One of the most frequently asked questions about daily hemodialysis is the effect upon vascular access. A number of articles have been published, mostly dealing with fistulae, and have reported the conclusion that there is either no effect or, surprisingly, that access survival improves with daily dialysis. 1–12 Studies reporting upon the effect of daily dialysis on grafts and central venous catheters are fewer, and no articles have dealt with the day to day complications and problems experienced by patients on daily dialysis.
During a 3 year study of daily home hemodialysis, the present authors performed a detailed analysis of the frequency of complications and adverse events and the longevity of fistulae, grafts, and central venous catheters. The authors compared the outcomes between the different types of access and also compared these with results reported for the various accesses with regular three times weekly hemodialysis.
Twenty-three patients on home hemodialysis participated in this study. Inclusion criteria included an expected patient survival of more than 6 months and a blood access that could deliver a blood flow of at least 300 ml/min. All patients signed a consent form approved by the local Institutional Review Board.
For the first 4,499 dialyses, patients completed an 11 page questionnaire before and after each dialysis. This included questions about pain, bleeding, and redness or swelling around the vascular access site. Problems were graded from 1 (no problem) to 3 (severe problem). Similarly, patients filled in a questionnaire at the end of every dialysis asking if they had had any problem with their access during dialysis. For the remainder of the study (4,710 further dialyses), patients completed a single page questionnaire after each dialysis.
During the entire study, patients were also asked at the time of each dialysis whether they had visited any outpatient facility, their dialysis training unit, or a doctor’s office, or if they had been hospitalized. If hospitalized, staff obtained the discharge records. Dialysis nurses or technicians also contacted patients weekly to review the questionnaires and then reported independently upon any adverse events that had occurred. Data related to vascular access were extracted from these reports.
Results were entered into and analyzed with the Statview-5 (SAS, Institute Inc., Cary, NC) program. Results are reported as mean ± SD unless otherwise stated. Analyses were performed with the chi-square test, Student’s unpaired t-test, Kaplan-Meier survival analyses with the Mantel- Cox log rank test, and Cox proportional hazard analysis with 95% confidence interval (CI). A p value of less than 5% was regarded as significant.
Twenty-three patients were studied (7 women and 16 men), with a mean age of 51 ± 14 years (range 23–80 years). Four patients had central venous catheters, five had grafts, and fourteen had native blood vessel fistulae. The cause of renal failure was diabetes in four patients, glomerulonephritis in five, hypertensive nephrosclerosis in six, and other kidney diseases in eight. Patient weight was 76 ± 21 kg (range 45–122 kg). Patients had been on end-stage renal disease (ESRD) therapy for 92 ± 80 months (range 8–298 months) and on home hemodialysis for 66 ± 80 months (range 1–251 months).
Daily Hemodialysis and Vascular Accesses
Total observation time on daily hemodialysis was 409 months with a mean of 18 ± 10 (range 7–36) months per patient. Information on blood access was available from the detailed records of 4,499 dialyses and from the adverse event records and hospitalization records of a further 4,710 dialyses, for a total of 9,209 dialyses. During daily hemodialysis, thirteen patients dialyzed five times per week, nine dialyzed six times, and one patient dialyzed four times per week, for a mean of 5.3 ± 0.5 dialyses weekly. Mean dialysis duration was 131 ± 29 minutes, and prescribed blood flow was 357 ± 61 ml/min. There was no significant difference in mean dialysis time or blood flow between the different accesses. Total observation time for fistulae was 254 months, with a mean of 18 ± 11 months per patient; for grafts it was 105 months (mean 21 ± 11 months), and for catheters it was 50 months (mean 13 ± 3 months). Details are shown in Tables 1 and 2.
Nineteen (83%) of the accesses were free of failure during the entire study, and only four (17%) had to be replaced. Two of the four catheters were replaced after 1 and 14 months of daily dialysis. One graft was replaced with a native fistula after 1 month, and this was then changed to an upper arm fistula after another 3 months of daily dialysis. One primary fistula was replaced after 22 months. Cumulative survival at 15 months was 100% for fistulae, 80% for grafts, and 25% for catheters. Cumulative survival at 3 years was 80% for fistulae and grafts, and no catheter lasted beyond 15 months (p = 0.041, Kaplan-Meier analysis with Mantel- Cox log rank test). The failure rate per year was 0.05 for fistulae, 0.11 for grafts, and 0.48 for catheters (p = 0.042, Student’s unpaired t-test, fistulae vs. others).
