Total Joint Arthroplasty in Patients with Chronic Renal Disease: Is It Worth the Risk?

The Journal of Arthroplasty xxx (2015) xxx–xxx

Contents lists available at ScienceDirect

The Journal of Arthroplasty journal homepage: www.arthroplastyjournal.org

Total Joint Arthroplasty in Patients with Chronic Renal Disease: Is It Worth the Risk? Lucian C. Warth, MD, Andrew J. Pugely, MD, Christopher T. Martin, MD, Yubo Gao, PhD, John J. Callaghan, MD University of Iowa, Iowa City, Iowa

a r t i c l e

i n f o

Article history: Received 19 August 2014 Accepted 2 December 2014 Available online xxxx Keywords: total joint arthroplasty chronic renal disease eGFR (estimated glomerular filtration rate) ACS NSQIP database

a b s t r a c t 26-27% of patients with end stage hip and knee arthritis requiring TJR have chronic renal disease. A multi-center, prospective clinical registry was queried for TJA's from 2006 to 2012, and 74,300 cases were analyzed. Renal impairment was quantified using estimated glomerular filtration rate (eGFR) to stratify each patient by stage of CRD (1–5). There was a significantly greater rate of overall complications in patients with moderate to severe CRD (6.1% vs. 7.6%, P b 0.001). In those with CRD (Stage 3–5), mortality was twice as high (0.26% vs. 0.48%, P b 0.001). Patients with Stage 4 and 5 CRD had a 213% increased risk of any complication (OR 2.13, 95% CI: 1.73–2.62). Surgeons may use these findings to discuss the risk–benefit ratio of elective TJR in patients with CRD. © 2015 Elsevier Inc. All rights reserved.

Maximizing the quality and value of medical and surgical care has become a top priority in the current healthcare arena. Federal legislature such as the Patient Protection and Affordable Care Act has squarely placed the onus for improvement on the physician and the healthcare institution by calling for healthcare performance measures with significant financial implications. The ramifications of such quality measures will continue to mold the economic landscape of healthcare in the United States. With this in mind it becomes increasingly important for orthopedic surgeons to define the metrics on which the public evaluates performance. Total joint arthroplasty (TJA) is among the most common and effective surgical interventions in the Unites States annually [1–3]. Many of these patients suffer from a high comorbidity burden including renal disease. Patients with chronic renal disease (CRD) have significant cardiovascular comorbidities and reduced long-term survival in proportion to the severity of renal disease [4]. Chronic renal disease has been previously associated with high complications after total joint arthroplasty [5–10]. While recognized as a risk factor, the degree to which severity of co-morbid renal dysfunction plays a role in negative perioperative outcomes has not been well quantified. As with any comorbid factor, without a full understanding of how renal disease

One or more of the authors of this paper have disclosed potential or pertinent conflicts of interest, which may include receipt of payment, either direct or indirect, institutional support, or association with an entity in the biomedical field which may be perceived to have potential conflict of interest with this work. For full disclosure statements refer to doi: http://dx.doi.org/10.1016/j.arth.2014.12.037. Disclosure: The American College of Surgeons National Surgical Quality Improvement Program and the hospitals participating in the ACS NSQIP are the sources of the data used herein; they have not verified and are not responsible for the statistical validity of the data analysis or the conclusions derived by the authors. Reprint requests: John J. Callaghan, MD, University of Iowa, Department of Orthopaedics, 200 Hawkins Drive, Iowa City, IA 52242.

severity relates to complications, costs, and outcomes, the adequacy of current national risk-adjustment schemes must be questioned. There are a number of studies which have evaluated the effect of end stage renal disease and need for dialysis on perioperative total joint arthroplasty outcomes [8]. Additionally, in recent years there has been increased attention to chronic renal disease as an independent risk factor associated with need for blood transfusion [11], pulmonary embolism [12], early infection, early revision [7], 30 day readmission [10], 30 day morbidity, and 90 day mortality among others [6]. The importance of adequate kidney function to successful outcomes in arthroplasty has been well delineated. That said, there has been a relative dearth of information evaluating the severity of renal disease as a continuous variable affecting short term outcomes and complications of lower extremity total joint arthroplasty. An accurate gradation of the spectrum of renal disease and how it relates to potential negative outcomes in elective total joint arthroplasty is important to help guide the clinician and most suitably inform and counsel the surgical patient. The American College of Surgeons (ACS) prospectively collects complications and readmissions data from over 480 hospitals throughout the country via the National Surgical Quality Improvement Program (NSQIP) [13]. Thus, the purpose of this study is to quantify the impact of increasing renal impairment on short-term systemic morbidity and mortality following TJA. Methods Data We queried the American College of Surgeons National Surgical Quality Improvement Program (ACS NSQIP) database for all adult patients undergoing elective total knee and hip arthroplasty from

