Abstract
Background: Hard-on-hard bearings offer the potential to improve the survivorship of total hip arthroplasty implants. However, the specific indications for the use of these advanced technologies remain controversial. The purpose of this study was to characterize the epidemiology of bearing surface utilization in total hip arthroplasty in the United States with respect to patient, hospital, geographic, and payer characteristics.
Methods: The Nationwide Inpatient Sample database was used to analyze bearing type and demographic characteristics associated with 112,095 primary total hip arthroplasties performed in the United States between October 1, 2005, and December 31, 2006. The prevalence of each type of total hip arthroplasty bearing was calculated for population subgroups as a function of age, sex, census region, payer class, and hospital type.
Results: The most commonly reported bearing was metal-on-polyethylene (51%) followed by metal-on-metal (35%) and ceramic-on-ceramic (14%). Metal-on-polyethylene bearings were most commonly reported in female Medicare patients who were sixty-five to seventy-four years old, while metal-on-metal and ceramic-on-ceramic bearings were most commonly reported in privately insured male patients who were less than sixty-five years old. Thirty-three percent of patients over sixty-five years old had a hard-on-hard bearing reported. There was substantial regional variation in bearing usage; the highest prevalence of metal-on-polyethylene bearings was reported in the Northeast and at nonteaching hospitals, and the highest prevalence of metal-on-metal bearings was reported in the South and at teaching hospitals.
Conclusions: The usage of total hip arthroplasty bearings varies considerably by patient characteristics, hospital type, and geographic location throughout the United States. Despite uncertain advantages in older patients, hard-on-hard bearings are commonly used in patients over the age of sixty-five years. Further study is necessary to define the appropriate indications for these advanced technologies in total hip arthroplasty.
Metal-on-conventional ultra-high molecular weight polyethylene bearings in total hip arthroplasty have produced excellent long-term results, with reports of implant survivorship of almost 80% at over twenty years of follow-up1. However, the consequences of bearing surface wear, including osteolysis and aseptic loosening, have been a leading cause of late revision total hip arthroplasty, especially in younger, more active patients2-7. Hard-on-hard bearing surfaces (ceramic-on-ceramic and metal-on-metal) offer the potential to decrease the prevalence of revision total hip arthroplasties caused by bearing surface-related failures. Numerous laboratory studies8-14 have demonstrated substantial reductions in wear rates associated with both ceramic-on-ceramic and metal-on-metal bearings relative to metal-on-conventional ultra-high molecular weight polyethylene bearings, and many clinical studies have demonstrated excellent short-term and mid-term functional outcomes in patients with ceramic-on-ceramic and metal-on-metal bearings15-22. Nonetheless, the specific indications for the use of these advanced technologies remain unclear, and their benefits in terms of reducing the rates of revision total hip arthroplasty have not yet been conclusively demonstrated, especially in light of promising clinical and laboratory reports of reduced wear and osteolysis with newer, highly cross-linked ultra-high molecular weight polyethylene bearings23,24. Furthermore, the cost-effectiveness of hard-on-hard bearings (which are more costly than metal-on-polyethylene bearings), particularly in older patients who are less likely to experience total hip arthroplasty failures related to bearing surface wear, remains unproven25.
Many factors influence a surgeon's choice of bearing surface for a particular patient, including age, activity level, and patient-specific preferences and risk factors. Other factors that may affect a surgeon's decision include familiarity with a particular implant or bearing surface, implant availability in their hospital, implant cost, marketing programs, and relationships with and influence from hip implant manufacturers.
Beginning in October 2005, the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) procedure codes included three new modifier codes to track bearing surface utilization: 00.74 (metal-on-polyethylene), 00.75 (metal-on-metal), and 00.76 (ceramic-on-ceramic). These coding changes allow identification of the specific bearing surface used in total hip arthroplasties in large administrative databases, and therefore have enabled a large-scale evaluation of bearing surface usage throughout the United States. The purpose of this study was to characterize the epidemiology of bearing surface utilization in total hip arthroplasty in the United States with respect to patient, hospital, geographic, and payer characteristics.
