Hemodialysis International 2015; 19:242–248
Catheter-related bacteremia and mortality in frequent nocturnal home hemodialysis Hui XUE,1,2 Nien-Chen LI,3 Eduardo LACSON, Jr,3 Steven M. BRUNELLI,4 Robert S. LOCKRIDGE5 1 Division of Nephrology, Department of Medicine, Kaiser Medical Center, San Diego, California, USA; Division of Hospital Medicine, Department of Medicine, University of California San Diego, San Diego, California, USA; 3Fresenius Medical Care, Waltham, Massachusetts, USA; 4Renal Division, Brigham and Women’s Hospital, Harvard Medical School, Boston, Massachusetts, USA; 5Lynchburg Nephrology Physicians, Lynchurg, Virginia, USA 2
Abstract Frequent nightly home hemodialysis (NHHD) has emerged as an attractive alternative to thrice weekly in-center hemodialysis, albeit with preponderant long-term hemodialysis catheter used. Sixty-three NHHD patients from University of Virginia Lynchburg Dialysis Facility were matched 1:2 with 121 conventional hemodialysis patients admitted to Fresenius Medical Care North America facilities from January 1, 2007 to December 31, 2010. Matching considered age (± 5 years), gender, race, dialysis vintage, and diabetes. The primary end-point was the combined incidence of bacteremia/sepsis, for up to 20 months or upon changing to a fistula/graft (with catheter removal), transferring to peritoneal dialysis (PD), or at the time of kidney transplant or death. No significant differences were observed in rate of fistula/graft conversion, transfer to PD, transplant, or death between NHHD and in-center hemodialysis (IHD) groups. For the first catheter used, the rate of catheter-related sepsis was not significantly different between the NHHD (1.77 per 100 patient months) and IHD (2.03 per 100 patient months; P = 0.21). Combining all catheters, the rate of bacteremia/sepsis per 100 patient months in the NHHD group was 1.51 and in the IHD group was 2.01 (P = 0.35). Median catheter lifespan for the first catheter was 5.6 (1.7∼19.0) for NHHD and 4.6 (2.7∼7.8) for the IHD group (P = 0.64), and for all catheters used was 5.2 (Q1∼Q3 = 1.5∼15.2) months in NHHD group, and 4.1 (2.0∼6.8) months in IHD group (P = 0.20). The rate of bacteremia and death is not different for up to 20 months in catheter users who dialyze via frequent NHHD vs. thrice weekly IHD. Key words: Nocturnal hemodialysis, home hemodialysis, catheter infections, catheter mortality, catheter survival
Correspondence to: H. Xue, MD, MMSC, Division of Nephrology, Department of Medicine, Kaiser Medical Center, 4510 Viewridge Ave, San Diego, CA 92123, U.S.A. E-mail: [email protected] Disclosures: As of this writing, authors N-CL and EL are employees of Fresenius Medical Care, North America. SMB is an employee of Davita and has served as an advisor to Amgen, C.B. Fleet Company and Proctor & Gamble. SMB has received speaking honoria from Fresenius Medical Care North America, and his spouse is employed by Astra Zeneca. RSL sits on the Machine Medical Advisory Board for Fresenius Medical Care North America (Sorbent Machine). HX has nothing to disclose. No external funding was used for this research.
© 2015 International Society for Hemodialysis DOI:10.1111/hdi.12245
242
Bacteremia in nightly home hemodialysis
INTRODUCTION In recent years, frequent nightly home hemodialysis (NHHD) has emerged as an alternative to thrice weekly in-center hemodialysis (IHD). Clinicians and investigators have postulated better survival with this modality due to increased clearance of uremic solutes and reduced peridialytic volume fluctuations. Observational studies of NHHD performed five to six times per week have shown markedly higher phosphorous clearance, reduced phosphate binder requirements,1–4 better nutritional status,4 health-related quality of life,2 sleep apnea,5 hypertension control,1,2,6,7 decreased left ventricular mass,1,2,6,7 and better survival compared with thrice weekly hemodialysis.8,9 In an observation study, the survival rate on NHHD was similar to that seen among deceased donor kidney transplant recipients.10 Two randomized trials to date have yielded opposing results. Culleton’s and colleagues11 showed improved blood pressure control, reduced antihypertensive medication requirement, and regression of left ventricular mass after only 6 months of being randomized to NHHD. The Frequent Hemodialysis Network (FHN) Nocturnal Trial confirmed better phosphate and hypertension control but did not show an improved survival or significant reduction of left ventricular mass.12 However, the latter has been criticized for failure to meet enrollment goals, inclusion of patients with significant residual kidney function, and poor compliance with assigned therapies.13,14 One concerning potential adverse sequela of NHHD is heightened risk for vascular access complications12 due to increased number and/or prolonged cannulations.15 Although arteriovenous fistula is the preferred dialysis access, a significant number of end-stage renal disease (ESRD) patients continue to dialyze via a tunneled central catheter, with 80% initiating dialysis with a catheter alone.16,17 Patients maintained on chronic hemodialysis via catheters having the highest risk for bacteremia, sepsis, hospitalization, and death18–25 compared with those dialyzing via arteriovenous fistula or grafts. Interestingly, 40% of FHN nocturnal patients dialyzed using a tunneled central catheter along with 50% of control patients.12 No conclusions could be drawn that increased frequency of use of a catheter resulted in more infections from the FHN nocturnal trial due to the small number of septic events that occurred during the study. Furthermore, in the largest and longest running NHHD program in the United States (3413 patient months consisting of over 12 years in 87 patients), patients used a dialysis catheter in 70% of all patient months.9 Until now, it was unknown if patients who received more frequent catheter manipulation while receiving NHHD would experience increased risk of
Hemodialysis International 2015; 19:242–248
bacteremic infection compared with those receiving thrice weekly IHD.