http://dx.doi.org/10.1016/j.arth.2014.12.037 0883-5403/© 2015 Elsevier Inc. All rights reserved.

Please cite this article as: Warth LC, et al, Total Joint Arthroplasty in Patients with Chronic Renal Disease: Is It Worth the Risk?, J Arthroplasty (2015), http://dx.doi.org/10.1016/j.arth.2014.12.037

2

L.C. Warth et al. / The Journal of Arthroplasty xxx (2015) xxx–xxx

2006 to 2012 using Current Procedural Terminology (CPT) codes 27447, 27446, and 27130. The ACS NSQIP is a large, multi-center, prospectively collected clinical registry, which is HIPPA compliant and fully deidentified [13]. This study was deemed exempt by the University of Iowa (UI) Institutional Review Board. No external funding was utilized. Patient characteristics were strictly defined within the NSQIP database. The over 200 reported variables were divided into the categories of demographic characteristics, preoperative comorbidities, preoperative laboratory values, operative variables, and 30-day complications. Notable demographic variables included age, gender, race, and functional status. Comorbidities were defined dichotomously, including body mass index (BMI), smoking history, steroid use, alcohol use, diabetes mellitus, chronic obstructive pulmonary disease (COPD), recent weight loss, coronary artery disease (CAD), renal failure, dialysis use, blood disorders, and recent treatment for cancer, among others. Key laboratory values included hematocrit, BUN, creatinine, albumin and others. Relevant operative variables included operative time, wound class, American Society of Anesthesiologists (ASA) class, intraoperative blood transfusions, and resident presence. NSQIP reports on nearly 30 short-term complications, including infectious, hematologic, cardiac, and neurologic with strict definitions. All patients over the age of 18, undergoing elective primary TKA or THA were included. Our query identified 86,172 cases in total, and 74,300 cases were analyzed after exclusion for incomplete data. Patients without appropriate documentation of pre-operative BUN/creatinine with which to estimate glomerular filtration were excluded from the study. Chronic kidney disease is a progressive loss in renal function over a period of months or years. The estimated glomerular filtration rate (eGFR) was calculated for each patient using the Modified Diet in Renal Disease equation (eGFR = 186.3 × serum creatinine — 1.154 × age — 0.203 × 1.212 [if patient is black] × 0.742 [if female]). This has been accepted as an accurate method in clinical practice to quantify renal function in adults [14]. Only a pre-operative serum creatinine value was used to determine eGFR in our analysis of retrospective NSQIP data. This represents a necessary assumption that eGFR at a single point in time is representative of a chronic condition. Patients were stratified into CKD class based on calculated eGFR values in accordance with National Kidney Foundation clinical guidelines [15]. Stage 1 CKD (normal) — eGFR ≥ 90; Stage 2 CKD (mild) — eGFR 60–89 mL/min/1.73 m2; Stage 3 CKD (moderate) — eGFR (30–59 mL/min/1.73 m 2 ); Stage 4 CKD (severe) — eGFR (15–29 mL/min/1.73 m 2); and Stage 5 CKD or established kidney failure — eGFR b 15 mL/min/1.73 m 2. Chronic kidney disease severity was then subdivided as normal to mild (CKD class 1 or 2), moderate (CKD class 3), and severe (CKD class 4 or 5). Statistical Analysis Propensity score analysis was used to develop matched cohorts of patients based on pre-operative comorbidities and the procedure type performed. The incidence of 30-day morbidity and mortality were then compared between patients with none or mild renal impairment, (Stage 1, 2: eGFR ≥ 60 mL/min/1.73 m 2), against those with moderate or severe disease (Stage 3,4,5: eGFR b 60). Categorical variables are presented as absolute number and percentage frequencies. Continuous data is presented as mean values with standard deviation. Univariate analysis was performed using chi-squared or the Fisher exact test for categorical variables, the t-test with equal variances for normally distributed continuous variables, and the Wilcoxon rank-sum test for non-normally distributed continuous and ordinal variables. Propensity matching was performed using the caliper method, matching case and control patients within 0.2 standard deviations of the propensity score. Results with a P b 0.05 were considered statistically significant. Separately, the morbidity risk associated with eGFR was analyzed as a continuous variable, with appropriate odds ratios calculated.

