Scholarly article on topic 'Outcomes of Ceramic Bearings After Revision Total Hip Arthroplasty in the Medicare Population'

Outcomes of Ceramic Bearings After Revision Total Hip Arthroplasty in the Medicare Population Academic research paper on "Economics and business"

CC BY-NC-ND
0
0
Share paper
Academic journal
The Journal of Arthroplasty
OECD Field of science
Keywords
{"revision total hip arthroplasty" / "ceramic bearings" / infection / dislocation / readmission / mortality}

Abstract of research paper on Economics and business, author of scientific article — Steven M. Kurtz, Edmund C. Lau, Doruk Baykal, Bryan D. Springer

Abstract Background The purpose of this study was to analyze the utilization and outcomes of ceramic bearings used in revision total hip arthroplasty (R-THA) in the Medicare population. Methods A total of 31,809 patients aged >65 years at the time of revision surgery who underwent R-THA between 2005 and 2013 were identified from the United States Medicare 100% national administrative claims database. Outcomes of interest included relative risk of readmission (90 days) or infection, dislocation, rerevision, or mortality at any time point after revision. Propensity scores were developed to adjust for selection bias in the choice of bearing type at revision surgery. Results The utilization of ceramic-on-polyethylene (C-PE) and ceramic-on-ceramic (COC) bearings in R-THA increased from 5.3% to 26.6% and from 1.8% to 2.5% in between 2005 and 2013, respectively. For R-THA patients treated with C-PE bearings, there was reduced risk of 90-day readmission (hazard ratio, HR: 0.90, P = .007). We also observed a trend for reduced risk of infection with C-PE (HR: 0.88) that did not reach statistical significance (P = .14). For R-THA patients treated with COC bearings, there was reduced risk of dislocation (HR: 0.76, P = .04). There was no significant difference in risk of rerevision or mortality for either the C-PE or COC bearing cohorts when compared with the metal-on-polyethylene bearing cohort. Conclusion Medicare patients treated in a revision scenario with ceramic bearings exhibit similar risk of rerevision, infection, or mortality as those treated with metal-on-polyethylene bearings. Conversely, we found an association between the use of specific ceramic bearings in R-THA and reduced risk of readmission (C-PE) and dislocation (COC).

Academic research paper on topic "Outcomes of Ceramic Bearings After Revision Total Hip Arthroplasty in the Medicare Population"

The Journal of Arthroplasty xxx (2016) 1—7

ELSEVIER

Contents lists available at ScienceDirect

The Journal of Arthroplasty

journal homepage: www.arthroplastyjournal.org

Original article

Outcomes of Ceramic Bearings After Revision Total Hip Arthroplasty in the Medicare Population

Steven M. Kurtz, PhD a' *, Edmund C. Lau, MS b, Doruk Baykal, PhD b, Bryan D. Springer, MD c

a Exponent, Inc, Philadelphia, Pennsylvania

b Exponent, Inc, Menlo Park, California

c OrthoCarolina Hip and Knee Center, Charlotte, North Carolina

ABSTRACT

Background: The purpose of this study was to analyze the utilization and outcomes of ceramic bearings used in revision total hip arthroplasty (R-THA) in the Medicare population.

Methods: A total of 31,809 patients aged >65 years at the time of revision surgery who underwent R-THA between 2005 and 2013 were identified from the United States Medicare 100% national administrative claims database. Outcomes of interest included relative risk of readmission (90 days) or infection, dislocation, rerevision, or mortality at any time point after revision. Propensity scores were developed to adjust for selection bias in the choice of bearing type at revision surgery.

Results: The utilization of ceramic-on-polyethylene (C-PE) and ceramic-on-ceramic (COC) bearings in R-THA increased from 5.3% to 26.6% and from 1.8% to 2.5% in between 2005 and 2013, respectively. For R-THA patients treated with C-PE bearings, there was reduced risk of 90-day readmission (hazard ratio, HR: 0.90, P = .007). We also observed a trend for reduced risk of infection with C-PE (HR: 0.88) that did not reach statistical significance (P = .14). For R-THA patients treated with COC bearings, there was reduced risk of dislocation (HR: 0.76, P = .04). There was no significant difference in risk of rerevision or mortality for either the C-PE or COC bearing cohorts when compared with the metal-on-polyethylene bearing cohort.