METHODS Matching Sixty-three hemodialysis patients from the NHHD program from 1997 to 2010 at University of Virginia Lynchburg Dialysis Facility in Virginia were matched to controls from 6285 IHD patients admitted to Fresenius Medical Care North America facilities in Virginia and surrounding states from January 1, 2007 to December 31, 2010. All patients consented for use of data for quality assessment and practice patterns at the time of HD enrollment at University of Virginia Lynchburg Dialysis Facility in Virginia and in all FMCNA facilities. This provider initiated that quality improvement study was deemed at the time of its conduct to be without any risk to subjects, covered by and consistent with quality assessment and improvement under treatment, payment and health care operations, and therefore did not require institutional review board approval or additional patient consent. Matching considered five variables: age (± 5 years), gender, race, dialysis vintage, and diabetes. Dialysis vintage was divided into seven categories: 1 day, >1 to 30 days, >1 to 3 months, >3 to 12 months, >1 to 2 years, >2 to 5 years, and >5 years. NHHD patients were matched to up to two eligible IHD patients on these bases. Fifty-eight NHHD patients were matched with 2 IHD patients while 5 NHHD patients could be matched with only 1 IHD patient, resulting in 63 NHHD patients and 121 IHD patients (2 × 58 + 1 × 5 = 121) included in the analysis. The primary end-point was the combined incidence of catheter-related bacteremia (i.e., positive blood culture) and/or a clinical diagnosis of sepsis (primary ICD-9 codes 790.7 and 995.91 and other related codes for sepsis), from a hospitalization episode. Follow-up was censored at 20 months for the first catheter because there were only 3 catheters left in place in the control group at 20 months.
Catheter lifespan Because only 2 out of 166 catheter episodes in the IHD group spanned 21 months or longer, all catheter lifespan in this study was cut off or censored at 20 months.
Analysis The baseline characteristic of the NHHD and IHD groups was compared using two-sided t tests for continuous vari-
243
Xue et al.
any catheter present for 21 months or longer were censored. NHHD and IHD groups were similar in the patient distribution in age, dialysis vintage, gender, race, and diabetes status (i.e., matching variables). Mean age was 52.8 (± 14.7), mean dialysis vintage was 38.5 (± 49.6) months, 57.1% were male, 57.1% were Caucasian, and 20.6% had diabetes (Table 1). Three additional covariates evaluated were hypertension, GN, and polycystic kidney disease as primary cause of ESRD. Hypertension was more common among patients in the IHD group (41.3%) compared with the NHHD group (20.6%; P = 0.005). There were 19.0% of NHHD patients that had GN compared with only 9.1% of IHD patients (P = 0.05). There was no significant difference in the frequency of polycystic kidney disease between groups. For the first catheter used, the rate of catheter-related sepsis was not significantly different between the NHHD group (1.77 per 100 patient months) and IHD group (2.03 per 100 patient months; P = 0.21) (Table 2), during the observation period. Censoring due to conversion to a permanent dialysis access, transfer to PD, transplant, and death was not statistically different between the two groups. Median catheter life for first catheter was 5.6 (Q1∼Q3 = 1.7∼19.0) months in the NHHD group compared with 4.6 (2.7∼7.8) months in the IHD group, not significantly different between these two (P = 0.64). Combining all catheters used during the study period, a total of 192 catheters were used in the NHHD group and 166 catheters in the IHD group. The rate of bacteremia/ sepsis in the NHHD group was 1.51 per 100 patient
ables and chi-square tests for categorical variables. The rate and number of catheter bacteremia/sepsis in each study cohort was monitored for 20 months. Two approaches for tracking these episodes were used; first, using only the first catheter of each patient, and second, including also any subsequent catheters for each patient during the study period. Patients were censored upon changing to a fistula/ graft (with catheter removal), transferring to peritoneal dialysis (PD), or at the time of kidney transplant. Because of skewed distribution, catheter lifespan was compared between NHHD and IHD groups using median and interquartile range, and tested with two-sample median test. The “time to bacteremia/sepsis” event was depicted using Kaplan–Meier survival curves. The ratio of hazard rates for these events for NHHD vs. IHD was calculated using Cox proportional hazard regression models, both as unadjusted models and with further adjustment for residual differences in age, gender, race, vintage, diabetes, as well as for primary cause of ESRD (including diabetes, hypertension, glomerulonephritis [GN], and polycystic kidney disease [PKD]). Sensitivity analyses that did not censor follow-up at 20 months yielded qualitatively similar results. Analyses were performed with SAS version 9.1 or higher (http:// www.sas.com).
RESULTS Sixty-three NHHD patients were matched with 121 IHD patients, with both groups followed for 20 months, and
Table 1 Demographics of NHHD and matched IHD patients Study group Variable
NHHD (n = 63)
Age (in years) Vintage (in months) Gender Race Diabetes Hypertensive GN PKD
IHD (n = 121)
Mean (SD) 52.8 (14.7) 38.5 (49.6) Male Female White Black Yes Yes Yes Yes
Mean (SD) 53.8 (14.1) 37.1 (46.1)
N
%
N
%
36 27 36 27 13 13 12 4
57.1 42.9 57.1 42.9 20.6 20.6 19.0 6.3
70 51 69 52 26 50 11 4
57.9 42.1 57.0 43.0 21.5 41.3 9.1 3.3
P value 0.65 0.84 0.93 0.99 0.89 0.005 0.0526 0.34
Subjects were matched for age (± 5 years), gender, race, dialysis vintage, and diabetes. GN = glomerulonephritis; IHD = in-center hemodialysis; NHHD = nightly home hemodialysis; PKD = Polycystic Kidney Disease; SD = standard deviation.
244
Hemodialysis International 2015; 19:242–248
Bacteremia in nightly home hemodialysis
Table 2 Disposition and event rate for first catheter treatment First initial catheter studied NHHD (n = 63) Disposition event
n
Sepsis Fistula/graft Transfer to PD Transplant Death None above Catheter duration in months without 20-month censor
10 0 0 3 0 50
%
IHD (n = 121)
Event rate
15.9 1.77 0 0 0 0 4.8 0.53 0 0.00 79.4 Median (Q1∼Q3) 5.6 (1.7∼19.0)
n 14 4 8 0 3 94
%
Event rate
11.6 2.03 3.3 2.53 6.6 1.14 0 0 2.5 0.40 77.7 Median (Q1∼Q3) 4.6 (2.7∼7.8)
P value 0.21 0.96 0.96 0.94 0.96 0.31 0.64
Note: All events after 20 months since the start of catheter are censored. Event rate is per 100 patient months. IHD = in-center hemodialysis; NHHD = nightly home hemodialysis; PD = peritoneal dialysis.