Results Using the ACS NSQIP database, we analyzed 74,300 patients undergoing TJA. Normal or mild CKD was present in 61,742 (83%), and moderate to severe CKD was present in 12,558 (17%) of patients. Baseline demographic and clinical information of the study population demonstrated notable differences with increasing CKD grade including age, gender, race, BMI, diabetes, smoking, coronary artery disease etc. (Table 1). In those with CRD (Stage 3–5), mortality was twice as high (0.26% vs. 0.48%, P b 0.001). The risk of morbidity increased dramatically with worsening CRD (eGFR: R2 = 0.77). As demonstrated in Fig. 1, the odds ratio for morbidity risk after total joint arthroplasty increased in a logarithmic fashion. Compared with patients without CRD, patients with Stage 4 and 5 CRD had a 213% increased risk of any complication (OR 2.13, 95% CI: 1.73–2.62). In the propensity score matched cohorts (1:1; normal/mild vs. moderate/severe), there was a significantly greater rate of overall complications in patients with moderate to severe renal impairment, as compared to patients with no or mild disease (6.1% vs. 7.6%, P b 0.001) (Table 2). A number of major morbidities were also higher in patients with CRD. For the majority of these outcomes, CRD patients had twice the complication rates compared with those without CRD (Table 2). Cardiovascular complications including cardiac arrest (P = 0.005) and myocardial infarction (P = 0.02) were significantly greater between the propensity matched cohorts. While systemic infectious complications including pneumonia (P = 0.001), UTIs (P b 0.001), sepsis (P = 0.01) and septic shock (P = 0.025) were all significantly greater in the CRD cohort, there were no differences in infectious surgical site complications such as superficial wound infection (P = 0.3), deep wound infection (P = 0.9), organ space infection (P = 0.8), wound dehiscence (P = 0.4), or any wound complication (P = 0.5). Thromboembolic complications including deep venous thrombosis (P = 0.4) and pulmonary embolism (P = 0.2) demonstrated no significant differences between the two groups. Additionally, we were unable to demonstrate any differences in neurologic complication rates including stroke (P = 0.6), coma N 24 hours (P = 0.5), and peripheral nerve injury (P = 0.2) between the two cohorts. Discussion Maximizing quality and value within healthcare has become a national priority. While the negative effect of comorbid disease on health outcomes has been previously explored, the methodology of risk-adjustment algorithms remains relatively underdeveloped. Few studies have evaluated the incremental effect of illness severity on outcomes. In this study we have evaluated the effect of worsening renal disease on short-term systemic outcomes after TJA. According to our findings, complications increase dramatically above a disease severity threshold. These findings have significant implications in the risk-adjustment process and merit further discussion. The existing orthopedic literature includes relatively few articles specifically focused on the role of chronic renal disease in short term perioperative patient outcomes. Several articles have identified impaired renal function as an independent risk factor for early 90-day mortality and 30-day morbidity after total joint arthroplasty [6–8,16,17]. The current literature routinely defines renal disease as a dichotomous variable without correlating the entire spectrum of renal dysfunction to possible comorbid outcomes. Recent studies have shown the prevalence of chronic kidney disease in an elective arthroplasty setting to be between 6% and 27% of patients [5,9,18]. Bozic et al [7,19] have previously shown that renal disease is an independent risk factor for early revision after primary hip and knee arthroplasty, as well as periprosthetic joint infection. In a comparatively small cohort of 526 patients, Ackland et al [5] found that patients with CKD had

Please cite this article as: Warth LC, et al, Total Joint Arthroplasty in Patients with Chronic Renal Disease: Is It Worth the Risk?, J Arthroplasty (2015), http://dx.doi.org/10.1016/j.arth.2014.12.037

L.C. Warth et al. / The Journal of Arthroplasty xxx (2015) xxx–xxx

3

Table 1 Demographics and Clinical Features by Severity of Chronic Kidney Disease.