Conclusion: Medicare patients treated in a revision scenario with ceramic bearings exhibit similar risk of rerevision, infection, or mortality as those treated with metal-on-polyethylene bearings. Conversely, we found an association between the use of specific ceramic bearings in R-THA and reduced risk of readmission (C-PE) and dislocation (COC). © 2016 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND

license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

ARTICLE INFO

Article history:

Received 24 November 2015 Received in revised form 23 February 2016 Accepted 24 February 2016 Available online xxx

Keywords:

revision total hip arthroplasty

ceramic bearings

infection

dislocation

readmission

mortality

Ceramic bearings, in which a ceramic femoral head articulates against either polyethylene (ceramic on polyethylene [C-PE]) or a ceramic acetabular component (ceramic on ceramic [COC]), have been used for >40 years in primary total hip arthroplasty (THA) as

Each author certifies that all investigations were conducted in conformity with ethical principles of research.

This work was performed at Exponent, Inc, Philadelphia, PA.

This study was supported with institutional funding from CeramTec.

One or more of the authors of this paper have disclosed potential or pertinent conflicts of interest, which may include receipt of payment, either direct or indirect, institutional support, or association with an entity in the biomedical field which may be perceived to have potential conflict of interest with this work. For full disclosure statements refer to http://dx.doi.org/10.1016/j.arth.2016.02.054.

* Reprint requests: Steven M. Kurtz, PhD, Exponent, Inc, 3440 Market Street, Suite 600, Philadelphia, PA 19104.

an alternative to metal-on-polyethylene (M-PE) bearings [1-3]. Contemporary ceramic bearings have well-documented, long-term clinical survivorship in primary THA [4,5]. Although ceramic components were initially adopted because of their improved wear resistance relative to M-PE [6,7], recent research has also shown that the use of ceramic mitigates the risk of taper corrosion [8]. Previous studies have also suggested that COC bearings may be associated with reduced risk of dislocation [9,10]. Balancing these advantages, ceramic bearings have well-known drawbacks, namely, their increased cost [11]; the potential risk of fracture [5,12], although substantially diminished for the current generation [13,14]; and, in the case of COC articulations, squeaking [15]. Due partly to concerns about squeaking, COC is currently a less popular bearing choice than C-PE among US surgeons [11].

In the past decade, there have been substantial changes in the usage of alternative bearings in primary THA due, first, to the widespread adoption of metal-on-metal (MOM) articulations up to

http://dx.doi.org/10.1016/j.arth.2016.02.054

0883-5403/© 2016 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

S.M. Kurtz et al. / The Journal of Arthroplasty xxx (2016) 1-7

2008, followed by their subsequent decline [16]. In the United States, MOM hips were used in 26%-32% of revision THAs between 2005 and 2008 because of concerns about dislocation, especially in elderly patients [16]. By contrast, comparatively little is known regarding the usage patterns for ceramic bearings during revision surgery in the past decade, especially after the decline of MOM usage that began in 2009.

The outcomes of different ceramic bearings during revision surgery likewise remain poorly understood. Previous studies on ceramic bearing usage during revision have focused on the outcomes of patients who were revised after the rare circumstances of a fracture of the femoral head [17] or who were revised because of squeaking [18]. Wong et al [19] studied 884 aseptic revisions of MOM hip resurfacing arthroplasties in Australia. They observed no difference in the rerevision rate as a function of the bearing surface (M-PE, MOM, COC, or C-PE); however, they cautioned that a larger sample size than what they examined would be likely be necessary to see differences, if there were any [19]. Jack et al [20], also from Australia, followed the outcomes of 165 acetabular cup revisions using COC bearings; however, this observational cohort study did not include a control group with an alternative bearing. In the United States, Cooper et al [21] described the treatment of patients who were revised for adverse locale tissue reactions from taper corrosion in M-PE bearings and later recommended that device components that do not include cobalt or chromium be used in the treatment of such patients at revision, using C-PE or COC bearings [22].

It remains unknown how the patient outcomes after revision using contemporary ceramic bearings compare with the outcomes for patients revised using M-PE bearings in the United States. Accordingly, we addressed the following related research questions: (1) what is the utilization of ceramic bearings for revision total hip arthroplasty (R-THA) in the Medicare population and how has it evolved over time; (2) does the use of C-PE bearings influence outcomes after R-THA as compared with M-PE; and (3) does the use of COC bearings influence outcomes after R-THA as compared with M-PE?

Methods

The Medicare 100% national administrative claims database was used to identify revision THA patients between October 1, 2005, and December 31, 2013. This set of data captures all fee-for-service claims submitted by hospitals for hip revision and other hospital-izations from this group of patients. Patients aged <65 years or beneficiaries enrolled in a health maintenance organization were excluded. A small number of beneficiaries residing outside of the 50 states were also excluded. Thus, our study considered the elderly Medicare population of revision hip arthroplasty patients.

