Healthcare Costs Associated with Hemodialysis Catheter–Related Infections: A Single‐Center Experience
In patients undergoing hemodialysis, catheter‐related bacteremia results in expensive hospitalizations. In our study, the mean cost was $23,451 per hospitalization. When itemized, housing (“bed‐related”) costs accounted for 66% of the total; laboratory costs accounted for 4%, radiologic costs accounted for 9%, and procedure‐related costs accounted for 21%. Hypoalbuminemia and bacteremia due to methicillin‐resistant Staphylococcus aureus (MRSA) are associated with higher healthcare costs; bacteremia due to MRSA is also associated with poor survival rates.
Received July 10, 2006; accepted August 4, 2006; electronically published April 12, 2007.
The Centers for Medicare and Medicaid Services launched the “Fistula First” National Vascular Access Improvement Initiative, which aims at placing arteriovenous fistulas (AVFs) in 50% of patients beginning hemodialysis and has a long‐range goal of maintaining AVFs in 40% of patients undergoing hemodialysis.1 In the United States, approximately one‐third of patients undergoing chronic hemodialysis currently undergo dialysis via an AVF, although a high proportion of patients undergoing chronic hemodialysis still rely on a catheter for vascular access.2 Rates of both AVF use and catheter use have increased over the past few years, and this parallel increase in catheter use has been attributed partly to the time required for AVF maturation. In addition to delivering relatively poor dialysis, hemodialysis catheters pose an enormous risk for catheter‐related bacteremia and subsequent infectious complications.3
Since the early 1990s, the number of vascular access procedures for hemodialysis and its associated cost have increased significantly.4 When vascular access procedures are compared, catheters are associated with the highest costs, whereas AVFs are the least likely to result in expensive complications.4 Patients who have hemodialysis catheters may have multiple medical problems that preclude creating primary or secondary AVFs, or may be using the catheter as the last option after the failure of multiple AVFs or arteriovenous grafts (AVGs). However, there is no doubt that the hospitalizations associated with catheter‐related bacteremia inflate total healthcare costs.
A recent report by Nisensson et al.5 examined the economic consequences of hospitalizations resulting from septicemia due to S. aureus, using claims data from the US Renal Data System. The authors derived the expenses on the basis of Medicare paid costs, and the study cohort was confined to patients with bacteremia due to S. aureus. Other studies have reported total healthcare costs for bacteremia due to methicillin‐sensitive and methicillin‐resistant strains of S. aureus.6‐8 In our study, we report the total and itemized healthcare costs at our facility that were associated with catheter‐related sepsis due to gram‐positive and gram‐negative bacteria. In addition, we investigated whether catheters are being used as a vascular access method to bridge the time prior to the maturation of an AVF.
Methods
In this retrospective study, we reviewed the medical records of all patients undergoing chronic hemodialysis who were admitted to our hospital between January and December 2005 and then identified the subset of patients admitted with a diagnosis of catheter‐related bacteremia, as defined by positive blood culture results for a hemodialysis patient with a tunneled catheter and in whom no other obvious source of infection was evident. Demographic information; clinical data; results of all laboratory, radiologic, and cardiac tests; and all surgical procedures that were done during the hospitalization for catheter‐related bacteremia were recorded and their costs calculated. Costs related to these tests and procedures were obtained from the hospital and a Decision Science Services database. The cost for each day of hospitalization was obtained from the hospital’s finance department. The patients were followed up after hospital discharge for a period of 3‐55 weeks, with death as the end point.
For defining whether catheters were the sole possibility for vascular access in a given patient, we divided the subjects into 5 groups. Group I was composed of patients for whom a primary AVF or AVG was not planned within 3 months of initiating hemodialysis. Group II was composed of patients awaiting maturation of an AVF or AVG as the primary site of vascular access. Group III comprised patients whose AVF and/or AVG had failed and whose physicians had made no plans for another arteriovenous conduit. Group IV comprised patients whose AVF and/or AVG had failed and were awaiting maturation of their next arteriovenous conduit. Group V was made up of patients who had catheters placed as the last option for vascular access after multiple AVFs and/or AVGs had failed.
Demographic and clinical data were presented as the mean ±SD or as the proportion of patients in each subgroup. The Fisher exact test, 2‐tailed Student t test, or Mann‐Whitney U test was used, as appropriate. Kaplan‐Meier survival curves were compared by the log‐rank test. A P value of .05 or less was considered statistically significant.
