (SPSS Inc, Chicago, IL). Proportions were tested for significance using the chi square test.
Of the 49 devices approved in 1994, 26 (53%) currently were available on the market; 22 devices (45%) no longer were available (Table 2). Of the 51 devices approved in 1999, 38 (75%) were still available for use. Devices from 1994 were less (p < 0.01) likely to be available than devices from 1999. Looking at implants only (devices meant to be left in the body such as prostheses or internal fixation devices), 17 of 29 (59%) were available 10 years after approval, and 31 of 34 (91%) were available 5 years after approval. Of the companies producing implants in 1994, 21% were no longer in business and had not been acquired 10 years later.
Devices introduced by large capitalization companies were more likely (p < 0.001) to be available than devices introduced by small and midcapitalization companies. Fourteen of 22 small capitalization companies (63%) were no longer in business after 10 years.
The most common reason for a device no longer to be on the market was that the company was out of business (Table 3).
Few devices had clinical performance problems (Table 3). Most were withdrawn because of slow sales or companies going out of business. For example, one device was a series of bone screws applicable to orthopaedic trauma surgery. Despite aggressive marketing, few screws were sold and the company ultimately decided to move on to other projects. In another example, a new company was formed to introduce a new humeral nail. After several years of slow sales, the company folded. The former owner was contacted and reported the device performed well without device-related problems, but the market was not interested in humeral nails. With the numbers available, none of the implant subspecialty areas (spine for example) was more or less likely to be available than the general group of devices.
Although 12 devices had device-related problems reported in the FDA's MAUDE database, 10 of these reports were for minor difficulties (primarily packaging problems), and no specific action was taken. Two devices had major problems resulting in recalls (Table 3). One device was a ceramic femoral head that was found to be prone to fracture and no longer is in use. One product was an external fixator that had problems with clamps loosening; the clamps were recalled and repaired and the product is still on the market. Both of these problems were noted in company surveys. Overall, only two of the 100 devices studied had any major device-related problems.
We analyzed the randomly selected cohort of devices introduced to the market 5 and 10 years ago. We found approximately 50% of devices no longer were available on the market, devices approved 10 years ago were less likely to be available than devices approved 5 years ago, and there were few reported device-related problems.
Our study has several limitations. We analyzed only devices approved by the FDA through the 510k and PMA processes; medications were not studied. We were able to analyze only a small percentage of the devices approved each year. The number of devices was chosen after power analysis. Our ability to detect rare events (such as recalls) is limited. A case-control study would be a more powerful design for analyzing rare events. Also, our design does not account for the relative market shares of the companies involved. Instead, we used a more epidemiologic approach with each new FDA-approved device as the unit in question because orthopaedic surgeons generally will evaluate each new technology introduced, not just those marketed by large companies. There is some inherent bias in determining whether a device is an orthopaedic device. The data for devices were reported by the companies, which may be biased. Finally, we considered availability on the market as a dichotomous (yes/no) variable, and did not assess the relative sales volume of the devices. A device may be available, but on a practical basis, be selling very low volumes.
Orthopaedic surgeons and their patients continue to benefit from rapidly improving orthopaedic technology.5 Devices and instruments available today allow surgical treatment of conditions that were difficult to treat, or were untreatable, just 20 years ago. Examples include hip arthroscopy, locked plating of fractures, and bone morpho-genetic protein for adjunctive fracture healing.6-8,11 These beneficial new technologies come from the complex symbiotic relationship between orthopaedic surgeons and industry.
However, each new technology introduced can result in new problems. For example, vertebroplasty resulted in a great clinical benefit to selected patients,9 but also resulted in unique complications of cement extrusion.10 Another more recent example is the two-incision minimally invasive hip replacement with special retractors and reamers. Original reports were excellent,1 but followup reports have been less enthusiastic.12 Much attention has been given to the careful introduction of new technology to the orthopaedic marketplace.4,13 It is necessary to balance excitement over new surgical options and protection of patients.
We designed our study to ascertain what proportion of new technology is available for use 5 and 10 years after introduction, and what proportion actually results in device-related problems. By randomly selecting devices from the entire group of devices approved each year, we reviewed a representative sample of the devices that an orthopaedic surgeon may see. The majority of new devices introduced each year is approved under the 510k process and has related predicate devices, unlike PMA devices which have no predicate device. The choice of 100 devices was based on a preliminary power analysis.
