- Email: [email protected]
Predictive Value of the Charlson Comorbidity Index in Kidney Transplantation
Predictive Value of the Charlson Comorbidity Index in Kidney Transplantation G. Grosso, D. Corona, A. Mistretta, D. Zerbo, N. Sinagra, A. Giaquinta, T. Tallarita, B. Ekser, A. Leonardi, R. Gula, P. Veroux, and M. Veroux ABSTRACT Background. Nonimmunologic factors have been recently implicated in worse outcomes after kidney transplantation, producing a need to predict the operative risk among kidney recipients. We assessed the predictive value of the Charlson comorbidity index (CCI) among kidney transplant recipients. Methods. A retrospective study of 223 first deceased-donor kidney transplantations performed from 2000 to 2007 evaluated the role of comorbidities. Results. About 50% of recipients displayed ⬎1 comorbid condition before transplantation; the most frequently reported was diabetes mellitus. Increasing CCI scores significantly affected graft and patient survivals. Crude analysis showed a significant association between CCI ⬎1 and risk of death (hazard ratio [HR], 3.87; 95% confidence interval [CI], 1.06 –14.06; P ⫽ .04). After adjustment for several covariates, high CCI values remained significantly predictive of posttransplantation outcomes with a HR for death of (12.53; 95% CI, 1.9 – 82.68; P ⫽ .009). Conclusions. Our predictive model showed a strong association of CCI and patient survival even after adjustment for several clinical covariates. CCI may be used to evaluate patients referred for kidney transplantation who display a significant burden of comorbid conditions that increase the risk of premature death or graft loss. he effects of population aging have led to an increased rate of end-stage renal disease (ESRD).1 Many studies have demonstrated an increased number of comorbid conditions among the elderly population with ESRD, assessing the roles of individual conditions such as diabetes,2 cardiovascular disease,3,4 and chronic obstructive pulmonary disease (COPD).5 Multiple comorbidity indices have been developed; most of them—the Index of Coexistent Disease (ICED) and the Khan, Davies, and Charlson indexes— have been applied to the ESRD population.6 –9 Given the heterogeneity of the clinical histories among transplant candidates, morbidity assessment should be considered to be an important aid to stratify patients according to a summary indicator of health status, offering a potential global advantage rather than considering the diseases individually. Moreover, nonimmunologic factors have been shown to correlate with posttransplantation outcomes more than immunologic and transplant-related factors.10 –12 Therefore, a comorbidity index may be used to predict posttransplantation outcomes. Finally, use of a common
T
instrument could be useful to better compare results across institutions. The Charlson comorbidity index (CCI)13 uses a simple weighted scoring system based on the presence or absence of comorbid conditions.14 The CCI has been validated in ESRD patients,8,9,15 seeking to predict the outcomes of patients on either peritoneal16 –18 or hemodialysis.19,20 Moreover, CCI has been used to assess operative risk From the Department “G.F. Ingrassia” (G.G., A.M.), Section of Hygiene and Public Health, and Vascular Surgery and Organ Transplant Unit, Department of Surgery, Transplantation, and Advanced Technologies (D.C., D.Z., N.S., A.G., T.T., B.E., A.L., R.G., P.V., M.V.), University Hospital of Catania, Catania, Italy; and Thomas E. Starzl Transplantation Institute (B.E.), University of Pittsburgh Medical Center, Pittsburgh, Pennsylvania. Address reprint requests to Massimiliano Veroux, MD, PhD, Vascular Surgery and Organ Transplant Unit, Department of Surgery, Transplantation, and Advanced Technologies, University Hospital of Catania, Via Santa Sofia, 83 95128 Catania, Italy. E-mail: [email protected]
© 2012 by Elsevier Inc. All rights reserved. 360 Park Avenue South, New York, NY 10010-1710
0041-1345/–see front matter http://dx.doi.org/10.1016/j.transproceed.2012.06.042
Transplantation Proceedings, 44, 1859 –1863 (2012)
1859
1860
among kidney transplant recipients,21,22 but only a few studies have assessed its role to predict posttransplantation outcomes.23–25 Therefore, the aim of the present study was to define the predictive value of CCI for graft loss or mortality among a population of kidney transplant recipients. METHODS We retrospectively reviewed the data from 223 recipients of first deceased-donor kidney transplantations performed from January 2002 to December 2007. Kidney recipients ⬍18 years old, donors ⬍18 years old, living-donor cases, and simultaneous kidneypancreas transplantations were excluded from the study. Demographic data included age and sex. Clinical characteristics included primary cause of ESRD and baseline patient comorbidities; data on post-transplantation outcomes, graft failures and patient deaths. Delayed graft function (DGF) was defined as the need for ⱖ1 dialysis session within 1 week after transplantation; primary kidney nonfunction (PNF) was defined as the complete lack of functionality such that the