TY - JOUR
AU - Herzog, Charles A.
AB - Abstract Background. Bacterial endocarditis is associated with poor long-term survival among dialysis patients. Renal transplant patients and those waiting list for renal transplantation are predisposed to developing bacterial endocarditis; data regarding incidence and outcomes are limited. Methods. Patients hospitalised for bacterial endocarditis were identified from patients transplanted or waiting list between 1995 and 2003. Transplant and waiting list cohorts were derived from the United States Renal Data System (USRDS) database. All patients had Medicare as primary payer. Long-term survival was estimated by the Kaplan–Meier method. Cox proportional hazards analysis was used to identify independent predictors of bacterial endocarditis. Results. During the study period, 282 renal transplant patients and 549 waiting list patients were hospitalised with bacterial endocarditis. Incidence rates of bacterial endocarditis per 1000 patient-years were 5.6 among waiting list patients, 2.6 among deceased-donor transplant recipients and 1.8 among living-donor transplant recipients. In-hospital mortality rates were 16.0% for the renal transplant cohort and 18.6% for the waiting list cohort. Two-year post-endocarditis survival rates were 58% for transplant patients and 41% for waiting list patients. The most powerful predictors of bacterial endocarditis among transplant patients were donor age, patient age, diabetic end-stage renal disease (ESRD) and prior dialysis time longer than 2 years. Conclusions. Renal transplant patients hospitalised with bacterial endocarditis sustain high in-hospital and long-term mortality rates. Waiting list patients are at higher risk of developing bacterial endocarditis than renal transplant recipients. dialysis, kidney, survival, transplant, wait-list Introduction The end-stage renal disease (ESRD) population continues to increase rapidly in the United States. In 2004, about 335 963 patients were receiving long-term dialysis and 104 364 patients were newly diagnosed with ESRD [1]. The prevalence of transplant patients in the United States increased from 108 517 in 2000 to 136 136 in 2004 [1]. The number of patients on the kidney transplant waiting list has consequently grown, with >5000 additional patients listed annually since 2003 [2]. Waiting list patients continue to experience high mortality rates and long waiting periods. In 2005, of ∼82 582 patients waiting list for renal transplantation, 4156 died (an unadjusted mortality rate of 70.2 per 1000 patient-years at risk) [2]. In contrast, survival of transplanted patients has consistently improved. For patients who received transplants between 2001 and 2004, unadjusted survival rates at 3 years were 88.1% for deceased-donor recipients and 94.5% for living-donor recipients [2]. Bacterial endocarditis is associated with poor survival among dialysis patients, with an in-hospital mortality rate of 24% and a 2-year survival rate of 25% [3]. A large majority of waiting list patients undergo dialysis and are likely at higher risk for developing endocarditis than the general population. Transplant recipients are predisposed to infections secondary to immunosuppression. The purpose of this study was to determine the incidence and survival of renal transplant and waiting list patients with bacterial endocarditis in the contemporary era. Using the United States Renal Data System (USRDS) database, we examined outcomes for 282 renal transplant and 549 waiting list patients hospitalised with bacterial endocarditis over a 9-year period. Subjects and methods All data were derived from the USRDS database. The accuracy of these data has been validated previously [4]. The transplant cohort included first-time, kidney-only recipients transplanted between 1 January 1995, and 31 December 2003, and followed for up to 3 years. As some people receive transplants without first being waiting list, not all patients in the cohort were previously waiting list. The waiting list cohort included patients first listed on a transplant waitlist during the same time period, and followed until transplantation, death, or 31 December 2003. The contribution to the waitlist analysis of those who subsequently received transplants was censored at the time of transplantation. Both cohorts are limited to patients with Medicare as primary payer. We identified an index hospitalisation for bacterial endocarditis (i.e. the