months, and the rate in IHD group was 2.01 per 100 patient months (P = 0.35) (Table 3). No significant differences were observed in rate of fistula/graft conversion, transfer to PD, transplant, or death between the two groups. The overall disposition of patients indicated very low death rates (2.1% vs. 1.8% for NHHD and IHD, respectively), very few transplants (none for IHD), and very few transfers to PD (none for NHHD). Kaplan–Meier curves of time to catheter-related bacteremia/sepsis for the first catheter is shown in Figure 1, with no significant difference between NHHD vs. IHD cohorts (Logrank P = 0.98). The drastic reduction in number of patients at risk for event by month 5 is due to short median survival time of 5.6 months (1.7∼19.0) for NHHD and 4.6 months (2.7∼7.8) for IHD group (Table 2). A similar Kaplan–Meier graph for time to catheter-related bacteremia/sepsis including all sequential catheters used
during 20 months of observation is shown in Figure 2 (P = 0.50), indicating that NHHD and IHD curves were not significantly different. Figure 3 shows adjusted hazard ratios (95% confidence interval) for catheter-related bacteremia/sepsis during catheter treatment for NHHD vs. IHD (reference) groups; no clinically meaningful or statistically significant difference was noted. The models were further adjusted for residual difference in age, hypertension and GN, in addition to the initial matching for gender, race, vintage, and diabetes status.
DISCUSSION Patients on maintenance hemodialysis in the United States have suffered from annual mortality rates of ∼18–20% for two decades.16,17 Although the survival benefit of frequent nocturnal hemodialysis is debated, our work has shown
Table 3 Disposition and event rate for all catheter treatments Study group NHHD (n = 192) Disposition event Sepsis Fistula/graft Transfer to PD Transplant Death None above Catheter duration in months without 20-month censor
n 23 1 0 10 4 155
%
Event rate
12.0 1.51 0.5 0.06 0 0 5.2 0.65 2.1 0.26 80.7 Median (Q1∼Q3) 5.2 (1.5∼15.2)
IHD (n = 166) n 17 6 8 0 3 134
%
Event rate
10.2 2.01 3.6 0.66 4.8 0.93 0 0 1.8 0.33 80.7 Median (Q1∼Q3) 4.1 (2.0∼6.8)
P value 0.35 0.08 0.95 0.95 0.99 0.50 0.20
Note: All events after 20 months since the start of catheter were censored. Event rate is per 100 patient months. IHD = in-center hemodialysis; NHHD = nightly home hemodialysis; PD = peritoneal dialysis.
Hemodialysis International 2015; 19:242–248
245
Survival curves
Xue et al.
3.0
1 0.9 0.8 0.7 0.6 0.5 0.4
2.5 2.0
0
5
10 Months to sepsis IHD
15
20
H 1.5 R 1.0
1.01
0.99
0.83
0.80
0.5
NHHD
0.0 Unadjusted
Number of patients at risk
Adjusted*
First catheter
Months
0
5
10
15
20
Total
IHD % NHHD %
121 56 63 39
61 28 35 22
22 10 25 16
9 4 22 14
3 1 16 10
216 161
Unadjusted
Adjusted*
All catheters
Sample type and model adjustment
Figure 3 Hazard ratio (95% confidence interval) for catheterrelated bacteremia/sepsis during catheter treatment for nightly home hemodialysis vs. in-center hemodialysis group. *Adjusted by age hypertension and glomerulonephritis.
IHD = in-center hemodialysis; NHHD = nightly home hemodialysis.
Figure 1 Kaplan–Meier curves for time to bacteremia/sepsis for each patient’s first catheter. IHD = in-center hemodialysis; NHHD = nightly home hemodialysis.
that for catheter-dependent patients, frequent catheter use does not predispose these NHHD-treated patients to increase infection risks or death as compared with the IHD patients. This study found that there were no significant differences in the bacteremia/sepsis rate for HD patients treated
Survival curves
1 0.9 0.8 0.7 0.6 0.5 0.4 0
5
10 Months to sepsis IHD
15
20
NHHD
Number of catheter administrations at risk Months
0
5
10
15
20
Total
IHD % NHHD %
166 59 192 42
76 27 105 23
24 9 70 15
10 4 51 11
3 1 37 8
279 455
IHD = in-center hemodialysis; NHHD = nightly home hemodialysis.
Figure 2 Kaplan–Meier curves for time to bacteremia/sepsis for all catheters used in 20 months. IHD = in-center hemodialysis; NHHD = nightly home hemodialysis.
246
using a central venous catheter between selected patients on nocturnal home HD and a matched cohort of in-center patients treated three times per week. However, both groups have event rates that are slightly lower than those reported by the Center for Disease Control National Healthcare Safety Network for 2007–2008 at 1.5 to 2.0 vs. 3.2/100 catheter using patient months,26 likely due to patient selection, since both NHHD and IHD study cohorts were both younger, and had less DM burden than the general IHD population. The overall disposition of patients indicated very low death rates, very few transplants, and very few transfers to PD. The limitations of this study include NHHD participants only from a single center, even though it is the largest and longest existing program in the United States. Additionally, its observational study design does not address mechanisms that result in the observed differences and is subject to residual confounding. A list of residual confounding factors include limited amount of laboratory values available in the data set, lack of socioeconomic indicator, and limited adjustment for comorbidities. Since the control subjects for the NHHD cohort was selected through matching for each of the five categories: age, gender, race, dialysis vintage, and diabetes, the matching minimizes some of the difference between the two comparison groups. However, one example stems from Lockridge and Kjellstrand9 who previously reported that education levels below graduation from high school was associated with a 6.53 hazard ratio for mortality for patients on NHHD, more significant than dialysis vintage, baseline renal disease, presence of congestive heart failure, and the sum of comorbidities. Unfortunately, education level was not available in this retrospective cohort analysis, therefore offering direction for future studies. An
Hemodialysis International 2015; 19:242–248
Bacteremia in nightly home hemodialysis
additional limitation is the 20-month study period cut-off, which may miss event capture. Lastly, there may be greater chance of incomplete hospitalization diagnosis data collection for the control IHD group, which was selected from 6285 patients with data captured electronically compared with the more focused manual data collection for the 63 NHHD patients. The potential for incomplete documentation of bacteremia/sepsis episodes in IHD patients could render the findings more in favor of NHHD; therefore, the observed null difference between the two groups may be a conservative estimate of truth. In an era with 27.7% of all prevalent ESRD patients still dialyzing via catheter alone,16 this study has shown that the rate of bacteremia and death is not different for catheter users who dialyze via frequent NHHD vs. thrice weekly IHD. Although Fistula First initiative continues to succeed in increasing the number of fistula users, there will always be a subset of patients who lack suitable blood vessels for arteriovenous access placement, cannot tolerate the physiologic shunt of a fistula/graft, and/or have emotional/psychological resistance to conversion to a permanent arteriovenous vascular access. In addition, patient compliance with physician and caregiver instructions and adherence to therapy remains key factors influencing survival that current studies have found difficulty accounting for. NHHD offers the advantage of being dialyzed at home and may appeal to more “independent” patients and/or very involved caregivers, thereby potentially increasing treatment compliance. It may also influence more careful attention to catheter care and preservation. Although future studies are needed to elucidate if NHHD offers survival advantage compared with IHD, we posit that for ESRD patients, in particular the catheter committed subset, NHHD is a viable renal replacement modality with the potential for improved catheter lifespan, improved phosphate and hypertension control,1,12 and in the hands of the country’s most durable NHHD program, come with at least comparable, if not potentially better infection risks than a matched cohort on IHD.