The companies generally were responsive, as in another study,2 and we were able to obtain information regarding 93% of the randomly selected devices. Of the devices approved 10 and 5 years ago, 45% and 25%, respectively, no longer were available on the market.
Clinical problems with the devices were rare. Only two devices had substantial device-related problems, and only one was removed from the market. The most common reason for a device becoming unavailable was the company was out of business.
These data clarify the performance of new orthopaedic devices. Although actual device-related problems are uncommon, approximately 50% of the implants and 50% of the surgical instruments approved each year no longer will be in use 10 years later. Our data indicate the natural forces of the orthopaedic marketplace will eliminate approximately 50% of all technologies introduced each year. The elimination is not because of poor clinical performance of the devices, but slow device sales and companies going out of business in a competitive medical device market. Devices introduced by larger companies tend to fare better than devices introduced by smaller companies.
Orthopaedic implants such as hips, knees and screws, are surgically implanted in patients' bodies and often are retained for 10 years or more. It is notable that 40% of the orthopaedic implants introduced no longer were available on the market after 10 years, and the companies that manufactured them frequently were out of business. Therefore surgeons should use more caution when adopting a new implant from a small company because a large proportion (63%) of smaller companies are not in business 10 years later when implant revision or removal may be recommended.
Approximately 50% of devices approved 10 years ago and 25% of devices approved 5 years ago no longer are available. Devices from smaller companies are more likely to be unavailable than devices from larger companies. The devices usually are unavailable because of slow sales or poor business performance of the company, not because of clinical problems with the devices.
1. Berger RA, Jacobs JJ, Meneghini RM, Della Valle C, Paprosky W, Rosenberg AG. Rapid rehabilitation and recovery with minimally invasive total hip arthroplasty. Clin Orthop Relat Res
. 2004;429: 239-247.
2. Bhattacharyya T, Tornetta P3rd, Healy WL, Einhorn TA. The validity of claims made in orthopaedic print advertisements. J Bone Joint Surg Am
3. Brown SL, Bright RA, Tavris DR. Medical device epidemiology and surveillance: patient safety is the bottom line. Expert Rev Med Devices
4. Callaghan JJ, Crowninshield RD, Greenwald AS, Lieberman JR, Rosenberg AG, Lewallen DG. Symposium: introducing technology into orthopaedic practice: How should it be done? J Bone Joint Surg Am
5. Crowninshield R. The orthopaedic profession and the orthopaedic industry: partners in quality patient care (Comment on J Bone Joint Surg Am
. 2003;85:168-170). J Bone Joint Surg Am
6. Einhorn TA. Clinical applications of recombinant human BMPs: early experience and future development. J Bone Joint Surg Am
. 2003;85 (suppl 3):82-88.
7. Einhorn TA, Majeska RJ, Mohaideen A, Kagel EM, Bouxsein ML, Turek TJ, Wozney JM. A single percutaneous injection of recombinant human bone morphogenetic protein-2 accelerates fracture repair. J Bone Joint Surg Am
8. Gosling T, Schandelmaier P, Marti A, Hufner T, Partenheimer A, Krettek C. Less invasive stabilization of complex tibial plateau fractures: a biomechanical evaluation of a unilateral locked screw plate and double plating. J Orthop Trauma
9. Hadjipavlou AG, Tzermiadianos MN, Katonis PG, Szpalski M. Percutaneous vertebroplasty and balloon kyphoplasty for the treatment of osteoporotic vertebral compression fractures and osteolytic tumours. J Bone Joint Surg Br
10. Harrington KD. Major neurological complications following percutaneous vertebroplasty with polymethylmethacrylate: a case report. J Bone Joint Surg Am
11. McCarthy JC. Hip arthroscopy: when it is and when it is not indicated. Instr Course Lect
12. Pagnano MW, Leone J, Lewallen DG, Hanssen AD. Two-incision THA had modest outcomes and some substantial complications. Clin Orthop Relat Res
13. Purvis JM, Alexander AH, Einhorn TA, Griffin LY. American Orthopaedic Association symposium: Evaluating the flood of orthopaedic media and marketing information. J Bone Joint Surg Am
14. Stannard JP, Harris HW, McGwin G Jr, Volgas DA, Alonso JE. Intramedullary nailing of humeral shaft fractures with a locking flexible nail. J Bone Joint Surg Am
15. Torrence ME. Data sources: use in the epidemiologic study of medical devices. Epidemiology
. 2002;13 (3 suppl):S10-S14.