first hospitalisation identified in the analysis) by searching Medicare claims for International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) code 421.0 occurring during the follow-up period. We identified infectious organisms by searching Medicare claims for ICD-9-CM codes 038.xx and 041.xx occurring during the same hospitalisation as the bacterial endocarditis claim. Baseline data included age, sex, race, etiology of ESRD and time on dialysis. In-hospital mortality was determined in the cohort of patients hospitalised for bacterial endocarditis after receiving a transplant or being waiting list for renal transplantation. Time to bacterial endocarditis was calculated from the date of transplant (transplant cohort) or listing date (waitlist cohort) to the date of the first bacterial endocarditis claim or censor date. Post-endocarditis survival time was calculated from the date of hospital admission for bacterial endocarditis to death or censor date. Patients in the transplant cohort were censored at the earliest of 3 years post-renal transplantation, 31 December 2003, or loss of Medicare coverage. Patients in the waiting list cohort were censored at the earliest of renal transplantation, 31 December 2003, or loss of Medicare coverage. Differences in baseline characteristics were assessed using chi-square tests. The cumulative incidence of bacterial endocarditis was determined and unadjusted survival was estimated by the Kaplan–Meier method [5]. A Cox proportional hazards model and hazard ratios were used to identify independent predictors of bacterial endocarditis [6]. The adjusted hazard ratios for developing bacterial endocarditis were estimated by a time-dependent Cox proportional hazards model adjusted for age, sex, race, etiology of ESRD and dialysis time. All statistical analyses were performed using SAS for Windows, version 9 (SAS Institute Inc., Cary, NC, USA). Results In the 9-year period between 1 January 1995 and 31 December 2003, 47 899 renal transplantation patients and 62 520 waiting list patients were identified from the USRDS database. An index hospitalisation with bacterial endocarditis was identified for 282 renal transplant patients and 549 waiting list patients. The median follow-up period was 35.2 months for transplant patients and 12.8 months for waiting list patients. Baseline characteristics of the renal transplant patients are listed in Table 1. Compared with transplant patients not hospitalised with bacterial endocarditis, those hospitalised with bacterial endocarditis were more likely to be between the ages of 45 and 64 years at the time of transplantation and black, and to have diabetic ESRD. Among hospitalised patients, 57% had been on dialysis for >3 years before transplantation, compared with 46% of patients not hospitalised with bacterial endocarditis. Table 1 Baseline characteristics of transplant patients with and without bacterial endocarditis   Endocarditis      No  Yes    Characteristics  n  %  n  %  P  All  47 617  99.4  282   0.6    Age at transplantation (year)          0.003     <45  14 204  29.8   61  21.6       45–64  19 582  41.1  142  50.4       65–74   8 884  18.7   56  19.9       ≥75   4 947  10.4   23   8.2    Sex          0.82     Female  19 054  40.0  111  39.4       Male  28 563  60.0  171  60.6    Race          0.001     White  30 343  63.7  156  55.3       Black  14 428  30.3  117  41.5       Asian   2 080   4.4    7   2.5       Native American     675   1.4    2   0.7       Other      91   0.2    0   0.0    ESRD etiology          0.028     Diabetes  12 041  25.3   87  30.9       Hypertension  10 984  23.1   76  27.0       Glomerulonephritis  12 043  25.3   57  20.2       Cystic kidney disease   3 485   7.3   14   5.0       Other   9 064  19.0   48  17.0    Time on dialysis (year)          0.0005     Preemptive   1 439   3.0    3   1.1       >0–1   5 753  12.1   17   6.0       >1–2   9 075  19.1   43  15.2       >2–3   9 292  19.5   58  20.6       ≥3  21 988  46.2  161  57.1      Endocarditis      No  Yes    Characteristics  n  %  n  %  P  All  47 617  99.4  282   0.6    Age at transplantation (year)          0.003     <45  14 204  29.8   61  21.6       45–64  19 582  41.1  142  50.4       65–74   8 884  18.7   56  19.9       ≥75   4 947  10.4   23   8.2    Sex          0.82     Female  19 054  40.0  111  39.4       Male  28 563  60.0  171  60.6    Race          0.001     White  30 343  63.7  156  55.3       Black  14 