recipient never discontinued dialysis sessions after transplantation. We collected a total of 9 comorbidities comprising the CCI according to Jassal et al.23 The CCI was calculated by assigning a weight of 2 points to diabetes mellitus (DM), cerebrovascular accident, and solid tumor or leukemia, and a weight of 1 to previous myocardial infarction, congestive heart failure, peripheral vascular disease, COPD, connective tissue disease, and mild liver disease. Given the nature of transplant recipients, we did not consider renal insufficiency to be a comorbid condition, because it was present in every patient. Baseline characteristics were described as the frequency of occurrence of categoric data and the mean ⫾ SD for continuous variables. Patients were divided into 2 groups according to CCI values with a cutoff point of CCI ⱕ1 versus CCI ⬎1 on the basis of the results of Jassal et al,23 who observed a substantial change in the hazard ratio (HR) for death with increasing CCI scores mostly between the lower rates of the score. Differences between the groups according to CCI value were tested using the chi-square test for categoric variables and 2-sample Student t test or analysis of variance for continuous variables. Patient and graft survivals were estimated by the Kaplan-Meier method. Unadjusted and multivariate analyse were performed using Cox proportional hazards modeling with backward stepwise selection of covariates to determine independent comorbid factors that predict patient and graft survival. All covariates with P values ⬍.1 in univariate analysis were entered in the multivariate model. To identify the role of CCI to predict posttransplantation outcomes, we performed another Cox proportional hazards analysis including CCI (unadjusted) and patient factors (age, sex, and body mass index), other pretransplantation scores (clinical status, as defined above), other comorbidity conditions not included in the previous analysis (hypertension and DGF), and donor factors (age and sex) in adjusted models. All statistical tests were 2 tailed. Data were entered into Microsoft Excel for Windows (Microsoft Corp, Redmond, Washington). Statistical analysis was performed using SPSS for Windows release 17.0 (SPSS, Chicago, Illinois).
RESULTS
Patient baseline characteristics were stratified by CCI as listed in Table 1. No differences were observed between patients with CCI ⱕ1 compared with CCI ⬎1, with the
GROSSO, CORONA, MISTRETTA ET AL Table 1. Baseline Demographics of Study Population of Kidney Transplant Recipients (n ⴝ 223)
Recipients Age (y), mean Male (%) BMI, kg/m2 (%) ⬎28 ⬎32 Primary cause of ESRD (%) Glomerulonephritis Tubulopathy Polycystic kidney Vasculopathy Others Duration of dialysis, y (%) ⬍3 3–5 ⬎5 Cold ischemia time, h (%) 0–24 25–36 ⬎36 Serum creatinine (mg/dL): At discharge At 6 mo At 1 y Donors Age (y), mean ⫾ SD Male (%)
CCI ⱕ1 (n ⫽ 113)
CCI ⬎1 (n ⫽ 120)
P Value
49.2 77
48.2 54.5
.485 .001 .661
19.2 10.6
24.5 10.6 .933
56.6 4.4 2.7 14.2 22.1
60 5.5 1.8 14.5 18.2
52 15 33
52 19.6 28.4
75.2 23.9 0.9
84.5 14.5 0.9
1.37 ⫾ 0.82 1.38 ⫾ 0.83 1.35 ⫾ 0.68
1.49 ⫾ 1.25 1.42 ⫾ 1.06 1.34 ⫾ 0.7
48.8 ⫾ 18.9 61.9
53.3 ⫾ 17.9 60
.618
.209
.426 .794 .949 .778 .069 .766
BMI, body mass index; CCI, Charlson comorbidity index; ESRD, end-stage renal disease.
exception of recipient sex showing a lower percentage of men among patients with CCI ⬎1 (77% vs 54.5%; P ⬍ .001). About 50% of recipients displayed ⬎1 comorbid condition before transplantation, with diabetes mellitus the most frequently reported one (Table 2). Survival analysis revealed a significant association between CCI and decreased posttransplantation patient survival (P ⫽ .003; Fig 1). Furthermore, Kaplan-Meier plots and log rank tests showed a significant association between graft survival and CCI score (P ⫽ .039; Fig 1). The independent contribution of each individual component of the CCI to predict patient death was explored by multivariate analysis. There was no correlation between any single comorbid condition and posttransplantation patient status among recipients. Adjusted analysis of graft loss risk showed a significant correlation with COPD (HR, 4.71; 95% CI, 1.07–20.83; P ⫽ .041). Unadjusted and adjusted HR for patient death according to CCI are listed in Table 3. Crude analysis showed a significant association between CCI ⬎1 and risk of death (HR, 3.87; 95% CI, 1.06 –14.06; P ⫽ .04). After adjusting for several covariates, high CCI values remained significantly predictive of posttransplantation outcome with an HR for death of 12.53 (95% CI, 1.9 – 82.68; P ⫽ .009). Graft
CHARLSON COMORBIDITY INDEX IN KIDNEY TRANSPLANTATION Table 2. Charlson Comorbidity Index (CCI) and Comorbidities Distribution Among Kidney Transplant Recipients (n ⴝ 223) n (%)
CCI distribution 0 1 2 3 4 5 6 Comorbidities distribution Congestive heart failure Coronary disease Diabetes mellitus Stroke Peripheral artery disease COPD Connective tissue disease Liver disease Malignancy
93 (41.7) 20 (9) 81 (36.3) 18 (8.1) 9 (4) 0 2 (0.9) 2 (0.9) 22 (9.9) 94 (42.2) 2 (0.9) 19 (8.5) 16 (7.2) 13 (5.8) 2 (0.9) 6 (2.7)
COPD, chronic pulmonary disease.