The Nationwide Inpatient Sample database was used to analyze bearing surface type and demographic characteristics associated with 112,095 primary total hip arthroplasties performed in the United States between October 1, 2005, and December 31, 2006. Hip resurfacing and revision total hip arthroplasty procedures were excluded from the study. The Nationwide Inpatient Sample is a stratified, statistically valid survey of hospitals conducted by the Healthcare Cost and Utilization Project26. Hospitals within the sampling frame are stratified according to census region, governance (e.g., public or private), location (rural or urban), teaching status, and bed size. Hospitals are randomly selected to achieve an approximate 20% sample of the universe of community hospitals in each stratum. Thus, the Nationwide Inpatient Sample can be weighted to produce national estimates from the 20% sample, with use of the provided sampling weights27. All discharge records from each of the selected hospitals are collected and form part of the Nationwide Inpatient Sample file for a given year. In 2006, the Nationwide Inpatient Sample had a sample size of approximately eight million records from 1045 hospitals in thirty-eight states, which represents approximately 20% of all discharges from community hospitals in the United States, regardless of payment source. Because of the large size of the database, the Nationwide Inpatient Sample is particularly well suited for epidemiological studies related to specific procedures or diseases in the national population.
The prevalence of each type of bearing surface used in total hip arthroplasty (metal-on-polyethylene, metal-on-metal, and ceramic-on-ceramic) was calculated with use of the Nationwide Inpatient Sample for population subgroups in the United States stratified by age, sex, race, diagnosis, census region, primary payer class, and hospital characteristics (including size, location [urban or rural], and teaching status). Bearing surface codes obtained in administrative databases such as the Nationwide Inpatient Sample are reported in the hospital discharge abstract and are obtained by hospital coding personnel from clinical records, including operative notes and discharge summaries. Hospital size (small, medium, or large) is defined by the Healthcare Cost and Utilization Project according to the number of short-term acute care beds in a hospital, and it varies according to census region, location (urban or rural), and teaching status. The cutoff points for bed size in the Nationwide Inpatient Sample are chosen so that approximately one-third of the hospitals in a given combination of region, location, and teaching status would fall within each bed size category (small, medium, or large). For example, in the Northeast, the bed sizes for large hospitals are =100, =200, and =425 for rural, urban nonteaching, and urban teaching institutions, respectively. While in the Midwest, large rural, urban nonteaching, and urban teaching hospitals have =50, =175, and =375 beds, respectively. Urban and rural are defined by the Healthcare Cost and Utilization Project on the basis of the location of the hospital. A hospital located in a metropolitan area of a core based statistical area as defined in the Census 2000 is considered urban, and one that is not in a metropolitan core based statistical area is rural. Additionally, bearing surface usage was stratified by patient age, sex, race, hospital type, hospital size, payer type, and U.S. census region. The proportion of patients in each demographic subgroup for whom bearing codes had been reported was also determined. Descriptive statistics were used to characterize the demographics and other characteristics of patients receiving hip arthroplasty with different bearing surfaces, and statistical tests of the differences in the proportion of use of the bearing codes (e.g., the proportion of female patients) were performed with use of the Rao-Scott chi-square test or likelihood ratio tests.
Source of Funding
Funding for this study was received from the Orthopaedic Research and Education Foundation.
There were 288,283 total hip arthroplasties reported in the Nationwide Inpatient Sample database between October 1, 2005 and December 31, 2006. Of those, 112,095 (38.9%) of the procedures had a unique optional modifier code for the type of bearing surface used. The proportion of procedures without the bearing surface modifier code was unchanged over the time period under study. In the last quarter of 2005 and in 2006, 60.8% and 60.7% of the total hip arthroplasties, respectively, did not include a bearing surface code. The average age for the entire population of patients who had a total hip arthroplasty with a known bearing surface was sixty-five years; 59,734 (53.5%) were female. Urban, nonteaching hospitals accounted for 54,846 (48.9%) of the total hip arthroplasties, 69,799 (62.3%) were performed in large (compared with medium and small) hospitals, and 55,778 (49.8%) were performed in Medicare patients. Procedures were dispersed throughout the four geographic census regions of the United States, with 25,971 procedures (23.2%) performed in the Northeast, 28,721 (25.6%) in the Midwest, 29,821 (26.6%) in the South, and 27,582 (24.6%) in the West (Table I).