Results
During the 12‐month study period, our hospital admitted a total of 258 patients undergoing hemodialysis; in this group, 33 patients had 40 hospitalizations for catheter‐related bacteremia. The mean age of these patients was 62, and all of them were men; 64% were African Americans and 30% were white. Twenty‐one (64%) of these patients had a history of diabetes and 31 (94%) had a history of hypertension. The median duration of dialysis prior to hospitalization was 6.5 months (interquartile range, 3‐16 months). The mean interval between placement of the catheter and hospitalization for catheter‐related bacteremia was 113 days, and the median was 76 days (interquartile range, 28‐180 days). In our series of patients, S. aureus was the most common pathogen isolated (from 46% of patients), followed by gram‐negative organisms (from 28%), Staphylococcus epidermidis (from 25%), and Enterococcus species (from 10%). Three patients had polymicrobial blood cultures. Four patients (10%) died of severe sepsis during hospitalization, and 7 more patients died during the follow‐up period.
The use of a hemodialysis catheter as a means of vascular access to bridge the time during the process of AVF maturation (in patient groups I and IV) was seen in only 25% of patients. The use of a catheter as the last option for vascular access was noted for only 10% of patients. A significant proportion of patients (38%) had no treatment plan for a primary AVF and/or AVG, and 28% of patients had had a recent AVF and/or AVG fail and had no treatment plan for creating another AVF and/or AVG.
Patients with catheter‐related bacteremia accounted for a total of 693 hospital‐days, a significant proportion (14%) of which were spent in the intensive care unit. The mean length of stay was 17 days, and the median length of stay was 11 days. The mean healthcare cost per patient admitted was $23,451, and the median was $12,123. Sixty‐six percent of this cost was associated with housing (“bed related”); laboratory costs accounted for 4% of the total, radiologic costs accounted for 9%, and procedure‐related expenses accounted for 21%.
On further analysis, we found that patients with catheter‐related bacteremia who had very low serum albumin levels at the time of admission (2.5 g/dL or less) used more healthcare resources, compared with patients whose serum albumin levels were above 2.5 g/dL (Table), and the study subjects were divided accordingly into group A and group B. The 2 groups of patients were similar with regard to age, ethnicity, history of diabetes, the organism causing their infection, duration of dialysis, and duration of hemodialysis catheter placement. Similarly, the reasons for using a hemodialysis catheter were comparable in both groups. However, the median length of stay was significantly longer for group A, compared with group B (14 vs 8 days; 
), and the median hospitalization‐related costs were also higher ($16,084 vs $9,570; 
). Although not significantly different, the 1‐year mortality rate was higher in group A, compared with those in group B (54% vs 20%).
MRSA bacteremia was noted in 15 patients, and compared with bacteremia caused by another organism, it was associated with a trend toward prolonged hospital stay (median, 13 vs 9 days). It was also associated with a trend toward higher healthcare costs (average additional cost, $5,806 per patient admitted). Catheter‐related bacteremia due to MRSA was significantly associated with poor patient survival rates (Figure) on Kaplan‐Meier analysis (
, by log‐rank test).
Figure. Comparison of Kaplan‐Meier survival curves for patients undergoing hemodialysis who have methicillin‐resistant Staphylococcus aureus bacteremia, compared with those who have bacteremia caused by other organisms.
Discussion
Hemodialysis catheters pose a risk of bacteremia and infectious complications that are associated with high morbidity and mortality,3 as well as prolonged hospitalization. In this study, we have calculated the actual expenses incurred by a Veterans Affairs hospital for a patient admitted with catheter‐related bacteremia and itemized these expenditures in broad categories. We have also identified clinical factors that are associated with higher total healthcare costs and defined the reasons for using a hemodialysis catheter at the time of hospitalization for catheter‐related bacteremia. The annual healthcare costs associated with catheter‐related bacteremia admissions were close to $1 million, and our findings were consistent with prior reports.5,8 Costs associated with loss of life increase the total economic burden. Following the Centers for Disease Control and Prevention recommendations to maximize the use of fistulas and grafts9 will have significant financial impact, in addition to reducing mortality and the emergence of antimicrobial resistance.
The recent increase in tunneled catheter use has been partly attributed to the increase in fistula creation rates and the time required for maturation of an arteriovenous access site. Hence, it was presumed that catheters were increasingly being used as a “bridge” method for vascular access prior to the first cannulation of an AVF. However, our study indicates that three‐fourths of patients (75%) using a catheter did not have a maturing fistula, and a catheter was being used as the access site of last resort in only 10% of patients. It is important to question whether a catheter is the sole possibility for vascular access.
Previous studies have shown a positive correlation between low serum albumin levels and catheter‐related bacteremia.10 In addition, hypoalbuminemia11 and a time‐varying decline in serum albumin12 levels are associated with mortality in patients undergoing hemodialysis. In our study, low serum albumin levels were associated with prolonged hospitalization and increased healthcare costs. There was also a trend toward a higher mortality rate in this subgroup. It has become a common practice to start intravenous antibiotic therapy for every episode of fever in patients undergoing hemodialysis via catheters, on the presumption that there is arteriovenous access–related bacteremia. Injudicious use of antibiotics has clearly led to a recent increase in the proportion of cases of bacteremia caused by MRSA. Comparing the pattern of endocarditis in patients undergoing hemodialysis over the last 15 years, we recently demonstrated a microbiologic shift to drug‐resistant pathogens.13 The results of this study are consistent with prior observations6,14,15 regarding the association of MRSA bacteremia with higher mortality risk, longer hospitalization, and increased healthcare costs.
