Advertisement

The Impact of Hepatitis C on Complication Rates After Revision Total Knee Arthroplasty: A Matched Cohort Study

Open AccessPublished:October 22, 2022DOI:https://doi.org/10.1016/j.artd.2022.09.010

      Abstract

      Background

      It is unclear if hepatitis C (HCV) negatively impacts outcomes of revision total knee arthroplasty (rTKA). The purpose of this study was to compare complication rates after rTKA for patients with HCV vs matched controls.

      Methods

      A retrospective cohort study was conducted using the PearlDiver database (PearlDiver Inc., Colorado Springs, CO). Patients with HCV who underwent rTKA (n = 1448) were matched 1:4 with controls (n = 5792) on age, sex, and several comorbidities. Rates of medical complications within 90 days and prothesis-related complications within 2 years postoperatively were compared with logistic regression for (1) patients with vs without HCV and (2) HCV patients who underwent aseptic vs septic rTKA.

      Results

      Relative to controls, patients with HCV exhibited significantly higher rates of medical complications (27.7% vs 20.9%; odds ratio [OR] 1.47), periprosthetic fractures (2.3% vs 1.1%; OR 2.20), all-cause repeat rTKA (11.7% vs 9.4%; OR 1.29), and repeat rTKA for prosthetic joint infection (PJI) (6.7% vs 3.6%; OR 1.92). Within the HCV cohort, HCV patients with initial septic rTKA exhibited significantly higher rates of medical complications (41.7% vs 22.7%; OR 2.39), all-cause subsequent rTKA (15.9% vs 10.2%; OR 1.67), and repeat rTKA for PJI (15.9% vs 3.4%; OR 5.39). Conversely, HCV patients with initial aseptic rTKA exhibited significantly higher rates of aseptic loosening (2.6% vs 7.4%; OR 0.33).

      Conclusions

      Patients with HCV exhibited significantly higher rates of medical and prosthesis-related complications after rTKA than controls. Among patients with HCV, initial septic rTKA was associated with significantly higher rates of medical complications, repeat rTKA, and PJI.

      Keywords

      Introduction

      Total knee arthroplasty (TKA) is a highly successful surgery performed in the United States (U.S.) [
      • Cram P.
      • Landon B.E.
      • Matelski J.
      • Ling V.
      • Stukel T.A.
      • Paterson J.M.
      • et al.
      Utilization and short-term outcomes of primary total hip and knee arthroplasty in the United States and Canada: an analysis of New York and Ontario administrative data.
      ]. More than 680,000 TKAs are performed each year in the U.S., and the annual TKA volume is projected to exceed 1.26 million by 2030 [
      • Kurtz S.
      • Ong K.
      • Lau E.
      • Mowat F.
      • Halpern M.
      Projections of primary and revision hip and knee arthroplasty in the United States from 2005 to 2030.
      ,
      • Singh J.A.
      • Yu S.
      • Chen L.
      • Cleveland J.D.
      Rates of total joint replacement in the United States: future projections to 2020-2040 using the national inpatient sample.
      ,
      • Sloan M.
      • Premkumar A.
      • Sheth N.P.
      Projected volume of primary total joint arthroplasty in the U.S., 2014 to 2030.
      ]. While TKA outcomes are predominantly excellent, the short-term risk of revision arthroplasty has remained relatively unchanged in recent years [
      • Brockman B.S.
      • Maupin J.J.
      • Thompson S.F.
      • Hollabaugh K.M.
      • Thakral R.
      Complication rates in total knee arthroplasty performed for osteoarthritis and post-traumatic arthritis: a comparison study.
      ,
      • Bozic K.J.
      • Kurtz S.M.
      • Lau E.
      • Ong K.
      • Chiu V.
      • Vail T.P.
      • et al.
      The epidemiology of revision total knee arthroplasty in the United States.
      ]. Consequently, the annual volume of revision TKA (rTKA) is also growing and is projected to surpass 128,000 procedures by 2030 [
      • Schwartz A.M.
      • Farley K.X.
      • Guild G.N.
      • Bradbury Jr., T.L.
      Projections and epidemiology of revision hip and knee arthroplasty in the United States to 2030.
      ]. Compared to primary TKA, rTKA is associated with a higher risk of complications and revision procedures [
      • Ong K.L.
      • Lau E.
      • Suggs J.
      • Kurtz S.M.
      • Manley M.T.
      Risk of subsequent revision after primary and revision total joint arthroplasty.
      ,
      • Jaffer A.K.
      • Barsoum W.K.
      • Krebs V.
      • Hurbanek J.G.
      • Morra N.
      • Brotman D.J.
      Duration of anesthesia and venous thromboembolism after hip and knee arthroplasty.
      ,
      • Mahomed N.N.
      • Barrett J.A.
      • Katz J.N.
      • Phillips C.B.
      • Losina E.
      • Lew R.A.
      • et al.
      Rates and outcomes of primary and revision total hip replacement in the United States medicare population.
      ,
      • Pulido L.
      • Parvizi J.
      • Macgibeny M.
      • Sharkey P.F.
      • Purtill J.J.
      • Rothman R.H.
      • et al.
      In hospital complications after total joint arthroplasty.
      ,
      • Zhan C.
      • Kaczmarek R.
      • Loyo-Berrios N.
      • Sangl J.
      • Bright R.A.
      Incidence and short-term outcomes of primary and revision hip replacement in the United States.
      ]. Risk factors for poor outcomes after rTKA include the quality of the index primary TKA, indication for rTKA, and comorbidities such as obesity, diabetes, and tobacco use [
      • Roman M.D.
      • Russu O.
      • Mohor C.
      • Necula R.
      • Boicean A.
      • Todor A.
      • et al.
      Outcomes in revision total knee arthroplasty (Review).
      ].
      Hepatitis C (HCV) affects more than 3 million Americans and approximately 3.3% of the orthopaedic patient population [
      • Hofmeister M.G.
      • Rosenthal E.M.
      • Barker L.K.
      • Rosenberg E.S.
      • Barranco M.A.
      • Hall E.W.
      • et al.
      Estimating prevalence of hepatitis C virus infection in the United States, 2013-2016.
      ,
      • Issa K.
      • Boylan M.R.
      • Naziri Q.
      • Perfetti D.C.
      • Maheshwari A.V.
      • Mont M.A.
      The impact of hepatitis C on short-term outcomes of total joint arthroplasty.
      ]. Prior studies have demonstrated patients with HCV who undergo TKA exhibit higher rates of 90-day medical complications and surgical complications including rTKA than noninfected patients [
      • Ross A.J.
      • Ross B.J.
      • Lee O.C.
      • Weldy J.M.
      • Sherman W.F.
      • Sanchez F.L.
      A missed opportunity; the impact of hepatitis C treatment prior to total knee arthroplasty on postoperative complications.
      ,
      • Cancienne J.M.
      • Kandahari A.M.
      • Casp A.
      • Novicoff W.
      • Browne J.A.
      • Cui Q.
      • et al.
      Complication rates after total hip and knee arthroplasty in patients with hepatitis C compared with matched control patients.
      ,
      • Cheng T.
      • Yang C.
      • Hao L.
      • Cheng X.
      • Hu J.
      • Ren W.
      • et al.
      Hepatitis C virus infection increases the risk of adverse outcomes following joint arthroplasty: a meta-analysis of observational studies.
      ,
      • Kildow B.J.
      • Politzer C.S.
      • DiLallo M.
      • Bolognesi M.P.
      • Seyler T.M.
      Short and long-term postoperative complications following total joint arthroplasty in patients with human immunodeficiency virus, hepatitis B, or hepatitis C.
      ,
      • Wei W.
      • Liu T.
      • Zhao J.
      • Li B.
      • Li S.
      • Liu J.
      Does the hepatitis C virus affect the outcomes of total joint arthroplasty? A meta-analysis of ten studies.
      ]. However, the impact of HCV on outcomes of rTKA has not been studied. As such, examination of HCV as a risk factor for postoperative complications following an rTKA is needed.
      The purpose of this study was to (1) analyze the impact of HCV on postoperative outcomes following an rTKA and (2) compare postoperative complication rates for HCV patients who underwent septic vs aseptic rTKAs. It was hypothesized that patients with HCV would exhibit significantly higher postoperative complication rates than matched controls and that HCV patients who underwent septic rTKAs would exhibit significantly higher rates of complications than HCV patients who underwent aseptic rTKAs.