Patient Characteristics
Although the average age of the entire population of total hip arthroplasty patients with a known bearing surface was sixty-five years, there were considerable differences in reported bearing surface use across age groups. The average age of the patients who received a metal-on-polyethylene bearing surface was sixty-eight years compared with an average age of sixty-one years for patients who had a metal-on-metal bearing surface and fifty-six years for patients managed with a ceramic-on-ceramic bearing surface. For patients who were sixty-five years old or more, 66.6% had a metal-on-polyethylene bearing surface, 27.4% had a metal-on-metal bearing surface and only 6% had a ceramic-on-ceramic bearing surface, in the total hip replacement. In contrast, for patients less than sixty-five years old, 41.8% had a metal-on-metal bearing surface, followed by 35.8% with a metal-on-polyethylene bearing surface, and 22.4% with a ceramic-on-ceramic bearing surface (Table II), and these differences were significant (p < 0.001).
There were also significant differences in the reported use of bearing surfaces between men and women. Although the order of the prevalence between men and women was the same (metal-on-polyethylene was first, metal-on-metal was second, and ceramic-on-ceramic was third), the majority of women (57.1%) had a metal-on-polyethylene bearing compared with only 44.5% of the men (p < 0.001). Metal-on-metal bearings were a close second for men (39.7%) but were considerably less common among women (30.1%). The use of a ceramic-on-ceramic bearing surface was 15.9% for men and 12.8% for women (Table II). All of these differences were significant (p < 0.001).
Payer Characteristics (Table II)
There were highly significant differences (p < 0.001) in reported bearing use by payer type. Among Medicare patients, 64.9% received metal-on-polyethylene bearings compared with 28.5% who had metal-on-metal bearings and 6.5% who had ceramic-on-ceramic bearings. In contrast, among patients with private insurance, a similar percentage of metal-on-polyethylene (37.4%) and metal-on-metal (40.6%) bearings were implanted, with the remaining 22% consisting of ceramic-on-ceramic bearings. Medicaid patients and patients with no insurance or other coverage represented only 7% of the patients with a total hip arthroplasty, and the reported bearing surface usage was similar to that of patients with private insurance, which may be explained by their similarities in age to privately insured patients. Furthermore, among patients over the age of sixty-five years, those who were privately insured had hard-on-hard bearings significantly more often than did those insured by Medicare (40.1% compared with 32.7%; p < 0.0001).
Hospital Characteristics
Among urban nonteaching hospitals, the use of metal-on-polyethylene bearings was reported almost twice as often as metal-on-metal bearings (55.2% and 29.6%, respectively). In rural hospitals and urban teaching hospitals, the use of metal-on-metal bearings was reported almost as commonly as the use of metal-on-polyethylene bearings (41.4% compared with 48.5% and 38.9% compared with 47.2%, respectively) (Table III). Overall, urban nonteaching hospitals reported bearing codes in 42% of the procedures compared with 34.6% by rural hospitals and 37.9% by urban teaching hospitals, although these reporting rates were not significantly different (p = 0.18). In large hospitals, metal-on-polyethylene bearings were implanted in >50% of the procedures, while in small and medium hospitals, metal-on-polyethylene (46.9% and 49.4%, respectively) and metal-on-metal (38% and 36.9%) bearings were implanted in more similar frequencies (Table III).
Regional Variations in Bearing Surface Usage (Table III)
There were significant regional variations in bearing surface usage across U.S. census regions. The dominant bearing surface in the Northeast was metal-on-polyethylene (60.5% of procedures), whereas the procedure volumes according to type of bearing surface in the Midwest, South, and West were much more evenly split between metal-on-polyethylene (48.9%, 48.2%, and 48.3%, respectively) and metal-on-metal (37.6%, 38.7%, and 37.7%). Overall, there were significant variations in the reporting of the bearing codes by census region (p = 0.0025). Bearing codes were reported more frequently in the Northeast (47.2%) and West (44.3%) compared with the South (32.7%) and Midwest (37.8%).