Unique, encrypted Medicare beneficiary identifiers were used to follow patients longitudinally throughout the study period. Patients' Medicare entitlement status and mortality were tracked using a linked "denominator" file provided by the Centers for Medicare and Medicaid Services that accompanied the analytic data sets. International Classification of Diseases, Ninth Revision, Clinical Modification (1CD-9-CM: 81.53, 00.70-00.73) procedure codes were used to identify hip revision patients. We did not distinguish between the type of revision surgery (ie, acetabular vs femoral revision) in assigning patients to the study. Previous research has suggested that outcomes after revision are not sensitive to the type of revision surgery [19].

Our focus was to investigate outcomes as a function of the bearing surface used in the revision, which was identified in the revision claim record using an 1CD-9-CM code of 00.74 (M-PE); 00.75 (MOM); 00.76 (COC); and 00.77 (C-PE). These bearing surface

codes were introduced in October 2005 for M-PE, MOM, and COC bearings. 1n October 2006, the code for C-PE was introduced. As a result, the C-PE cohort has one less year of follow-up than the other bearing surface cohorts in this study. Between 2006 and 2013, about 31%-33% of revisions recorded in the Medicare database included a bearing surface code (Table 1). We investigated the difference in patient characteristics between those with known vs unknown bearings to better understand the study population. We observed a slight but significant difference in the patient characteristics that received a bearing code in the Medicare database and those that were uncoded. Overall, female patients, older patients, those needing Medicare buy-in (indicative of lower socioeconomic status), those with greater comorbidities (based on their Charlson score [23]), and patients residing in the South census region all had a lower probability of having their bearing type coded. Conversely, a higher total hospital charge was associated with a higher likelihood of having the bearing type coded.

Because MOM bearings are no longer widely used, we focused our research on the comparison of outcomes in patients with known M-PE, C-PE, and COC bearings at the time of revision THA. Outcomes of interest included 90-day readmission for any reason, periprosthetic joint infection, dislocation, rerevision, or death at any time point after the index revision procedure during the study period. Hospital readmission was determined by the appearance of any inpatient claims within 90 days of discharge from the index revision surgery, indicating a rehospitalization episode. The rate of 90-day hospital readmission is a quality measure defined by the Affordable Care Act of 2010 [24]. Periprosthetic joint infection was identified using an 1CD-9-CM diagnosis code of 996.66 [25], whereas dislocation was identified using 1CD-9-CM diagnosis codes of 718.35, 835.00-835.03, and 996.42 (effective October 2005) [26]. Rerevision was identified using the same revision codes listed previously to assemble the study cohorts, and death was identified using the previously mentioned denominator file accompanying the inpatient analytical data set. We used the Kaplan-Meier approach to inspect the crude (unadjusted) survivorship of the M-PE, C-PE, and COC cohorts for each of the outcomes of interest.

Propensity scores were developed to adjust for selection bias in the choice of bearing type at revision surgery. As discussed in a recent review [27], propensity scores were used to treat large data sets of retrospective registry data, such as are available via Medicare, for selection bias. The application of propensity scores represents an approach to treating Medicare data like a randomized clinical study by effectively balancing patient factors known to be

Table 1

Summary of Revision THAs Coded by Bearing Type in the Medicare Population (2005-2013).

Year THA Revisions Coded, Uncoded, Total % Coded

Coded by Bearing Subtotala Subtotal

C-PE COC M-PE

2005 0 50 959 1353 16,007 17,360 7.8

2006 64 189 3455 5114 11,001 16,115 31.7

2007 276 91 3400 5183 10,520 15,703 33.0

2008 279 68 3118 4856 10,429 15,285 31.8

2009 302 82 3039 4662 10,345 15,007 31.1

2010 400 75 3089 4626 10,499 15,125 30.6

2011 638 85 3201 4910 11,021 15,931 30.8

2012 922 101 3242 5014 10,900 15,914 31.5

2013 1384 128 3172 5203 11,309 16,512 31.5

Total 4265 869 26,675

THA, total hip arthroplasty; C-PE, ceramic-on-polyethylene; COC, ceramic-on-ceramic; M-PE, metal-on-polyethylene.

a 1n addition to C-PE, COC, and M-PE, this subtotal also includes those revisions that were coded with a metal-on-metal bearing.