      Material and methods

      Data source and study design

      Patient records were queried from the PearlDiver Mariner Database (PearlDiver Inc., Colorado Springs, CO), a commercially available administrative claims database with deidentified patient data. The database contains the medical records of approximately 144 million patients across the U.S. from 2010 through Q3 of 2020 which are collected by an independent data aggregator. Researchers identify patients and outcomes using the Current Procedural Technology (CPT) and International Classification of Diseases, Ninth and Tenth Revision (ICD-9/ICD-10), codes on insurance claims. For this study, the “MKnee” data set was analyzed which contains records of a subset of patients with diagnoses and procedures localized to the knee. All health insurance payors are represented including commercial, private, and government plans. Institutional review board exemption was granted as provided data were deidentified and compliant with the Health Insurance Portability and Accountability Act. This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
      A retrospective cohort study was conducted to investigate the impact of HCV on complication rates following rTKA. Patients who underwent aseptic rTKA were identified by claims containing procedural codes for partial or total revision knee arthroplasty (eg, CPT-27486, CPT-27487) without associated diagnosis codes for prosthetic joint infection (PJI). One-stage septic rTKA (ie, revision for PJI) was defined by procedural codes for rTKA paired with diagnosis codes for PJI. For 2-stage septic revisions, the first stage of the procedure was defined by procedural codes for implant removal (eg, CPT-27488) with concomitant insertion of an antibiotic cement spacer (eg, CPT-11981). The second stage was defined by procedural codes for TKA (eg, CPT-27447) with concomitant removal of the antibiotic spacer (eg, CPT-11982) [
      • Samuel L.T.
      • Grits D.
      • Acuña A.J.
      • Piuzzi N.S.
      • Higuera-Rueda C.A.
      • Kamath A.F.
      Work relative value units do not adequately support the burden of infection management in revision knee arthroplasty.
      ]. Patients who underwent 2-stage septic revisions were identified by a claim for each stage, with the claim for the second stage following the claim for the first stage. For these patients, postoperative complications were tracked from the date of the second stage. CPT, ICD-9, and ICD-10 procedural codes were used to define all procedures.
      In order to limit potential transfer bias due to patients leaving or joining the data set during the study period, only patients with continuous database enrollment for at least 1 year prior and 2 years after the index rTKA were included. Pediatric patients and patients infected with hepatitis B or human immunodeficiency virus were excluded.
      Patients with HCV were identified by claims containing ICD-9/ICD-10 diagnosis codes for acute, chronic, and/or unspecified HCV infection before or at the time of the index rTKA. Prior validation studies analyzing the accuracy of coding for the presence of HCV have reported a positive predictive value between 88% and 94%, while the negative predictive value for patients without documented HCV is 90% to 93% [
      • Kramer J.R.
      • Davila J.A.
      • Miller E.D.
      • Richardson P.
      • Giordano T.P.
      • El-Serag H.B.
      The validity of viral hepatitis and chronic liver disease diagnoses in Veterans affairs administrative databases.
      ,
      • Niu B.
      • Forde K.A.
      • Goldberg D.S.
      Coding algorithms for identifying patients with cirrhosis and hepatitis B or C virus using administrative data.
      ]. All codes used to define inclusion and exclusion criteria are provided in Appendix A.

      Demographic data and clinical characteristics

      Baseline demographic data including age, sex, body mass index (BMI), and U.S. region were obtained. Clinical characteristics queried included length of stay (LOS) during the initial rTKA and the prevalence of diabetes mellitus, tobacco use, chronic kidney disease, hypertension, and obesity. The distribution of indications for the initial rTKA (aseptic or septic) was also obtained.

      Outcomes

      Rates of medical complications during the index hospital encounter and within 90 days postoperatively were obtained. Medical complications queried included inpatient readmissions, deep vein thrombosis (DVT), pulmonary embolism, urinary tract infection (UTI), acute kidney injury (AKI), and blood transfusion. The diagnosis and procedural codes used to define these complications are provided in Appendix B.
      Prosthesis-related complications were evaluated at 2 years postoperatively. Specific complications queried included manipulation under anesthesia and/or lysis of adhesions for knee stiffness, all-cause subsequent rTKA, PJI, aseptic loosening, and periprosthetic fracture. All-cause subsequent rTKA included revision of the femoral and/or tibial components, implant removal, and/or insertion of an antibiotic spacer. PJI was defined as repeat rTKA for infection (1-stage or 2-stage) using the same criteria for septic revisions as outlined above. Rates of all-cause repeat rTKA, re-revision for PJI, aseptic loosening, and periprosthetic fracture were queried after the 90-day global postoperative period in order to minimize the possibility that indications for initial rTKA were counted as complications during routine postoperative follow-ups. The codes used to define these complications are provided in Appendix A and Appendix B.

      Statistical analysis

      Statistical analyses were performed using the R statistical software (Version 4.1.0; R Project for Statistical Computing) integrated within the PearlDiver software with an α level set to 0.05. In order to reduce confounding bias, exact matching without replacement was performed to generate similar patient cohorts. HCV patients were matched at a 1:4 ratio with noninfected controls on the following parameters: age, sex, diabetes mellitus, tobacco use, obesity, chronic kidney disease, and hypertension.
      Categorical variables were compared with a chi-square test, and continuous variables were compared with Welch’s t test or the Mann-Whitney U test. Rates of postoperative complications were compared using multivariable logistic regression for (1) patients with HCV vs controls for all rTKA, (2) initial aseptic rTKA, and (3) initial septic rTKA and (4) for patients who underwent initial aseptic vs septic rTKA within the HCV cohort. Odds ratios (OR) with the corresponding 95% confidence intervals (CIs) were calculated for each outcome.

      Results

      Study population

      A total of 51,548 patients who underwent rTKAs were identified, including 1462 (2.8%) patients with HV. After 1:4 matching, 1448 HCV patients were matched with 5792 noninfected controls (Table 1). Initial septic rTKA was significantly more common in the HCV cohort (26.5% vs 17.8%, P < .001). Patients with HCV had a significantly longer mean LOS (4.24 vs 3.29 days, P < .001). There were significant differences in region and BMI data between the 2 cohorts although BMI data were only available for 26.8% of HCV patients and 32.0% of controls.
      Table 1Baseline demographic data and clinical characteristics of rTKA cohorts (matched 1:4).
      CharacteristicsHCV (n = 1448)No HCV (n = 5792)P value
      Age (y), mean ± SD59.03 ± 7.4959.18 ± 7.56.485
      Female sex, n (%)741 (51.2)2955 (51.0).939
      U.S. region, n (%)
      Region data available for 99% of included patients.
       Northeast321 (22.2)1038 (17.9)<.001
       South590 (40.7)2148 (37.1).012
       Midwest324 (22.4)1871 (32.3)<.001
       West208 (14.4)705 (12.2).028
      BMI, n (%)
      BMI data available for 388 (26.8%) HCV patients and 1856 (32.0%) controls.
       <3070 (18.0)218 (11.7).001
       30-3593 (24.0)381 (20.5).149
       35-4067 (17.3)452 (24.4).003
       >40158 (40.7)805 (43.4).366
      Comorbidities, n (%)
       Diabetes mellitus828 (57.2)3305 (57.1).957
       Obesity946 (65.3)3789 (65.4).975
       Tobacco use1101 (76.0)4408 (76.1).984
       Chronic kidney disease443 (30.6)1771 (30.6)1
       Hypertension1338 (92.4)5357 (92.5).956
      Revision indication, n (%)
       Aseptic1064 (73.5)4761 (82.2)<.001
       Septic384 (26.5)1031 (17.8)
      LOS (d), Mean ± SD
       All rTKA4.24 ± 3.513.29 ± 2.21<.001
       Aseptic3.72 ± 2.853.09 ± 1.99<.001
       Septic5.15 ± 4.453.79 ± 2.63<.001
      SD, standard deviation.
      Bolded P values indicate statistically significant results.
      a Region data available for 99% of included patients.
      b BMI data available for 388 (26.8%) HCV patients and 1856 (32.0%) controls.

      HCV vs controls, all rTKA

      Within 90 days following rTKA, rates of at least 1 medical complication were significantly higher in the HCV cohort relative to controls (27.7% vs 20.9%; OR 1.47; 95% CI, 1.29-1.68). This included significantly higher rates of AKI, UTI, blood transfusions, and inpatient readmissions (all P < .05; Table 2).
      Table 2Postoperative complication rates after rTKA for patients with HCV vs controls.
      ComplicationHCV (n = 1448)No HCV (n = 5792)Statistical analysis (ref. group, HCV cohort)
      n%n%OR (95% CI)P value
      90 D
       Any medical complication
      The number of patients with at least 1 medical or joint complication.
      40127.7%121220.9%1.47 (1.29-1.68)<.001
       DVT5
      For the sake of protecting patients’ identities, the PearlDiver software does not report exact patient counts when defined groups have <11 patients. In such instances, a cohort size of 5 (median between 1-10) was assigned although the software uses the real patient counts for the statistical analysis.
      0.3%430.7%0.78 (0.34-1.59).539
       PE110.8%490.8%0.94 (0.46-1.76).867
       AKI926.4%2764.8%1.43 (1.11-1.82).005
       UTI805.5%2354.1%1.42 (1.08-1.84).010
       Transfusion14610.1%3556.1%1.68 (1.37-2.06)<.001
       Inpatient readmission24216.7%70312.1%1.47 (1.25-1.72)<.001
      2 Y
       Any joint complication
      The number of patients with at least 1 medical or joint complication.
      29020.0%99417.2%1.21 (1.04-1.40).012
       MUA/LoA664.6%2524.4%1.03 (0.77-1.35).847
       Subsequent rTKA16911.7%5429.4%1.29 (1.07-1.55).007
       PJI976.7%2093.6%1.92 (1.49-2.50)<.001
       Aseptic loosening896.1%3305.7%1.07 (0.83-1.36).603
       Periprosthetic fracture342.3%631.1%2.20 (1.42-3.34)<.001
      DVT, deep vein thrombosis; LoA, lysis of adhesions; MUA, manipulation under anesthesia; PE, pulmonary embolism; ref., reference.
      Bolded OR/CI/P values indicate statistically significant results.
      a The number of patients with at least 1 medical or joint complication.
      b For the sake of protecting patients’ identities, the PearlDiver software does not report exact patient counts when defined groups have <11 patients. In such instances, a cohort size of 5 (median between 1-10) was assigned although the software uses the real patient counts for the statistical analysis.
      Rates of at least 1 prosthesis-related complication were also significantly higher in the HCV cohort within 2 years after the rTKA (20.0% vs 17.2%; OR 1.21; 95% CI, 1.04-1.40), including significantly higher rates of periprosthetic fracture, all-cause repeat rTKA, and repeat rTKA for PJI (all P < .05). Among patients who underwent a repeat rTKA, PJI was significantly more common in the HCV cohort (57.4% vs 38.6%, P < .001) (Fig. 1a).
      Figure thumbnail gr1
      Figure 1Among patients who underwent a subsequent rTKA, comparison of incidence of PJI (ie, subsequent septic rTKA) for (a) all HCV patients vs controls, (b) HCV patients with an initial aseptic rTKA vs controls with an initial aseptic rTKA, (c) HCV patients with an initial septic rTKA vs controls with an initial septic rTKA, and (d) HCV patients with an initial aseptic rTKA vs HCV patients with an initial septic rTKA.