survival analysis revealed a similar crude risk of graft loss (HR, 2.44; 95% CI, 1.19 –5.02; P ⫽ .015) but no association was found after adjusting for covariates (data not shown). DISCUSSION
We have presented the comorbidity conditions of 223 kidney transplant recipients. Not surprisingly, we observed that 50% of patients had ⱖ1 associated disease; the number of patients with multiple baseline comorbidities who received a transplant at our institution has increased over the past several years. Given the controversies concerning the role of a single comorbidity to predict renal transplantation outcomes, we adopted a modified version of the CCI, a
1861
simple comorbidity score that has been validated among patients with ESRD15 to assess the effects of pretransplantation comorbid conditions on posttransplantation patient and graft survivals. Patients with CCI ⬎1 showed a 12.53fold increased risk of death after kidney transplantation, but CCI was not significantly associated with graft survival. Our study demonstrated no significant association between a single point in the score with the risk of patient death; however, a cutoff point significantly defined a higher risk of patient death. The nonlinear correlation between CCI and outcomes may be explained by several comorbid conditions that are often related to each other, namely, cluster disease involving the cardiovascular, respiratory, connective and autoimmune systems. Furthermore, various comorbid conditions may lead to similar posttransplantation outcomes, playing the same role in patient death. For example, diabetes mellitus and COPD may increase the risk of infection and pneumonia. Therefore, we considered patients with lower (ⱕ1) versus higher rates of comorbidity according to results reported by Jassal et al23 and methodology adopted by Volk et al.26 We showed changes in the clinical characteristics of kidney transplant patients in our center over time, including a greater percentage with substantial burdens of comorbid disease, increasing the importance of a comorbidity assessment. Our data suggested that the CCI was a simple tool to evaluate comorbidity among patients referred for transplantation. Indeed, the analysis of baseline comorbidity has often been used in ESRD patients8,9,15 and kidney transplant recipients to assess operative risk,21,22 showing that increased preoperative comorbidity was related to worse patient survival. Thus, preoperative comorbidity assessment can be used to better identify a group of patients at higher
Fig 1. Actuarial Kaplan-Meier analysis demonstrated that (A) graft and (B) patient survivals are significantly affected by a higher Charlson comorbidity index (CCI).
1862
GROSSO, CORONA, MISTRETTA ET AL Table 3. Independent Comorbidity Predictor of Kidney Transplant Patient Death by Multivariate Analysis
Crude CCI ⬎1 Adjusted model Adjusted model Adjusted model Adjusted model Adjusted model
1: 2: 3: 4: 5:
⫹ recipient age, sex, BMI ⫹ MELD, clinical score model 1 ⫹ hepatocellular carcinoma/DGF, hypertension model 3 ⫹ donor age, sex, ischemia time model 2 ⫹ model 4
HR (95% CI)
P Value
3.87 (1.06–14.06) 3.57 (0.91–14.02) 5.41 (1.16–25.29) 3.55 (0.91–13.95) 6.03 (1.21–30.11) 12.53 (1.9–82.68)
.04 .069 .032 .069 .029 .009
BMI, body mass index; CCI, Charlson comorbidity index; CI, confidence interval; DGF, delayed graft function; HR, hazard ratio; MELD, Model for End-Stage Liver Disease.