428  30.3  117  41.5       Asian   2 080   4.4    7   2.5       Native American     675   1.4    2   0.7       Other      91   0.2    0   0.0    ESRD etiology          0.028     Diabetes  12 041  25.3   87  30.9       Hypertension  10 984  23.1   76  27.0       Glomerulonephritis  12 043  25.3   57  20.2       Cystic kidney disease   3 485   7.3   14   5.0       Other   9 064  19.0   48  17.0    Time on dialysis (year)          0.0005     Preemptive   1 439   3.0    3   1.1       >0–1   5 753  12.1   17   6.0       >1–2   9 075  19.1   43  15.2       >2–3   9 292  19.5   58  20.6       ≥3  21 988  46.2  161  57.1    View Large The unadjusted incidence of bacterial endocarditis was 5.6 per 1000 patient-years among waiting list patients, 2.6 per 1000 patient-years among deceased-donor renal transplant recipients and 1.8 per 1000 patient-years among living-donor renal transplants recipients (Table 2). As illustrated in Figure 1, incidence of bacterial endocarditis was lowest among living-donor recipients and second lowest among deceased-donor recipients. Incidence was highest among the waiting list patients, and continued to increase with longer time on the list. The in-hospital mortality rate was 16.0% for the renal transplant cohort and 18.6% for the waiting list cohort (Table 2). Table 2 In-hospital mortality and long-term survival of renal transplant patients and waiting list patients with bacterial endocarditis   Cohort    Transplant  Waiting list  Mortality and survival  n = 47 899  n = 62 520  Hospitalised for bacterial endocarditis, n   282   549  Bacterial endocarditis incidence (per 1000 patient-years)   2.72   5.57  In-hospital death (%)   16.0   18.6  Survival (year) (%)         0.5   71.2   61.1     1   66.1   52.4     2   57.8   41.4     3   54.2   33.4    Cohort    Transplant  Waiting list  Mortality and survival  n = 47 899  n = 62 520  Hospitalised for bacterial endocarditis, n   282   549  Bacterial endocarditis incidence (per 1000 patient-years)   2.72   5.57  In-hospital death (%)   16.0   18.6  Survival (year) (%)         0.5   71.2   61.1     1   66.1   52.4     2   57.8   41.4     3   54.2   33.4  View Large Fig. 1 View largeDownload slide Cumulative incidence (percent) of bacterial endocarditis in waiting list patients, living donor renal transplant recipients and deceased donor renal transplant recipients. Fig. 1 View largeDownload slide Cumulative incidence (percent) of bacterial endocarditis in waiting list patients, living donor renal transplant recipients and deceased donor renal transplant recipients. Post-endocarditis survival rates at 2 years were 58% for the transplant cohort and 41% for the waiting list cohort (Figure 2). Survival rates in the two groups were significantly different (P < 0.0001 by the log-rank test). Fig. 2 View largeDownload slide Survival of renal transplant patients and waiting list patients with bacterial endocarditis. Fig. 2 View largeDownload slide Survival of renal transplant patients and waiting list patients with bacterial endocarditis. The Cox proportional hazards analysis was used to identify independent predictors of endocarditis in the renal transplant cohort (Table 3). The most powerful predictors of bacterial endocarditis were donor age, patient age, diabetic ESRD and prior dialysis time. The unadjusted survival of waiting list patients and renal transplant recipients was significantly different; however, in a Cox model used to examine predictors of mortality in patients with endocarditis, none of the variables was independently associated with all-cause death (data not shown). Table 3 Cox proportional hazards model of independent predictors of bacterial endocarditis   Hazard ratio    Covariate  (95% CI)  P  Donor age 50–65 years (versus 18–34)  1.64 (1.18–2.29)  0.003  Living donor (versus Deceased)  0.91 (0.64–1.29)  NS  Age 0–44 years (versus 45–64)  0.66 (0.48–0.90)  0.010  Black (versus white)  1.24 (0.93–1.64)  NS  Male (versus female)  1.02 (0.80–1.30)  NS  Diabetic ESRD (versus glomerulonephritis)  1.54 (1.09–2.17)  0.01  Prior dialysis time (year)         <1  1.00 (ref)  (ref)     1–2  1.48 (0.84–2.60)  NS     2–3  1.86 (1.07–3.24)  0.029     ≥3  2.24 (1.32–3.80)  0.003    Hazard ratio    Covariate  (95% CI)  P  Donor age 50–65 years (versus 18–34)  1.64 (1.18–2.29)  0.003  Living donor (versus Deceased)  0.91 (0.64–1.29)  NS  Age 0–44 years (versus 45–64)  0.66 (0.48–0.90)  0.010  Black (versus white)  1.24 (0.93–1.64)  NS  Male (versus female)  1.02 (0.80–1.30)  NS  Diabetic ESRD (versus glomerulonephritis)  1.54 (1.09–2.17)  0.01  Prior dialysis time (year)         <1  1.00 (ref)  (ref)     1–2  1.48 (0.84–2.60)  NS     2–3  1.86 (1.07–3.24)  0.029     ≥3  2.24 (1.32–3.80)  0.003  NS, not significant. View Large There were differences in the distribution of infectious organisms among transplant and waitlist patients. For transplant patients with endocarditis, Streptococcal species were identified in 33%, Staphylococcal species in 29%, other organism in 8% and unknown organism in 33%. For waiting list patients with endocarditis, Streptococcal species were identified in 13%, Staphylococcal species in 47%, other organism in 8% and unknown organism in 35%. (As more than one pathogen was identified for some patients, total can exceed 100% by the patient group.) Discussion Solid organ transplant recipients have been found to have about a 170-fold risk of endocarditis compared with the general population [7]. In small clinical studies, the overall mortality from endocarditis among transplant recipients has been reported to be in the range of 44–57% [7–9]. Limited data are available regarding the long-term survival of renal transplant recipients with bacterial endocarditis in the contemporary era. To the best of our knowledge, there are no published long-term survival data regarding waiting list patients with bacterial endocarditis. This study highlights the poor survival rates of renal transplant patients with bacterial endocarditis. Our findings complement those of Abbott et al., who reported a 2-year mortality of 44.2% among renal transplant recipients with bacterial endocarditis versus 8.2% among transplant recipients without bacterial endocarditis between 1994 and 1997 [10]. The 2-year post-endocarditis survival of patients transplanted between 1995 and 2003 was ∼58%, relatively unchanged since the findings reported by Abbott et al. Despite advances in the medical field, dialysis patient survival after hospitalisation for bacterial endocarditis has changed little over more than a decade [3]. In-hospital mortality rates for the waiting list cohort were higher than those for the transplant cohort (18.6% versus 16.0%, respectively) [3]. Chu et al. previously identified early independent predictors of in-hospital mortality with bacterial endocarditis in the general population, namely diabetes mellitus, Staphylococcus aureus as causative organism, APACHE (acute physiology, age and chronic health evaluation) II score and embolic events [11]. Interestingly, although the in-hospital mortality rate for dialysis patients with bacterial endocarditis is significantly higher at ∼23.5% [3], in-hospital mortality for transplant patients is similar to that described for the general population [11,12]. Transplant patients are predisposed to infections secondary to immunosuppression. In the view of some, higher immunosuppressive therapy, especially during episodes of acute rejection, is more likely to enhance the risk of septicaemia [7]. However, little information is currently available regarding specific risk factors or modes of infection leading to endocarditis in the transplant population. Interestingly, Paterson et al. found a lesser incidence of endocarditis in the early post-transplantation period among renal transplant recipients compared with liver and heart transplant recipients, leading to the postulation that renal transplant recipients may receive lesser cumulative doses of immunosuppression therapy because dialysis is an available option to cope with graft rejection [8]. We found that the duration of dialysis before transplantation was an independent predictor of bacterial endocarditis and that patients who received preemptive renal transplantation had the lowest chance of developing endocarditis. Dialysis vintage was also identified as an independent risk factor for endocarditis in the study by Abbott et al. [10]. Factors predisposing dialysis patients to death from endocarditis after transplantation, or to infection while on the transplant waitlist, could not be elucidated from the current study. Type of vascular access (e.g. catheter) is likely an independent predictor of risk of infection and endocarditis, but our data lack vascular access information. (We found no difference in endocarditis risk for haemodialysis compared to peritoneal dialysis patients, but the number of peritoneal dialysis