joint registries as continuous surveillance systems: the experience of the catalan arthroplasty...

7
Joint Registries as Continuous Surveillance Systems: The Experience of the Catalan Arthroplasty Register (RACat) Alejandro Allepuz, MD, MPH a, b, c , Olga Martínez, MSc a , Cristian Tebé, MSc, MPH a, c , Joan Nardi, MD d , Frederic Portabella, MD e , Mireia Espallargues, MD, MPH, PhD a, f On behalf of the Catalan Arthroplasty Register (RACat) a Department of Health, Catalan Agency for Quality and Healthcare Assessment; Catalan Health Service, Barcelona, Spain b Àmbit d'Atenció Primària Costa de Ponent, Catalan Institute of Health, l'Hospitalet de Llobregat, Spain c Red de Investigación en Servicios Sanitarios en Enfermedades Crónicas (REDISSEC), Valencia, Spain d Hospital Universitari de la Vall d'Hebron, Barcelona, Spain e Hospital Universitari de Bellvitge, L'Hospitalet de Llobregat, Catalonia, Spain f CIBER de Epidemiología y Salud Pública (CIBERESP), Spain abstract article info Article history: Received 24 May 2013 Accepted 30 July 2013 Keywords: hip arthroplasty knee arthroplasty registries outcomes assessment The aim was to present results on prosthesis performance in Catalonia for the period 20052010. All publicly funded hospitals submit in an electronic format data on hip and knee arthroplasties: patients' insurance identication number, hospital, joint (hip/knee), type of arthroplasty (primary/revision), side (right/left), date of surgery and prosthesis (manufacturer name and catalogue number). A standard survival analysis based on KaplanMeier estimation was carried out. Fifty-two hospitals have sent information to the RACat which has data on 36,951 knee and 26,477 hip arthroplasties. Cumulative prostheses revision risks at 3 years were 3.3% (95% CI: 3.13.6) for knee, 2.9% (95% CI: 2.53.3) for total hip and 2.5% (95% CI: 2.03.1) for partial hip. When compared to other registries a higher risk of revision was observed. © 2014 Elsevier Inc. All rights reserved. Primary knee arthroplasty rates increased vefold between 1994 and 2005 in Catalonia (Spain) and at the end of this period, Catalonia was the Spanish region with the highest joint arthroplasty rates [1,2]. This relevant increase could be partly explained by a specic Governmental plan to reduce surgical waiting lists starting in 1999, although changes observed on patients' demographic characteristics operated on primary knee arthroplasties reveal that surgical in- dications could have also broadened in Catalonia [2], which might contribute to an increase in the number of revision arthroplasties performed in the future as more patients are operated on. Arthroplasty registries have emerged as an appropriate method to systematically obtain information on arthroplasty outcomes with the advantage of its naturalistic approach: they evaluate prosthesis performance in normal conditions of use, implanted by orthopedic surgeons with different skills and in a large variety of hospitals [35]. Furthermore, joint arthroplasty registries utility has been highlighted as a method for post-marketing surveillance [5,6]. Besides, joint registries could be a useful tool to assess hospitals' and surgeons' performance allowing for benchmarking between them in order to improve outcomes [7]. The evolution of knee arthroplasty rates and its likely impact on revision surgery, led in 2005 to the development and implementation of an arthroplasty register in Catalonia. The aim of this paper was to present its functioning and results on data quality, and prostheses characteristics and performance for the period 20052010. Patients and Methods Catalonia is an autonomous community in the northeast part of Spain that had a population of 7,565,603 inhabitants at the start of 2012. In Spain, the National Healthcare Service (NHS) is nanced by the general taxes and medical visits and hospital admissions are fully covered by the NHS. The Catalan Arthroplasty Register (RACat) structure includes the following: a three-member Steering Committee to monitor the execution of the project, an Advisory Committee to supervise and assess the process of development and analysis of data, and a Plenary Council where all the register collaborators at each healthcare center are represented [8]. At the start of the project it was decided that hospital participation would be on a voluntary basis and that only knee and hip prostheses were going to be included, although in the future it is planned to incorporate other types of prostheses. Hospitals started to send information by the end of 2005. The Journal of Arthroplasty 29 (2014) 484490 The Conict of Interest statement associated with this article can be found at http:// dx.doi.org/10.1016/j.arth.2013.07.048. Reprint requests: Alejandro Allepuz, MD, MPH, Department of Health, Catalan Agency for Health Information, Assessment and Quality, Catalan Health Service, Catalonia, Carrer de Roc Boronat, 81-95 2ª Planta, 08005 Barcelona, Spain. 0883-5403/2903-0008$36.00/0 see front matter © 2014 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.arth.2013.07.048 Contents lists available at ScienceDirect The Journal of Arthroplasty journal homepage: www.arthroplastyjournal.org

Upload: mireia

Post on 05-Jan-2017

216 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Joint Registries as Continuous Surveillance Systems: The Experience of the Catalan Arthroplasty Register (RACat)

The Journal of Arthroplasty 29 (2014) 484–490

Contents lists available at ScienceDirect

The Journal of Arthroplasty

j ourna l homepage: www.ar throp lasty journa l .o rg

Joint Registries as Continuous Surveillance Systems: The Experience of the CatalanArthroplasty Register (RACat)

Alejandro Allepuz, MD, MPH a,b,c, Olga Martínez, MSc a, Cristian Tebé, MSc, MPH a,c, Joan Nardi, MD d,Frederic Portabella, MD e, Mireia Espallargues, MD, MPH, PhD a,f

On behalf of the Catalan Arthroplasty Register (RACat)a Department of Health, Catalan Agency for Quality and Healthcare Assessment; Catalan Health Service, Barcelona, Spainb Àmbit d'Atenció Primària Costa de Ponent, Catalan Institute of Health, l'Hospitalet de Llobregat, Spainc Red de Investigación en Servicios Sanitarios en Enfermedades Crónicas (REDISSEC), Valencia, Spaind Hospital Universitari de la Vall d'Hebron, Barcelona, Spaine Hospital Universitari de Bellvitge, L'Hospitalet de Llobregat, Catalonia, Spainf CIBER de Epidemiología y Salud Pública (CIBERESP), Spain

a b s t r a c ta r t i c l e i n f o

The Conflict of Interest statement associated with thdx.doi.org/10.1016/j.arth.2013.07.048.