S.M. Kurtz et al. / The Journal of Arthroplasty xxx (2016) 1-7

related to bearing surface selection. Specifically, the propensity score calculates a patient's chance of receiving a C-PE or COC implant, given certain patient and hospital factors. The actual bearing material received is independent of this propensity score. If different types of bearings were implanted for patients having identical propensity, the choice of bearing material can be thought of as randomly assigned. This ensures that the outcomes associated with each bearing type are not confounded by patient factors. The propensity score was calculated for each patient using the following predictors: age, sex, region, race, Medicare buy-in (a proxy for socioeconomic status), Charlson Comorbidity Score, revision calendar year, length of stay, charge amount, hospital volume, surgeon volume, principal diagnosis, hospital location (urban or rural), hospital type (eg, public, private), size of the hospital, and 2-way interactions among age, gender, race, Charlson score, hospital size, and hospital type.

Cox regression incorporating propensity score stratification (10 levels) was then used to evaluate the impact of bearing surface selection on outcomes, after adjusting for patient-, hospital-, and surgeon-related factors. The Cox model was stratified into 10 propensity strata. The Cox model combined the likelihood functions from each stratum and estimated an overall hazard ratio and corresponding confidence intervals. The Cox regression model incorporated the main study variables: bearing type (C-PE, COC, or M-PE) and the following potential confounding variables: patient age; race; census region; patient diagnosis of diabetes, heart disease or obesity; patient's Charlson Comorbidity Index; hospital type, location, and size; hospital procedure volume; surgeon procedure volume; total hospital charges; length of stay; Medicare buy-in; operating room charges; and calendar year. All statistical analyses were performed using SAS version 9.4 (Cary, NC).

Results

A total of 31,809 Medicare patients who underwent R-THA between 2005 and 2013 with known bearing types were identified from the Medicare 100% inpatient sample administrative database, including 4,265 patients who received C-PE, 869 patients who received COC, and 26,675 patients who received M-PE bearings (Table 1). The relative usage of ceramic bearings varied over time (Fig. 1). In 2007, the first calendar year in which all 4 bearing codes (including MOM) were fully implemented; C-PE and COC usage was 5.3% and 1.8%, respectively. By 2013, C-PE and COC bearings usage increased to 26.6% and 2.5%, respectively (Fig. 1). During this time

"S- an

2004 2006 2008 2010 2012 2014

Fig. 1. Reported bearing usage of C-PE, COC, and M-PE in revision total hip arthroplasty in the Medicare population between 2005 and 2013. C-PE, ceramic on polyethylene; COC, ceramic on ceramic; M-PE, metal on polyethylene.

period, the relative usage of MOM bearings declined from 27.3% to 10% among R-THAs.

The R-THA patients in this study were 60% female, on average (±standard deviation) 75 (±10) years old, 94% white, and 50% had no comorbidities (corresponding to a Charlson score of 0, Table 2). The usage of C-PE implants in R-THA was the highest in the 65- to 69-year cohort (21% of revision operations in that cohort) and lowest for the 85+ year cohort (7% of revision operations). The opposite trend was observed for M-PE implants: 76% of patients in the 65- to 69-year cohort received M-PE implants compared with 90% of the 85+ year cohort. Percentage of patients receiving COC implants, on the other hand, was uniform among all age cohorts. Utilization of C-PE, COC, and M-PE was comparable in male and female patients. Although patients with bearing codes were dominantly white in this study, the utilization of ceramic bearings was homogeneous across races. For instance, 13% of white patients, 14% of black patients, and 10% of patients of unknown/other races received C-PE bearings, whereas the utilization of COC bearings were 3%, 4%, and 4% for white, black, and unknown/other races, respectively. While patients in the Midwest received the largest number of M-PE implants (7615), patients in the South received the

Table 2

Overall Patient Demographics.

Demographic C-PE COC M-PE Total

N % N % N %

Age (y)

65-69 1705 21 270 3 6206 76 8181

70-74 1050 15 188 3 5884 83 7122

75-79 753 11 155 2 6187 87 7095

80-84 488 9 154 3 5006 89 5648

85+ 269 7 102 3 3392 90 3763

Gender

M 1712 14 308 2 10,545 84 12,565

F 2553 13 561 3 16,130 84 19,244

Gender, age (y)