      HCV vs controls, initial aseptic rTKA

      Within 90 days after the initial aseptic rTKA, HCV patients exhibited significantly higher rates of medical complications than controls (22.7% vs 17.5%; OR 1.41; 95% CI, 1.19-1.65), including AKI, UTI, blood transfusions, and inpatient readmissions (all P < .05; Table 3).
      Table 3Postoperative complication rates after aseptic primary rTKAs for patients with HCV vs controls.
      ComplicationHCV (n = 1064)No HCV (n = 4761)Statistical analysis (ref. group, HCV cohort)
      n%n%OR (95% CI)P value
      90 D
       Any medical complication
      The number of patients with at least 1 medical or joint complication.
      24122.7%83217.5%1.41 (1.19-1.65)<.001
       DVT5
      For the sake of protecting patients’ identities, the PearlDiver software does not report exact patient counts when defined groups have <11 patients. In such instances, a cohort size of 5 (median between 1-10) was assigned although the software uses the real patient counts for the statistical analysis.
      0.5%270.6%0.74 (0.22-1.92).598
       PE5
      For the sake of protecting patients’ identities, the PearlDiver software does not report exact patient counts when defined groups have <11 patients. In such instances, a cohort size of 5 (median between 1-10) was assigned although the software uses the real patient counts for the statistical analysis.
      0.5%390.8%0.96 (0.41-1.97).926
       AKI565.3%1593.3%1.70 (1.23-2.32).001
       UTI615.7%1773.7%1.62 (1.19-2.18).002
       Transfusion747.0%2224.7%1.48 (1.12-1.94).005
       Inpatient readmission13712.9%4639.7%1.38 (1.12-1.69).002
      2 Y
       Any joint complication
      The number of patients with at least 1 medical or joint complication.
      20719.5%79916.8%1.18 (1.01-1.40).047
       MUA/LoA535.0%2094.4%1.12 (0.81-1.51).485
       Subsequent rTKA10810.2%4178.8%1.18 (0.93-1.47).157
       PJI363.4%871.8%1.87 (1.24-2.75).002
       Aseptic loosening797.4%2956.2%1.20 (0.92-1.55).171
       Periprosthetic fracture262.4%521.1%2.28 (1.39-3.65).001
      DVT, deep vein thrombosis; LoA, lysis of adhesions; MUA, manipulation under anesthesia; PE, pulmonary embolism; ref., reference.
      Bolded OR/CI/P values indicate statistically significant results.
      a The number of patients with at least 1 medical or joint complication.
      b For the sake of protecting patients’ identities, the PearlDiver software does not report exact patient counts when defined groups have <11 patients. In such instances, a cohort size of 5 (median between 1-10) was assigned although the software uses the real patient counts for the statistical analysis.
      Rates of periprosthetic fractures (2.4% vs 1.1% OR 2.28; 95% CI, 1.39-3.65) and repeat rTKA for incident PJI (3.4% vs 1.8%; OR 1.87; 95% CI, 1.24-2.75) were significantly higher in the HCV cohort within 2 years postoperatively. Among the identified patients with repeat revision procedures during the 2-year follow-up, PJI was a significantly more common indication for subsequent rTKA in the HCV cohort (33.3% vs 20.9%, P = .009) (Fig. 1b).

      HCV vs controls, initial septic rTKA

      Within the subgroup of included patients who underwent an initial septic rTKA (ie, revision for PJI), patients with HCV exhibited a significantly higher rate of blood transfusions than controls (18.8% vs 12.9%; OR 1.56; 95% CI, 1.12-2.17) (Table 4). Rates of all other medical complications were comparable (all P > .05).
      Table 4Postoperative complication rates after a septic primary rTKA for patients with HCV vs controls.
      ComplicationHCV (n = 384)No HCV (n = 1031)Statistical analysis (ref. group, HCV cohort)
      n%n%OR (95% CI)P value
      90 D
       Any medical complication
      The number of patients with at least 1 medical or joint complication.
      16041.7%38036.9%1.27 (0.99-1.62).057
       DVT5
      For the sake of protecting patients’ identities, the PearlDiver software does not report exact patient counts when defined groups have <11 patients. In such instances, a cohort size of 5 (median between 1-10) was assigned although the software uses the real patient counts for the statistical analysis.
      1.3%161.6%0.65 (0.18-1.84).477
       PE5
      For the sake of protecting patients’ identities, the PearlDiver software does not report exact patient counts when defined groups have <11 patients. In such instances, a cohort size of 5 (median between 1-10) was assigned although the software uses the real patient counts for the statistical analysis.
      1.3%5
      For the sake of protecting patients’ identities, the PearlDiver software does not report exact patient counts when defined groups have <11 patients. In such instances, a cohort size of 5 (median between 1-10) was assigned although the software uses the real patient counts for the statistical analysis.
      0.5%1.03 (0.22-3.68).970
       AKI369.4%11711.3%0.87 (0.57-1.29).514
       UTI194.9%585.6%0.97 (0.55-1.64).920
       Transfusion7218.8%13312.9%1.56 (1.12-2.17).008
       Inpatient readmission10527.3%24023.3%1.28 (0.98-1.69).075
      2 Y
       Any joint complication
      The number of patients with at least 1 medical or joint complication.
      8321.6%19518.9%1.18 (0.88-1.58).271
       MUA/LoA133.4%434.2%0.73 (0.36-1.37).362
       Subsequent rTKA6115.9%12512.1%1.40 (1.01-1.96).046
       PJI6115.9%12211.8%1.41 (1.02-1.96).039
       Aseptic loosening102.6%353.4%0.76 (0.35-1.51).471
       Periprosthetic fracture5
      For the sake of protecting patients’ identities, the PearlDiver software does not report exact patient counts when defined groups have <11 patients. In such instances, a cohort size of 5 (median between 1-10) was assigned although the software uses the real patient counts for the statistical analysis.
      1.3%111.1%2.08 (0.78-5.41).138
      DVT, deep vein thrombosis; LoA, lysis of adhesions; MUA, manipulation under anesthesia; PE, pulmonary embolism; ref., reference.
      Bolded OR/CI/P values indicate statistically significant results.
      a The number of patients with at least 1 medical or joint complication.
      b For the sake of protecting patients’ identities, the PearlDiver software does not report exact patient counts when defined groups have <11 patients. In such instances, a cohort size of 5 (median between 1-10) was assigned although the software uses the real patient counts for the statistical analysis.
      At 2 years, patients with HCV exhibited significantly higher rates of all-cause repeat rTKA (15.9% vs 12.1%; OR 1.40; 95% CI, 1.01-1.96) and re-revisions for PJI (15.9% vs 11.8%; OR 1.41; 95% CI, 1.02-1.96). Notably, all 61 (100%) repeat rTKA procedures in the HCV cohort and 122 of 125 (97.6%) in the control cohort were performed for PJI (P = .552) (Fig. 1c).

      HCV cohort, initial aseptic rTKA vs initial septic rTKA

      Within the HCV cohort, patients who underwent an initial septic rTKA exhibited significantly higher rates of at least 1 medical complication (41.7% vs 22.7%; OR 2.39; 95% CI, 1.86-3.06), including AKI, blood transfusions, and inpatient readmissions (all P < .05; Table 5).
      Table 5Postoperative complication rates after aseptic vs septic primary rTKAs among patients with HCV.
      ComplicationHCV aseptic rTKA (n = 1064)HCV septic rTKA (n = 384)Statistical analysis (ref. group, septic rTKA)
      n%n%OR (95% CI)P value
      90 D
       Any medical complication
      The number of patients with at least 1 medical or joint complication.
      24122.7%16041.7%2.39 (1.86-3.06)<.001
       DVT5
      For the sake of protecting patients’ identities, the PearlDiver software does not report exact patient counts when defined groups have <11 patients. In such instances, a cohort size of 5 (median between 1-10) was assigned although the software uses the real patient counts for the statistical analysis.
      0.5%5
      For the sake of protecting patients’ identities, the PearlDiver software does not report exact patient counts when defined groups have <11 patients. In such instances, a cohort size of 5 (median between 1-10) was assigned although the software uses the real patient counts for the statistical analysis.
      1.3%2.79 (0.66-11.85).164
       PE5
      For the sake of protecting patients’ identities, the PearlDiver software does not report exact patient counts when defined groups have <11 patients. In such instances, a cohort size of 5 (median between 1-10) was assigned although the software uses the real patient counts for the statistical analysis.
      0.5%5
      For the sake of protecting patients’ identities, the PearlDiver software does not report exact patient counts when defined groups have <11 patients. In such instances, a cohort size of 5 (median between 1-10) was assigned although the software uses the real patient counts for the statistical analysis.
      1.3%1.04 (0.23-3.61).960
       AKI565.3%369.4%1.81 (1.15-2.79).009
       UTI615.7%194.9%0.86 (0.49-1.42).590
       Transfusion747.0%7218.8%2.93 (2.06-4.17)<.001
       Inpatient readmission13712.9%10527.3%2.55 (1.91-3.39)<.001
      2 Y
       Any joint complication
      The number of patients with at least 1 medical or joint complication.
      20719.5%8321.6%1.12 (0.85-1.52).453
       MUA/LoA535.0%133.4%0.62 (0.31-1.12).145
       Subsequent rTKA10810.2%6115.9%1.67 (1.19-2.34).003
       PJI363.4%6115.9%5.39 (3.53-8.36)<.001
       Aseptic loosening797.4%102.6%0.33 (0.17-0.65).001
       Periprosthetic fracture262.4%5
      For the sake of protecting patients’ identities, the PearlDiver software does not report exact patient counts when defined groups have <11 patients. In such instances, a cohort size of 5 (median between 1-10) was assigned although the software uses the real patient counts for the statistical analysis.
      1.3%0.85 (0.36-1.81).706
      DVT, deep vein thrombosis; LoA, lysis of adhesions; MUA, manipulation under anesthesia; PE, pulmonary embolism; ref., reference.
      Bolded OR/CI/P values indicate statistically significant results.
      a The number of patients with at least 1 medical or joint complication.
      b For the sake of protecting patients’ identities, the PearlDiver software does not report exact patient counts when defined groups have <11 patients. In such instances, a cohort size of 5 (median between 1-10) was assigned although the software uses the real patient counts for the statistical analysis.
      Rates of all-cause repeat rTKA (15.9% vs 10.2%; OR 1.67; 95% CI, 1.19-2.34) and re-revisions for PJI (15.9% vs 3.4%; OR 5.39; 95% CI, 3.53-8.36) at 2-year follow-up were significantly higher among HCV patients who underwent an initial septic rTKA. PJI was a significantly more common indication for a repeat rTKA in patients who underwent an initial septic rTKA (100% vs 33.3%, P < .001) (Fig. 1d). Patients with HCV who underwent an initial aseptic rTKA exhibited a significantly higher rate of aseptic loosening (2.6% vs 7.4%; OR 0.33; 95% CI, 0.17-0.65).