patients is small, making the comparison difficult.) One can nevertheless postulate that previous bacteraemia during dialysis, indolent infection at the site of vascular access, or harbouring of particular bacteria such as Staphylococcus aureus might predispose to subsequent infection during periods of intense immunosuppression [13,14]. We found a suggestive difference in the bacteriology of transplant and waiting list patients. Of waiting list patients, 47% apparently had Staphylococcal species and 13% had Streptococcal species identified as infectious organisms, compared to 29% of transplant patients with Staphylococcal species and 33% with Streptococcal species. These patterns may reflect the heightened risk of dialysis patients to bacteraemia from skin flora (as would be expected with a greater preponderance of Staphylococcal infections). Older age (both donor and recipient) and diabetic ESRD were independent predictors of bacterial endocarditis in our study, presumably related to weaker immune constitution and resultant predisposition to systemic infection. Compared with the general population, valvular disease has not been conclusively linked as an independent predictor of endocarditis in transplant patients. Although Abbott et al. found that previous hospitalisation for valvular heart disease was a strong independent risk factor for endocarditis in transplant recipients (odds ratio 25.81, confidence interval 11.28–59.07) [10], other investigators have not found a significant association between valvular disease and endocarditis and have concluded that previous valvular disease is not as relevant in the context of an immuno-compromised host [7]. The present analysis did not address this issue. The phenomenon of self-selection has also been implicated, suggesting that patients with severe valvular disease might not be considered candidates for transplantation [8]. Patients waiting list for transplantation had about twice the incidence of bacterial endocarditis in this study compared with transplant patients (5.57 versus 2.72 per 1000 patient-years). As illustrated in Figure 1, the incidence of bacterial endocarditis in waiting list patients continues to increase whereas the incidence in transplant patients tends to plateau with time. Moreover, the long-term mortality of waiting list patients with bacterial endocarditis was much higher than that of the transplant population. Thus, waiting list patients have a survival rate between patients on chronic dialysis (lowest survival rates) and renal transplant recipients (highest survival rates). This finding may be explained by the fact that within the ESRD population healthier patients are waiting list for transplantation and long-term survival is better among waiting list patients who eventually undergo transplantation, as described by Wolfe et al. [15]. Also, not all waiting list patients receive dialysis before transplantation, and thus might be at lower risk of subsequent development of endocarditis. The limitations of this study should be noted. The USRDS database includes little clinical data. The data pertaining to type of bacterial pathogen are derived from claims data, and no information related to specific pathogens was available for a third of the patients. Although endocarditis is most commonly of bacterial origin in transplant patients, various organisms including fungi (especially Aspergillus) can be responsible for endocarditis in immunosuppressed individuals, particularly in the early post-transplant period [7–9]. Thus, the overall incidence of endocarditis is likely underestimated in this study, as non-bacterial causes of endocarditis were not considered. Likewise, because fungal endocarditis often carries a much poorer prognosis the post-endocarditis mortality is also likely underestimated in this study. Finally, this study included only patients with Medicare as primary payer; hence, the entire renal transplant and waiting list populations were not evaluated. In conclusion, renal transplant patients hospitalised with bacterial endocarditis sustain high in-hospital and long-term mortality rates. Waiting list patients are at higher risk of developing bacterial endocarditis compared with renal transplant recipients. Additional measures are necessary to reduce bacteraemia and implement risk reduction strategies to prevent bacterial endocarditis and the attendant morbidity and mortality in these vulnerable patients. The