Secondary Logo

Journal Logo

Clinical Research

Higher Charlson Comorbidity Index Scores are Associated With Readmission After Orthopaedic Surgery

Voskuijl, Timothy MD1; Hageman, Michiel MD1; Ring, David MD, PhD1,a

Author Information
Clinical Orthopaedics and Related Research: May 2014 - Volume 472 - Issue 5 - p 1638-1644
doi: 10.1007/s11999-013-3394-8
  • Free



Hospital readmissions are used as an indicator of quality of care [3, 5, 24]. Readmissions are costly [13], and recent legislation introduced financial penalties for high risk-adjusted 30-day readmission for certain diagnoses by establishing the Hospital Readmissions Reduction Program [6, 16, 19]. One center developed an intervention program to reduce readmissions after orthopaedic surgery [14].

There is a relationship between comorbidities and readmissions after orthopaedic trauma surgery [23, 35]. In orthopaedic surgery, comorbidities have been identified as predictors of mortality after arthroplasty, trauma surgery, and oncologic surgery [4, 9, 18, 20, 21, 25, 27, 36]. Complications after ankle fractures, spinal metastasis surgery, and knee or hip arthroplasty also are associated with comorbidities [2, 12, 31-34] as are infections after orthopaedic trauma, spine surgery, knee or hip arthroplasty [8, 17, 22, 36], and transfusion after knee or shoulder arthroplasty [12, 29].

The Charlson Comorbidity Index (CCI) [7] and its recent modification [26] are comorbidity-based measures developed to predict mortality within 1 year of hospital admission in patients without trauma. The updated CCI scores 12 comorbidities with various weightings giving a maximum score of 24 [26]. A higher CCI indicates a higher risk of death within 1 year. The CCI also has been evaluated as a predictor for outcomes after orthopaedic trauma surgery, postoperative adverse events after general surgery, and readmission for geriatric and general medicine patients [2, 9, 17, 30, 32-34, 37].

We sought to answer the following: Is an increased score on the CCI associated with (1) readmission, (2) surgical site infection or other adverse events, (3) transfusion risk, or (4) mortality after orthopaedic surgery?

Patients and Methods

The Research Patient Data Registry is a centralized clinical data registry covering a patient base of greater than 4.0 million insured and uninsured patients from Massachusetts General Hospital and Brigham Women's Hospital, Boston. It comprises diagnoses, medications, laboratory scores, procedures codes, demographic information (sex, age, race, marital status) and visit entries. We used the Research Patient Data Registry to identify patients having orthopaedic surgery. A total of 30,126 orthopaedic surgeries met the following inclusion criteria: (1) surgery performed between 2008 and 2011, and (2) no prior orthopaedic surgery within 30 days of the operation. The included surgeries were performed with an equal distribution of males and females, with the majority of the patients being between 40 and 59 years old (Table 1).

Table 1:

The International Classification of Diseases, 9th Revision, Clinical Modification (ICD-9-CM) was used to code diagnoses and procedures. Surgeries were excluded if data were missing for diagnosis codes (194 cases) or procedure codes (1270 cases). Surgery-related adverse events and infections (Appendix 1) and comorbidities as listed in the updated and reweighted CCI [26] (Appendix 2) also were classified using ICD-9 codes.

Statistical Analysis

Baseline characteristics of study patients were summarized with frequencies and percentages for categorical variables and as mean ± SD for continuous variables. Factors significantly associated with readmission, adverse event, surgical site infection, transfusion, and mortality were assessed in bivariate analysis using Student's t-tests and chi-square tests (Table 2). Variables entered in these analyses were indicators for the ICD-9 procedure group, the surgeon, the patient's age group, sex, race and marital status, the timing of the operation, duration of operation, length of hospital stay, and orthopaedic subspecialty performing surgery (arthroplasty, foot and ankle, hand, shoulder, spine, sports, pediatrics, trauma, and oncology). For analysis of postoperative surgery-related adverse events and infections within 30 days of surgery, we excluded operations that were performed for surgery-related adverse events or infection within 30 days of another operation. Ordinary least squares regression analyses then were used to determine factors associated with readmission, adverse event, surgical site infection, transfusion, and mortality within 30 days of discharge, from the CCI and the other explanatory variables, which were significantly related to our dependent variables in the bivariate analyses. We decided against using logistic regressions for computational reasons. The coefficient of the ordinary least squares regression indicates the direction of one unit increase of the response variable in relation to explanatory variable. The R-square means the percentage of the variance of the explanatory variable is accounted by the response variables in the model.