Manuscript received January 2014; revised October 2014.
REFERENCES 1 Pierratos A. Nocturnal hemodialysis: Dialysis for the new millennium. CMAJ. 1999; 161:1137. 2 Lockridge RS Jr, Spencer M, Craft V, et al. Nightly home hemodialysis: Five and one-half years of experience in Lynchburg, Virginia. Hemodial Int. 2004; 8:61–69.
Hemodialysis International 2015; 19:242–248
3 Mucsi I, Hercz G, Uldall R, Ouwendyk M, Francoeur R, Pierratos A. Control of serum phosphate without any phosphate binders in patients treated with nocturnal hemodialysis. Kidney Int. 1998; 53:1399–1404. 4 Sikkes ME, Kooistra MP, Weijs PJ. Improved nutrition after conversion to nocturnal home hemodialysis. J Ren Nutr. 2009; 19:494–499. 5 Hanly PJ, Pierratos A. Improvement of sleep apnea in patients with chronic renal failure who undergo nocturnal hemodialysis. N Engl J Med. 2001; 344:102– 107. 6 Chan CT, Floras JS, Miller JA, Richardson RM, Pierratos A. Regression of left ventricular hypertrophy after conversion to nocturnal hemodialysis. Kidney Int. 2002; 61:2235–2239. 7 Chan CT, Jain V, Picton P, Pierratos A, Floras JS. Nocturnal hemodialysis increases arterial baroreflex sensitivity and compliance and normalizes blood pressure of hypertensive patients with end-stage renal disease. Kidney Int. 2005; 68:338–344. 8 Johansen KL, Zhang R, Huang Y, et al. Survival and hospitalization among patients using nocturnal and short daily compared to conventional hemodialysis: A USRDS study. Kidney Int. 2009; 76:984–990. 9 Lockridge RS, Kjellstrand CM. Nightly home hemodialysis: Outcome and factors associated with survival. Hemodial Int. 2011; 15:211–218. 10 Pauly RP, Gill JS, Rose CL, et al. Survival among nocturnal home haemodialysis patients compared to kidney transplant recipients. Nephrol Dial Transplant. 2009; 24:2915–2919. 11 Culleton BF, Walsh M, Klarenbach SW, et al. Effect of frequent nocturnal hemodialysis vs conventional hemodialysis on left ventricular mass and quality of life: A randomized controlled trial. JAMA. 2007; 298:1291– 1299. 12 Rocco MV, Lockridge RS Jr, Beck GJ, et al. The effects of frequent nocturnal home hemodialysis: The Frequent Hemodialysis Network Nocturnal Trial. Kidney Int. 2011; 80:1080–1091. 13 Stokes JB. Nocturnal hemodialysis: Analysis following the Frequent Hemodialysis Network trial. Semin Dial. 2011; 24:614–620. 14 Misra M, Twardowski ZJ. Benefits of frequent nocturnal home hemodialysis. Kidney Int. 2012; 82:114–115, author reply 5. 15 Suri RS, Larive B, Sherer S, et al. Risk of vascular access complications with frequent hemodialysis. J Am Soc Nephrol. 2013; 24:498–505. 16 USRDS. 2012 Annual Data Report: Atlas of Chronic Kidney Disease and End-Stage Renal Disease in the United States. Bethesda (MD): National Institute of Health; 2012. 17 USRDS. 2013 Annual Data Report: Atlas of Chronic Kidney Disease and End-Stage Renal Disease in the United States. Bethesda (MD): National Institute of Health; 2013.
247
Xue et al.
18 Bradbury BD, Fissell RB, Albert JM, et al. Predictors of early mortality among incident US hemodialysis patients in the Dialysis Outcomes and Practice Patterns Study (DOPPS). Clin J Am Soc Nephrol. 2007; 2:89–99. 19 Lacson E Jr, Wang W, Lazarus JM, Hakim RM. Change in vascular access and hospitalization risk in long-term hemodialysis patients. Clin J Am Soc Nephrol. 2010; 5:1996–2003. 20 Ishani A, Collins AJ, Herzog CA, Foley RN. Septicemia, access and cardiovascular disease in dialysis patients: The USRDS Wave 2 study. Kidney Int. 2005; 68:311–318. 21 Astor BC, Eustace JA, Powe NR, Klag MJ, Fink NE, Coresh J. Type of vascular access and survival among incident hemodialysis patients: The Choices for Healthy Outcomes in Caring for ESRD (CHOICE) Study. J Am Soc Nephrol. 2005; 16:1449–1455.
248
22 Gulati S, Sahu KM, Avula S, Sharma RK, Ayyagiri A, Pandey CM. Role of vascular access as a risk factor for infections in hemodialysis. Ren Fail. 2003; 25:967– 973. 23 Dhingra RK, Young EW, Hulbert-Shearon TE, Leavey SF, Port FK. Type of vascular access and mortality in U.S. hemodialysis patients. Kidney Int. 2001; 60:1443–1451. 24 Mazonakis E, Stirling C, Booth KL, McClenahan J, Heron N, Geddes CC. The influence of comorbidity on the risk of access-related bacteremia in chronic hemodialysis patients. Hemodial Int. 2009; 13:6–10. 25 Lacson E Jr, Wang W, Lazarus JM, Hakim RM. Change in vascular access and mortality in maintenance hemodialysis patients. Am J Kidney Dis. 2009; 54:912–921. 26 National Healthcare Safety Network–Dialysis Surveillance Report. Atlanta (GA): Center for Disease Control; 2008.
Hemodialysis International 2015; 19:242–248
Copyright of Hemodialysis International is the property of Wiley-Blackwell and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.