Demographics Age, mean (SD) Female Gender (%) Race/Ethnicity (%) White Black Other Pre-Operative Health and Comorbidities BMI (kg/m2), mean (SD) Recent Weight Loss (%) Diabetes Mellitus (%) Smoking (%) Alcohol (%) Chronic Obstructive Pulmonary Disease (%) Coronary Artery Disease (MI or CHF) (%) Peripheral Vascular Disease (%) Hx of TIA (%) Dialysis (%) Steroids (%) Bleeding Disorder (%) Pre-Operative Blood Transfusion (%) Open Wound or Wound Infection (%) Radiation therapy (%) Chemotherapy (%) Other Recent Operation (%) Pre-Operative Sepsis (%) Pre-Operative Laboratory Values WBC, mean (SD) Hematocrit, mean (SD) Platelets, mean (SD) Creatinine, mean (SD) Serum albumin, mean (SD) INR, mean (SD)

Mild or Normal CKD (n = 61,742)

Matched Mild or Normal CKD (n = 12,558)

Moderate to Severe CKD (N = 12,558)

Un-adjusted P Value

Propensity Matched P Value**

65.24 (10.76) 58.28

70.67 (10.21) 68.08

72.58 (9.30) 70.62

b0.0001 b0.0001 b0.0001

b0.0001 b0.0001 0.0009

78.93 6.82 14.25

80.27 5.15 14.58

82.11 4.62 13.27

31.63 (7.06) 0.26 13.94 11.04 3.2 3.59 0.26 0.54 2.44 0.01 2.81 2.41 0.1 0.58 0.06 0.2 0.32 0.32

31.71 (7.32) 0.31 22.62 7.66 2.19 5.09 0.51 0.54 3.56 0.07 4.14 3.97 0.16 0.8 0.11 0.27 0.51 0.41

32.02 (7.09) 0.29 24.44 6.48 1.54 5.59 0.73 1.39 4.78 0.92 4.44 4.36 0.28 0.98 0.13 0.34 0.19 0.4

b0.0001 0.5238 b0.0001 b0.0001 b0.0001 b0.0001 b0.0001 b0.0001 b0.0001 b0.0001 b0.0001 b0.0001 b0.0001 b0.0001 0.0857 0.0338 0.0731 0.1322

0.0008 0.8183 0.0007 0.0003 0.0069 0.077 0.0245 b0.0001 0.0006 b0.0001 0.2368 0.1215 0.0429 0.1218 0.8331 0.4716 0.002 0.9244

7.00 (2.12) 40.68 (4.06) 247.4 (68.21) 0.83 (0.17) 4.12 (0.41) 1.03 (0.24)

7.15 (2.41) 39.38 (4.51) 246 (69.27) 0.82 (0.16) 4.04 (0.42) 1.04 (0.22)

7.35 (2.31) 38.75 (4.40) 243.4 (71.41) 1.39 (0.87) 4.03 (0.44) 1.06 (0.35)

b0.0001 b0.0001 b0.0001 b0.0001 b0.0001 b0.0001

b0.0001 b0.0001 0.0043 b0.0001 0.0537 b0.0001

increased pulmonary, infectious, and cardiovascular complications while experiencing an increased length of stay of 4 days. Miric et al [9,18] found an increased risk for 90-day readmission for both THA and TKA patients in the setting of CRD. Pugely et al [10] corroborated these findings, showing that dialysis, high serum BUN, and elevated serum creatinine were predictive risk factors for 30-day readmission. The current study supports an increased risk of cardiac and systemic infectious complications, as well as overall morbidity and mortality, but is unique in that we demonstrate that complications increase dramatically above a disease severity threshold, and provides a more complete understanding of how comorbidity relates to outcomes. To our knowledge, no previous studies have analyzed renal disease as it relates to joint arthroplasty outcomes in a continuous fashion. Renal disease has previously been shown to have a dose–response relationship to be associated with operative mortality and complications associated with increasing severity of CKD in the general surgery literature [20]. Our findings are particularly noteworthy in that we

Fig. 1. The odds ratio for morbidity risk after total joint arthroplasty increased in a logarithmic fashion (R2 = 0.77). The red lines show 95% confidence intervals.