authors thank USRDS colleagues Shane Nygaard for manuscript preparation and Nan Booth for manuscript editing. The Cardiovascular Special Studies Center of the United States Renal Data System is supported by Contract No. HHSN267200715003C (National Institute of Diabetes and Digestive and Kidney Diseases, National Institutes of Health, Bethesda, MD, USA). Conflict of interest statement. The data reported here have been supplied by the United States Renal Data System (USRDS). The interpretation and reporting of these data are the responsibility of the authors and in no way should be seen as an official policy or interpretation of the US government. The authors have no conflict of interest with the subject matter of this manuscript. References 1 US Renal Data System. USRDS. ,  2006  2006 Annual Data Report. National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases, Bethesda, MD 2 US Organ Procurement and Transplantation Network and the Scientific Registry of Transplant Recipients. ,  2007(pg.  7- 12)  Transplant Statistics: 2006 Annual Data Report. US Organ Procurement and Transplantation Network and the Scientific Registry of Transplant Recipients 2006–2007 Department of Health and Human Services, Health Resources and Services Administration, Healthcare Systems Bureau, Division of Transplantation 3 Shroff GR,  Herzog CA,  Ma JZ,  Collins AJ.  Long-term survival of dialysis patients with bacterial endocarditis in the United States,  Am J Kidney Dis ,  2004, vol.  44 (pg.  1077- 1082) Google Scholar CrossRef Search ADS PubMed  4 US Renal Data System. USRDS. ,  1992  1992 Annual Data Report. The National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases, Bethesda, MD 5 Kaplan EL,  Meier P.  Nonparametric estimation from incomplete observations,  J Am Stat Assoc ,  1958, vol.  53 (pg.  457- 481) Google Scholar CrossRef Search ADS   6 Cox DR.  Regression models and life-tables,  J R Stat Soc[R] ,  1972, vol.  34 (pg.  187- 202) 7 Ruttmann E,  Bonatti H,  Legit C, et al.  Severe endocarditis in transplant recipients—an epidemiologic study,  Transpl Int ,  2005, vol.  18 (pg.  690- 696) Google Scholar CrossRef Search ADS PubMed  8 Paterson DL,  Dominguez EA,  Chang FY, et al.  Infective endocarditis in solid organ transplant recipients,  Clin Infect Dis ,  1998, vol.  26 (pg.  689- 694) Google Scholar CrossRef Search ADS PubMed  9 Bishara J,  Robenshtok E,  Weinberger M, et al.  Infective endocarditis in renal transplant recipients,  Transpl Infect Dis ,  1999, vol.  1 (pg.  138- 143) Google Scholar CrossRef Search ADS PubMed  10 Abbott KC,  Duran M,  Hypolite I, et al.  Hospitalisations for bacterial endocarditis after renal transplantation in the United States,  J Nephrol ,  2001, vol.  14 (pg.  353- 360) Google Scholar PubMed  11 Chu VH,  Cabell CH,  Benjamin DKJr, et al.  Early predictors of in-hospital death in infective endocarditis,  Circulation ,  2004, vol.  109 (pg.  1745- 1749) Google Scholar CrossRef Search ADS PubMed  12 Hoen B,  Alla F,  Selton-Suty C, et al.  Changing profile of infective endocarditis: results of a 1-year survey in France,  JAMA ,  2002, vol.  288 (pg.  75- 81) Google Scholar CrossRef Search ADS PubMed  13 Kessler M,  Hoen B,  Mayeux D, et al.  Bacteraemia in patients on chronic hemodialysis. A multicenter prospective survey,  Nephron ,  1993, vol.  64 (pg.  95- 100) Google Scholar CrossRef Search ADS PubMed  14 Robinson DL,  Fowler VG,  Sexton DJ, et al.  Bacterial endocarditis in hemodialysis patients,  Am J Kidney Dis ,  1997, vol.  30 (pg.  521- 524) Google Scholar CrossRef Search ADS PubMed  15 Wolfe RA,  Ashby VB,  Milford EL, et al.  Comparison of mortality in all patients on dialysis, patients on dialysis awaiting transplantation, and recipients of a first cadaveric transplant,  N Engl J Med ,  1999, vol.  341 (pg.  1725- 1730) Google Scholar CrossRef Search ADS PubMed  © The Author [2008]. Published by Oxford University Press on behalf of ERA-EDTA. All rights reserved. For Permissions, please e-mail: journals.permissions@oxfordjournals.org Oxford University Press
TI - Outcomes of renal transplant and waiting list patients with bacterial endocarditis in the United States
JO - Nephrology Dialysis Transplantation
DO - 10.1093/ndt/gfn006
DA - 2008-02-26
UR - https://www.deepdyve.com/lp/oxford-university-press/outcomes-of-renal-transplant-and-waiting-list-patients-with-bacterial-g9vVsVCJvu
SP - 2381
EP - 2385
VL - 23
IS - 7
DP - DeepDyve
ER -