Catheter-related bacteremia and mortality in frequent nocturnal home hemodialysis Hui XUE,1,2 Nien-Chen LI,3 Eduardo LACSON, Jr,3 Steven M. BRUNELLI,4 Robert S. LOCKRIDGE5 1 Division of Nephrology, Department of Medicine, Kaiser Medical Center, San Diego, California, USA; Division of Hospital Medicine, Department of Medicine, University of California San Diego, San Diego, California, USA; 3Fresenius Medical Care, Waltham, Massachusetts, USA; 4Renal Division, Brigham and Women’s Hospital, Harvard Medical School, Boston, Massachusetts, USA; 5Lynchburg Nephrology Physicians, Lynchurg, Virginia, USA 2
Abstract Frequent nightly home hemodialysis (NHHD) has emerged as an attractive alternative to thrice weekly in-center hemodialysis, albeit with preponderant long-term hemodialysis catheter used. Sixty-three NHHD patients from University of Virginia Lynchburg Dialysis Facility were matched 1:2 with 121 conventional hemodialysis patients admitted to Fresenius Medical Care North America facilities from January 1, 2007 to December 31, 2010. Matching considered age (± 5 years), gender, race, dialysis vintage, and diabetes. The primary end-point was the combined incidence of bacteremia/sepsis, for up to 20 months or upon changing to a fistula/graft (with catheter removal), transferring to peritoneal dialysis (PD), or at the time of kidney transplant or death. No significant differences were observed in rate of fistula/graft conversion, transfer to PD, transplant, or death between NHHD and in-center hemodialysis (IHD) groups. For the first catheter used, the rate of catheter-related sepsis was not significantly different between the NHHD (1.77 per 100 patient months) and IHD (2.03 per 100 patient months; P = 0.21). Combining all catheters, the rate of bacteremia/sepsis per 100 patient months in the NHHD group was 1.51 and in the IHD group was 2.01 (P = 0.35). Median catheter lifespan for the first catheter was 5.6 (1.7∼19.0) for NHHD and 4.6 (2.7∼7.8) for the IHD group (P = 0.64), and for all catheters used was 5.2 (Q1∼Q3 = 1.5∼15.2) months in NHHD group, and 4.1 (2.0∼6.8) months in IHD group (P = 0.20). The rate of bacteremia and death is not different for up to 20 months in catheter users who dialyze via frequent NHHD vs. thrice weekly IHD. Key words: Nocturnal hemodialysis, home hemodialysis, catheter infections, catheter mortality, catheter survival
Correspondence to: H. Xue, MD, MMSC, Division of Nephrology, Department of Medicine, Kaiser Medical Center, 4510 Viewridge Ave, San Diego, CA 92123, U.S.A. E-mail: [email protected] Disclosures: As of this writing, authors N-CL and EL are employees of Fresenius Medical Care, North America. SMB is an employee of Davita and has served as an advisor to Amgen, C.B. Fleet Company and Proctor & Gamble. SMB has received speaking honoria from Fresenius Medical Care North America, and his spouse is employed by Astra Zeneca. RSL sits on the Machine Medical Advisory Board for Fresenius Medical Care North America (Sorbent Machine). HX has nothing to disclose. No external funding was used for this research.
© 2015 International Society for Hemodialysis DOI:10.1111/hdi.12245
242
Bacteremia in nightly home hemodialysis
INTRODUCTION In recent years, frequent nightly home hemodialysis (NHHD) has emerged as an alternative to thrice weekly in-center hemodialysis (IHD). Clinicians and investigators have postulated better survival with this modality due to increased clearance of uremic solutes and reduced peridialytic volume fluctuations. Observational studies of NHHD performed five to six times per week have shown markedly higher phosphorous clearance, reduced phosphate binder requirements,1–4 better nutritional status,4 health-related quality of life,2 sleep apnea,5 hypertension control,1,2,6,7 decreased left ventricular mass,1,2,6,7 and better survival compared with thrice weekly hemodialysis.8,9 In an observation study, the survival rate on NHHD was similar to that seen among deceased donor kidney transplant recipients.10 Two randomized trials to date have yielded opposing results. Culleton’s and colleagues11 showed improved blood pressure control, reduced antihypertensive medication requirement, and regression of left ventricular mass after only 6 months of being randomized to NHHD. The Frequent Hemodialysis Network (FHN) Nocturnal Trial confirmed better phosphate and hypertension control but did not show an improved survival or significant reduction of left ventricular mass.12 However, the latter has been criticized for failure to meet enrollment goals, inclusion of patients with significant residual kidney function, and poor compliance with assigned therapies.13,14 One concerning potential adverse sequela of NHHD is heightened risk for vascular access complications12 due to increased number and/or prolonged cannulations.15 Although arteriovenous fistula is the preferred dialysis access, a significant number of end-stage renal disease (ESRD) patients continue to dialyze via a tunneled central catheter, with 80% initiating dialysis with a catheter alone.16,17 Patients maintained on chronic hemodialysis via catheters having the highest risk for bacteremia, sepsis, hospitalization, and death18–25 compared with those dialyzing via arteriovenous fistula or grafts. Interestingly, 40% of FHN nocturnal patients dialyzed using a tunneled central catheter along with 50% of control patients.12 No conclusions could be drawn that increased frequency of use of a catheter resulted in more infections from the FHN nocturnal trial due to the small number of septic events that occurred during the study. Furthermore, in the largest and longest running NHHD program in the United States (3413 patient months consisting of over 12 years in 87 patients), patients used a dialysis catheter in 70% of all patient months.9 Until now, it was unknown if patients who received more frequent catheter manipulation while receiving NHHD would experience increased risk of
Hemodialysis International 2015; 19:242–248
bacteremic infection compared with those receiving thrice weekly IHD.