risk for death, who could benefit from more intense preoperative assessment and tailored immunosuppression. Our predictive model showed a strong association of CCI with patient survival even after adjustment for several covariates. Defining new determinants of transplant outcome should be useful as alternative for clinical decision making. Little evidence about CCI suggests that it could be useful for this purpose. Moreover, other nonclinical determinants that have been proposed, such as psychologic factors, disability, and resource use, should be involved in development of a more comprehensive model to govern distribution of donor organs.27 The results of our study must be considered in light of several limitations. First, it was a single-center retrospective cohort study. We serve a population with homogeneous characteristics, thus appropriate for to many European but possibly not other countries. On the other hand, regarding the retrospective nature of the study, data obtained from our database were highly detailed, so we think that there was only minimal bias. However, prospective studies related to the onset and duration of comorbidities before transplantation may help to improve our understanding of their role on survival. Second, the study sample was relatively small. Although our ability to perform multivariate analysis was not limited, and the findings remained significant even in the most comprehensive models, we could not adjust for pretransplantation characteristics, such as HLA-mismatch or posttransplantation outcomes, such as delayed graft function. Further studies are needed to determine the potential interactions between comorbid diseases and such covariates on patient survival, preferably in a multicenter study with a large patient cohort. Finally, as reported by Wu et al,24 adoption of a comorbidity index grouping several different diseases together may produce bias due to heterogeneity of data. However, the population referred for kidney transplantation presented a similar pattern of disease, so the validation of an index must assess if such lack of precision affects its implementation in a defined population. In conclusion, CCI could be applied to the evaluation of patients referred for kidney transplantation; a significant burden of comorbid conditions increased the risk of premature death or graft loss.
REFERENCES 1. Wolfe RA, Roys EC, Merion RM: Trends in organ donation and transplantation in the United States, 1999 –2008. Am J Transplant 10:961, 2010 2. Kronson JW, Gillingham KJ, Sutherland DE, et al: Renal transplantation for type II diabetic patients compared with type I diabetic patients and patients over 50 years old: a single-center experience. Clin Transplant 14:226, 2000 3. Woo YM, McLean D, Kavanagh D, et al: The influence of pre-operative electrocardiographic abnormalities and cardiovascular risk factors on patient and graft survival following renal transplantation. J Nephrol 15:380, 2002 4. Moghimi M, Falaknazi K, Bagheri N: Impact of cardiovascular risk factors on the outcome of renal transplantation. Saudi J Kidney Dis Transpl 21:438, 2010 5. Johnson CP, Kuhn EM, Hariharan S, et al: Pre-transplant identification of risk factors that adversely affect length of stay and charges for renal transplantation. Clin Transplant 13:168, 1999 6. Imamura K, McKinnon M, Middleton R, et al: Reliability of a comorbidity measure: the Index of Co-existent Disease (ICED). J Clin Epidemiol 50:1011, 1997 7. Powe NR: Patients, populations and policy: patient outcomes in chronic kidney disease. Trans Am Clin Climatol Assoc 112:224, 2001 8. van Manen JG, Korevaar JC, Dekker FW, et al: How to adjust for comorbidity in survival studies in ESRD patients: a comparison of different indices. Am J Kidney Dis 40:82, 2002 9. van Manen JG, Korevaar JC, Dekker FW, et al: Adjustment for comorbidity in studies on health status in ESRD patients: which comorbidity index to use? J Am Soc Nephrol 14:478, 2003 10. Gill JS, Abichandani R, Kausz AT, et al: Mortality after kidney transplant failure: the impact of non-immunologic factors. Kidney Int 62:1875, 2002 11. Gill JS, Abichandani R, Khan S, et al: Opportunities to improve the care of patients with kidney transplant failure. Kidney Int 61:2193, 2002 12. Veroux M, Grosso G, Corona D, et al: Age is an important predictor of kidney transplantation outcome. Nephrol Dial Transplant 27:1663, 2012 13. Charlson ME, Pompei P, Ales KL, et al: A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis 40:373, 1987 14. Charlson M, Szatrowski TP, Peterson J, et al: Validation of a combined comorbidity index. J Clin Epidemiol 47:1245, 1994 15. Hemmelgarn BR, Manns BJ, Quan H, et al: Adapting the Charlson comorbidity index for use in patients with ESRD. Am J Kidney Dis 42:125, 2003 16. Fried L, Bernardini J, Piraino B: Charlson comorbidity index as a predictor of outcomes in incident peritoneal dialysis patients. Am J Kidney Dis 37:337, 2001 17. Fried L, Bernardini J, Piraino B: Comparison of the Charlson comorbidity index and the Davies score as a predictor of outcomes in PD patients. Perit Dial Int 23:568, 2003
CHARLSON COMORBIDITY INDEX IN KIDNEY TRANSPLANTATION 18. Beddhu S, Zeidel ML, Saul M, et al: The effects of comorbid conditions on the outcomes of patients undergoing peritoneal dialysis. Am J Med 112:696, 2002 19. di Iorio B, Cillo N, Cirillo M, et al: Charlson comorbidity index is a predictor of outcomes in incident hemodialysis patients and correlates with phase angle and hospitalization. Int J Artif Organs 27:330, 2004 20. Miskulin DC, Martin AA, Brown R, et al: Predicting 1 year mortality in an outpatient haemodialysis population: a comparison of comorbidity instruments. Nephrol Dial Transplant 19:413, 2004 21. Kasiske BL, Cangro CB, Hariharan S, et al: The evaluation of renal transplantation candidates: clinical practice guidelines. Am J Transplant 1(suppl 2):3, 2001 22. Humar A, Kerr SR, Ramcharan T, et al: Peri-operative cardiac morbidity in kidney transplant recipients: incidence and risk factors. Clin Transplant 15:154, 2001