Innovation in bearing surfaces used in total hip arthroplasty has resulted in increased implant choices for patients and surgeons. Hard-on-hard bearings, including metal-on-metal and ceramic-on-ceramic, offer the potential to reduce the number of wear-related failures, particularly in younger, more active patients. As a result, these bearings are often referred to and perceived by patients, surgeons, hospitals, and payers as so-called premium bearings, given their higher cost and potential for greater implant longevity. However, the choice of bearing surface is dependent on many factors, and the specific indications for and contraindications to hard-on-hard bearings remain unclear.
Hard-on-hard bearings have been associated with very low wear rates in the laboratory setting8-14. However, despite numerous reports of excellent clinical outcomes at short-term and midterm follow-up15-22, the benefits of hard-on-hard bearings in terms of enhancing implant longevity and reducing long-term rates of revision total hip arthroplasty have not yet been conclusively demonstrated. Furthermore, concerns have been raised about the safety of these new technologies, including ceramic fracture and so-called squeaking with ceramic-on-ceramic bearings28,29 as well as metal ion hypersensitivity and systemic toxicity with metal-on-metal bearings30-32. Despite limited evidence regarding the long-term benefits of hard-on-hard bearings in terms of improved implant longevity, and despite higher costs associated with their use, in the first fifteen months since the inception of the optional modifier codes for bearing surfaces in total hip arthroplasty, approximately half of all procedures with known bearing usage were hard-on-hard bearings.
Our study demonstrates that bearing usage in total hip arthroplasty varies considerably by age, sex, hospital type, and geographic region throughout the United States. Some of this variation appears to be appropriate, e.g., more frequent use of metal-on-polyethylene bearings in patients over the age of sixty-five years and more frequent use of metal-on-metal bearings in younger, male patients. However, some of the variation, including the more frequent use of metal-on-metal bearings in urban teaching hospitals, small and medium hospitals, and in the South, cannot be explained by differences in age or disease prevalence alone. This is consistent with previous reports of widespread regional variation in total joint arthroplasty utilization that cannot be explained by regional differences in disease prevalence alone33. Furthermore, despite uncertain advantages in older patients, the use of hard-on-hard bearings is commonly reported in patients who are over sixty-five years old. We also found that patients over sixty-five years old who were privately insured had hard-on-hard bearings more frequently than did those insured by Medicare, which may reflect the influence of reimbursement on clinical practice.
Beginning in October 2006, an additional ICD-9 procedure modifier code for ceramic-on-polyethylene bearings was added to the ICD-9-CM codes. However, this was not considered in our study because of the relatively small sample size (1591 procedures), which corresponded to only 0.5% of all total hip arthroplasties during the study period. In addition, the shorter effective period for the ceramic-on-polyethylene bearing code (three months) was not comparable with the effective period for the other bearing codes (fifteen months).
Our findings are limited by the use of an administrative database, in which bearing surface type is an optional modifier code that can be reported in conjunction with the primary procedure code (primary or revision total hip arthroplasty), which introduces a potential source of bias into our study. Since their introduction in October 2005, these optional bearing codes were used for only 40% of primary total hip arthroplasties in the United States, and their usage did not appear to increase over time within the study period. The limited reporting of the codes may be attributed to the codes being optional and unnecessary for reimbursement purposes. Furthermore, hospital administrative coding personnel may not report these codes because of a lack of knowledge of bearing types or insufficient information provided in the surgeons' clinical documentation (e.g., operative reports) to discern specific bearing types. It is also possible that hospitals that do not use multiple bearing surfaces might be more likely to underreport the bearing type compared with hospitals that use a mix of two or three bearing surfaces. However, bearing surface type did not appear to be overreported or underreported in any systematic way (such as by hospital type) that would undermine or call into question the validity of our results. Additionally, there are no financial or other incentives to report certain types of bearing surfaces more commonly than others. Therefore, we believe that the likelihood that reporting bias influenced the results of our study in any meaningful way is extremely low. Also, these limitations are somewhat mitigated by the large number of procedures (>112,000) from across the United States that had a code for bearing surface type and were included in our analysis, which substantially increases the external validity and generalizability of our findings. As additional data and experience are accumulated with the use of administrative claims to characterize bearing surfaces in total hip arthroplasties, a validation study to verify the accuracy of the claims data would be appropriate.