Reprint requests: Alejandro Allepuz, MD, MPH, DAgency for Health Information, Assessment and QuCatalonia, Carrer de Roc Boronat, 81-95 2ª Planta, 08005

0883-5403/2903-0008$36.00/0 – see front matter © 20http://dx.doi.org/10.1016/j.arth.2013.07.048

Article history:Received 24 May 2013Accepted 30 July 2013

Keywords:hip arthroplastyknee arthroplastyregistriesoutcomes assessment

The aim was to present results on prosthesis performance in Catalonia for the period 2005–2010. All publiclyfunded hospitals submit in an electronic format data on hip and knee arthroplasties: patients' insuranceidentification number, hospital, joint (hip/knee), type of arthroplasty (primary/revision), side (right/left),date of surgery and prosthesis (manufacturer name and catalogue number). A standard survival analysisbased on Kaplan–Meier estimation was carried out. Fifty-two hospitals have sent information to the RACatwhich has data on 36,951 knee and 26,477 hip arthroplasties. Cumulative prostheses revision risks at 3 yearswere 3.3% (95% CI: 3.1–3.6) for knee, 2.9% (95% CI: 2.5–3.3) for total hip and 2.5% (95% CI: 2.0–3.1) for partialhip. When compared to other registries a higher risk of revision was observed.

is article can be found at http://

epartment of Health, Catalanality, Catalan Health Service,Barcelona, Spain.

14 Elsevier Inc. All rights reserved.

© 2014 Elsevier Inc. All rights reserved.

Primary knee arthroplasty rates increased fivefold between 1994and 2005 in Catalonia (Spain) and at the end of this period, Cataloniawas the Spanish region with the highest joint arthroplasty rates [1,2].This relevant increase could be partly explained by a specificGovernmental plan to reduce surgical waiting lists starting in 1999,although changes observed on patients' demographic characteristicsoperated on primary knee arthroplasties reveal that surgical in-dications could have also broadened in Catalonia [2], which mightcontribute to an increase in the number of revision arthroplastiesperformed in the future as more patients are operated on.

Arthroplasty registries have emerged as an appropriate method tosystematically obtain information on arthroplasty outcomes with theadvantage of its naturalistic approach: they evaluate prosthesisperformance in normal conditions of use, implanted by orthopedicsurgeons with different skills and in a large variety of hospitals [3–5].Furthermore, joint arthroplasty registries utility has been highlightedas a method for post-marketing surveillance [5,6]. Besides, jointregistries could be a useful tool to assess hospitals' and surgeons'

performance allowing for benchmarking between them in order toimprove outcomes [7].

The evolution of knee arthroplasty rates and its likely impact onrevision surgery, led in 2005 to the development and implementationof an arthroplasty register in Catalonia. The aim of this paper was topresent its functioning and results on data quality, and prosthesescharacteristics and performance for the period 2005–2010.

Patients and Methods

Catalonia is an autonomous community in the northeast part ofSpain that had a population of 7,565,603 inhabitants at the start of2012. In Spain, the National Healthcare Service (NHS) is financed bythe general taxes and medical visits and hospital admissions are fullycovered by the NHS. The Catalan Arthroplasty Register (RACat)structure includes the following: a three-member Steering Committeeto monitor the execution of the project, an Advisory Committee tosupervise and assess the process of development and analysis of data,and a Plenary Council where all the register collaborators at eachhealthcare center are represented [8]. At the start of the project it wasdecided that hospital participation would be on a voluntary basis andthat only knee and hip prostheses were going to be included, althoughin the future it is planned to incorporate other types of prostheses.Hospitals started to send information by the end of 2005.

Page 2: Joint Registries as Continuous Surveillance Systems: The Experience of the Catalan Arthroplasty Register (RACat)

485A. Allepuz et al. / The Journal of Arthroplasty 29 (2014) 484–490

Data Collection and Management

To ensure register viability a minimum data set was establishedprioritizing information already available in the hospitals' informationsystems to avoid that RACat implementation was burdensome forsurgeons. The minimum of information for a register to work wasalready available in electronic format in all hospitals [9]: patient'sinsurance identificationnumber, joint (kneeorhip), type of arthroplasty(primary or revision), side of operation (right or left), date of surgeryand prosthesis data (manufacturer and catalogue number; thisinformation is also sent for cement in cemented implants).

The information on the type of arthroplasty, date of surgery andprosthesis is routinely collected in hospitals by the Department ofAccounting, and side of operation is also gathered as part of the datainformed to the Catalan Health Service (CHS). The Department ofInformatics at each hospital prepares an electronic file including theminimum of information required by our register and they send it tothe RACat through the CHS Web application. The RACat database ispart of the CHS information system, allowing us to accomplish withthe Spanish Personal Data Protection Act. RACat's data quality isreviewed twice a year and queries are sent back to the collaborator ofthe register at each hospital who is responsible for solving them.

The patient's insurance identification number (unique for eachperson who lives in Catalonia) allows us to incorporate to the RACatinformation from the CHS on age, sex and vital status (death date). Onthe other hand, in Spain it ismandatory for hospitals to sendaminimumof information from each patient discharged, including, among other,data on themain diagnosis (reason for surgery) and comorbidities.Maindiagnosis and comorbidities are coded through the InternationalClassification of Diseases, Ninth Revision (ICD-9) according to thedischarge report of the physician responsible for the patient. We alsoincorporate these data to the RACat linking the information through thepatient's insurance identification number and date of admission.