M, 65-69 720 20 111 3 2691 76 3522

M, 70-74 437 14 70 2 2520 83 3027

M, 75-79 297 11 58 2 2437 87 2792

M, 80-84 176 9 50 2 1779 89 2005

M, 85+ 82 7 19 2 1118 92 1219

F, 65-69 985 21 159 3 3515 75 4659

F, 70-74 613 15 118 3 3364 82 4095

F, 75-79 456 11 97 2 3750 87 4303

F, 80-84 312 9 104 3 3227 89 3643

F, 85+ 187 7 83 3 2274 89 2544

White 4034 13 796 3 25,205 84 30,035

Black 179 14 53 4 1022 81 1254

Other/unknown 52 10 20 4 448 86 520

Census region

Midwest 896 10 179 2 7615 88 8690

North East 767 11 193 3 5745 86 6705

South 1517 16 325 4 7440 80 9282

West 1085 15 172 2 5875 82 7132

Charlson score

0 2464 15 476 3 13,888 83 16,828

1-2 1453 13 291 3 9808 85 11,552

3-4 280 11 74 3 2282 87 2636

5+ 68 9 28 4 697 88 793

Medicare buy-in

No buy-in 3985 14 778 3 24,522 84 29,285

With buy-in 280 11 91 4 2153 85 2524

Reason for revisiona

1nfection 223 13 41 2 1423 84 1687

Dislocation 482 9 138 3 4680 88 5300

Loosening 783 10 159 2 6625 88 7567

Other 2491 16 454 3 12,173 81 15,118

M, male; F, female; C-PE, ceramic-on-polyethylene; COC, ceramic-on-ceramic; M-PE, metal-on-polyethylene. a Reason for revision is only listed for most common revisions (>1%).

S.M. Kurtz et al. / The Journal of Arthroplasty xxx (2016) 1-7

largest number of C-PE (1517) and COC (325) implants. Among infection, dislocation, and loosening as reasons for revision, loosening occurred most frequently among all 3 bearing types.

For R-THA patients treated with C-PE bearings, there was reduced risk of 90-day readmission (hazard ratio [HR]: 0.90,95% CI: 0.84-0.96, P = .007; Fig. 2). We also observed a trend for reduced risk of infection with C-PE (HR: 0.88, 95% CI: 0.74-1.04) that did not reach statistical significance (P = .14). Based on Kaplan-Meier analysis, the crude (unadjusted) survivorship at 5 years, using rerevision as an end point (with 95% CIs), was 83.7% (82.8%-84.6%) for M-PE, 82.2% (79.0%-84.9%) for C-PE, respectively. After propensity score stratification and adjustment for confounders, there was no significant difference in risk of rerevision (P = .99) or mortality (P = .51) for the C-PE bearing cohorts when compared with M-PE. When recipients of C-PE and M-PE are compared, reason for revision was a risk factor for all 5 outcomes (death, dislocation, rerevision, infection, readmission) analyzed in this study (Fig. 2). Charlson score and length of stay were risk factors for all outcomes except rerevision. Finally, race was a risk factor for all outcomes except infection.

For R-THA patients treated with COC, there was reduced risk of dislocation (HR: 0.76, 95% CI: 0.58-0.99, P = .04; Fig. 3). Based on Kaplan-Meier analysis, the crude (unadjusted) survivorship at 5 years, using rerevision as an end point (with 95% CIs), was 85.0%

(79.7%-88.9%) for the COC cohort. After propensity score stratification and adjustment for confounders, there was no significant difference in risk of rerevision (P = .31) or mortality (P = .45) for the COC bearing cohorts when compared with the M-PE cohort. When C-PE and M-PE cohorts were compared, reason for revision was a risk factor for all 5 outcomes (death, rerevision, infection, readmission) analyzed in this study (Fig. 3). Charlson score and length of stay were risk factors for all outcomes except rerevision. Race and year of implantation were risk factors for all outcomes except infection. Finally, age was a risk factor for all outcomes except dislocation.

Overall, age was the highest relative importance predictor of mortality, whereas reason for revision was the predictor with highest relative importance for dislocation, infection, readmission, rerevision for COC bearings for both COC and C-PE bearings (Figs 2 and 3).

Discussion

In this study of all comers for revision total hip surgery in the elderly Medicare population, we asked how the use of ceramic bearings changed over time and whether the type of ceramic bearing influenced outcomes relative to M-PE. Between 2006 and 2013, we observed an increase in the reported usage of C-PE

Fig. 2. Relative importance of patient, clinical, and institution factors on risk of mortality, dislocation, infection, 90-day readmission, and re-revision after revision total hip arthroplasty using C-PE vs M-PE bearings. The effect size for each factor is judged by the relative magnitude of the model Wald chi-squared statistic. LOS, length of stay; OR, operating room.

S.M. Kurtz et al. / The Journal of Arthroplasty xxx (2016) 1-7

Fig. 3. Relative importance of patient, clinical, and institution factors on risk of mortality, dislocation, infection, 90-day readmission, and re-revision after revision total hip arthroplasty using COC vs M-PE bearings. The effect size of each factor is judged by the relative magnitude of the model Wald chi-squared statistic.

bearings in revision surgeries for Medicare beneficiaries. We found no evidence to suggest that ceramic bearings were associated with worse outcomes than M-PE bearings when used in revisions. Conversely, we found support for our hypotheses that ceramic bearings may improve certain outcomes after revision surgery, such as 90-day readmission, dislocation, and perhaps infection; however, the results were bearing- and outcome-specific.