      Discussion

      The present study provides novel data illustrating significantly higher postoperative complication rates after an rTKA in patients with HCV than those in age-, sex-, and comorbidity-matched controls. Patients with HCV exhibited a significantly longer mean LOS during the index rTKA, as well as significantly higher rates of 90-day medical complications and joint complications within 2 years postoperatively. These results align with data from prior studies analyzing HCV in primary arthroplasty [
      • Ross A.J.
      • Ross B.J.
      • Lee O.C.
      • Weldy J.M.
      • Sherman W.F.
      • Sanchez F.L.
      A missed opportunity; the impact of hepatitis C treatment prior to total knee arthroplasty on postoperative complications.
      ,
      • Cancienne J.M.
      • Kandahari A.M.
      • Casp A.
      • Novicoff W.
      • Browne J.A.
      • Cui Q.
      • et al.
      Complication rates after total hip and knee arthroplasty in patients with hepatitis C compared with matched control patients.
      ,
      • Cheng T.
      • Yang C.
      • Hao L.
      • Cheng X.
      • Hu J.
      • Ren W.
      • et al.
      Hepatitis C virus infection increases the risk of adverse outcomes following joint arthroplasty: a meta-analysis of observational studies.
      ,
      • Kildow B.J.
      • Politzer C.S.
      • DiLallo M.
      • Bolognesi M.P.
      • Seyler T.M.
      Short and long-term postoperative complications following total joint arthroplasty in patients with human immunodeficiency virus, hepatitis B, or hepatitis C.
      ,
      • Wei W.
      • Liu T.
      • Zhao J.
      • Li B.
      • Li S.
      • Liu J.
      Does the hepatitis C virus affect the outcomes of total joint arthroplasty? A meta-analysis of ten studies.
      ]. The increased complication risk in this population is multifactorial and driven by factors such as viral-mediated immune dysfunction and thrombocytopenia [
      • Wei W.
      • Liu T.
      • Zhao J.
      • Li B.
      • Li S.
      • Liu J.
      Does the hepatitis C virus affect the outcomes of total joint arthroplasty? A meta-analysis of ten studies.
      ,
      • Best M.J.
      • Buller L.T.
      • Klika A.K.
      • Barsoum W.K.
      Increase in perioperative complications following primary total hip and knee arthroplasty in patients with hepatitis C without cirrhosis.
      ]. Given that an rTKA is associated with higher complication rates than a primary TKA, these data add to existing literature by suggesting HCV exacerbates this increased complication risk relative to noninfected controls.
      Recent epidemiologic analyses have identified PJI as the most common indication for rTKA (20.4%-25.2%) [
      • Bozic K.J.
      • Kurtz S.M.
      • Lau E.
      • Ong K.
      • Chiu V.
      • Vail T.P.
      • et al.
      The epidemiology of revision total knee arthroplasty in the United States.
      ,
      • Delanois R.E.
      • Mistry J.B.
      • Gwam C.U.
      • Mohamed N.S.
      • Choksi U.S.
      • Mont M.A.
      Current epidemiology of revision total knee arthroplasty in the United States.
      ]. In this study, 19.5% of initial rTKA procedures were performed for PJI, illustrating that infection remains a major cause of failure after a primary TKA. Previous studies have reported higher rates of complications including subsequent revisions after an initial septic rTKA [
      • Belt M.
      • Hannink G.
      • Smolders J.
      • Spekenbrink-Spooren A.
      • Schreurs B.W.
      • Smulders K.
      Reasons for revision are associated with rerevised total knee arthroplasties: an analysis of 8,978 index revisions in the Dutch Arthroplasty Register.
      ,
      • Boddapati V.
      • Fu M.C.
      • Mayman D.J.
      • Su E.P.
      • Sculco P.K.
      • McLawhorn A.S.
      Revision total knee arthroplasty for periprosthetic joint infection is associated with increased postoperative morbidity and mortality relative to noninfectious revisions.
      ,
      • Dai W.L.
      • Lin Z.M.
      • Shi Z.J.
      • Wang J.
      Outcomes following revision total knee arthroplasty septic versus aseptic failure: a national propensity-score-matched comparison.
      ]. Belt et al. reported a 16% rate of all-cause repeat rTKA at 1 year after a septic rTKA, while lower re-revision rates were observed after an aseptic rTKA for indications such as loosening (3%) [
      • Belt M.
      • Hannink G.
      • Smolders J.
      • Spekenbrink-Spooren A.
      • Schreurs B.W.
      • Smulders K.
      Reasons for revision are associated with rerevised total knee arthroplasties: an analysis of 8,978 index revisions in the Dutch Arthroplasty Register.
      ]. In this study, initial septic revisions were significantly more common in the HCV cohort. Furthermore, among patients with HCV, PJI was a significantly more common indication for a repeat rTKA. These data suggest that patients with HCV are at an increased risk of septic failure after both primary TKAs and rTKAs.
      In subgroup analyses based on the indication for an initial rTKA, patients with HCV who underwent an initial aseptic rTKA exhibited significantly higher rates of medical complications relative to controls. Among patients who underwent an initial septic rTKA, however, only rates of transfusions were significantly higher for patients with HCV. These data suggest that HCV is an important risk factor for medical complications after a aseptic rTKA. Conversely, given that a septic rTKA itself is a strong risk factor for poor short-term outcomes [
      • Belt M.
      • Hannink G.
      • Smolders J.
      • Spekenbrink-Spooren A.
      • Schreurs B.W.
      • Smulders K.
      Reasons for revision are associated with rerevised total knee arthroplasties: an analysis of 8,978 index revisions in the Dutch Arthroplasty Register.
      ,
      • Boddapati V.
      • Fu M.C.
      • Mayman D.J.
      • Su E.P.
      • Sculco P.K.
      • McLawhorn A.S.
      Revision total knee arthroplasty for periprosthetic joint infection is associated with increased postoperative morbidity and mortality relative to noninfectious revisions.
      ,
      • Dai W.L.
      • Lin Z.M.
      • Shi Z.J.
      • Wang J.
      Outcomes following revision total knee arthroplasty septic versus aseptic failure: a national propensity-score-matched comparison.
      ], HCV may only marginally increase the risk of medical complications after a septic rTKA relative to controls who are also at high risk. At 2-year follow-up, patients with HCV also exhibited significantly higher rates of re-revisions for PJI and periprosthetic fracture, which are both common modes of failure after an rTKA [
      • Belt M.
      • Hannink G.
      • Smolders J.
      • Spekenbrink-Spooren A.
      • Schreurs B.W.
      • Smulders K.
      Reasons for revision are associated with rerevised total knee arthroplasties: an analysis of 8,978 index revisions in the Dutch Arthroplasty Register.
      ,
      • Evangelopoulos D.S.
      • Ahmad S.S.
      • Krismer A.M.
      • Albers C.E.
      • Hoppe S.
      • Kleer B.
      • et al.
      Periprosthetic infection: major cause of early failure of primary and revision total knee arthroplasty.
      ,
      • Meyer J.A.
      • Zhu M.
      • Cavadino A.
      • Coleman B.
      • Munro J.T.
      • Young S.W.
      Infection and periprosthetic fracture are the leading causes of failure after aseptic revision total knee arthroplasty.
      ]. As such, these findings suggest patients with HCV have an increased risk of both aseptic and septic failure after rTKAs.
      Within the HCV cohort, patients who underwent an initial septic rTKA exhibited significantly higher rates of 90-day medical complications, all-cause repeat rTKAs, and re-revisions for PJI than patients with an initial aseptic rTKA. This result is consistent with prior literature reporting higher complication rates after a septic rTKA [
      • Belt M.
      • Hannink G.
      • Smolders J.
      • Spekenbrink-Spooren A.
      • Schreurs B.W.
      • Smulders K.
      Reasons for revision are associated with rerevised total knee arthroplasties: an analysis of 8,978 index revisions in the Dutch Arthroplasty Register.
      ]. Interestingly, patients with HCV who underwent an initial aseptic rTKA exhibited a significantly higher rate of aseptic loosening. In addition to recurrent infection, aseptic loosening is a major etiology of failure after rTKAs. Kienzle et al. reported high rates of aseptic loosening (22%) after a septic rTKA at 7.3-year follow-up [
      • Kienzle A.
      • Walter S.
      • von Roth P.
      • Fuchs M.
      • Winkler T.
      • Müller M.
      High rates of aseptic loosening after revision total knee arthroplasty for periprosthetic joint infection.
      ] although rates were <5% at 2 years which is comparable to the rate of 2.6% found in this study. Belt et al. demonstrated that indications for subsequent revisions are most commonly recurrences of the initial rTKA indication [
      • Belt M.
      • Hannink G.
      • Smolders J.
      • Spekenbrink-Spooren A.
      • Schreurs B.W.
      • Smulders K.
      Reasons for revision are associated with rerevised total knee arthroplasties: an analysis of 8,978 index revisions in the Dutch Arthroplasty Register.
      ]. Therefore, our data showing higher rates aseptic loosening among patients with HCV who underwent an initial aseptic rTKA may simply reflect higher rates of loosening at the time of the index rTKA.