Data on prosthesis characteristics are obtained linking data sent byhospitals on the prosthesis (manufacturer and catalogue number)with our prostheses catalogue, which is built from the informationsent by all the manufacturers commercialising prostheses in Catalo-nia. This catalogue includes data on type of component, brand andsome structural characteristics (whether the component is cementedor not and bearing surface).

Register Management and Funding

The RACat is managed by the CAQHA, a public company of theDepartment of Health attached to the CHS. To increase efficiencywhen implementing the RACat, it was decided to integrate the RACatdatabase in the information system of the CHS who is responsible forits maintenance whereas the CAQHA is responsible for data collection,data management and data analysis. One full-time (project manager)and three part-time (two epidemiologists and a statistician) CAQHAprofessionals are responsible for the project. Hospitals do not receive aspecific budget for data processing.

Data Quality

Data quality is assessed through: data coverage, the percentage ofprostheses used in primary operations that can be classified into anyof the prostheses types (Table 2) and the percentage of operationswhose side is informed.

Coverage of the register is calculated through the percentage ofarthroplasties the RACat has information on over the total number ofarthroplasties performed in Catalonia according to the hospitalminimum discharge data set (as mentioned, in Spain it is mandatoryfor all hospitals to send a minimum of information from eachdischarge; in consequence the hospital minimum discharge data sethas information on all the arthroplasties performed in Catalonia).

Prostheses identification depends on the information sent by thehospital: manufacturer name and catalogue number from eachcomponent of the prosthesis. The prosthesis cannot be classified ifthe hospital has not sent any of these data, if data includetranscription errors that preclude its identification or if these dataare not included in our prostheses catalogue. To overcome this lastsituation and keep our catalogue up to date, all catalogue numbersthat are not identified are sent to the manufacturer in order to updatenewmarketed prostheses or new catalogue numbers from prosthesesalready commercialized. If the manufacturer cannot identify thecatalogue number, it is considered a query that should be solved bythe collaborator of the register in each hospital. In the case ofidentifying a new manufacturer from information submitted by thehospital, the new company is contacted by the project manager inorder to obtain information on their products.

Prostheses Classification

Our prostheses catalogue includes for each catalogue number theprostheses component it belongs to (hip: stem, head, inlay or cup;knee: tibial, femoral, inlay or patella) and the fixation technique(cemented or uncemented). In hip prostheses we have information onthe type of stem (monoblock or modular), head (bipolar, unipolar,conventional total hip arthroplasty [THA] or resurfacing) and cup(conventional THA or resurfacing). In the case of knee, for eachcomponent (tibial, femoral, inlay and patella) we have also informa-tion on the prostheses type they belong to: cruciate retaining (CR),posterior stabilized (PS), constrained, hinge, unicompartmental orfemoropatellar. Hip and knee prostheses are classified into differenttypes (Table 2) depending on the implanted components. Weconsider a prosthesis as CR or PS depending on the inlay description.

Analysis

A descriptive analysis was performed on data quality and patients'characteristics: age, sex, more frequent comorbidities according toElixhauser classification algorithm [10], reason for surgery andprostheses characteristics (prosthesis type, fixation technique, bearingsurface). Burden of revision, representing the percentage of revisionarthroplasties performed over the total number of arthroplasties, wasalso calculated. For survival analysis revision, operations where aprosthesis component was exchange, removed or added were consid-ered a failure of the primary prosthesis. Only the first revision of eachprimary operation was included in the analysis. Survival analysiscomprised global and according to bearing surface cumulative revisionrisk rates estimation at 1 and 3 years after primary operation, applyingthe Kaplan–Meier method. Cox proportional regression modeling wasused to analyze differences in revision risk according to fixationtechnique adjusted by age and sex for the most common types ofprostheses in our register (knee: CR, PS; hip: total conventional).

Results

Fifty-two publicly funded hospitals are sending information onknee and hip arthroplasties to the RACat. For the period 2005–2010the RACat has information on 36,951 knee arthroplasties (33,639primary operations and 3312 revisions; burden of revision 9.0%) and26,477 hip arthroplasties (23,762 primary operations and 2715revision; burden of revision 10.2%). Osteoarthritis was the mostcommon reason for surgery and infections represented 14% of revisionsurgery (Table 1).

Data Quality

The quality of information has improved between 2005 and 2010.Data coverage rose from 24.4% to 78.3% in knee arthroplasty and from

Page 3: Joint Registries as Continuous Surveillance Systems: The Experience of the Catalan Arthroplasty Register (RACat)

Table 1Patients' Characteristics.

Knee Patients Hip Patients

Primary Revision Total Partial Revision

Total patients in RACat linked to the minimum discharge data seta 30,837 2696 13,809 7040 2045Women (%) 72.1 72.1 53.6 77.1 58.4Mean age in years (SD) 72.0 (7.6) 72.3 (8.3) 68.8 (12.3) 84.3 (7.2) 72.3 (11.9)Reason for surgery (%)b

Osteoarthritis 98.2 – 77.9 – –

Hip fracture – – 13.4 95.8 –

Avascular necrosis – – 4.9 0.2Mechanical complicationsc – 60.0 – – 68.5Aseptic loosening – 12.6 – – 8.7Infection – 17.4 – – 10.4Dislocation – 0.8 2.5Periprosthetic fracture – 0.3 1.3Other reasons 1.8 8.9 3.8 4.0 8.6More frequent comorbidities (%)b

Hypertension 52.2 50.7 41.2 52.3 40.1Diabetes mellitus without complications 14.2 15.1 11.1 18.1 12.3Chronic pulmonary obstructive disease 7.7 8.3 7.5 10.3 8.5Depression 7.7 8.0 5.8 9.4 5.8Obesity 9.6 9.8 4.9 1.3 4.6Length of stay (days)b

Median 8 9 8 11 12Interquartile range 3 6 4 7 11Discharge to short-term centers (%)b 6.4 7.2 9.7 29.5 14.1

a 90.7% of knee and 86.5% of hip arthroplasties sent by hospitals could be linked to the minimum discharge database.b Information obtained from the hospital minimum discharge database. ICD-9 codes for reason for surgery: hip fracture: 820xx; osteoarthritis: 715xx; avascular necrosis: 733.4x;

aseptic loosening: 996.41; mechanical complications: 996.4x (except for 996.41), 996.5x, 996.7x; infection: 996.6x, 998.5x.c Mechanical complications: generic codes, type of complication not specified.