1863
23. Jassal SV, Schaubel DE, Fenton SS: Baseline comorbidity in kidney transplant recipients: a comparison of comorbidity indices. Am J Kidney Dis 46:136, 2005 24. Wu C, Evans I, Joseph R, et al: Comorbid conditions in kidney transplantation: association with graft and patient survival. J Am Soc Nephrol 16:3437, 2005 25. Moore J, He X, Liu X, et al: Mortality prediction after kidney transplantation: comparative clinical use of 7 comorbidity indices. Exp Clin Transplant 9:32, 2011 26. Volk ML, Hernandez JC, Lok AS, et al: Modified Charlson comorbidity index for predicting survival after liver transplantation. Liver Transpl 13:1515, 2007 27. Krakauer H, Bailey RC, Lin MJ. Beyond survival: the burden of disease in decision making in organ transplantation. Am J Transplant 4:1555, 2004
T
instrument could be useful to better compare results across institutions. The Charlson comorbidity index (CCI)13 uses a simple weighted scoring system based on the presence or absence of comorbid conditions.14 The CCI has been validated in ESRD patients,8,9,15 seeking to predict the outcomes of patients on either peritoneal16 –18 or hemodialysis.19,20 Moreover, CCI has been used to assess operative risk From the Department “G.F. Ingrassia” (G.G., A.M.), Section of Hygiene and Public Health, and Vascular Surgery and Organ Transplant Unit, Department of Surgery, Transplantation, and Advanced Technologies (D.C., D.Z., N.S., A.G., T.T., B.E., A.L., R.G., P.V., M.V.), University Hospital of Catania, Catania, Italy; and Thomas E. Starzl Transplantation Institute (B.E.), University of Pittsburgh Medical Center, Pittsburgh, Pennsylvania. Address reprint requests to Massimiliano Veroux, MD, PhD, Vascular Surgery and Organ Transplant Unit, Department of Surgery, Transplantation, and Advanced Technologies, University Hospital of Catania, Via Santa Sofia, 83 95128 Catania, Italy. E-mail: [email protected]
© 2012 by Elsevier Inc. All rights reserved. 360 Park Avenue South, New York, NY 10010-1710
0041-1345/–see front matter http://dx.doi.org/10.1016/j.transproceed.2012.06.042
Transplantation Proceedings, 44, 1859 –1863 (2012)
1859
1860
among kidney transplant recipients,21,22 but only a few studies have assessed its role to predict posttransplantation outcomes.23–25 Therefore, the aim of the present study was to define the predictive value of CCI for graft loss or mortality among a population of kidney transplant recipients. METHODS We retrospectively reviewed the data from 223 recipients of first deceased-donor kidney transplantations performed from January 2002 to December 2007. Kidney recipients ⬍18 years old, donors ⬍18 years old, living-donor cases, and simultaneous kidneypancreas transplantations were excluded from the study. Demographic data included age and sex. Clinical characteristics included primary cause of ESRD and baseline patient comorbidities; data on post-transplantation outcomes, graft failures and patient deaths. Delayed graft function (DGF) was defined as the need for ⱖ1 dialysis session within 1 week after transplantation; primary kidney nonfunction (PNF) was defined as the complete lack of functionality such that the recipient never discontinued dialysis sessions after transplantation. We collected a total of 9 comorbidities comprising the CCI according to Jassal et al.23 The CCI was calculated by assigning a weight of 2 points to diabetes mellitus (DM), cerebrovascular accident, and solid tumor or leukemia, and a weight of 1 to previous myocardial infarction, congestive heart failure, peripheral vascular disease, COPD, connective tissue disease, and mild liver disease. Given the nature of transplant recipients, we did not consider renal insufficiency to be a comorbid condition, because it was present in every patient. Baseline characteristics were described as the frequency of occurrence of categoric data and the mean ⫾ SD for continuous variables. Patients were divided into 2 groups according to CCI values with a cutoff point of CCI ⱕ1 versus CCI ⬎1 on the basis of the results of Jassal et al,23 who observed a substantial change in the hazard ratio (HR) for death with increasing CCI scores mostly between the lower rates of the score. Differences between the groups according to CCI value were tested using the chi-square test for categoric variables and 2-sample Student t test or analysis of variance for continuous variables. Patient and graft survivals were estimated by the Kaplan-Meier method. Unadjusted and multivariate analyse were performed using Cox proportional hazards modeling with backward stepwise selection of covariates to determine independent comorbid factors that predict patient and graft survival. All covariates with P values ⬍.1 in univariate analysis were entered in the multivariate model. To identify the role of CCI to predict posttransplantation outcomes, we performed another Cox proportional hazards analysis including CCI (unadjusted) and patient factors (age, sex, and body mass index), other pretransplantation scores (clinical status, as defined above), other comorbidity conditions not included in the previous analysis (hypertension and DGF), and donor factors (age and sex) in adjusted models. All statistical tests were 2 tailed. Data were entered into Microsoft Excel for Windows (Microsoft Corp, Redmond, Washington). Statistical analysis was performed using SPSS for Windows release 17.0 (SPSS, Chicago, Illinois).