METHODS Matching Sixty-three hemodialysis patients from the NHHD program from 1997 to 2010 at University of Virginia Lynchburg Dialysis Facility in Virginia were matched to controls from 6285 IHD patients admitted to Fresenius Medical Care North America facilities in Virginia and surrounding states from January 1, 2007 to December 31, 2010. All patients consented for use of data for quality assessment and practice patterns at the time of HD enrollment at University of Virginia Lynchburg Dialysis Facility in Virginia and in all FMCNA facilities. This provider initiated that quality improvement study was deemed at the time of its conduct to be without any risk to subjects, covered by and consistent with quality assessment and improvement under treatment, payment and health care operations, and therefore did not require institutional review board approval or additional patient consent. Matching considered five variables: age (± 5 years), gender, race, dialysis vintage, and diabetes. Dialysis vintage was divided into seven categories: 1 day, >1 to 30 days, >1 to 3 months, >3 to 12 months, >1 to 2 years, >2 to 5 years, and >5 years. NHHD patients were matched to up to two eligible IHD patients on these bases. Fifty-eight NHHD patients were matched with 2 IHD patients while 5 NHHD patients could be matched with only 1 IHD patient, resulting in 63 NHHD patients and 121 IHD patients (2 × 58 + 1 × 5 = 121) included in the analysis. The primary end-point was the combined incidence of catheter-related bacteremia (i.e., positive blood culture) and/or a clinical diagnosis of sepsis (primary ICD-9 codes 790.7 and 995.91 and other related codes for sepsis), from a hospitalization episode. Follow-up was censored at 20 months for the first catheter because there were only 3 catheters left in place in the control group at 20 months.
Catheter lifespan Because only 2 out of 166 catheter episodes in the IHD group spanned 21 months or longer, all catheter lifespan in this study was cut off or censored at 20 months.
Analysis The baseline characteristic of the NHHD and IHD groups was compared using two-sided t tests for continuous vari-
243
Xue et al.
any catheter present for 21 months or longer were censored. NHHD and IHD groups were similar in the patient distribution in age, dialysis vintage, gender, race, and diabetes status (i.e., matching variables). Mean age was 52.8 (± 14.7), mean dialysis vintage was 38.5 (± 49.6) months, 57.1% were male, 57.1% were Caucasian, and 20.6% had diabetes (Table 1). Three additional covariates evaluated were hypertension, GN, and polycystic kidney disease as primary cause of ESRD. Hypertension was more common among patients in the IHD group (41.3%) compared with the NHHD group (20.6%; P = 0.005). There were 19.0% of NHHD patients that had GN compared with only 9.1% of IHD patients (P = 0.05). There was no significant difference in the frequency of polycystic kidney disease between groups. For the first catheter used, the rate of catheter-related sepsis was not significantly different between the NHHD group (1.77 per 100 patient months) and IHD group (2.03 per 100 patient months; P = 0.21) (Table 2), during the observation period. Censoring due to conversion to a permanent dialysis access, transfer to PD, transplant, and death was not statistically different between the two groups. Median catheter life for first catheter was 5.6 (Q1∼Q3 = 1.7∼19.0) months in the NHHD group compared with 4.6 (2.7∼7.8) months in the IHD group, not significantly different between these two (P = 0.64). Combining all catheters used during the study period, a total of 192 catheters were used in the NHHD group and 166 catheters in the IHD group. The rate of bacteremia/ sepsis in the NHHD group was 1.51 per 100 patient
ables and chi-square tests for categorical variables. The rate and number of catheter bacteremia/sepsis in each study cohort was monitored for 20 months. Two approaches for tracking these episodes were used; first, using only the first catheter of each patient, and second, including also any subsequent catheters for each patient during the study period. Patients were censored upon changing to a fistula/ graft (with catheter removal), transferring to peritoneal dialysis (PD), or at the time of kidney transplant. Because of skewed distribution, catheter lifespan was compared between NHHD and IHD groups using median and interquartile range, and tested with two-sample median test. The “time to bacteremia/sepsis” event was depicted using Kaplan–Meier survival curves. The ratio of hazard rates for these events for NHHD vs. IHD was calculated using Cox proportional hazard regression models, both as unadjusted models and with further adjustment for residual differences in age, gender, race, vintage, diabetes, as well as for primary cause of ESRD (including diabetes, hypertension, glomerulonephritis [GN], and polycystic kidney disease [PKD]). Sensitivity analyses that did not censor follow-up at 20 months yielded qualitatively similar results. Analyses were performed with SAS version 9.1 or higher (http:// www.sas.com).
RESULTS Sixty-three NHHD patients were matched with 121 IHD patients, with both groups followed for 20 months, and
Table 1 Demographics of NHHD and matched IHD patients Study group Variable
NHHD (n = 63)
Age (in years) Vintage (in months) Gender Race Diabetes Hypertensive GN PKD
IHD (n = 121)
Mean (SD) 52.8 (14.7) 38.5 (49.6) Male Female White Black Yes Yes Yes Yes
Mean (SD) 53.8 (14.1) 37.1 (46.1)
N
%
N
%
36 27 36 27 13 13 12 4
57.1 42.9 57.1 42.9 20.6 20.6 19.0 6.3
70 51 69 52 26 50 11 4
57.9 42.1 57.0 43.0 21.5 41.3 9.1 3.3
P value 0.65 0.84 0.93 0.99 0.89 0.005 0.0526 0.34
Subjects were matched for age (± 5 years), gender, race, dialysis vintage, and diabetes. GN = glomerulonephritis; IHD = in-center hemodialysis; NHHD = nightly home hemodialysis; PKD = Polycystic Kidney Disease; SD = standard deviation.
244
Hemodialysis International 2015; 19:242–248
Bacteremia in nightly home hemodialysis
Table 2 Disposition and event rate for first catheter treatment First initial catheter studied NHHD (n = 63) Disposition event
n
Sepsis Fistula/graft Transfer to PD Transplant Death None above Catheter duration in months without 20-month censor
10 0 0 3 0 50
%
IHD (n = 121)
Event rate
15.9 1.77 0 0 0 0 4.8 0.53 0 0.00 79.4 Median (Q1∼Q3) 5.6 (1.7∼19.0)
n 14 4 8 0 3 94
%
Event rate
11.6 2.03 3.3 2.53 6.6 1.14 0 0 2.5 0.40 77.7 Median (Q1∼Q3) 4.6 (2.7∼7.8)
P value 0.21 0.96 0.96 0.94 0.96 0.31 0.64
Note: All events after 20 months since the start of catheter are censored. Event rate is per 100 patient months. IHD = in-center hemodialysis; NHHD = nightly home hemodialysis; PD = peritoneal dialysis.