      Limitations

      There are several limitations to this study. First, by only evaluating complications within 2 years, this analysis is limited to short-term outcomes. Furthermore, because continuous database enrollment for 2 years after an rTKA was required for inclusion, patients who died within 2 years after the surgery were excluded. Therefore, these results may not be applicable to patients with a high perioperative mortality risk. This limitation is notable given that the mortality risk after an rTKA is not insignificant, especially for patients undergoing a septic rTKA and those with a greater comorbidity burden, both of which are more common in patients with HCV [
      • Boddapati V.
      • Fu M.C.
      • Mayman D.J.
      • Su E.P.
      • Sculco P.K.
      • McLawhorn A.S.
      Revision total knee arthroplasty for periprosthetic joint infection is associated with increased postoperative morbidity and mortality relative to noninfectious revisions.
      ,
      • Sinclair S.T.
      • Orr M.N.
      • Rothfusz C.A.
      • Klika A.K.
      • McLaughlin J.P.
      • Piuzzi N.S.
      Understanding the 30-day mortality burden after revision total knee arthroplasty.
      ,
      • Drain N.P.
      • Bertolini D.M.
      • Anthony A.W.
      • Feroze M.W.
      • Chao R.
      • Onyekweli T.
      • et al.
      High mortality after total knee arthroplasty periprosthetic joint infection is related to preoperative morbidity and the disease process but not treatment.
      ]. Additionally, the possibility of coding errors is inherent with any analysis of administrative claims data. Such instances are rare and made up only 0.7% of Medicare and Medicaid payments in 2021 [
      Centers for Medicare and Medicaid Services
      2021 Medicare fee-for-service supplemental improper payment data.
      ]. However, in a recent validation study comparing billing records against operative reports, Roof et al. found that ICD-10 procedural coding for rTKA is often imprecise [
      • Roof M.A.
      • Lygrisse K.
      • Keitel L.
      • Siddiqi A.
      • Emara A.
      • Piuzzi N.S.
      • et al.
      How accurate is ICD-10 coding for revision total knee arthroplasty?.
      ]. Therefore, it must be acknowledged that including ICD-10 procedural codes in this study consequently introduced coding bias. The impact of this limitation was mitigated through nuanced definitions of aseptic and septic rTKAs (Appendix A), as well as the concomitant use of both CPT and ICD-9 procedural codes which have demonstrated high accuracy in validation studies [
      • Daneshvar P.
      • Forster A.J.
      • Dervin G.F.
      Accuracy of administrative coding in identifying hip and knee primary replacements and revisions.
      ,
      • Singh J.A.
      • Ayub S.
      Accuracy of VA databases for diagnoses of knee replacement and hip replacement.
      ]. With respect to aseptic loosening, it is possible that some patients who underwent a septic rTKA had clinically significant implant loosening attributed to infection. It is also possible that some patients who underwent an aseptic rTKA had occult infections during the initial and/or subsequent revision that were undiagnosed. Given that loosening is common in the setting of PJI [
      • Dapunt U.
      • Radzuweit-Mihaljevic S.
      • Lehner B.
      • Haensch G.M.
      • Ewerbeck V.
      Bacterial infection and implant loosening in hip and knee arthroplasty: evaluation of 209 cases.
      ,
      • Kempthorne J.T.
      • Ailabouni R.
      • Raniga S.
      • Hammer D.
      • Hooper G.
      Occult infection in aseptic joint loosening and the diagnostic role of implant sonication.
      ,
      • Parvizi J.
      • Tarity T.D.
      • Steinbeck M.J.
      • Politi R.G.
      • Joshi A.
      • Purtill J.J.
      • et al.
      Management of stiffness following total knee arthroplasty.
      ], such cases may have influenced the results.
      It is possible that some patients had HCV but were never diagnosed and, therefore, could have been included in the control cohort. An additional limitation is that the database does not contain information regarding the viral load which prevents comprehensive characterization of HCV patients’ clinical status at the time of rTKAs. It was also not possible to identify HCV genotypes or the degree of liver damage (eg, Child-Pugh score). The HCV and control cohorts also differed significantly with respect to U.S. regional distribution, which may reflect geographic differences in HCV prevalence. Region was controlled for in the logistic regression analyses to mitigate the impact of this possible confounder. Lastly, although exact matching and multivariable regression were used, other confounders could have influenced the results. For example, there are certain confounding variables that may be more common in the HCV cohort such as intravenous drug and alcohol abuse, homelessness, and low socioeconomic status, all of which are only partially available or entirely inaccessible in the database. BMI data were also not universally available for all included patients, and therefore, the adjustment for BMI was incomplete.
      To the authors’ knowledge, this is the first analysis of complications after rTKAs in patients with HCV. As these patients often have more comorbidities than most patients undergoing an rTKA and higher comorbidity burdens are associated with an increased complication risk [
      • Cochrane N.H.
      • Wellman S.S.
      • Lachiewicz P.F.
      Early infection after aseptic revision knee arthroplasty: prevalence and predisposing risk factors.
      ,
      • Hussein I.H.
      • Zalikha A.K.
      • Tuluca A.
      • Crespi Z.
      • El-Othmani M.M.
      Epidemiology of obese patients undergoing revision total knee arthroplasty: understanding demographics, comorbidities, and propensity weighted analysis of inpatient outcomes.
      ,
      • Ross P.E.
      • Young J.R.
      • O’Connor C.M.
      • Anoushiravani A.A.
      • DiCaprio M.R.
      Perioperative management of hepatitis C in patients undergoing total joint arthroplasty.
      ,
      • Jain N.B.
      • Guller U.
      • Pietrobon R.
      • Bond T.K.
      • Higgins L.D.
      Comorbidities increase complication rates in patients having arthroplasty.
      ], medical optimization before an rTKA is critical in this population. Recent analyses have reported improved outcomes after a primary arthroplasty in patients with HCV who received preoperative antiviral treatment [
      • Ross A.J.
      • Ross B.J.
      • Lee O.C.
      • Weldy J.M.
      • Sherman W.F.
      • Sanchez F.L.
      A missed opportunity; the impact of hepatitis C treatment prior to total knee arthroplasty on postoperative complications.
      ,
      • Novikov D.
      • Feng J.E.
      • Anoushiravani A.A.
      • Vigdorchik J.M.
      • Lajam C.M.
      • Seyler T.M.
      • et al.
      Undetectable hepatitis C viral load is associated with improved outcomes following total joint arthroplasty.
      ,
      • Bendich I.
      • Takemoto S.
      • Patterson J.T.
      • Monto A.
      • Barber T.C.
      • Kuo A.C.
      Preoperative treatment of hepatitis C is associated with lower prosthetic joint infection rates in US veterans.
      ,
      • Schwarzkopf R.
      • Novikov D.
      • Anoushiravani A.A.
      • Feng J.E.
      • Vigdorchik J.
      • Schurko B.
      • et al.
      The preoperative management of hepatitis C may improve the outcome after total knee arthroplasty.
      ]. It is likely that similar interventions before an rTKA also decrease the complication risk in this population. Future studies are needed to investigate this hypothesis.

      Conclusions

      Patients with HCV exhibited significantly higher rates of postoperative medical complications, subsequent rTKA, PJI, and periprosthetic fracture relative to matched controls. Within the HCV cohort, patients who underwent an initial septic rTKA exhibited significantly higher rates of medical complications, all-cause repeat rTKAs, and re-revisions for PJI than patients with an initial aseptic rTKA. These data suggest that, similar to primary TKAs, HCV is a risk factor for poor outcomes following an rTKA.

      Conflicts of interest

      Dr. G. N. Guild is a paid consultant for and receives research support from Smith & Nephew and has stock or stock options in Total Joint Orthopaedics. All other authors declare no potential conflicts of interest.
      For full disclosure statements refer to https://doi.org/10.1016/j.artd.2022.09.010.