Table 2:
Example of bivariable analysis*



A total of 913 of 30,129 patients (3.0%) were readmitted within 30 days after discharge.

The CCI accounted for 10% of the variation in readmissions. Every point increase in CCI score adds an additional 0.45% risk in readmission for patients undergoing arthroplasty, 0.63% for patients undergoing trauma surgery, 0.90% risk for patients undergoing spine surgery (all p < 0.001), and −0.27% for patients undergoing hand and upper extremity surgery (p = 0.092) (Table 3). As a reference every year increase in age adds an additional 0.03% risk in readmission for all patients.

Table 3:
Additional risk on readmission with every increasing CCI score (%) per orthopaedic subspecialty*

Surgical Site Infections and Other Surgery-related Adverse Events

Surgical site infections and other surgery-related adverse events within 30 days of surgery occurred in 393 of 28,604 (1.4%) (Table 4) and 417 of 29,678 (1.4%) (Table 5) patients, respectively. An increase in the CCI was not associated with changes in the likelihood of these events for all surgery types.

Table 4:
Additional risk on surgery-related adverse events with every increasing CCI score (%) per orthopaedic subspecialty
Table 5:
Additional risk on surgical site infections with every increasing CCI score (%) per orthopaedic subspecialty

Blood Transfusion

Blood transfusion was prescribed after 211 of 30,129 (0.7%) orthopaedic surgeries. With every point increase in CCI score, the risk of transfusion was increased by 0.11% after arthroplasty (p = 0.044), 0.69% after spine surgery, 0.45% after trauma surgery, and 0.34% after oncologic surgery (all p < 0.001). The CCI explained 7.5% of the variation in blood transfusion after orthopaedic surgery (Table 6).

Table 6:
Additional risk on transfusion with every increasing CCI score (%) per orthopaedic subspecialty


Fifty-six of 30,129 patients (0.19%) died within 30 days after discharge. For every point increase in CCI score, mortality increased 0.25% after shoulder surgery, 0.24% after trauma surgery, and 0.33% after oncologic surgery. The CCI explained 9.9% of the variation in mortality after orthopaedic surgery (Table 7).

Table 7:
Additional risk on mortality with every increasing CCI score (%) per orthopaedic subspecialty


The ability to anticipate complications of treatment is essential to the effective practice of surgery. With the economic penalties associated with readmission after surgery with the Hospital Readmissions Reduction Program [6, 16, 19], there now are financial and professional incentives to improve our ability to anticipate complications of treatment. The CCI has been used as a predictor for adverse events [2, 9, 17, 30, 32-34, 37]. Our study addressed the association of comorbidities as measured by the CCI on readmission and adverse outcomes after orthopaedic surgeries. We evaluated the correlation between CCI and risk of readmission, surgical site infection, other adverse events, risk of transfusion, and mortality after orthopaedic surgery.

This study should be interpreted in light of several limitations. The data are from one registry representing a few years at two major teaching hospitals in the same area and therefore may not be generalizable. Our analyses on readmission did account for length of stay, although the relation of readmission and hospital stay has been questioned [11, 28]. We used ICD-9 codes to calculate the CCI (Appendix 2) and detect adverse events and surgical site infections (Appendix 1). ICD-9 codes may be inaccurate or incomplete in a large database, but the rate of errors seems acceptably low. The main drawback of using the CCI as a predictor of readmission or the other outcomes analyzed is that different conditions or combinations of conditions may lead to an equal index score. This blunt scoring system hinders the development of risk management protocols which should focus on specific conditions or interactions. In addition, we could not account for the severity of injury or complexity of surgery, both of which may influence surgery outcomes.

Readmission after surgery was associated with CCI for spine and trauma surgeries and arthroplasties but explained only a small amount of the variation in readmission. This is consistent with studies of medical patients [30, 37]. A retrospective study of readmission of general medicine patients admitted from the emergency department revealed a hazard ratio of 1.42 (95% CI, 1.07-1.89) for patients with a Charlson score of 2 to 4 and a hazard ratio of 1.93 (95% CI, 1.37-2.73) for patients with a Charlson score higher than 4 [30]. A prospective study on geriatric patients revealed that 3.1% of the variation in readmission in their study-population was explained by the CCI [37].