486 A. Allepuz et al. / The Journal of Arthroplasty 29 (2014) 484–490

21.0% to 71.6% in hip. Information on side of operation also improvedin both knee and hip arthroplasties from 62.3% to 97.8% and from60.0% to 99.4% respectively. The percentage of prostheses used inprimary operations that were identified also increased from 41.8% to88.1% in knee arthroplasty and from 41.3% to 80.7% in the hip.Regarding information on sex, age, date of admission and date ofsurgery completeness is 100% for the whole period.

Prostheses Characteristics

PS and total conventional prostheses are the most frequentlyused prostheses for knee and hip arthroplasty respectively, andcemented implants are the most common. Age and sex distribution

Table 2Primary Knee and Hip Prosthesis Characteristics, Period 2005–2010.a

Type of Prosthesis N %

Fixation Techniq

Cemented Hybrid

KneeTotal knee arthroplasty 26,467 97.4 74.0 21.9Cruciate retaining 12,709 46.8 59.5 34.5Posterior stabilized 13,155 48.4 85.6 11.8Constrained 545 2.0 96.9 0.2Hinge 54 0.2 100 –

Tumoral 4 0.01 25.0 25.0Partial knee arthroplasty 689 2.5 98.6 –

Femoro-patellar 110 0.4 100 –

Unicompartmental 579 2.1 98.4 –

HipPartial hip arthroplastyb 5939 33.4 53.8 –

Unipolar monoblock 2734 15.4 13.3 –

Unipolar modular 418 2.3 99.5 –

Bipolar 2787 15.7 86.6 –

Total hip arthroplasty 11,834 66.6 10.7 27.0Conventional 11,460 64.5 11.1 25.2Resurfacing 374 2.1 20.1 79.9

a Information on prostheses that were classified: knee 27,156 of 33,639 primary knee arb Partial hip arthroplasty: when only a stem was implanted.

in primary arthroplasties followed what was expected, being morefrequent in women and using more conservative techniques inyounger ages, except for resurfacing prostheses that were implantedmainly in young men (Table 2). Bearing surface in hip prostheseswas mainly metal-on-poly (73.6%) followed by ceramic-on-poly(12.5%), metal-on-metal (7.3%) and ceramic-on-ceramic (6.6%).Patellar button was used in 38.3% of total primary knees. A relevantnumber of different prostheses commercial brands (cemented anduncemented prostheses were considered separately) have beeninformed to the register. In 2010, 91 different knee prosthesescommercial brands were used and in the case of hip prosthesesthere were 101 and 119 different stem and cups commercial brandsimplanted (Table 3).

ue Age Group

Uncemented b65 65–74 75–84 ≥85 Women (%)

4.2 17.0 43.6 37.6 1.8 72.06.0 16.3 45.0 37.0 1.6 72.32.6 17.6 42.7 37.9 1.8 71.42.9 17.3 35.6 42.2 5.0 77.8– 16.7 27.8 44.4 11.0 75.9

50.0 75.0 25.0 – – 1001.4 51.1 34.1 13.9 0.9 67.8– 66.4 25.5 8.2 – 71.81.6 48.2 35.8 15.0 1.0 67.0

46.2 1.6 6.5 42.1 49.9 76.686.6 0.4 3.0 33.6 63.0 77.30.5 1.9 6.5 45.7 45.9 74.613.4 2.7 9.8 49.9 37.6 76.262.3 32.1 32.0 32.4 3.5 52.963.7 30.1 32.9 33.5 3.6 54.1– 95.7 4.0 0.3 – 14.4

throplasty prosthesis; hip 17,773 of 23,762 primary hip arthroplasty prosthesis.

Page 4: Joint Registries as Continuous Surveillance Systems: The Experience of the Catalan Arthroplasty Register (RACat)

Table 3Number of Primary Knee and Hip Prostheses Commercial Brands Used and Range byCenter.

Total Number of CommercialBrands

Range of Number ofCommercial Brands by Center

Period 2005–2010 Year 2010 Period 2005–2010 Year 2010

Knee prostheses 125 91 1–29 1–16Hip prosthesesStem 139 101 1–28 1–17Cup 161 119 1–28 2–18Stem–cup 621 360 1–62 3–32

487A. Allepuz et al. / The Journal of Arthroplasty 29 (2014) 484–490

Prostheses Performance

Information on 24,009 primary knee arthroplasties of 33,639 and16,798 primary hip arthroplasties of 23,762 was included for survivalanalysis. Cases were excluded because of poor data quality (prosthesisnot classified or absence of operation side [Fig. 1]). Median follow-upfor primary knee arthroplasties was 2.4 years (25% of patients hadmore than 3.9 follow-up years and 25% had less than 1.1 years) and2.0 years for hip primary arthroplasties (25% of patients had morethan 3.4 follow-up years and 25% had less than 0.9 years). Cumulativeknee revision risk were 1.1% (95% CI: 1.0–1.3) at 1 year and 3.3% (95%CI: 3.1–3.6) at 3 years. In the case of hip prostheses, total hiparthroplasty cumulative risks of revision were 1.9% (95% CI: 1.7–2.2)at 1 year and 2.9% (95% CI: 2.5–3.3) at 3 years, whereas for partialarthroplasties, they were 1.6% (95% CI: 1.3–2.0) and 2.5% (95% CI: 2.0–3.1) at 1 and 3 years respectively. Risk of revision of themost frequenttypes of prostheses in our register is presented in Table 4.