RESULTS
Patient baseline characteristics were stratified by CCI as listed in Table 1. No differences were observed between patients with CCI ⱕ1 compared with CCI ⬎1, with the
GROSSO, CORONA, MISTRETTA ET AL Table 1. Baseline Demographics of Study Population of Kidney Transplant Recipients (n ⴝ 223)
Recipients Age (y), mean Male (%) BMI, kg/m2 (%) ⬎28 ⬎32 Primary cause of ESRD (%) Glomerulonephritis Tubulopathy Polycystic kidney Vasculopathy Others Duration of dialysis, y (%) ⬍3 3–5 ⬎5 Cold ischemia time, h (%) 0–24 25–36 ⬎36 Serum creatinine (mg/dL): At discharge At 6 mo At 1 y Donors Age (y), mean ⫾ SD Male (%)
CCI ⱕ1 (n ⫽ 113)
CCI ⬎1 (n ⫽ 120)
P Value
49.2 77
48.2 54.5
.485 .001 .661
19.2 10.6
24.5 10.6 .933
56.6 4.4 2.7 14.2 22.1
60 5.5 1.8 14.5 18.2
52 15 33
52 19.6 28.4
75.2 23.9 0.9
84.5 14.5 0.9
1.37 ⫾ 0.82 1.38 ⫾ 0.83 1.35 ⫾ 0.68
1.49 ⫾ 1.25 1.42 ⫾ 1.06 1.34 ⫾ 0.7
48.8 ⫾ 18.9 61.9
53.3 ⫾ 17.9 60
.618
.209
.426 .794 .949 .778 .069 .766
BMI, body mass index; CCI, Charlson comorbidity index; ESRD, end-stage renal disease.
exception of recipient sex showing a lower percentage of men among patients with CCI ⬎1 (77% vs 54.5%; P ⬍ .001). About 50% of recipients displayed ⬎1 comorbid condition before transplantation, with diabetes mellitus the most frequently reported one (Table 2). Survival analysis revealed a significant association between CCI and decreased posttransplantation patient survival (P ⫽ .003; Fig 1). Furthermore, Kaplan-Meier plots and log rank tests showed a significant association between graft survival and CCI score (P ⫽ .039; Fig 1). The independent contribution of each individual component of the CCI to predict patient death was explored by multivariate analysis. There was no correlation between any single comorbid condition and posttransplantation patient status among recipients. Adjusted analysis of graft loss risk showed a significant correlation with COPD (HR, 4.71; 95% CI, 1.07–20.83; P ⫽ .041). Unadjusted and adjusted HR for patient death according to CCI are listed in Table 3. Crude analysis showed a significant association between CCI ⬎1 and risk of death (HR, 3.87; 95% CI, 1.06 –14.06; P ⫽ .04). After adjusting for several covariates, high CCI values remained significantly predictive of posttransplantation outcome with an HR for death of 12.53 (95% CI, 1.9 – 82.68; P ⫽ .009). Graft
CHARLSON COMORBIDITY INDEX IN KIDNEY TRANSPLANTATION Table 2. Charlson Comorbidity Index (CCI) and Comorbidities Distribution Among Kidney Transplant Recipients (n ⴝ 223) n (%)
CCI distribution 0 1 2 3 4 5 6 Comorbidities distribution Congestive heart failure Coronary disease Diabetes mellitus Stroke Peripheral artery disease COPD Connective tissue disease Liver disease Malignancy
93 (41.7) 20 (9) 81 (36.3) 18 (8.1) 9 (4) 0 2 (0.9) 2 (0.9) 22 (9.9) 94 (42.2) 2 (0.9) 19 (8.5) 16 (7.2) 13 (5.8) 2 (0.9) 6 (2.7)
COPD, chronic pulmonary disease.