In our study, the CCI was not related to repeat surgery for adverse events or surgical site infection. In other words, adverse events judged to be related to the surgery based on expert review were not related to the CCI. Prior studies that included medical adverse events found associations with the CCI. One 10-year study of patients undergoing primary THA revealed increased CCI score as a predictor of infection, dislocation, revision, perioperative fracture, neurologic injury, and thromboembolic events [33]. In another study, the risk of deep venous thrombosis or pulmonary embolism (DVT/PE) after surgery for ankle fracture was increased in patients with higher CCI scores [32]. An analysis of thromboembolic adverse events after THA and TKA exposed higher CCI score as a risk factor (odds ratio, 1.2; 95% CI, 1.1-1.4) for DVT/PE in patients undergoing TKAs but not for patients undergoing THAs [31, 34]. One retrospective study analyzing patients after oncologic orthopaedic surgery found the CCI was a predictor of medical complications (eg, pneumonia, respiratory failure) [2]. Higher CCI scores were found to significantly increase the risk of infection at 10 years after spine surgery (hazard ratio, 2.48; 95% CI, 1.93-3.19) [17]. Some comorbidities, in particular diabetes, have an important influence on infections after orthopaedic trauma surgery [8, 22].

CCI was associated with increased use of transfusion in patients undergoing arthroplasty, spine, trauma, or oncologic orthopaedic treatment in our study. Some studies of transfusion after orthopaedic surgery do not involve comorbidities in their risk analyses [1, 15], despite that some comorbidities (especially cardiovascular) commonly lead to increased need for transfusion. A prospective study of geriatric patients with primary hip or knee arthroplasty revealed coronary artery disease as a predictor of transfusion. The association between transfusion after shoulder arthroplasty and the number of comorbidities has been inconsistent [10, 12, 29].

In our study, the CCI was associated with a higher risk of death within 30 days of discharge for patients with trauma, undergoing shoulder surgery, or receiving oncologic treatment. The association of the CCI with mortality after trauma is consistent with prior studies in which higher in-hospital mortality was found in patients with hip fractures with higher CCI scores [9, 25].

This data set indicates that comorbidities as measured by the CCI have a significant influence on readmission, transfusion, and mortality after orthopaedic surgery, more so for certain subspecialties, but are not useful as a predictor of surgical site infection or surgery-related adverse events. To determine if these findings are generalizable, multicenter studies could be useful, although our conclusions are drawn from a robust data registry providing considerable power. However, since multiple comorbidities or combinations have the same weight, the CCI may not be helpful in developing risk management protocols focused on specific conditions or interactions, for which a more precise tool is needed. Using Current Procedural Terminology codes and other methods, more detailed analyses for common procedures in each orthopaedic subspecialty could be made, for example, by addressing different kinds of arthroplasty or severity of injury.