The sex- and age-adjusted revision risk for CR knee arthroplastieswas not different between uncemented and cemented prostheses(HR: 0.8; 95% CI: 0.4–1.4; P = 0.40) whereas it was lower for hybridimplants versus the cemented ones (HR: 0.5; 95% CI: 0.4–0.7;p b 0.01) (Fig. 2). In the case of PS arthroplasties uncementedimplants showed a higher risk of revision compared to the cementedones (HR: 2.1; 95% CI: 1.2–3.6; P = 0.01), but no differences wereobserved between hybrid and cemented implants (HR: 1.4; 95% CI:0.9–2.0; P = 0.09) (Fig. 3). In total conventional hip arthroplasties, nostatistically significant differences were observed according tofixation technique (uncemented vs. cemented HR: 0.8 95% CI: 0.5–1.2; P = 0.26; hybrid vs. cemented HR: 0.8; 95% CI: 0.5–1.3; P =0.39) (Fig. 4). No statistically significant differences on risk of revisionwere found in hip prostheses between bearing surfaces. Unadjusted

Fig. 1. Primary knee and hip arthropla

cumulative risk varied between 1.2% (95% CI: 0.6–2.4) for metal-on-metal and 2.0% (95% CI: 1.7–2.3) for metal-on-poly at 1 year, and 2.4%(95% CI: 1.5–3.7) for ceramic-on-poly and 2.9% (95% CI: 2.5–3.3) formetal-on-poly at 3 years.

Discussion

The Catalan Arthroplasty Register (RACat) is the first joint registerset up in Spain to assess prosthesis performance [11], allowing forbenchmarking between Catalan hospitals that could benefit outcomes[7] aside from contributing to post-marketing surveillance. InCatalonia, knee and hip prosthesis risk of revision at 3 years (knee:3.3% [95% CI: 3.1–3.6]; hip: 2.9% [95% CI: 2.5–3.3]) is higher than thatreported by the England and Wales register (knee: 2.68% [95% CI:2.61–2.74]; hip: 2.32% [95% CI: 2.25–2.38]); and it is likely that at7 years knee revision risk will be higher than that reported by theNorwegian (5.2% [95% CI: 4.8–5.6]) or the Kaiser Permanente (3.7%[95% CI: 3.4–4.0]) registries [12]. In fact at the moment, the burden ofknee revision in Catalonia (9.0%) is higher than that of Italy (6.4%),England and Wales (6.0%) and Sweden (5.2%) [13–15], which fullyjustifies the RACat implementation.

The higher knee and hip risk of revision observed in Cataloniacould be explained by different factors such as implant selection,although caution is necessary when comparing raw data. In Norway94.7% of total knee arthroplasties were performedwithout resurfacingof the patella, but in the Kaiser Permanente register 98.3% of total kneearthroplasties were performed with patellar resurfacing and unce-mented implants were used in 16.2% and 14.9% of arthroplasties inNorway and Kaiser Permanente, respectively. However, higherprosthesis survival was reported by Kaiser Permanente [12]. Regard-ing other European countries, in Denmark patellar button was used in76% of the total knee arthroplasties whereas in Norway and Sweden itwas used in only 11% and 14% of those, respectively [16], and in theEngland and Wales register, it was reported in 32% of primary kneearthroplasties [13].

In knee arthroplasty, uncemented or hybrid fixation was also morecommon in Denmark (22%) than in Norway (14%) and Sweden (2%),which had the lowest risk of revision [16]. In England and Walesuncemented or hybrid implants represented 15% of the total [13]. InCatalonia, patellar button utilization is within the range of othercountries (38.3%), although its effect on prosthesis survival is not clear[17]. In the case of fixation technique, uncemented or hybrid implantswere more frequent (26%) when compared to other registries that

sties included in survival analysis.

Page 5: Joint Registries as Continuous Surveillance Systems: The Experience of the Catalan Arthroplasty Register (RACat)

Table 4Risk of Revision at 3 Years of the Most Common Types of Prostheses According Fixation Technique.

Type of Prosthesis

Cemented Hybrid Uncemented

N (revisions) Risk of Revision (95% CI) N (revisions) Risk of Revision (95% CI) N (revisions) Risk of Revision (95% CI)

KneeCruciate retaining 6634 (166) 3.7 (3.2–4.4) 3910 (65) 2.0 (1.5–2.6) 693 (11) 2.6 (1.4–4.7)Posterior stabilized 9962 (267) 3.5 (3.1–4.0) 1314 (28) 4.7 (3.1–7.1) 289 (14) 6.7 (3.8–11.7)HipTotal conventional (any reason for surgery) 1199 (33) 3.0 (2.1–4.2) 2736 (65) 2.8 (2.2–3.6) 6789 (159) 2.8 (2.3–3.3)Total conventionala (ostheoarthritis indication) 670 (16) 2.4 (1.5–3.9) 1715 (36) 2.1 (1.5–2.9) 4816 (124) 2.6 (2.2–3.0)Total conventionala (hip fracture indication) 180 (9) 4.9 (2.6–9.1) 362 (20) 5.5 (3.6–8.3) 396 (14) 3.5 (2.1–5.8)Partial arthroplasty (hip fracture indication) 2311 (41) 1.8 (1.3–2.4) – – 1945 (57) 2.9 (2.3–3.8)

Only cases with operation side informed were included (included/total of prostheses). Cruciate retaining: cemented 6634/7562; hybrid 3910/4385; uncemented 693/762. Posteriorstabilized: cemented 9962/11,261; hybrid 1314/1552; uncemented 289/342. Total conventional: cemented 1199/1272; hybrid 2736/2888; uncemented 6789/7300.95% CI: 95% confidence interval.

a Only primary operations with information on reason for surgery were included.