survival analysis revealed a similar crude risk of graft loss (HR, 2.44; 95% CI, 1.19 –5.02; P ⫽ .015) but no association was found after adjusting for covariates (data not shown). DISCUSSION
We have presented the comorbidity conditions of 223 kidney transplant recipients. Not surprisingly, we observed that 50% of patients had ⱖ1 associated disease; the number of patients with multiple baseline comorbidities who received a transplant at our institution has increased over the past several years. Given the controversies concerning the role of a single comorbidity to predict renal transplantation outcomes, we adopted a modified version of the CCI, a
1861
simple comorbidity score that has been validated among patients with ESRD15 to assess the effects of pretransplantation comorbid conditions on posttransplantation patient and graft survivals. Patients with CCI ⬎1 showed a 12.53fold increased risk of death after kidney transplantation, but CCI was not significantly associated with graft survival. Our study demonstrated no significant association between a single point in the score with the risk of patient death; however, a cutoff point significantly defined a higher risk of patient death. The nonlinear correlation between CCI and outcomes may be explained by several comorbid conditions that are often related to each other, namely, cluster disease involving the cardiovascular, respiratory, connective and autoimmune systems. Furthermore, various comorbid conditions may lead to similar posttransplantation outcomes, playing the same role in patient death. For example, diabetes mellitus and COPD may increase the risk of infection and pneumonia. Therefore, we considered patients with lower (ⱕ1) versus higher rates of comorbidity according to results reported by Jassal et al23 and methodology adopted by Volk et al.26 We showed changes in the clinical characteristics of kidney transplant patients in our center over time, including a greater percentage with substantial burdens of comorbid disease, increasing the importance of a comorbidity assessment. Our data suggested that the CCI was a simple tool to evaluate comorbidity among patients referred for transplantation. Indeed, the analysis of baseline comorbidity has often been used in ESRD patients8,9,15 and kidney transplant recipients to assess operative risk,21,22 showing that increased preoperative comorbidity was related to worse patient survival. Thus, preoperative comorbidity assessment can be used to better identify a group of patients at higher
Fig 1. Actuarial Kaplan-Meier analysis demonstrated that (A) graft and (B) patient survivals are significantly affected by a higher Charlson comorbidity index (CCI).
1862
GROSSO, CORONA, MISTRETTA ET AL Table 3. Independent Comorbidity Predictor of Kidney Transplant Patient Death by Multivariate Analysis
Crude CCI ⬎1 Adjusted model Adjusted model Adjusted model Adjusted model Adjusted model
1: 2: 3: 4: 5:
⫹ recipient age, sex, BMI ⫹ MELD, clinical score model 1 ⫹ hepatocellular carcinoma/DGF, hypertension model 3 ⫹ donor age, sex, ischemia time model 2 ⫹ model 4
HR (95% CI)
P Value
3.87 (1.06–14.06) 3.57 (0.91–14.02) 5.41 (1.16–25.29) 3.55 (0.91–13.95) 6.03 (1.21–30.11) 12.53 (1.9–82.68)
.04 .069 .032 .069 .029 .009
BMI, body mass index; CCI, Charlson comorbidity index; CI, confidence interval; DGF, delayed graft function; HR, hazard ratio; MELD, Model for End-Stage Liver Disease.
risk for death, who could benefit from more intense preoperative assessment and tailored immunosuppression. Our predictive model showed a strong association of CCI with patient survival even after adjustment for several covariates. Defining new determinants of transplant outcome should be useful as alternative for clinical decision making. Little evidence about CCI suggests that it could be useful for this purpose. Moreover, other nonclinical determinants that have been proposed, such as psychologic factors, disability, and resource use, should be involved in development of a more comprehensive model to govern distribution of donor organs.27 The results of our study must be considered in light of several limitations. First, it was a single-center retrospective cohort study. We serve a population with homogeneous characteristics, thus appropriate for to many European but possibly not other countries. On the other hand, regarding the retrospective nature of the study, data obtained from our database were highly detailed, so we think that there was only minimal bias. However, prospective studies related to the onset and duration of comorbidities before transplantation may help to improve our understanding of their role on survival. Second, the study sample was relatively small. Although our ability to perform multivariate analysis was not limited, and the findings remained significant even in the most comprehensive models, we could not adjust for pretransplantation characteristics, such as HLA-mismatch or posttransplantation outcomes, such as delayed graft function. Further studies are needed to determine the potential interactions between comorbid diseases and such covariates on patient survival, preferably in a multicenter study with a large patient cohort. Finally, as reported by Wu et al,24 adoption of a comorbidity index grouping several different diseases together may produce bias due to heterogeneity of data. However, the population referred for kidney transplantation presented a similar pattern of disease, so the validation of an index must assess if such lack of precision affects its implementation in a defined population. In conclusion, CCI could be applied to the evaluation of patients referred for kidney transplantation; a significant burden of comorbid conditions increased the risk of premature death or graft loss.