1. Ahmed I, Chan JK, Jenkins P, Brenkel I, Walmsley P. Estimating the transfusion risk following total knee arthroplasty. Orthopedics. 2012;35:e1465-e1471 10.3928/01477447-20120919-13.
2. Arrigo RT, Kalanithi P, Cheng I, Alamin T, Carragee EJ, Mindea SA, Boakye M, Park J. Charlson score is a robust predictor of 30-day complications following spinal metastasis surgery. Spine (Phila Pa 1976). 2011;36:E1274-E1280 10.1097/BRS.0b013e318206cda3.
3. Ashton CM, Kuykendall DH, Johnson ML, Wray NP, Wu L. The association between the quality of inpatient care and early readmission. Ann Intern Med. 1995;122:415-421 10.7326/0003-4819-122-6-199503150-00003.
4. Aynardi M, Jacovides CL, Huang R, Mortazavi SM, Parvizi J. Risk factors for early mortality following modern total hip arthroplasty. J Arthroplasty. 2013;28:517-520 10.1016/j.arth.2012.06.040.
5. Benbassat J, Taragin M. Hospital readmissions as a measure of quality of health care: advantages and limitations. Arch Intern Med. 2000;160:1074-1081 10.1001/archinte.160.8.1074.
6. Burke RE, Coleman EA. Interventions to decrease hospital readmissions: keys for cost-effectiveness. JAMA Intern Med. 2013;173:695-698 10.1001/jamainternmed.2013.171.
7. Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis. 1987;40:373-383 10.1016/0021-9681(87)90171-8.
8. Folk JW, Starr AJ, Early JS. Early wound complications of operative treatment of calcaneus fractures: analysis of 190 fractures. J Orthop Trauma. 1999;13:369-372 10.1097/00005131-199906000-00008.
9. Gabbe BJ, Magtengaard K, Hannaford AP, Cameron PA. Is the Charlson Comorbidity Index useful for predicting trauma outcomes? Acad Emerg Med. 2005;12:318-321 10.1111/j.1553-2712.2005.tb01950.x.
10. Hardy JC, Hung M, Snow BJ, Martin CL, Tashjian RZ, Burks RT, Greis PE. Blood transfusion associated with shoulder arthroplasty. J Shoulder Elbow Surg. 2013;22:233-239 10.1016/j.jse.2012.04.013.
11. Hendren S, Morris AM, Zhang W, Dimick J. Early discharge and hospital readmission after colectomy for cancer. Dis Colon Rectum. 2011;54:1362-1367 10.1097/DCR.0b013e31822b72d3.
12. Higuera CA, Elsharkawy K, Klika AK, Brocone M, Barsoum WK. 2010 Mid-America Orthopaedic Association Physician in Training Award: predictors of early adverse outcomes after knee and hip arthroplasty in geriatric patients. Clin Orthop Relat Res. 2011;469:1391-14003069283 10.1007/s11999-011-1804-3.
13. Jencks SF, Williams MV, Coleman EA. Rehospitalizations among patients in the Medicare fee-for-service program. N Engl J Med. 2009;360:1418-2148 10.1056/NEJMsa0803563.
14. Jordan CJ, Goldstein RY, Michels RF, Hutzler L, Slover JD, Bosco JA 3rd, Comprehensive program reduces hospital readmission rates after total joint arthroplasty. Am J Orthop (Belle Mead NJ). 2012;41:E147-E151.
15. Kadar A, Chechik O, Steinberg E, Reider E, Sternheim A. Predicting the need for blood transfusion in patients with hip fractures. Int Orthop. 2013;37:693-700 10.1007/s00264-013-1795-73609993.
16. Kocher RP, Adashi EY. Hospital readmissions and the Affordable Care Act: paying for coordinated quality care. JAMA. 2011;306:1794-1795 10.1001/jama.2011.1561.
17. Kurtz SM, Lau E, Ong KL, Carreon L, Watson H, Albert T, Glassman S. Infection risk for primary and revision instrumented lumbar spine fusion in the Medicare population. J Neurosurg Spine. 2012;17:342-347 10.3171/2012.7.SPINE12203.
18. Mantilla CB, Wass CT, Goodrich KA, Johanns CJ, Kool ML, Zhu X, Corredor JA, Warner DO, Joyner MJ, Berry DJ, Schroeder DR, Sprung J. Risk for perioperative myocardial infarction and mortality in patients undergoing hip or knee arthroplasty: the role of anemia. Transfusion. 2011;51:82-91 10.1111/j.1537-2995.2010.02797.x.
19. McCarthy D, Johnson MB, Audet AM. Recasting readmissions by placing the hospital role in community context. JAMA. 2013;309:351-352 10.1001/jama.2013.1.
20. Memtsoudis SG, Ma Y, Chiu YL, Walz JM, Voswinckel R, Mazumdar M. Perioperative mortality in patients with pulmonary hypertension undergoing major joint replacement. Anesth Analg. 2010;111:1110-1116 10.1213/ANE.0b013e3181f43149.
21. Memtsoudis SG, Pumberger M, Ma Y, Chiu YL, Fritsch G, Gerner P, Poultsides L, Valle AG. Epidemiology and risk factors for perioperative mortality after total hip and knee arthroplasty. J Orthop Res. 2012;30:1811-18213407319 10.1002/jor.22139.