      Appendix A. Supplementary Data

      Appendix

      Appendix: PearlDiver Codes .
      Appendix ACodes used to define inclusion and exclusion criteria.
      CriteriaCode(s)
      Hepatitis CICD-9-D-07051, ICD-9-D-07054, ICD-9-D-07070, ICD-10-D-B1710, ICD-10-D-B182, ICD-10-D-B1920
      Aseptic rTKA
       rTKACPT-27486, CPT-27487, ICD-9-P-0080, ICD-9-P-0081, ICD-9-P-0082, ICD-9-P-0083, ICD-9-P-0084, ICD-10-P-0SWC0JZ, ICD-10-P-0SWD0JZ, ICD-10-P-0SWC0JC, ICD-10-P-0SWCXJZ, ICD-10-P-0SWD0JC, ICD-10-P-0SWV0JZ, ICD-10-P-0SWDXJZ, ICD-10-P-0SWW0JZ, ICD-10-P-0SWC09Z, ICD-10-P-0SWT0JZ, ICD-10-P-0SWD09Z, ICD-10-P-0SWU0JZ
       PJI (excluded)ICD-9-D-99666, ICD-10-D-M01X61, ICD-10-D-M01X62, ICD-10-D-M01X69, ICD-10-D-T8453XA, ICD-10-D-T8453XD, ICD-10-D-T8453XS, ICD-10-D-T8454XA, ICD-10-D-T8454XD, ICD-10-D-T8454XS
      1-Stage septic rTKA
       rTKACPT-27487, ICD-9-P-0080, ICD-10-P-0SWC0JZ, ICD-10-P-0SWCXJZ, ICD-10-P-0SWD0JZ, ICD-10-P-0SWDXJZ
       PJIICD-9-D-99666, ICD-10-D-M01X61, ICD-10-D-M01X62, ICD-10-D-M01X69, ICD-10-D-T8453XA, ICD-10-D-T8453XD, ICD-10-D-T8453XS, ICD-10-D-T8454XA, ICD-10-D-T8454XD, ICD-10-D-T8454XS
      2-Stage septic rTKA
       First stage
      Hardware removalCPT-27488, ICD-9-P-8006, ICD-10-P-0SPC0JZ, ICD-10-P-0SPD0JZ
      Spacer insertionCPT-11981, ICD-9-P-8456, ICD-10-P-0SHC08Z, ICD-10-P-0SHD08Z, ICD-10-P-0SRC0EZ, ICD-10-P-0SRD0EZ
       Second stage
      TKACPT-27447, ICD-9-P-8154, ICD-10-P-0SRC0J9, ICD-10-P-0SRC0JA, ICD-10-P-0SRC0JZ, ICD-10-P-0SRD0J9, ICD-10-P-0SRD0JA, ICD-10-P-0SRD0JZ
      Spacer removalCPT-11982, ICD-9-P-8457, ICD-10-P-0SPC08Z, ICD-10-P-0SPC0EZ, ICD-10-P-0SPD08Z, ICD-10-P-0SPD0EZ
      Exclusion criteria
       Hepatitis BICD-9-D-07020, ICD-9-D-07021, ICD-9-D-07022, ICD-9-D-07023, ICD-9-D-07030, ICD-9-D-07031, ICD-9-D-07032, ICD-9-D-07033, ICD-9-D-V0261, ICD-10-D-B160, ICD-10-D-B161, ICD-10-D-B162, ICD-10-D-B169, ICD-10-D-B180, ICD-10-D-B181, ICD-10-D-B1910, ICD-10-D-B1911, ICD-10-D-Z2251
       HIVICD-9-D-042, ICD-9-D-07953, ICD-10-D-B20, ICD-10-D-Z21, ICD-10-D-B9735
      Comorbidities
       Tobacco useICD-9-D-3051, ICD-9-D-V1582, ICD-10-D-F17220, ICD-10-D-F17221, ICD-10-D-F17223, ICD-10-D-F17228, ICD-10-D-F17229, ICD-10-D-F17290, ICD-10-D-F17291, ICD-10-D-F17293, ICD-10-D-F17298, ICD-10-D-F17299, ICD-10-D-Z720
       Diabetes mellitusICD-9-D-24900:ICD-9-D-25099, ICD-9-D-7902, ICD-9-D-79021, ICD-9-D-79022, ICD-9-D-79029, ICD-9-D-7915, ICD-9-D-7916, ICD-10-D-E090:ICD-10-D-E139
       ObesityICD-9-D-2780, ICD-9-D-27800, ICD-9-D-27801, ICD-9-D-27802, ICD-9-D-27803, ICD-10-D-E660:ICD-10-D-E669
       Chronic kidney diseaseICD-9-D-585, ICD-9-D-5851, ICD-9-D-5852, ICD-9-D-5853, ICD-9-D-5854, ICD-9-D-5855, ICD-9-D-5856, ICD-9-D-5859, ICD-9-D-7925, ICD-10-D-N18:ICD-10-D-N189
       HypertensionICD-9-D-4010:ICD-9-D-4059, ICD-10-D-I10:ICD-10-D-I159
      Appendix BCodes used to define complications.
      ComplicationCode(s)
      DVTICD-9-D-4532, ICD-9-D-4533, ICD-9-D-4534, ICD-9-D-45382, ICD-9-D-45384, ICD-9-D-45385, ICD-9-D-45386, ICD-10-D-I26:ICD-10-D-I2699
      PEICD-9-D-4151:ICD-9-D-4159, ICD-10-D-I26:ICD-10-D-I269
      Blood transfusionICD-9-P-9904, ICD-10-P-3023, ICD-10-P-30230AZ, ICD-10-P-30230G0, ICD-10-P- 30230G2, ICD-10-P-30230G3, ICD-10-P-30230G4, ICD-10-P-30230H0, ICD-10-P-30230H1, ICD-10-P- 30230J0, ICD-10-P-30230J1, ICD-10-P-30230K0, ICD-10-P-30230K1, ICD-10-P-30230L0, ICD-10-P-30230L1, ICD- 10-P-30230M0, ICD-10-P-30230M1, ICD-10-P-30230N0, ICD-10-P-30230N1, ICD-10-P-30230P0, ICD-10-P- 30230P1, ICD-10-P-30230Q0, ICD-10-P-30230Q1, ICD-10-P-30230R0, ICD-10-P-30230R1, ICD-10-P- 30230S0, ICD-10-P-30230S1, ICD-10-P-30230T0, ICD-10-P-30230T1, ICD-10-P-30230V0, ICD-10-P- 30230V1, ICD-10-P-30230W0, ICD-10-P-30230W1, ICD-10-P-30230X0, ICD-10-P-30230X2, ICD-10-P- 30230X3, ICD-10-P-30230X4, ICD-10-P-30230Y0, ICD-10-P-30230Y2, ICD-10-P-30230Y3, ICD-10-P- 30230Y4, ICD-10-P-30233AZ, ICD-10-P-30233G0, ICD-10-P-30233G2, ICD-10-P-30233G3, ICD-10-P- 30233G4, ICD-10-P-30233H0, ICD-10-P-30233H1, ICD-10-P-30233J0, ICD-10-P-30233J1, ICD-10-P- 30233K0, ICD-10-P-30233K1, ICD-10-P-30233L0, ICD-10-P-30233L1, ICD-10-P-30233M0, ICD-10-P- 30233M1, ICD-10-P-30233N0, ICD-10-P-30233N1, ICD-10-P-30233P0, ICD-10-P-30233P1, ICD-10-P- 30233Q0, ICD-10-P-30233Q1, ICD-10-P-30233R0, ICD-10-P-30233R1, ICD-10-P-30233S0, ICD-10-P- 30233S1, ICD-10-P-30233T0, ICD-10-P-30233T1, ICD-10-P-30233V0, ICD-10-P-30233V1, ICD-10-P- 30233W0, ICD-10-P-30233W1, ICD-10-P-30233X0, ICD-10-P-30233X2, ICD-10-P-30233X3, ICD-10-P- 30233X4, ICD-10-P-30233Y0, ICD-10-P-30233Y2, ICD-10-P-30233Y3, ICD-10-P-30233Y4, ICD-10-P- 30240AZ, ICD-10-P-30240G0, ICD-10-P-30240G2, ICD-10-P-30240G3, ICD-10-P-30240G4, ICD-10-P- 30240H0, ICD-10-P-30240H1, ICD-10-P-30240J0, ICD-10-P-30240J1, ICD-10-P-30240K0, ICD-10-P- 