Another limitation of our investigation is that we were unable to track longitudinal bearing surface utilization because of the limited time since the introduction and use of these codes. As a result, we have no information regarding how bearing surface utilization in the United States has changed over time. Since bearing surface utilization is not static, but rather constantly changing in response to many clinical and market factors, it will be important to compare our results with future studies that evaluate bearing surface utilization at different points in time.
Until a national joint replacement registry is created and validated for the United States, use of administrative data provides the most effective mechanism to characterize and evaluate patterns of implant utilization in joint replacement on a nationwide basis. The findings from our study suggest that there is currently wide variation in bearing surface utilization in the United States. Further investigation is necessary to define the appropriate indications for the use of hard-on-hard bearings in patients managed with a total hip arthroplasty. 
Berry DJ, Harmsen WS, Cabanela ME, Morrey BF. Twenty-five-year survivorship of two thousand consecutive primary Charnley total hip replacements: factors affecting survivorship of acetabular and femoral components. J Bone Joint Surg Am.2002;84:171-7.84171
2002
[PubMed]
Beaulé PE, Campbell P, Mirra J, Hooper JC, Schmalzried TP. Osteolysis in a cementless, second generation metal-on-metal hip replacement. Clin Orthop Relat Res.2001;386:159-65.386159
2001
[CrossRef]
Digas G, Thanner J, Nivbrant B, Röhrl S, Ström H, Kärrholm J. Increase in early polyethylene wear after sterilization with ethylene oxide: radiostereometric analyses of 201 total hips. Acta Orthop Scand.2003;74:531-41.74531
2003
[CrossRef]
Duffy P, Sher JL, Partington PF. Premature wear and osteolysis in an HA-coated, uncemented total hip arthroplasty. J Bone Joint Surg Br.2004;86:34-8.8634
2004
Engh CA, Engh CA Jr, Claus A. The relationship between lysis and loosening in failed anatomic medullary locking components. Orthopedics.2001;24:855-6.24855
2001
Hernandez JR, Keating EM, Faris PM, Meding JB, Ritter MA. Polyethylene wear in uncemented acetabular components. J Bone Joint Surg Br.1994;76:263-6.76263
1994
Kim YG, Kim SY, Park BC, Kim PT, Ihn JC, Kim ID. Uncemented Harris-Galante total hip arthroplasty in patients with osteonecrosis of the femoral head. A 10-16-year follow-up study. Acta Orthop.2005;76:42-8.7642
2005
[CrossRef]
Affatato S, Goldoni M, Testoni M, Toni A. Mixed oxides prosthetic ceramic ball heads. Part 3: effect of the ZrO2 fraction on the wear of ceramic on ceramic hip joint prostheses. A long-term in vitro wear study. Biomaterials.2001;22:717-23.22717
2001
[CrossRef]
Anissian HL, Stark A, Gustafson A, Good V, Clarke IC. Metal-on-metal bearing in hip prosthesis generates 100-fold less wear debris than metal-on-polyethylene. Acta Orthop Scand.1999;70:578-82.70578
1999
[CrossRef]
Clarke IC, Good V, Williams P, Schroeder D, Anissian L, Stark A, Oonishi H, Schuldies J, Gustafson G. Ultra-low wear rates for rigid-on-rigid bearings in total hip replacements. Proc Inst Mech Eng [H].2000;214:331-47.214331
2000
Dowson D. Tribological principles in metal-on-metal hip joint design. Proc Inst Mech Eng [H].2006;220:161-71.220161
2006
Isaac GH, Thompson J, Williams S, Fisher J. Metal-on-metal bearings surfaces: materials, manufacture, design, optimization, and alternatives. Proc Inst Mech Eng [H].2006;220:119-33.220119
2006