22. Miller AG, Margules A, Raikin SM. Risk factors for wound complications after ankle fracture surgery. J. Bone Joint Surg Am. 2012;94:2047-2052 10.2106/JBJS.K.01088.
23. Moore L, Thomas Stelfox H, Turgeon AF, Nathens AB, Sage NL, Emond M, Bourgeois G, Lapointe J, Gagné M. Rates, patterns, and determinants of unplanned readmission after traumatic injury: a multicentre cohort study. Ann Surg. 2013 Mar 8. [Epub ahead of print].
24. Navaneethan U, Gutierrez NG, Jegadeesan R, Venkatesh PG, Butt M, Sanaka MR, Vargo JJ, Parsi MA. Delay in performing ERCP and adverse events increase the 30-day readmission risk in patients with acute cholangitis. Gastrointest Endosc. 2013;78:81-90 10.1016/j.gie.2013.02.003.
25. Neuhaus V, King J, Hageman MG, Ring DC. Charlson comorbidity indices and in-hospital deaths in patients with hip fractures. Clin Orthop Relat Res. 2013;471:1712-1719 10.1007/s11999-012-2705-9.
26. Quan H, Li B, Couris CM, Fushimi K, Graham P, Hider P, Januel JM, Sundararajan V. Updating and validating the Charlson comorbidity index and score for risk adjustment in hospital discharge abstracts using data from 6 countries. Am J Epidemiol. 2011;173:676-682 10.1093/aje/kwq433.
27. Sanders RD, Bottle A, Jameson SS, Mozid A, Aylin P, Edger L, Ma D, Reed MR, Walters M, Lees KR, Maze M. Independent preoperative predictors of outcomes in orthopedic and vascular surgery: the influence of time interval between an acute coronary syndrome or stroke and the operation. Ann Surg. 2012;255:901-907 10.1097/SLA.0b013e31824c438d.
28. Schneider EB, Hyder O, Brooke BS, Efron J, Cameron JL, Edil BH, Schulick RD, Choti MA, Wolfgang CL, Pawlik TM. Patient readmission and mortality after colorectal surgery for colon cancer: impact of length of stay relative to other clinical factors. J Am Coll Surg. 2012;214:390-398 10.1016/j.jamcollsurg.2011.12.025.
29. Schumer RA, Chae JS, Markert RJ, Sprott D, Crosby LA. Predicting transfusion in shoulder arthroplasty. J Shoulder Elbow Surg. 2010;19:91-96 10.1016/j.jse.2009.05.001.
30. Shu CC, Lin YF, Hsu NC, Ko WJ. Risk factors for 30-day readmission in general medical patients admitted from the emergency department: a single centre study. Intern Med J. 2012;42:677-682 10.1111/j.1445-5994.2011.02562.x.
31. Singh JA, Jensen MR, Harmsen WS, Gabriel SE, Lewallen DG. Cardiac and thromboembolic complications and mortality in patients undergoing total hip and total knee arthroplasty. Ann Rheum Dis. 2011;70:2082-20883315837 10.1136/ard.2010.148726.
32. SooHoo NF, Eagan M, Krenek L, Zingmond DS. Incidence and factors predicting pulmonary embolism and deep venous thrombosis following surgical treatment of ankle fractures. Foot Ankle Surg. 2011;17:259-262 10.1016/j.fas.2010.08.009.
33. Soohoo NF, Farng E, Lieberman JR, Chambers L, Zingmond DS. Factors that predict short-term complication rates after total hip arthroplasty. Clin Orthop Relat Res. 2010;468:2363-23712914297 10.1007/s11999-010-1354-0.
34. SooHoo NF, Lieberman JR, Ko CY, Zingmond DS. Factors predicting complication rates following total knee replacement. J Bone Joint Surg Am. 2006;88:480-485 10.2106/JBJS.E.00629.
35. Teixeira A, Trinquart L, Raphael M, Bastianic T, Chatellier G, Holstein J. Outcomes in older patients after surgical treatment for hip fracture: a new approach to characterise the link between readmissions and the surgical stay. Age Ageing. 2009;38:584-589 10.1093/ageing/afp124.
36. Viens NA, Hug KT, Marchant MH, Cook C, Vail TP, Bolognesi MP. Role of diabetes type in perioperative outcomes after hip and knee arthroplasty in the United States. J Surg Orthop Adv. 2012;21:253-260 10.3113/JSOA.2012.0253.
37. Zekry D, Loures Valle BH, Graf C, Michel JP, Gold G, Krause KH, Herrmann FR. Prospective comparison of 6 comorbidity indices as predictors of 1-year post-hospital discharge institutionalization, readmission, and mortality in elderly individuals. J Am Med Dir Assoc. 2012;13:272-278 10.1016/j.jamda.2010.11.011.

Appendix 1. ICD-9 codes for variables used in this study*


Appendix 2. Comorbidities as listed in the updated CCI with corresponding ICD 9 codes as used in this study*

© 2014 Lippincott Williams & Wilkins, Inc.