We would like to highlight several limitations of our study for the reader. First, our analysis was based on a retrospective analysis of administrative billing data, which was limited to the 1CD-9-CM classification of procedures and diagnoses. Because the Medicare data set does not include clinical information, it was not possible for us to evaluate clinical factors such as soft tissue damage, the presence of metal-related pathology, or osteolysis in our study. We attempted to include and adjust for revision procedure complexity and difficulty due to patient and clinical factors by considering the patients' Charlson Comorbidity Index and length of stay as proxies. Furthermore, our analysis methodology including propensity scores was designed to adjust for selection bias in the assignment of ceramic bearings in the comparison with M-PE bearings and overcome the limitation of a restrospective nonrandomized study design.

Second, our study was limited to 31%-33% revision patients in Medicare with known billing codes, which are optional and not required for hospital reimbursement [28]. We addressed this limitation using propensity scores to adjust for selection bias

among the patients who were coded for bearing type. Third, the patient population was limited to those >65 years in age who were covered by Medicare; our findings may not necessarily apply to younger patients. Fourth, we included all comers to revision in this analysis and did not subclassify the treated population into septic vs aseptic revisions, for example. Fifth, our analysis was limited to outcomes recorded in an inpatient setting, which may have underestimated the risk of dislocation because not all of these events may be treated with clinical intervention requiring an overnight stay.

Sixth, the bearing codes for both types of ceramic bearings and the control (M-PE) bearings are general and do not distinguish between the types of polyethylene formulations, different types of ceramic biomaterials, or head size that were used clinically during the study period. 1n the first decade of the 2000s, many different formulations of highly cross-linked and thermally stabilized polyethylene were clinically introduced, including second-generation materials [29]. In addition, the type of ceramics available in the United States also varied in the time period of this study, with the increased adoption of zirconia-toughened alumina after 2003 [30]. Furthermore, changes in femoral head size were clinically introduced during this time period to improve joint stability and reduce dislocation risk [26]. Thus, the granularity of the administrative bearing codes limits our ability to answer questions about specific formulations of bearing materials and head size.

S.M. Kurtz et al. / The Journal of Arthroplasty xxx (2016) 1—7

Nevertheless, these limitations are offset by the use of the largest (100%) nationally representative data set available for the elderly population, in which ceramic bearings were used in about a quarter of revisions for any reason. The smallest cohort in our study (for revision COC patients) is approximately the same size as the entire study population of revision resurfacings considered by Wong et al [19] in their analysis of the Australian registry data. Because of the sample sizes necessary to identify potentially subtle trends in administrative data, very large data sets, such as orthopedic registries or the Medicare data set we used, are well suited to examining the outcomes after revision surgery.

The utilization of alternative bearings has been previously examined in the context of primary THA and as a result of concerns with MOM. Although the usage of MOM bearings has previously been studied in revision surgeries, we are aware of no previous utilization studies of ceramic bearings in revision for US patients, making comparison of our results difficult. Clinicians were attracted to alternative bearings because of the larger head sizes that could be achieved with MOM to improve joint stability and reduce dislocation risk. Since that time, larger diameter ceramic heads (up to 44 mm in diameter) are now clinically available in the United States. Also, there is greater understanding based on international registry data that head sizes >36 mm diameter may not necessarily provide improved dislocation risk in clinical practice. Finally, concerns with taper corrosion using cobalt-chromium femoral heads may also be playing a role in surgeon decisions to increasingly adopt ceramic heads in both a primary and revision scenario. Although it is not possible for us to identify from claims data which of the aforementioned trends are responsible for the increase in ceramic bearing usage among Medicare beneficiaries undergoing revision, the trends are temporally coincident with the reduction of utilization of MOM bearings and, to a growing extent, reduced usage of M-PE bearings as well.

Few studies have examined the rates of THA rerevision for large populations [19,31]. Based on the Australian registry, Wong et al [19] found a 26% rerevision rate at 10 years, with no significant effect of the bearing surface. Examining the elderly Medicare population, Ong et al [31] reported 81% survivorship at 5 years after revision, which is comparable to the survivorship for the 3 cohorts recorded in the present study. The difference in lower rerevision rates between the study by Wong et al and the Medicare studies is most likely due to differences in the patient mix in the 2 studies. In the study by Wong et al [19], the patients were all revisions of hip resurfacing performed in Australia, which according to the 2014 registry report [5] was most often performed in male patients aged <65 years.