REFERENCES 1. Wolfe RA, Roys EC, Merion RM: Trends in organ donation and transplantation in the United States, 1999 –2008. Am J Transplant 10:961, 2010 2. Kronson JW, Gillingham KJ, Sutherland DE, et al: Renal transplantation for type II diabetic patients compared with type I diabetic patients and patients over 50 years old: a single-center experience. Clin Transplant 14:226, 2000 3. Woo YM, McLean D, Kavanagh D, et al: The influence of pre-operative electrocardiographic abnormalities and cardiovascular risk factors on patient and graft survival following renal transplantation. J Nephrol 15:380, 2002 4. Moghimi M, Falaknazi K, Bagheri N: Impact of cardiovascular risk factors on the outcome of renal transplantation. Saudi J Kidney Dis Transpl 21:438, 2010 5. Johnson CP, Kuhn EM, Hariharan S, et al: Pre-transplant identification of risk factors that adversely affect length of stay and charges for renal transplantation. Clin Transplant 13:168, 1999 6. Imamura K, McKinnon M, Middleton R, et al: Reliability of a comorbidity measure: the Index of Co-existent Disease (ICED). J Clin Epidemiol 50:1011, 1997 7. Powe NR: Patients, populations and policy: patient outcomes in chronic kidney disease. Trans Am Clin Climatol Assoc 112:224, 2001 8. van Manen JG, Korevaar JC, Dekker FW, et al: How to adjust for comorbidity in survival studies in ESRD patients: a comparison of different indices. Am J Kidney Dis 40:82, 2002 9. van Manen JG, Korevaar JC, Dekker FW, et al: Adjustment for comorbidity in studies on health status in ESRD patients: which comorbidity index to use? J Am Soc Nephrol 14:478, 2003 10. Gill JS, Abichandani R, Kausz AT, et al: Mortality after kidney transplant failure: the impact of non-immunologic factors. Kidney Int 62:1875, 2002 11. Gill JS, Abichandani R, Khan S, et al: Opportunities to improve the care of patients with kidney transplant failure. Kidney Int 61:2193, 2002 12. Veroux M, Grosso G, Corona D, et al: Age is an important predictor of kidney transplantation outcome. Nephrol Dial Transplant 27:1663, 2012 13. Charlson ME, Pompei P, Ales KL, et al: A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis 40:373, 1987 14. Charlson M, Szatrowski TP, Peterson J, et al: Validation of a combined comorbidity index. J Clin Epidemiol 47:1245, 1994 15. Hemmelgarn BR, Manns BJ, Quan H, et al: Adapting the Charlson comorbidity index for use in patients with ESRD. Am J Kidney Dis 42:125, 2003 16. Fried L, Bernardini J, Piraino B: Charlson comorbidity index as a predictor of outcomes in incident peritoneal dialysis patients. Am J Kidney Dis 37:337, 2001 17. Fried L, Bernardini J, Piraino B: Comparison of the Charlson comorbidity index and the Davies score as a predictor of outcomes in PD patients. Perit Dial Int 23:568, 2003
CHARLSON COMORBIDITY INDEX IN KIDNEY TRANSPLANTATION 18. Beddhu S, Zeidel ML, Saul M, et al: The effects of comorbid conditions on the outcomes of patients undergoing peritoneal dialysis. Am J Med 112:696, 2002 19. di Iorio B, Cillo N, Cirillo M, et al: Charlson comorbidity index is a predictor of outcomes in incident hemodialysis patients and correlates with phase angle and hospitalization. Int J Artif Organs 27:330, 2004 20. Miskulin DC, Martin AA, Brown R, et al: Predicting 1 year mortality in an outpatient haemodialysis population: a comparison of comorbidity instruments. Nephrol Dial Transplant 19:413, 2004 21. Kasiske BL, Cangro CB, Hariharan S, et al: The evaluation of renal transplantation candidates: clinical practice guidelines. Am J Transplant 1(suppl 2):3, 2001 22. Humar A, Kerr SR, Ramcharan T, et al: Peri-operative cardiac morbidity in kidney transplant recipients: incidence and risk factors. Clin Transplant 15:154, 2001
1863
23. Jassal SV, Schaubel DE, Fenton SS: Baseline comorbidity in kidney transplant recipients: a comparison of comorbidity indices. Am J Kidney Dis 46:136, 2005 24. Wu C, Evans I, Joseph R, et al: Comorbid conditions in kidney transplantation: association with graft and patient survival. J Am Soc Nephrol 16:3437, 2005 25. Moore J, He X, Liu X, et al: Mortality prediction after kidney transplantation: comparative clinical use of 7 comorbidity indices. Exp Clin Transplant 9:32, 2011 26. Volk ML, Hernandez JC, Lok AS, et al: Modified Charlson comorbidity index for predicting survival after liver transplantation. Liver Transpl 13:1515, 2007 27. Krakauer H, Bailey RC, Lin MJ. Beyond survival: the burden of disease in decision making in organ transplantation. Am J Transplant 4:1555, 2004