30240K1, ICD-10-P-30240L0, ICD-10-P-30240L1, ICD-10-P-30240M0, ICD-10-P-30240M1, ICD-10-P- 30240N0, ICD-10-P-30240N1, ICD-10-P-30240P0, ICD-10-P-30240P1, ICD-10-P-30240Q0, ICD-10-P- 30240Q1, ICD-10-P-30240R0, ICD-10-P-30240R1, ICD-10-P-30240S0, ICD-10-P-30240S1, ICD-10-P- 30240T0, ICD-10-P-30240T1, ICD-10-P-30240V0, ICD-10-P-30240V1, ICD-10-P-30240W0, ICD-10-P- 30240W1, ICD-10-P-30240X0, ICD-10-P-30240X2, ICD-10-P-30240X3, ICD-10-P-30240X4, ICD-10-P- 30240Y0, ICD-10-P-30240Y2, ICD-10-P-30240Y3, ICD-10-P-30240Y4, ICD-10-P-30243AZ, ICD-10-P- 30243G0, ICD-10-P-30243G2, ICD-10-P-30243G3, ICD-10-P-30243G4, ICD-10-P-30243H0, ICD-10-P- 30243H1, ICD-10-P-30243J0, ICD-10-P-30243J1, ICD-10-P-30243K0, ICD-10-P-30243K1, ICD-10-P-30243L0, ICD- 10-P-30243L1, ICD-10-P-30243M0, ICD-10-P-30243M1, ICD-10-P-30243N0, ICD-10-P-30243N1, ICD-10-P- 30243P0, ICD-10-P-30243P1, ICD-10-P-30243Q0, ICD-10-P-30243Q1, ICD-10-P-30243R0, ICD-10-P- 30243R1, ICD-10-P-30243S0, ICD-10-P-30243S1, ICD-10-P-30243T0, ICD-10-P-30243T1, ICD-10-P- 30243V0, ICD-10-P-30243V1, ICD-10-P-30243W0, ICD-10-P-30243W1, ICD-10-P-30243X0, ICD-10-P- 30243X2, ICD-10-P-30243X3, ICD-10-P-30243X4, ICD-10-P-30243Y0, ICD-10-P-30243Y2, ICD-10-P- 30243Y3, ICD-10-P-30243Y4, ICD-10-P-30250G0, ICD-10-P-30250G1, ICD-10-P-30250H0, ICD-10-P- 30250H1, ICD-10-P-30250J0, ICD-10-P-30250J1, ICD-10-P-30250K0, ICD-10-P-30250K1, ICD-10-P-30250L0, ICD- 10-P-30250L1, ICD-10-P-30250M0, ICD-10-P-30250M1, ICD-10-P-30250N0, ICD-10-P-30250N1, ICD-10-P- 30250P0, ICD-10-P-30250P1, ICD-10-P-30250Q0, ICD-10-P-30250Q1, ICD-10-P-30250R0, ICD-10-P- 30250R1, ICD-10-P-30250S0, ICD-10-P-30250S1, ICD-10-P-30250T0, ICD-10-P-30250T1, ICD-10-P- 30250V0, ICD-10-P-30250V1, ICD-10-P-30250W0, ICD-10-P-30250W1, ICD-10-P-30250X0, ICD-10-P- 30250X1, ICD-10-P-30250Y0, ICD-10-P-30250Y1, ICD-10-P-30253G0, ICD-10-P-30253G1, ICD-10-P- 30253H0, ICD-10-P-30253H1, ICD-10-P-30253J0, ICD-10-P-30253J1, ICD-10-P-30253K0, ICD-10-P- 30253K1, ICD-10-P-30253L0, ICD-10-P-30253L1, ICD-10-P-30253M0, ICD-10-P-30253M1, ICD-10-P- 30253N0, ICD-10-P-30253N1, ICD-10-P-30253P0, ICD-10-P-30253P1, ICD-10-P-30253Q0, ICD-10-P- 30253Q1, ICD-10-P-30253R0, ICD-10-P-30253R1, ICD-10-P-30253S0, ICD-10-P-30253S1, ICD-10-P- 30253T0, ICD-10-P-30253T1, ICD-10-P-30253V0, ICD-10-P-30253V1, ICD-10-P-30253W0, ICD-10-P- 30253W1, ICD-10-P-30253X0, ICD-10-P-30253X1, ICD-10-P-30253Y0, ICD-10-P-30253Y1, ICD-10-P- 30260G0, ICD-10-P-30260G1, ICD-10-P-30260H0, ICD-10-P-30260H1, ICD-10-P-30260J0, ICD-10-P- 30260J1, ICD-10-P-30260K0, ICD-10-P-30260K1, ICD-10-P-30260L0, ICD-10-P-30260L1, ICD-10-P- 30260M0, ICD-10-P-30260M1, ICD-10-P-30260N0, ICD-10-P-30260N1, ICD-10-P-30260P0, ICD-10-P- 30260P1, ICD-10-P-30260Q0, ICD-10-P-30260Q1, ICD-10-P-30260R0, ICD-10-P-30260R1, ICD-10-P- 30260S0, ICD-10-P-30260S1, ICD-10-P-30260T0, ICD-10-P-30260T1, ICD-10-P-30260V0, ICD-10-P- 30260V1, ICD-10-P-30260W0, ICD-10-P-30260W1, ICD-10-P-30260X0, ICD-10-P-30260X1, ICD-10-P- 30260Y0, ICD-10-P-30260Y1, ICD-10-P-30263G0, ICD-10-P-30263G1, ICD-10-P-30263H0, ICD-10-P- 30263H1, ICD-10-P-30263J0, ICD-10-P-30263J1, ICD-10-P-30263K0, ICD-10-P-30263K1, ICD-10-P-30263L0, ICD- 10-P-30263L1, ICD-10-P-30263M0, ICD-10-P-30263M1, ICD-10-P-30263N0, ICD-10-P-30263N1, ICD-10-P- 30263P0, ICD-10-P-30263P1, ICD-10-P-30263Q0, ICD-10-P-30263Q1, ICD-10-P-30263R0, ICD-10-P- 30263R1, ICD-10-P-30263S0, ICD-10-P-30263S1, ICD-10-P-30263T0, ICD-10-P-30263T1, ICD-10-P- 30263V0, ICD-10-P-30263V1, ICD-10-P-30263W0, ICD-10-P-30263W1, ICD-10-P-30263X0, ICD-10-P-30263X1, ICD-10-P-30263Y0, ICD-10-P-30263Y1, ICD-10-P-30273H1, ICD-10-P-30273J1, ICD-10-P-30273K1, ICD-10-P- 30273L1, ICD-10-P-30273M1, ICD-10-P-30273N1, ICD-10-P-30273P1, ICD-10-P-30273Q1, ICD-10-P- 30273R1, ICD-10-P-30273S1, ICD-10-P-30273T1, ICD-10-P-30273V1, ICD-10-P-30273W1, ICD-10-P- 30277H1, ICD-10-P-30277J1, ICD-10-P-30277K1, ICD-10-P-30277L1, ICD-10-P-30277M1, ICD-10-P- 30277N1, ICD-10-P-30277P1, ICD-10-P-30277Q1, ICD-10-P-30277R1, ICD-10-P-30277S1, ICD-10-P- 30277T1, ICD-10-P-30277V1, ICD-10-P-30277W1, ICD-10-P-30280B1, ICD-10-P-30283B1
      Urinary tract infectionICD-9-D-5990, ICD-10-D-N390
      Acute kidney injuryICD-9-D-5845, ICD-9-D-5846, ICD-9-D-5847, ICD-9-D-5848, ICD-9-D-5849, ICD-10-D-N17:ICD-10-D-N179
      MUA/LoACPT-27570, CPT-29884
      Periprosthetic fractureICD-9-D-99644, ICD-10-D-M9712XA, ICD-10-D-T84042A, ICD-10-D-T84043A, ICD-10-D-M9711XA
      Aseptic looseningICD-9-D-99641, ICD-10-D-T84032A, ICD-10-D-T84032D, ICD-10-D-T84032S, ICD-10-D-T84033A, ICD-10-D-T84033D, ICD-10-D-T84033S
      Secondary rTKAAny codes for rTKA (aseptic or septic) as outlined in Table S1
      PJISame criteria outlined in Table S1 for septic rTKA
      DVT, deep vein thrombosis; LoA, lysis of adhesions; MUA, manipulation under anesthesia; PE, pulmonary embolism.