488 A. Allepuz et al. / The Journal of Arthroplasty 29 (2014) 484–490

could have influenced risk of revision due to its inferior survival [18].However, in our register fixation technique affected differently therisk of revision on CR and PS implants, with CR hybrid implantsshowing a lower risk of revision compared to the cemented ones, anduncemented PS implants presenting a higher risk of revision whencompared to cemented implants. The study of Gandhi et al [18]excluded hybrid implants and a posterior Cochrane review found thatcemented tibial components had an increased risk of subsequentloosening compared to the cementless components [19]. On the otherhand, studies that have analyzed differences on knee arthroplastysurvival between CR and PS prostheses are not conclusive, whichcould not explain our results [20]. Other aspects such as mobilebearing are not currently available in our register.

Data quality of RACat was the main issue at the start of theproject. Missing information was related to difficulties on dataextraction by the Departments of Informatics, but feedback on dataquality has incentived its improvement along the study period. Theinformation that was excluded from the analysis reduces ourstatistical power, but, from our point of view, our results were notbiased since missing data were related to how information on eachprosthesis was stored in the hospital's information system. Despitethe limitations on data quality at the start of the project, the methoddesigned for collecting information, probably ensures the projectsurvivorship and reduces the resources needed for its implementa-tion. In the case of coverage, it is not as high as in other registries(78.3% for knee and 71.6% for hip in 2010), which might limit therepresentativeness of RACat's results. Australia reached 93% coveragein 2010 [21], the England and Wales register reached 99.1% in thissame year [13], and the Swedish registers, hip and knee, had over95% coverage [15,22]. However, it should be taken into account thatonly publicly funded hospitals are participating in the project at the

0.00

0.02

0.04

0.06

0.08

0.10

0 1 2 3 4Years since primary knee replacement

Cemented Uncemented Hybrid

Adjusted by age and sex

Fig. 2. Cruciate-retaining prostheses risk of revision according to fixation technique.

moment representing 80%–85% of surgical activity. Moreover, manysurgeons combine public and private activity, which limits thedifferent surgeons' effect on prostheses results, although otherfactors such as the hospital or prosthesis selection subjected todifferent budget constrains between the public and private sectorsare not controlled. The available follow-up time represents currentlya limitation to assess prosthesis performance. This could partlyexplain the absence of prosthesis survival differences among bearingsurfaces where a higher risk for metal-on-metal was expected [23].Other factors that might have influenced these results are the lownumber of resurfacing prostheses implanted and differences inprosthesis selection. In Catalonia two of the three resurfacingprostheses with the higher risk of revision [23], Bionik (Orthody-namics, Lübeck, Germany) and Icon (International Orthopaedics,Geisingen, Germany), are not used and the third one, ASR (ArticularSurface Replacement [DePuy]), has been implanted in only 50patients (data not shown). When interpreting the results, someaspects should be taken into account. In the case of bearing surfaces,at the moment, we cannot differentiate highly cross-linked polyeth-ylene, which have shown better results than historical polyethylene[24]. On the other hand, generic codes are frequently used whencoding discharges, which limits the interpretation on the reason forsurgery and could explain the low frequency of periprostheticfractures and dislocations observed when compared to otherregistries [25,26]. It is interesting to highlight the high percentageof partial hip arthroplasties performed in Catalonia [26]. However, itwas not possible to analyze if this was related to a higher frequencyof hip fractures with a hip arthroplasty indication or if a highernumber of patients with hip fracture undergo an arthroplasty inCatalonia. Future studies should address this aspect in order to assessthe indication of partial hip arthroplasty.

0.00

0.02

0.04

0.06

0.08

0.10

0.12

0.14

0.16

0 1 2 3 4Years since primary knee replacementCemented Uncemented Hybrid

Adjusted by age and sex

Fig. 3. Posterior-stabilized prostheses risk of revision according to fixation technique.

Page 6: Joint Registries as Continuous Surveillance Systems: The Experience of the Catalan Arthroplasty Register (RACat)

0.00

0.02

0.04

0.06

0 1 2 3 4Years since primary hip replacementCemented Uncemented Hybrid

Adjusted by age and sex

Fig. 4. Total conventional hip prostheses risk of revision according to fixationtechnique.

489A. Allepuz et al. / The Journal of Arthroplasty 29 (2014) 484–490

The Catalan experience follows what others have done before [27]and it is the first arthroplasty register in the Spanish National HealthSystem. Our current results suggest that if risk of revision slopecontinues diverging from that of other registries, the burden ofrevision in the future might affect resources allocation [28]. However,it is expected that a systematic registering of prostheses andperformance assessment feedback could contribute to improveresults.

Acknowledgments

The Catalan Society of Orthopaedic Surgery and Traumatology, theCatalan Health Service (CHS) and the Catalan Agency for Quality andHealthcare Assessment (CAQHA) are the Catalan institutions respon-sible for the Catalan Arthroplasty Register (RACat).