For the C-PE cohort, the reduced 90-day admission rates and trend for reduced risk of infection were independent findings. Recent studies presented at national conferences suggest that ceramics may be more resistant to infection than cobalt-chromium surfaces [32-35], which would help explain these results viewed here. Further analysis is needed to better understand the association between infection, early readmission, and the use of C-PE bearings.

Previous studies have reported that COC bearings have a lower risk of dislocation than M-PE bearings in primary THA [9,10] and revision THA [36]. Hernigou et al [9] specifically addressed this topic with primary THA, comparing the risk of dislocation in C-PE and COC bearings that were implanted between 1972 and 1982. Interestingly, they noted biologic factors that differed between the C-PE and COC bearings, which enabled significantly greater capsular thickening and, hypothetically, greater long-term dislocation resistance in the COC cohort. Also, the C-PE incorporated historical, gamma-air-sterilized polyethylene for the acetabular liner that would generate biologically active wear particles.

In summary, our results indicate that, after adjusting for selection bias and various confounding patient-, surgeon-, and hospital-related factors, Medicare patients treated in a revision scenario with ceramic bearings exhibit similar risk of rerevision or mortality as those treated with M-PE bearings. Conversely, we found an association between the use of specific ceramic bearings in R-THA and reduced risk of readmission (C-PE) and dislocation (COC). The findings of this study support further research into the association between ceramic bearings in R-THA and lower risk of hospital readmission, dislocation, and, potentially, infection.

References

1. Boutin P. [Alumina and its use in surgery of the hip. (Experimental study)]. Presse Med 1971;79(14):639.

2. Boutin P. [Total arthroplasty of the hip by fritted aluminum prosthesis. Experimental study and 1st clinical applications]. Revue de chirurgie orthopedique et reparatrice de l'appareil moteur 1972;58(3):229.

3. Shikata T, Oonishi H, Hashimato Y, et al. Wear resistance of irradiated UHMW polyethylenes to Al2O3 ceramics in total hip prostheses. Transactions of the 3rd Annual Meeting of the Society for Biomaterials: 118.1977.

4. National Joint Registry for England and Wales. 11th Annual NJR Report. London: NJR; 2014.

5. Australian Orthopaedic Association National Joint Replacement Registry. Annual Report. Adelaide: AOA; 2014.

6. Meftah M, Klingenstein GG, Yun RJ, et al. Long-term performance of ceramic and metal femoral heads on conventional polyethylene in young and active patients: a matched-pair analysis. J Bone Joint Surg Am 2013;95(13):1193.

7. Wang S, Zhang S, Zhao Y. A comparison of polyethylene wear between cobalt-chrome ball heads and alumina ball heads after total hip arthroplasty: a 10-year follow-up. J Orthop Surg Res 2013;8:20.

8. Kurtz SM, Kocagoz SB, Hanzlik JA, et al. Do ceramic femoral heads reduce taper fretting corrosion in hip arthroplasty? A retrieval study. Clin Orthop Relat Res 2013;471(10):3270.

9. Hernigou P, Homma Y, Pidet O, et al. Ceramic-on-ceramic bearing decreases the cumulative long-term risk of dislocation. Clin Orthop Relat Res 2013;471(12):3875.

10. Hu D, Tie K, Yang X, et al. Comparison of ceramic-on-ceramic to metal-on-polyethylene bearing surfaces in total hip arthroplasty: a meta-analysis of randomized controlled trials. J Orthop Surg Res 2015;10(1):22.

11. Mendenhall S. Hospital resources and implant cost management—a 2013 update. Orthop Network News 2014;25(3):9.

12. Lee GC, Kim RH. Reliability of ceramic heads in over 5.7 million hip replacements. Trans American Assoc Hip and Knee Surgeons (AAHKS) 2014;15.

13. Massin P, Lopes R, Masson B, et al. Does Biolox Delta ceramic reduce the rate of component fractures in total hip replacement? Orthop Traumatol Surg Res 2014;100(6 Suppl):S317.

14. Australian Orthopaedic Association National Joint Replacement Registry. Annual Report. Adelaide: AOA; 2015.

15. Owen DH, Russell NC, Smith PN, et al. An estimation of the incidence of squeaking and revision surgery for squeaking in ceramic-on-ceramic total hip replacement: a meta-analysis and report from the Australian Orthopaedic Association National Joint Registry. Bone Joint J 2014;96-B(2):181.