months, and the rate in IHD group was 2.01 per 100 patient months (P = 0.35) (Table 3). No significant differences were observed in rate of fistula/graft conversion, transfer to PD, transplant, or death between the two groups. The overall disposition of patients indicated very low death rates (2.1% vs. 1.8% for NHHD and IHD, respectively), very few transplants (none for IHD), and very few transfers to PD (none for NHHD). Kaplan–Meier curves of time to catheter-related bacteremia/sepsis for the first catheter is shown in Figure 1, with no significant difference between NHHD vs. IHD cohorts (Logrank P = 0.98). The drastic reduction in number of patients at risk for event by month 5 is due to short median survival time of 5.6 months (1.7∼19.0) for NHHD and 4.6 months (2.7∼7.8) for IHD group (Table 2). A similar Kaplan–Meier graph for time to catheter-related bacteremia/sepsis including all sequential catheters used
during 20 months of observation is shown in Figure 2 (P = 0.50), indicating that NHHD and IHD curves were not significantly different. Figure 3 shows adjusted hazard ratios (95% confidence interval) for catheter-related bacteremia/sepsis during catheter treatment for NHHD vs. IHD (reference) groups; no clinically meaningful or statistically significant difference was noted. The models were further adjusted for residual difference in age, hypertension and GN, in addition to the initial matching for gender, race, vintage, and diabetes status.
DISCUSSION Patients on maintenance hemodialysis in the United States have suffered from annual mortality rates of ∼18–20% for two decades.16,17 Although the survival benefit of frequent nocturnal hemodialysis is debated, our work has shown
Table 3 Disposition and event rate for all catheter treatments Study group NHHD (n = 192) Disposition event Sepsis Fistula/graft Transfer to PD Transplant Death None above Catheter duration in months without 20-month censor
n 23 1 0 10 4 155
%
Event rate
12.0 1.51 0.5 0.06 0 0 5.2 0.65 2.1 0.26 80.7 Median (Q1∼Q3) 5.2 (1.5∼15.2)
IHD (n = 166) n 17 6 8 0 3 134
%
Event rate
10.2 2.01 3.6 0.66 4.8 0.93 0 0 1.8 0.33 80.7 Median (Q1∼Q3) 4.1 (2.0∼6.8)
P value 0.35 0.08 0.95 0.95 0.99 0.50 0.20
Note: All events after 20 months since the start of catheter were censored. Event rate is per 100 patient months. IHD = in-center hemodialysis; NHHD = nightly home hemodialysis; PD = peritoneal dialysis.
Hemodialysis International 2015; 19:242–248
245
Survival curves
Xue et al.
3.0
1 0.9 0.8 0.7 0.6 0.5 0.4
2.5 2.0
0
5
10 Months to sepsis IHD
15
20
H 1.5 R 1.0
1.01
0.99
0.83
0.80
0.5
NHHD
0.0 Unadjusted
Number of patients at risk
Adjusted*
First catheter
Months
0
5
10
15
20
Total
IHD % NHHD %
121 56 63 39
61 28 35 22
22 10 25 16
9 4 22 14
3 1 16 10
216 161
Unadjusted
Adjusted*
All catheters
Sample type and model adjustment
Figure 3 Hazard ratio (95% confidence interval) for catheterrelated bacteremia/sepsis during catheter treatment for nightly home hemodialysis vs. in-center hemodialysis group. *Adjusted by age hypertension and glomerulonephritis.
IHD = in-center hemodialysis; NHHD = nightly home hemodialysis.
Figure 1 Kaplan–Meier curves for time to bacteremia/sepsis for each patient’s first catheter. IHD = in-center hemodialysis; NHHD = nightly home hemodialysis.
that for catheter-dependent patients, frequent catheter use does not predispose these NHHD-treated patients to increase infection risks or death as compared with the IHD patients. This study found that there were no significant differences in the bacteremia/sepsis rate for HD patients treated
Survival curves
1 0.9 0.8 0.7 0.6 0.5 0.4 0
5
10 Months to sepsis IHD
15
20
NHHD
Number of catheter administrations at risk Months
0
5
10
15
20
Total
IHD % NHHD %
166 59 192 42
76 27 105 23
24 9 70 15
10 4 51 11
3 1 37 8
279 455
IHD = in-center hemodialysis; NHHD = nightly home hemodialysis.
Figure 2 Kaplan–Meier curves for time to bacteremia/sepsis for all catheters used in 20 months. IHD = in-center hemodialysis; NHHD = nightly home hemodialysis.
246
using a central venous catheter between selected patients on nocturnal home HD and a matched cohort of in-center patients treated three times per week. However, both groups have event rates that are slightly lower than those reported by the Center for Disease Control National Healthcare Safety Network for 2007–2008 at 1.5 to 2.0 vs. 3.2/100 catheter using patient months,26 likely due to patient selection, since both NHHD and IHD study cohorts were both younger, and had less DM burden than the general IHD population. The overall disposition of patients indicated very low death rates, very few transplants, and very few transfers to PD. The limitations of this study include NHHD participants only from a single center, even though it is the largest and longest existing program in the United States. Additionally, its observational study design does not address mechanisms that result in the observed differences and is subject to residual confounding. A list of residual confounding factors include limited amount of laboratory values available in the data set, lack of socioeconomic indicator, and limited adjustment for comorbidities. Since the control subjects for the NHHD cohort was selected through matching for each of the five categories: age, gender, race, dialysis vintage, and diabetes, the matching minimizes some of the difference between the two comparison groups. However, one example stems from Lockridge and Kjellstrand9 who previously reported that education levels below graduation from high school was associated with a 6.53 hazard ratio for mortality for patients on NHHD, more significant than dialysis vintage, baseline renal disease, presence of congestive heart failure, and the sum of comorbidities. Unfortunately, education level was not available in this retrospective cohort analysis, therefore offering direction for future studies. An
Hemodialysis International 2015; 19:242–248
Bacteremia in nightly home hemodialysis
additional limitation is the 20-month study period cut-off, which may miss event capture. Lastly, there may be greater chance of incomplete hospitalization diagnosis data collection for the control IHD group, which was selected from 6285 patients with data captured electronically compared with the more focused manual data collection for the 63 NHHD patients. The potential for incomplete documentation of bacteremia/sepsis episodes in IHD patients could render the findings more in favor of NHHD; therefore, the observed null difference between the two groups may be a conservative estimate of truth. In an era with 27.7% of all prevalent ESRD patients still dialyzing via catheter alone,16 this study has shown that the rate of bacteremia and death is not different for catheter users who dialyze via frequent NHHD vs. thrice weekly IHD. Although Fistula First initiative continues to succeed in increasing the number of fistula users, there will always be a subset of patients who lack suitable blood vessels for arteriovenous access placement, cannot tolerate the physiologic shunt of a fistula/graft, and/or have emotional/psychological resistance to conversion to a permanent arteriovenous vascular access. In addition, patient compliance with physician and caregiver instructions and adherence to therapy remains key factors influencing survival that current studies have found difficulty accounting for. NHHD offers the advantage of being dialyzed at home and may appeal to more “independent” patients and/or very involved caregivers, thereby potentially increasing treatment compliance. It may also influence more careful attention to catheter care and preservation. Although future studies are needed to elucidate if NHHD offers survival advantage compared with IHD, we posit that for ESRD patients, in particular the catheter committed subset, NHHD is a viable renal replacement modality with the potential for improved catheter lifespan, improved phosphate and hypertension control,1,12 and in the hands of the country’s most durable NHHD program, come with at least comparable, if not potentially better infection risks than a matched cohort on IHD.