      References

        • Cram P.
        • Landon B.E.
        • Matelski J.
        • Ling V.
        • Stukel T.A.
        • Paterson J.M.
        • et al.
        Utilization and short-term outcomes of primary total hip and knee arthroplasty in the United States and Canada: an analysis of New York and Ontario administrative data.
        Arthritis Rheum. 2018; 70: 547-554
        • Kurtz S.
        • Ong K.
        • Lau E.
        • Mowat F.
        • Halpern M.
        Projections of primary and revision hip and knee arthroplasty in the United States from 2005 to 2030.
        J Bone Joint Surg Am. 2007; 89: 780-785
        • Singh J.A.
        • Yu S.
        • Chen L.
        • Cleveland J.D.
        Rates of total joint replacement in the United States: future projections to 2020-2040 using the national inpatient sample.
        J Rheumatol. 2019; 46: 1134-1140
        • Sloan M.
        • Premkumar A.
        • Sheth N.P.
        Projected volume of primary total joint arthroplasty in the U.S., 2014 to 2030.
        J Bone Joint Surg Am. 2018; 100: 1455-1460
        • Brockman B.S.
        • Maupin J.J.
        • Thompson S.F.
        • Hollabaugh K.M.
        • Thakral R.
        Complication rates in total knee arthroplasty performed for osteoarthritis and post-traumatic arthritis: a comparison study.
        J Arthroplasty. 2020; 35: 371-374
        • Bozic K.J.
        • Kurtz S.M.
        • Lau E.
        • Ong K.
        • Chiu V.
        • Vail T.P.
        • et al.
        The epidemiology of revision total knee arthroplasty in the United States.
        Clin Orthop Relat Res. 2010; 468: 45-51
        • Schwartz A.M.
        • Farley K.X.
        • Guild G.N.
        • Bradbury Jr., T.L.
        Projections and epidemiology of revision hip and knee arthroplasty in the United States to 2030.
        J Arthroplasty. 2020; 35: S79-S85
        • Ong K.L.
        • Lau E.
        • Suggs J.
        • Kurtz S.M.
        • Manley M.T.
        Risk of subsequent revision after primary and revision total joint arthroplasty.
        Clin Orthop Relat Res. 2010; 468: 3070-3076
        • Jaffer A.K.
        • Barsoum W.K.
        • Krebs V.
        • Hurbanek J.G.
        • Morra N.
        • Brotman D.J.
        Duration of anesthesia and venous thromboembolism after hip and knee arthroplasty.
        Mayo Clin Proc. 2005; 80: 732-738
        • Mahomed N.N.
        • Barrett J.A.
        • Katz J.N.
        • Phillips C.B.
        • Losina E.
        • Lew R.A.
        • et al.
        Rates and outcomes of primary and revision total hip replacement in the United States medicare population.
        J Bone Joint Surg Am. 2003; 85: 27-32
        • Pulido L.
        • Parvizi J.
        • Macgibeny M.
        • Sharkey P.F.
        • Purtill J.J.
        • Rothman R.H.
        • et al.
        In hospital complications after total joint arthroplasty.
        J Arthroplasty. 2008; 23: 139-145
        • Zhan C.
        • Kaczmarek R.
        • Loyo-Berrios N.
        • Sangl J.
        • Bright R.A.
        Incidence and short-term outcomes of primary and revision hip replacement in the United States.
        J Bone Joint Surg Am. 2007; 89: 526-533
        • Roman M.D.
        • Russu O.
        • Mohor C.
        • Necula R.
        • Boicean A.
        • Todor A.
        • et al.
        Outcomes in revision total knee arthroplasty (Review).
        Exp Ther Med. 2022; 23: 29
        • Hofmeister M.G.
        • Rosenthal E.M.
        • Barker L.K.
        • Rosenberg E.S.
        • Barranco M.A.
        • Hall E.W.
        • et al.
        Estimating prevalence of hepatitis C virus infection in the United States, 2013-2016.
        Hepatology. 2019; 69: 1020-1031
        • Issa K.
        • Boylan M.R.
        • Naziri Q.
        • Perfetti D.C.
        • Maheshwari A.V.
        • Mont M.A.
        The impact of hepatitis C on short-term outcomes of total joint arthroplasty.
        J Bone Joint Surg Am. 2015; 97: 1952-1957
        • Ross A.J.
        • Ross B.J.
        • Lee O.C.
        • Weldy J.M.
        • Sherman W.F.
        • Sanchez F.L.
        A missed opportunity; the impact of hepatitis C treatment prior to total knee arthroplasty on postoperative complications.
        J Arthroplasty. 2021; 37: 709-713.e2
        • Cancienne J.M.
        • Kandahari A.M.
        • Casp A.
        • Novicoff W.
        • Browne J.A.
        • Cui Q.
        • et al.
        Complication rates after total hip and knee arthroplasty in patients with hepatitis C compared with matched control patients.
        J Am Acad Orthop Surg. 2017; 25: e275-e281
        • Cheng T.
        • Yang C.
        • Hao L.
        • Cheng X.
        • Hu J.
        • Ren W.
        • et al.
        Hepatitis C virus infection increases the risk of adverse outcomes following joint arthroplasty: a meta-analysis of observational studies.
        Orthop Traumatol Surg Res. 2021; 108: 102947
        • Kildow B.J.
        • Politzer C.S.
        • DiLallo M.
        • Bolognesi M.P.
        • Seyler T.M.
        Short and long-term postoperative complications following total joint arthroplasty in patients with human immunodeficiency virus, hepatitis B, or hepatitis C.
        J Arthroplasty. 2018; 33: S86-S92
        • Wei W.
        • Liu T.
        • Zhao J.
        • Li B.
        • Li S.
        • Liu J.
        Does the hepatitis C virus affect the outcomes of total joint arthroplasty? A meta-analysis of ten studies.
        J Orthop Sci. 2019; 24: 822-829
        • Samuel L.T.
        • Grits D.
        • Acuña A.J.
        • Piuzzi N.S.
        • Higuera-Rueda C.A.
        • Kamath A.F.
        Work relative value units do not adequately support the burden of infection management in revision knee arthroplasty.
        J Bone Joint Surg Am. 2020; 102: 230-236
        • Kramer J.R.
        • Davila J.A.
        • Miller E.D.
        • Richardson P.
        • Giordano T.P.
        • El-Serag H.B.
        The validity of viral hepatitis and chronic liver disease diagnoses in Veterans affairs administrative databases.
        Aliment Pharmacol Ther. 2008; 27: 274-282
        • Niu B.
        • Forde K.A.
        • Goldberg D.S.
        Coding algorithms for identifying patients with cirrhosis and hepatitis B or C virus using administrative data.
        Pharmacoepidemiol Drug Saf. 2015; 24: 107-111
        • Best M.J.
        • Buller L.T.
        • Klika A.K.
        • Barsoum W.K.
        Increase in perioperative complications following primary total hip and knee arthroplasty in patients with hepatitis C without cirrhosis.
        J Arthroplasty. 2015; 30: 663-668
        • Delanois R.E.
        • Mistry J.B.
        • Gwam C.U.
        • Mohamed N.S.
        • Choksi U.S.
        • Mont M.A.
        Current epidemiology of revision total knee arthroplasty in the United States.
        J Arthroplasty. 2017; 32: 2663-2668
        • Belt M.
        • Hannink G.
        • Smolders J.
        • Spekenbrink-Spooren A.
        • Schreurs B.W.
        • Smulders K.
        Reasons for revision are associated with rerevised total knee arthroplasties: an analysis of 8,978 index revisions in the Dutch Arthroplasty Register.
        Acta Orthop. 2021; 92: 597-601
        • Boddapati V.
        • Fu M.C.
        • Mayman D.J.
        • Su E.P.
        • Sculco P.K.
        • McLawhorn A.S.
        Revision total knee arthroplasty for periprosthetic joint infection is associated with increased postoperative morbidity and mortality relative to noninfectious revisions.
        J Arthroplasty. 2018; 33: 521-526
        • Dai W.L.
        • Lin Z.M.
        • Shi Z.J.
        • Wang J.
        Outcomes following revision total knee arthroplasty septic versus aseptic failure: a national propensity-score-matched comparison.
        J Knee Surg. 2021; 34: 1227-1236
        • Evangelopoulos D.S.
        • Ahmad S.S.
        • Krismer A.M.
        • Albers C.E.
        • Hoppe S.
        • Kleer B.
        • et al.
        Periprosthetic infection: major cause of early failure of primary and revision total knee arthroplasty.
        J Knee Surg. 2019; 32: 941-946
        • Meyer J.A.
        • Zhu M.
        • Cavadino A.
        • Coleman B.
        • Munro J.T.
        • Young S.W.
        Infection and periprosthetic fracture are the leading causes of failure after aseptic revision total knee arthroplasty.
        Arch Orthop Trauma Surg. 2021; 141: 1373-1383
        • Kienzle A.
        • Walter S.
        • von Roth P.
        • Fuchs M.
        • Winkler T.
        • Müller M.
        High rates of aseptic loosening after revision total knee arthroplasty for periprosthetic joint infection.
        JB JS Open Access. 2020; 5: e20.00026
        • Sinclair S.T.
        • Orr M.N.
        • Rothfusz C.A.
        • Klika A.K.
        • McLaughlin J.P.
        • Piuzzi N.S.
        Understanding the 30-day mortality burden after revision total knee arthroplasty.
        Arthroplast Today. 2021; 11: 205-211
        • Drain N.P.
        • Bertolini D.M.
        • Anthony A.W.
        • Feroze M.W.
        • Chao R.
        • Onyekweli T.
        • et al.
        High mortality after total knee arthroplasty periprosthetic joint infection is related to preoperative morbidity and the disease process but not treatment.
        J Arthroplasty. 2022; 37: 1383-1389
        • Centers for Medicare and Medicaid Services
        2021 Medicare fee-for-service supplemental improper payment data.
        www.cms.gov
        Date: 2021
        ([accessed 29.03.22])
        • Roof M.A.
        • Lygrisse K.
        • Keitel L.
        • Siddiqi A.
        • Emara A.
        • Piuzzi N.S.
        • et al.
        How accurate is ICD-10 coding for revision total knee arthroplasty?.
        J Arthroplasty. 2021; 36: 3950-3958
        • Daneshvar P.
        • Forster A.J.
        • Dervin G.F.
        Accuracy of administrative coding in identifying hip and knee primary replacements and revisions.
        J Eval Clin Pract. 2012; 18: 555-559
        • Singh J.A.
        • Ayub S.
        Accuracy of VA databases for diagnoses of knee replacement and hip replacement.
        Osteoarthritis Cartilage. 2010; 18: 1639-1642
        • Dapunt U.
        • Radzuweit-Mihaljevic S.
        • Lehner B.
        • Haensch G.M.
        • Ewerbeck V.
        Bacterial infection and implant loosening in hip and knee arthroplasty: evaluation of 209 cases.
        Materials (Basel). 2016; 9: 871
        • Kempthorne J.T.
        • Ailabouni R.
        • Raniga S.
        • Hammer D.
        • Hooper G.
        Occult infection in aseptic joint loosening and the diagnostic role of implant sonication.
        Biomed Res Int. 2015; 2015: 946215
        • Parvizi J.
        • Tarity T.D.
        • Steinbeck M.J.
        • Politi R.G.
        • Joshi A.
        • Purtill J.J.
        • et al.
        Management of stiffness following total knee arthroplasty.
        J Bone Joint Surg Am. 2006; 88 Suppl 4: 175-181
        • Cochrane N.H.
        • Wellman S.S.
        • Lachiewicz P.F.
        Early infection after aseptic revision knee arthroplasty: prevalence and predisposing risk factors.
        J Arthroplasty. 2022; 37: S281-S285
        • Hussein I.H.
        • Zalikha A.K.
        • Tuluca A.
        • Crespi Z.
        • El-Othmani M.M.
        Epidemiology of obese patients undergoing revision total knee arthroplasty: understanding demographics, comorbidities, and propensity weighted analysis of inpatient outcomes.
        J Am Acad Orthop Surg Glob Res Rev. 2022; 6: e21.00263
        • Ross P.E.
        • Young J.R.
        • O’Connor C.M.
        • Anoushiravani A.A.
        • DiCaprio M.R.
        Perioperative management of hepatitis C in patients undergoing total joint arthroplasty.
        JBJS Rev. 2021; 9: e20.00223
        • Jain N.B.
        • Guller U.
        • Pietrobon R.
        • Bond T.K.
        • Higgins L.D.
        Comorbidities increase complication rates in patients having arthroplasty.
        Clin Orthop Relat Res. 2005; 435: 232-238
        • Novikov D.
        • Feng J.E.
        • Anoushiravani A.A.
        • Vigdorchik J.M.
        • Lajam C.M.
        • Seyler T.M.
        • et al.
        Undetectable hepatitis C viral load is associated with improved outcomes following total joint arthroplasty.
        J Arthroplasty. 2019; 34: 2890-2897
        • Bendich I.
        • Takemoto S.
        • Patterson J.T.
        • Monto A.
        • Barber T.C.
        • Kuo A.C.
        Preoperative treatment of hepatitis C is associated with lower prosthetic joint infection rates in US veterans.
        J Arthroplasty. 2019; 34: S319-S326
        • Schwarzkopf R.
        • Novikov D.
        • Anoushiravani A.A.
        • Feng J.E.
        • Vigdorchik J.
        • Schurko B.
        • et al.
        The preoperative management of hepatitis C may improve the outcome after total knee arthroplasty.
        Bone Joint J. 2019; 101-B: 667-674