16. Kurtz SM, Ong KL, Lau E, et al. Prevalence of metal-on-metal bearings in the United States. In: Kurtz SM, Greenwald AS, Mihalko WM, Lemons J, editors. Metal-on-Metal Total Hip Replacement Devices, STP 1560. Conshohocken, PA: ASTM; 2013. 3.

17. Hannouche D, Delambre J, Zadegan F, et al. Is there a risk in placing a ceramic head on a previously implanted trunion? Clin Orthop Relat Res 2010;468(12):3322.

18. Matar WY, Restrepo C, Parvizi J, et al. Revision hip arthroplasty for ceramic-on-ceramic squeaking hips does not compromise the results. J Arthroplasty 2010;25(6 Suppl):81.

19. Wong JM, Liu YL, Graves S, et al. What is the rerevision rate after revising a hip resurfacing arthroplasty? Analysis from the AOANJRR. Clin Orthop Relat Res 2015;473:3458.

20. Jack CM, Molloy DO, Walter WL, et al. The use of ceramic-on-ceramic bearings in isolated revision of the acetabular component. Bone Joint J 2013;95-B(3):333.

21. Cooper HJ, Della Valle CJ, Berger RA, et al. Corrosion at the head-neck taper as a cause for adverse local tissue reactions after total hip arthroplasty. J Bone Joint Surg Am 2012;94(18):1655.

22. Cooper HJ. The local effects of metal corrosion in total hip arthroplasty. Orthop Clin North Am 2014;45(1):9.

23. Deyo RA, Cherkin DC, Ciol MA. Adapting a clinical comorbidity index for use with 1CD-9-CM administrative databases. J Clin Epidemiol 1992;45(6):613.

24. Saucedo JM, Marecek GS, Wanke TR, et al. Understanding readmission after primary total hip and knee arthroplasty: who's at risk? J Arthroplasty 2014;29(2):256.

25. Kurtz SM, Lau E, Watson H, et al. Economic burden of periprosthetic joint infection in the United States. J Arthroplasty 2012;27(8 Suppl):61.

26. Malkani AL, Ong KL, Lau E, et al. Early- and late-term dislocation risk after primary hip arthroplasty in the Medicare population. J Arthroplasty 2010;25(6 Suppl):21.

S.M. Kurtz et al. / The Journal of Arthroplasty xxx (2016) 1-7 7

27. Inacio MC, Chen Y, Paxton EW, et al. Statistics in brief: an introduction to the use of propensity scores. Clin Orthop Relat Res 2015;473:2722.

28. Bozic K, Ong K, Lau E, et al. The comparative effectiveness of total hip bearing surfaces in the Medicare population. Trans American Assoc Hip and Knee Surgeons (AAHKS) 2009;19:1.

29. Kurtz SM, Patel JD. The clinical performance of highly cross-linked UHMWPE in hip replacements. 1n: Kurtz SM, editor. The UHMWPE biomaterials handbook: ultra-high molecular weight polyethylene in total joint replacement and medical devices. 3rd ed. Oxford, UK: Elsevier, Inc; 2016. p. 57.

30. Kurtz SM, Ong K. Contemporary total hip arthroplasty: alternative bearings. 1n: Kurtz SM, editor. The UHMWPE biomaterials handbook: ultra-high molecular weight polyethylene in total joint replacement and medical devices. 3rd ed. Oxford, UK: Elsevier, 1nc; 2016. p. 72.

31. Ong KL, Lau E, Suggs J, Kurtz SM, Manley MT. Risk of subsequent revision after primary and revision total joint arthroplasty. Clin Orthop Relat Res 2010;468(11):3070.

32. Streicher RM, Porporati AA, Preuss R, Leto A. Can ceramic femoral heads reduce the revision burden in THA? Transactions of the Japanese Hip Society: Abstract. 2014.

33. Smith L, Alijanipour P, Restrepo C, Maltenfort M, Parvizi J, Malkani A. Periprosthetic joint infection: could the bearing surface play a role? Transactions of the Eastern Orthopaedic Association: Abstract. 2014.

34. Graves SE, Lorimer M, Bragdon C, Muratoglu O, Malchau H. Reduced risk of revision for infection when a ceramic bearing surface is used. Abstract, Transactions 28th ISTA Congress: Abstract. 2015.

35. Pitto R, Baker K, Sedel L. Periprosthetic joint infection in total hip arthroplasty. Does the bearing surface play a role? A registry study. Transactions of the International Hip Society: Abstract. 2015.

36. Hernigou P, Roussignol X, Delambre J, Poignard A, Flouzat-Lachaniette CH. Ceramic-on-ceramic THA associated with fewer dislocations and less muscle degeneration by preserving muscle progenitors. Clin Orthop Relat Res 2015;473:3762.