Manuscript received January 2014; revised October 2014.
REFERENCES 1 Pierratos A. Nocturnal hemodialysis: Dialysis for the new millennium. CMAJ. 1999; 161:1137. 2 Lockridge RS Jr, Spencer M, Craft V, et al. Nightly home hemodialysis: Five and one-half years of experience in Lynchburg, Virginia. Hemodial Int. 2004; 8:61–69.
Hemodialysis International 2015; 19:242–248
3 Mucsi I, Hercz G, Uldall R, Ouwendyk M, Francoeur R, Pierratos A. Control of serum phosphate without any phosphate binders in patients treated with nocturnal hemodialysis. Kidney Int. 1998; 53:1399–1404. 4 Sikkes ME, Kooistra MP, Weijs PJ. Improved nutrition after conversion to nocturnal home hemodialysis. J Ren Nutr. 2009; 19:494–499. 5 Hanly PJ, Pierratos A. Improvement of sleep apnea in patients with chronic renal failure who undergo nocturnal hemodialysis. N Engl J Med. 2001; 344:102– 107. 6 Chan CT, Floras JS, Miller JA, Richardson RM, Pierratos A. Regression of left ventricular hypertrophy after conversion to nocturnal hemodialysis. Kidney Int. 2002; 61:2235–2239. 7 Chan CT, Jain V, Picton P, Pierratos A, Floras JS. Nocturnal hemodialysis increases arterial baroreflex sensitivity and compliance and normalizes blood pressure of hypertensive patients with end-stage renal disease. Kidney Int. 2005; 68:338–344. 8 Johansen KL, Zhang R, Huang Y, et al. Survival and hospitalization among patients using nocturnal and short daily compared to conventional hemodialysis: A USRDS study. Kidney Int. 2009; 76:984–990. 9 Lockridge RS, Kjellstrand CM. Nightly home hemodialysis: Outcome and factors associated with survival. Hemodial Int. 2011; 15:211–218. 10 Pauly RP, Gill JS, Rose CL, et al. Survival among nocturnal home haemodialysis patients compared to kidney transplant recipients. Nephrol Dial Transplant. 2009; 24:2915–2919. 11 Culleton BF, Walsh M, Klarenbach SW, et al. Effect of frequent nocturnal hemodialysis vs conventional hemodialysis on left ventricular mass and quality of life: A randomized controlled trial. JAMA. 2007; 298:1291– 1299. 12 Rocco MV, Lockridge RS Jr, Beck GJ, et al. The effects of frequent nocturnal home hemodialysis: The Frequent Hemodialysis Network Nocturnal Trial. Kidney Int. 2011; 80:1080–1091. 13 Stokes JB. Nocturnal hemodialysis: Analysis following the Frequent Hemodialysis Network trial. Semin Dial. 2011; 24:614–620. 14 Misra M, Twardowski ZJ. Benefits of frequent nocturnal home hemodialysis. Kidney Int. 2012; 82:114–115, author reply 5. 15 Suri RS, Larive B, Sherer S, et al. Risk of vascular access complications with frequent hemodialysis. J Am Soc Nephrol. 2013; 24:498–505. 16 USRDS. 2012 Annual Data Report: Atlas of Chronic Kidney Disease and End-Stage Renal Disease in the United States. Bethesda (MD): National Institute of Health; 2012. 17 USRDS. 2013 Annual Data Report: Atlas of Chronic Kidney Disease and End-Stage Renal Disease in the United States. Bethesda (MD): National Institute of Health; 2013.
247
Xue et al.
18 Bradbury BD, Fissell RB, Albert JM, et al. Predictors of early mortality among incident US hemodialysis patients in the Dialysis Outcomes and Practice Patterns Study (DOPPS). Clin J Am Soc Nephrol. 2007; 2:89–99. 19 Lacson E Jr, Wang W, Lazarus JM, Hakim RM. Change in vascular access and hospitalization risk in long-term hemodialysis patients. Clin J Am Soc Nephrol. 2010; 5:1996–2003. 20 Ishani A, Collins AJ, Herzog CA, Foley RN. Septicemia, access and cardiovascular disease in dialysis patients: The USRDS Wave 2 study. Kidney Int. 2005; 68:311–318. 21 Astor BC, Eustace JA, Powe NR, Klag MJ, Fink NE, Coresh J. Type of vascular access and survival among incident hemodialysis patients: The Choices for Healthy Outcomes in Caring for ESRD (CHOICE) Study. J Am Soc Nephrol. 2005; 16:1449–1455.
248
22 Gulati S, Sahu KM, Avula S, Sharma RK, Ayyagiri A, Pandey CM. Role of vascular access as a risk factor for infections in hemodialysis. Ren Fail. 2003; 25:967– 973. 23 Dhingra RK, Young EW, Hulbert-Shearon TE, Leavey SF, Port FK. Type of vascular access and mortality in U.S. hemodialysis patients. Kidney Int. 2001; 60:1443–1451. 24 Mazonakis E, Stirling C, Booth KL, McClenahan J, Heron N, Geddes CC. The influence of comorbidity on the risk of access-related bacteremia in chronic hemodialysis patients. Hemodial Int. 2009; 13:6–10. 25 Lacson E Jr, Wang W, Lazarus JM, Hakim RM. Change in vascular access and mortality in maintenance hemodialysis patients. Am J Kidney Dis. 2009; 54:912–921. 26 National Healthcare Safety Network–Dialysis Surveillance Report. Atlanta (GA): Center for Disease Control; 2008.
Hemodialysis International 2015; 19:242–248
Copyright of Hemodialysis International is the property of Wiley-Blackwell and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.