Our study has some inherent limitations. A large fraction of veterans undergo hemodialysis at private healthcare units and they are admitted to our facility only for inpatient care. It is impossible for us to calculate the incidence of catheter‐related sepsis per patient‐year or per catheter‐day on the basis of catheter‐related bacteremia admission data. Second, even though we have factored in the duration of dialysis and catheter placement, as well as the reasons for catheter use, other unknown inflammatory and noninflammatory medical conditions could have led to low serum albumin levels and possibly influenced the duration of stay and resultant expenditure. Third, all study subjects were men, and the results may be different when the sex distribution is different. Finally, the lack of statistical difference between groups could be caused by a type II error resulting from the small sample size.
To conclude, our study has confirmed the tremendous overall economic burden of catheter‐related bacteremia and specifically identified hypoalbuminemia and MRSA bacteremia as risk factors associated with higher healthcare costs.
Acknowledgments
Potential conflicts of interest. All authors report no conflicts of interest relevant to this article.
References
- 1. US Department of Health and Human Services, Centers for Medicare and Medicaid Services. CMS launches “fistula first” initiative to improve care and quality of life for hemodialysis patients. Available at: http://www.cms.hhs.gov/apps/media/press/release.asp?Counter=1007. Accessed July 4, 2006.
- 2. Pisoni RL, Young EW, Dykstra DM, et al. Vascular access use in Europe and the United States: results from the DOPPS. Kidney Int 2002; 61:305‐316.
- 3. Nori US, Manoharan A, Thornby JI, Yee J, Parasuraman R, Ramanathan V. Mortality risk factors in chronic haemodialysis patients with infective endocarditis. Nephrol Dial Transplant 2006; 21:2184‐2190.
- 4. National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases. USRDS 2005 Annual Data Report. Bethesda, MD. Available at: http://www.usrds.org/adr.htm. Accessed July 4, 2006.
- 5. Nissenson AR, Dylan ML, Griffiths RI, et al. Clinical and economic outcomes of Staphylococcus aureus septicemia in ESRD patients receiving hemodialysis. Am J Kidney Dis 2005; 46:301‐308.
- 6. Reed SD, Friedman JY, Engemann JJ, et al. Costs and outcomes among hemodialysis‐dependent patients with methicillin‐resistant or methicillin‐susceptible Staphylococcus aureus bacteremia. Infect Control Hosp Epidemiol 2005; 26:175‐183.
- 7. Engemann JJ, Carmeli Y, Cosgrove SE, et al. Adverse clinical and economic outcomes attributable to methicillin resistance among patients with Staphylococcus aureus surgical site infection. Clin Infect Dis 2003; 36:592‐598.
- 8. Engemann JJ, Friedman JY, Reed SD, et al. Clinical outcomes and costs due to Staphylococcus aureus bacteremia among patients receiving long‐term hemodialysis. Infect Control Hosp Epidemiol 2005; 26:534‐539.
- 9. Department of Health and Human Services. Centers for Disease Control and Prevention. 12 steps to prevent antimicrobial resistance among dialysis patients. Available at: http://www.cdc.gov/drugresistance/healthcare/dialysis/12steps_dialysis.htm. Accessed August 2, 2006.
- 10. Unver S, Atasoyu EM, Evrenkaya TR, Ardic N, Ozyurt M. Risk factors for the infections caused by temporary double‐lumen hemodialysis catheters. Arch Med Res 2006; 37:348‐352.
- 11. Cooper BA, Penne EL, Bartlett LH, Pollock CA. Protein malnutrition and hypoalbuminemia as predictors of vascular events and mortality in ESRD. Am J Kidney Dis 2004; 43:61‐66.
- 12. Kalantar‐Zadeh K, Kilpatrick RD, Kuwae N, et al. Revisiting mortality predictability of serum albumin in the dialysis population: time dependency, longitudinal changes and population‐attributable fraction. Nephrol Dial Transplant 2005; 20:1880‐1888.
- 13. Ramanathan V, Parasuraman R, Manoharan A, Nori US. Epidemiologic trend of infective endocarditis in hemodialysis patients. J Am Soc Nephrol 2004; 15:402A.
- 14. Kopp BJ, Nix DE, Armstrong EP. Clinical and economic analysis of methicillin‐susceptible and ‐resistant Staphylococcus aureus infections. Ann Pharmacother 2004; 38:1377‐1382.
- 15. Cosgrove SE, Qi Y, Kaye KS, Harbarth S, Karchmer AW, Carmeli Y. The impact of methicillin resistance in Staphylococcus aureus bacteremia on patient outcomes: mortality, length of stay, and hospital charges. Infect Control Hosp Epidemiol 2005; 26:166‐174.