have been able to demonstrate a logarithmic increase in morbidity after TJA with increasing renal dysfunction as defined by eGFR. Additionally we have identified several significant post-operative complications in a large, well validated, and highly generalizable database. Previously identified differences seen after arthroplasty in the patient with CRD including increased length of stay and increased rates of readmission have significant financial and cost-savings implications. Table 2 Post-Operative Outcomes by Severity of Chronic Renal Disease (CRD). Characteristic

Matched Normal or Mild CKD

Moderate to Severe CKD

P Value

Superficial Wound Infection Deep Wound Infection Organ Space Infection Wound Dehiscence Any Wound Complication Pneumonia Unplanned Intubation Deep Venous Thrombosis Pulmonary Embolism Renal Insufficiency Acute Renal Failure Urinary Tract Infection Stroke Coma N 24 hours Peripheral Nerve Injury Cardiac Arrest Myocardial Infarction Bleeding Transfusions Graft/Implant Failure Sepsis Septic Shock Reoperation Mortality Any Complication

0.8 0.22 0.16 0.19 1.27 0.44 0.15 0.8 0.49 0.14 0.04 1.6 0.15 0 0.02 0.1 0.33 16.16 0.03 0.37 0.12 1.46 0.26 6.1

0.7 0.33 0.18 0.24 1.37 0.74 0.53 0.9 0.61 0.34 0.32 2.13 0.18 0.02 0.06 0.25 0.53 19.62 0.01 0.66 0.24 1.75 0.48 7.64

0.343 0.089 0.757 0.414 0.472 0.002 b0.001 0.371 0.202 0.001 b0.001 0.002 0.639 0.500 0.180 0.005 0.021 b0.001 0.375 0.001 0.025 0.063 0.005 b0.001

Please cite this article as: Warth LC, et al, Total Joint Arthroplasty in Patients with Chronic Renal Disease: Is It Worth the Risk?, J Arthroplasty (2015), http://dx.doi.org/10.1016/j.arth.2014.12.037

4

L.C. Warth et al. / The Journal of Arthroplasty xxx (2015) xxx–xxx

Patients with intrinsic kidney disease have significant cardiovascular comorbidities and reduced long term survival in proportion to the severity of renal disease [4]. Attention should be paid by the surgeon and medical staff to minimize further insult in at risk patients. This may best be achieved with effective pre-operative risk stratification [21], optimizing perioperative fluid hydration status, and avoiding potentially nephrotoxic agents such as non-steroidal anti-inflammatories in the perioperative period. With an improved understanding of CRD in arthroplasty patients, it will be important to monitor the effect of multimodal pain control regimens incorporating potentially nephrotoxic pain adjuncts such as ketorolac and celecoxib in at risk individuals. Further research is warranted to better understand the incremental effect of illness severity on outcomes after major elective orthopedic surgery. It will be increasingly important to define these risks so that they can be conveyed to policymakers who define the quality metrics by which our outcomes are being judged. Although the large numbers and rigorous methodology of the NSQIP database are relative strengths, as with any database study, we identify several inherent limitations. Approximately 14% of our cohort was excluded from evaluation for missing data. It was beyond the scope of this paper to address functional outcomes or patient satisfaction scores. Additionally, relevant demographic factors such as insurance and socioeconomic status were not included.

Conclusions In our evaluation of ACS NSQIP data we have defined the incidence of several significant complications after total joint arthroplasty which are significantly associated with increasing renal disease. In using data from a stringently collected and nationally accepted registry with a large cohort, we feel that our findings are highly generalizable. The quality of care which institutions and surgeons provide is being publically scrutinized based on short-term patient outcomes. While Medicare riskadjustment models account for renal disease as a dichotomous variable, they fail to quantify the severity of disease and its effect on outcome. Our data has shown higher complication rates in those with significant renal disease. Surgeons may use these findings to discuss the risk– benefit ratio of operating on patients with significant CRD, particularly in elective cases. Policymakers should use these findings to develop risk-adjustment algorithms which appropriately adjust for the severity of renal disease.