Saikko V, Ahlroos T, Calonius O, Keränen J. Wear simulation of total hip prostheses with polyethylene against CoCr, alumina and diamond-like carbon. Biomaterials.2001;22:1507-14.221507
2001
[CrossRef]
Wang FC, Brockett C, Williams S, Udofia I, Fisher J, Jin ZM. Lubrication and friction prediction in metal-on-metal hip implants. Phys Med Biol.2008;53:1277-93.531277
2008
[CrossRef]
Dorr LD, Wan Z, Longjohn DB, Dubois B, Murken R. Total hip arthroplasty with use of the Metasul metal-on-metal articulation. Four to seven-year results. J Bone Joint Surg Am.2000;82:789-98.82789
2000
Bizot P, Banallec L, Sedel L, Nizard R. Alumina-on-alumina total hip prostheses in patients 40 years of age or younger. Clin Orthop Relat Res.2000;379:68-76.37968
2000
[CrossRef]
D'Antonio JA, Capello W, Manley M, Bierbaum B. New experience with alumina-on-alumina ceramic bearings for total hip arthroplasty. J Arthroplasty.2002;17:390-7.17390
2002
[CrossRef]
Garcia-Cimbrelo E, Garcia-Rey E, Murcia-Mazón A, Blanco-Pozo A, Martí E. Alumina-on-alumina in THA: a multicenter prospective study. Clin Orthop Relat Res.2008;466:309-16.466309
2008
[CrossRef]
Hamadouche M, Boutin P, Daussange J, Bolander ME, Sedel L. Alumina-on-alumina total hip arthroplasty: a minimum 18.5-year follow-up study. J Bone Joint Surg Am.2002;84:69-77.8469
2002
MacDonald SJ, Mehin R. Metal on metal: clinical results with modern implants. Sem Arthroplasty.2003;14:123-30.14123
2003
Sedel L, Nizard RS, Kerboull L, Witvoet J. Alumina-alumina hip replacement in patients younger than 50 years old. Clin Orthop Relat Res.1994;298:175-83.298175
1994
Yoo JJ, Kim YM, Yoon KS, Koo KH, Song WS, Kim HJ. Alumina-on-alumina total hip arthroplasty. A five-year minimum follow-up study. J Bone Joint Surg Am.2005;87:530-5.87530
2005
[CrossRef]
Atienza C Jr, Maloney WJ. Highly cross-linked polyethylene bearing surfaces in total hip arthroplasty. J Surg Orthop Adv.2008;17:27-33.1727
2008
Manning DW, Chiang PP, Martell JM, Galante JO, Harris WH. In vivo comparative wear study of traditional and highly cross-linked polyethylene in total hip arthroplasty. J Arthroplasty.2005;20:880-6.20880
2005
[CrossRef]
Bozic KJ, Morshed S, Silverstein MD, Rubash HE, Kahn JG. Use of cost-effectiveness analysis to evaluate new technologies in orthopaedics. The case of alternative bearing surfaces in total hip arthroplasty. J Bone Joint Surg Am.2006;88:706-14.88706
2006
[CrossRef]
Yang CC, Kim RH, Dennis DA. The squeaking hip: a cause for concern-disagrees. Orthopedics.2007;30:739-42.30739
2007
Willmann G. Ceramic femoral head retrieval data. Clin Orthop Relat Res.2000;379:22-8.37922
2000
[CrossRef]
Visuri T, Pukkala E, Paavolainen P, Pulkkinen P, Riska EB. Cancer risk after metal on metal and polyethylene on metal total hip arthroplasty. Clin Orthop Relat Res.1996;329 Suppl:S280-9.329S280
1996
[CrossRef]
Davies AP, Willert HG, Campbell PA, Learmonth ID, Case CP. An unusual lymphocytic perivascular infiltration in tissues around contemporary metal-on-metal joint replacements. J Bone Joint Surg Am.2005;87:18-27.8718
2005
[CrossRef]
Jacobs JJ, Skipor AK, Patterson LM, Hallab NJ, Paprosky WG, Black J, Galante JO. Metal release in patients who have had a primary total hip arthroplasty. A prospective, controlled, longitudinal study. J Bone Joint Surg Am.1998;80:1447-58.801447
1998
Losina E, Kessler CL, Wright EA, Creel AH, Barrett JA, Fossel AH, Katz JN. Geographic diversity of low-volume hospitals in total knee replacement: implications for regionalization policies. Med Care.2006;44:637-45.44637
2006
[CrossRef]