Catalan Arthroplasty Register (RACat) members and participant-ing hospitals:

Steering Committee. Subdirector of CatSalut: Dr. Francesc Brosa;President of SCCOT, Hospital de Bellvitge: Dr. Federico Portabella;Director AIAQS: Dr. Josep Argimon. Advisory Committee. President ofthe Advisory Committee, Hospital Vall d'Hebron, Pla Director deMalalties Reumàtiques i de l'Aparell Locomotor: Dr. Joan Nardi.Hospital de Mataró: Dr. Jaume Auleda; Hospital Vall d'Hebron: Dr.Enric Cáceres; Hospital Joan XXIII de Tarragona: Dr. Josep Giné;Hospital de Blanes: Dr. Ramon Oller; Hospital Sta. Maria de Lleida: Dr.Francesc Pallisó; Hospital Clínic de Barcelona: Dr. Santiago Suso;Hospital de Granollers: Dr. Alejandro Yunta; Coordinadora Llistesd'Espera, Divisió de Registres de Demanda i Activitat–CatSalut: Dra.Sílvia Cutillas, Divisió de Gestió dels Registres de Demanda i Activitat–CatSalut: Dra. Montse Bustins; Divisió de Compra de ServeisAssistencials i Assignació Poblacional–CatSalut: Dra. Carme Casas;Gerent Consorci Sanitari Barcelona: Dr. Jaume Estany; SubdirectorHealthcare Quality Area–AIAQS: Dra. Mireia Espallargues; researchersof AIAQS: Sra. Vicky Serra-Sutton, Dr. Alejandro Allepuz; projectmanager—AIAQS: Ms. Olga Martínez.

Participating hospitals. Centre Hospitalari-ALTHAIA, Clínica Girona,Clínica Plató, Fundació Privada, Clínica de Ponent, Corporació SanitàriaParc Taulí, F.G.S. Hospital de la Santa Creu i Sant Pau, Fundació PrivadaHospital de Mollet, Fundació Sanitària d'Igualada F.P., Fundació SantHospital de la Seu Urgell, H. Sant Joan Despí–Moisès Broggi, H.Universitari Vall d'Hebron, Hospital Clínic i Provincial de Barcelona,Hospital Comarcal Móra d'Ebre, Hospital Comarcal d'Amposta,Hospital Comarcal de Blanes, Hospital Comarcal de l'Alt Penedès,Hospital Comarcal del Pallars, Hospital Dos de Maig–CSI, HospitalGeneral L'Hospitalet–CSI, Hospital General de Granollers, HospitalGeneral de Vic, Hospital Municipal de Badalona, Hospital Mútua deTerrassa, Hospital Provincial Santa Caterina, Hospital Residència Sant

Camil, Hospital Sant Bernabé, Hospital Sant Jaume d'Olot, HospitalSant Joan de Déu de Martorell, Hospital Sant Rafael, Hospital SantaMaria, Hospital Universitari Arnau de Vilanova, Hospital UniversitariGermans Trias i Pujol, Hospital Universitari Sagrat Cor, HospitalUniversitari Sant Joan de Reus, Hospital Universitari de Bellvitge,Hospital Universitari de Girona Dr. Josep Trueta, Hospital Universitaride Tarragona Joan XXIII, Hospital de Campdevànol, Hospital deFigueres, Hospital de Mataró, Hospital de Palamós, Hospital dePuigcerdà, Hospital de Sant Boi–Parc Sanitari St Joan de Déu, Hospitalde Sant Celoni Fundació Privada, Hospital de Sant Jaume Calella,Hospital de Sant Joan de Déu d'Esplugues Llobregat, Hospital de SantPau i Santa Tecla, Hospital de Terrassa, Hospital de Tortosa Verge de laCinta, Hospital de Viladecans, Hospital de l'Esperit Sant, Hospital delVendrell, IMAS, Pius Hospital de Valls.

Manufacturer companies: aap Implantate AG; Active ImplantsCorporation; Adler Ortho; Amplitude; Arthrosurface; Aston Medical;B. Braun Surgical, SA; Bioimpianti; Biomet; Biomed; Biotechni; CGDBSRL; Conformis, Inc; ConMed Corporation; Corin Medical; CousinBiotech; De Soutter Medical; Dedienne Sante; Downs Surgical; EarthyMedical International; ESKA Implants AG; European Medical ContractM.; F.I.I. FH Orthopedics; Finsbury Orthopaedics; Global MedicalImplants, SL; Groupe Lépine; Hit Medica; I Ceram (Sa Mil); ImplantsIndustrie; Io, J.R.I.; Johnson & Johnson, SA; Lafitt; Lima Implantes, SL;Mathys; Medin, SA; Merete; New2DM; OHST Medizintechnik AG;Omni Life Science, INC; Orthodynamic Orthopedics; OTHESIO Im-plants; Permedica; PETER BREHM GMBH; PSB Exactech; Sanortho;Scanos Medical España; Science et Médecine; Serf, SA; Small BoneInnovations, Inc; Smith & Nephew, SA; Socinser 21, SA; SomepicTechnologie; Stanmore Implants; Stryker Ibérica; Surgical Medibér-ica; Surgival CO, SA; Symbios; Tantum; Tecres; Teknimed; TornierEspaña, SL; Traiber España, SA; Transysteme, SA; United OrthopedicCorporation; Waldemar Link España, SA; Wright Medical Technology;Zimmer, SA.

Provider companies: 3M España, SA; A&T Soluciones Médicas, SL;A2 C; Acuña y Fombona, SA; Alomedic, SL; Bio-implants Medical, SL;Bosch Ortopèdics, SL; Catimp; Distrauma, SL; Euroimplant Medical;Eurotrauma, SL; Grifols; Hospitak, SL; Hospitrauma, SL; HR Fungibles,SL; Intermedic; Karey Ortho, SA; Kinetics Plus, SL; Lifante; M. Kor;Material Médico, SL; MBA; Medcomtech; Medical Service; MeditramOrthopaedic, SL; Orbimed, SA; Palex Medical, SA; Polymedic 2000, SA;Prim Suministros; Prognomed, SA; Scanos Medical España; StemcupMedical Products AG; Subministraments Medics Lleida, SL; Sucesoresde Pedro Molina, SA; Técnicas Médicas MAB, SA; Tramedic, SA;Transplant Services Foundation; Vortrom, SRL; Wescott Medical.

References

1. Allepuz A, Serra-Sutton V, Espallargues M, et al. Hip and knee arthroplasties inCatalonia [Spain] from 1994 to 2005. Gac Sanit 2008;22:534.