References 1. Learmonth ID, Young C, Rorabeck C. The operation of the century: Total hip replacement. Lancet 2007;370(9597):1508. 2. Kurtz S, Ong K, Lau E, et al. Projections of primary and revision hip and knee arthroplasty in the United States from 2005 to 2030. J Bone Joint Surg Am 2007; 89(4):780. 3. Fehring TK, Odum SM, Troyer JL, et al. Joint replacement access in 2016: A supply side crisis. J Arthroplasty 2010;25(8):1175. 4. Go AS, Chertow GM, Fan D, et al. Chronic kidney disease and the risks of death, cardiovascular events, and hospitalization. N Engl J Med 2004;351(13):1296. 5. Ackland GL, Moran N, Cone S, et al. Chronic kidney disease and postoperative morbidity after elective orthopedic surgery. Anesth Analg 2011;112(6):1375. 6. Belmont PJ, Goodman GP, Hamilton W, et al. Morbidity and mortality in the thirtyday period following total hip arthroplasty: Risk factors and incidence. J Arthroplasty 2014;29(10):2025. 7. Bozic KJ, Lau E, Kurtz S, et al. Patient-related risk factors for periprosthetic joint infection and postoperative mortality following total hip arthroplasty in medicare patients. J Bone Joint Surg Am 2012;94(9):794. 8. Deegan BF, Richard RD, Bowen TR, et al. Impact of chronic kidney disease stage on lower-extremity arthroplasty. Orthopedics 2014;37(7):e613. 9. Miric A, Inacio MC, Namba RS. The effect of chronic kidney disease on total hip arthroplasty. J Arthroplasty 2014;29(6):1225. 10. Pugely AJ, Callaghan JJ, Martin CT, et al. Incidence of and risk factors for 30-day readmission following elective primary total joint arthroplasty: Analysis from the ACSNSQIP. J Arthroplasty 2013;28(9):1499. 11. Augustin ID, Yeoh TY, Sprung J, et al. Association between chronic kidney disease and blood transfusions for knee and hip arthroplasty surgery. J Arthroplasty 2013;28(6): 928. 12. Memtsoudis SG, Besculides MC, Gaber L, et al. Risk factors for pulmonary embolism after hip and knee arthroplasty: A population-based study. Int Orthop 2009;33(6):1739. 13. Ingraham AM, Richards KE, Hall BL, et al. Quality improvement in surgery: The American College of Surgeons national surgical quality improvement program approach. Adv Surg 2010;44:251. 14. Levey AS, Bosch JP, Lewis JB, et al. A more accurate method to estimate glomerular filtration rate from serum creatinine: A new prediction equation. modification of diet in renal disease study group. Ann Intern Med 1999;130(6):461. 15. National Kidney Foundation. K/DOQI clinical practice guidelines for chronic kidney disease: Evaluation, classification, and stratification. Am J Kidney Dis 2002;39(2 Suppl. 1):S1. 16. Jamsen E, Puolakka T, Eskelinen A, et al. Predictors of mortality following primary hip and knee replacement in the aged. A single-center analysis of 1,998 primary hip and knee replacements for primary osteoarthritis. Acta Orthop 2013;84(1):44. 17. Memtsoudis SG, Della Valle AG, Besculides MC, et al. Risk factors for perioperative mortality after lower extremity arthroplasty: A population-based study of 6,901,324 patient discharges. J Arthroplasty 2010;25(1):19. 18. Miric A, Inacio MC, Namba RS. Can total knee arthroplasty be safely performed in patients with chronic renal disease? Acta Orthop 2014;85(1):71. 19. Bozic KJ, Lau E, Ong K, et al. Risk factors for early revision after primary TKA in medicare patients. Clin Orthop Relat Res 2014;472(1):232. 20. Patel VI, Lancaster RT, Mukhopadhyay S, et al. Impact of chronic kidney disease on outcomes after abdominal aortic aneurysm repair. J Vasc Surg 2012;56(5):1206. 21. Radcliff KE, Orozco FR, Quinones D, et al. Preoperative risk stratification reduces the incidence of perioperative complications after total knee arthroplasty. J Arthroplasty 2012;27(8 Suppl.):77.

Please cite this article as: Warth LC, et al, Total Joint Arthroplasty in Patients with Chronic Renal Disease: Is It Worth the Risk?, J Arthroplasty (2015), http://dx.doi.org/10.1016/j.arth.2014.12.037