2. Allepuz A, Serra-Sutton V, Espallargues M, et al. Hip and knee replacement in theSpanish National Health System. Rev Esp Cir Ortop Traumatol 2009;53:290.

3. Horan FT. Joint registries. J Bone Joint Surg Br 2010;92:749.4. Labek G, Neumann D, Agreiter M, et al. Impact of implant developers on published

outcome and reproducibility of cohort-based clinical studies in arthroplasty. J BoneJoint Surg Am 2011;93(Suppl 3):55.

5. SedrakyanA, Paxton EW, Phillips C, et al. The International ConsortiumofOrthopaedicRegistries: overview and summary. J Bone Joint Surg Am 2011;93(Suppl 3):1.

6. Carr AJ, Robertsson O, Graves S, et al. Knee replacement. Lancet 2012;379:1331.7. Ivers N, Jamtvedt G, Flottorp S, et al. Audit and feedback: effects on professional

practice and healthcare outcomes. Cochrane Database Syst Rev 2012;6 CD000259.8. Serra-Sutton V, Martínez O, Allepuz A, et al. Catalan Arthroplasty Register. Results

in hip and knee 2005–2008. Barcelona: Catalan Agency for Health TechnologyAssessment and Research. Catalan Health Service. Department of Health. General-itat de Catalunya. available at: http://www.gencat.cat/salut/depsan/units/aatrm/pdf/racat_hip_knee2005-2008_cahta2010en.pdf; 2010.

9. Robertsson O. Knee arthroplasty registers. J Bone Joint Surg Br 2007;89:1.10. Elixhauser A, Steiner C, Harris DR, et al. Comorbidity measures for use with

administrative data. Med Care 1998;36:8.11. Allepuz A, Serra-Sutton V, Martínez O, et al. Developing the methodology and pilot

implementation of prosthetic implant registers in the Spanish National HealthSystem. Madrid: Plan de Calidad del Sistema Nacional de Salud. Ministerio deSanidad y Consumo. Agència d'Avaluació de Tecnologia i Recerca Mèdiques de

Page 7: Joint Registries as Continuous Surveillance Systems: The Experience of the Catalan Arthroplasty Register (RACat)

490 A. Allepuz et al. / The Journal of Arthroplasty 29 (2014) 484–490

Catalunya; 2007. Informes de Evaluación de Tecnologías Sanitarias, AATRM2006/04. Available in Spanish at http://www.gencat.cat/salut/depsan/units/aatrm/pdf/in0604es.pdf.

12. Paxton EW, Furnes O, Namba RS, et al. Comparison of the Norwegian kneearthroplasty register and a United States arthroplasty registry. J Bone Joint Surg Am2011;93(Suppl 3):20.

13. National Joint Registry (NJR) for England and Wales. 8th Annual Report 2011.Hemel Hempstead (United Kingdom): National Joint Registry Centre; 2011.

14. Dati complessivi interventi di protesi d'anca, di ginocchio e di spalla in EmiliaRomagna 2000–2010. Bologna (Italy): Istituti Ortopedici Rizzoli. Servizio SanitarioRegionale Emilia-Romagna; 2011.

15. The Swedish Knee Arthroplasty Registry (SKAR). Annual Report 2011. Lund(Sweden): Swedish Knee Arthroplasty Registry. Lund University Hospital; 2011.

16. Havelin LI, Robertsson O, Fenstad AM, et al. A Scandinavian experience of registercollaboration: the Nordic Arthroplasty Register Association (NARA). J Bone JointSurg Am 2011;93(Suppl 3):13.

17. Lygre SH, Espehaug B, Havelin LI, et al. Failure of total knee arthroplasty with orwithout patella resurfacing. Acta Orthop 2011;82:282.

18. Gandhi R, Tsvetkov D, Davey JR, et al. Survival and clinical function of cemented anduncemented prostheses in total knee replacement: a meta-analysis. J Bone JointSurg Br 2009;91:889.

19. Nakama GY, Peccin MS, Almeida GJ, Lira Neto Ode A, Queiroz AA, Navarro RD.Cemented, cementless or hybrid fixation options in total knee arthroplasty for

osteoarthritis and other non-traumatic diseases. Cochrane Database Syst Rev2012;10 CD006193.

20. Jacobs WC, Clement DJ, Wymenga AB. Retention versus sacrifice of the posteriorcruciate ligament in total knee replacement for treatment of osteoarthritis andrheumatoid arthritis. Cochrane Database Syst Rev 2005;4 CD004803.

21. Australian Orthopeadic Association National Joint Replacement Registry. AnnualReport. Adelaide: AOA; 2011.

22. Garellick G, Kärrholm J, Rogmark C, et al. The Swedish Hip Arthroplasty Register.Annual Report 2010. Göteborg (Sweden): Joint Replacement Unit. SahlgrenskaUniversity Hospital: Swedish Orthopaedic Association; 2011.

23. Graves SE, Rothwell A, Tucker K, et al. A multinational assessment of metal-on-metal bearings in hip replacement. J Bone Joint Surg Am 2011;93(Suppl 3):43.

24. Kurtz SM, Gawel HA, Patel JD. History and systematic review of wear and osteolysisoutcomes for first-generation highly crosslinked polyethylene. Clin Orthop RelatRes 2011;469:2262.

25. National Joint Registry (NJR) for England andWales. 8th annual report 2011. HemelHempstead, United Kingdom: National Joint Registry Centre; 2012.

26. Australian Orthopeadic Association National Joint Replacement Registry. Annualreport. Adelaide: AOA; 2012.

27. Serra-Sutton V, Allepuz A, Espallargues M, et al. Artroplasty registers: a review ofinternational experiences. Int J Technol Assess Health Care 2009;25:63.

28. Lavernia C, Lee DJ, Hernandez VH. The increasing financial burden of knee revisionsurgery in the United States. Clin Orthop Relat Res 2006;446:221.