Adverse Events Requiring Interventions
Fourteen (61%) of the accesses needed no interventions. There were twenty-seven events in nine accesses that required hospitalization or an outpatient intervention. Details are shown in Table 3. Of the fistulae, eleven were free of events during the entire study, whereas three required interventions. Only one of the five grafts was event free, and two of the four catheters required intervention (p = 0.029, fistula vs. other accesses, chi-square analysis). Details are in Table 3.
Six patients were hospitalized seven times for a total of 13 days for access problems. There were three times as many events per year for patients with a graft compared with patients with a fistula, and almost four times as many events for patients with catheters when compared with those with fistulae (p = 0.080, fistulae vs. other accesses, Student’s unpaired t-test).
In Cox proportional hazards analyses, catheters had an increased hazard for failure when compared with the subcutaneous accesses [hazard ratio (HR) = 12, CI 1.03–133, p = 0.047), and fistulae had considerably less risk of needing interventions compared with grafts and catheters (HR = 0.15, CI 0.03–0.75, p = 0.021). There were no hazards associated with patient age, months on ESRD therapy, body weight, gender, or frequency of dialysis per week.
Patient Reported Problems
Patients with fistulae reported sixteen times as many problems between dialyses compared with patients with catheters and eight times as many compared with patients with grafts (p < 0.0001, chi-square). Almost 90% of these problems were complaints about pain and redness around the fistula site. During dialysis, the fewest problems were reported with grafts (0.9% of all dialyses), the most with catheters (9.1%), and with fistulae this was 2.7%. (p < 0.0001, chi-square). The problems are detailed in Table 4.
Perhaps the most interesting finding in this study was the rarity of vascular access problems with daily hemodialysis. Nineteen of the twenty-three accesses never failed, and fourteen of the patients never needed any intervention for access problems during a cumulative observation period of 409 patient months. Only twenty-seven adverse events requiring intervention were reported in nine accesses during more than 9,200 dialyses and 34 patient years of observation. The number of dialyses and thus access attachments in his study was 9,209. It corresponds to the number of access attachments done during 60 patient years of regular, three times weekly dialysis.
The patients reported a total of only 91 problems during 4,499 dialyses.
As expected, fistulae had the best survival, and the worst survival was with catheters. Compared with fistulae, twice as many replacements were required for grafts and ten times as many for catheters. Fistulae were also associated with the fewest adverse events requiring intervention. There were more than twice as many adverse events associated with grafts and three times as many with catheters.
Of interest is the fact that patients with fistulae reported more problems between dialyses, eight times more than with catheters and sixteen times as many as with grafts. Redness and pain were the two most common complaints. During dialysis, however, there were more problems with catheters and fewer with grafts than with fistulae.
Comparison with Other Studies of Daily Hemodialysis
Central venous catheters.
Lockridge et al.1 reported that four of seven patients followed for 49 months needed replacement of a catheter used for long nightly hemodialysis. Similarly, two of the four of the patients with catheters in the present study needed replacement during the 50 month observation period. The Toronto group, led by Pierratos, had a 38% 1 year survival of catheters used by 30 patients for long nightly hemodialysis 2 compared with the 70% 1 year survival in our patients. Ting, to the contrary, found a lower complication rate than we did in his short daily hemodialysis program. Only one of six catheters needed replacement during a mean follow up of 2 years per catheter, and the event rate of 0.52 events per year was lower than the 1.44 events per year that our patients with catheters experienced. 3
Vos and Kooistra, studying short daily hemodialysis, reported that two of eleven patients observed for a mean of 18 months lost their grafts. 4 This is identical to the present finding that one of five grafts observed for 21 months was lost. Ting reported 1.39 events per year in seven patients with PTFE grafts, similar to the present incidence of 1.37 events per year. 3
The largest study of fistulae and short daily hemodialysis is that by Woods and colleagues. 5 In 72 patients followed for 2,864 months, there were 13 failures or 0.05 failures per year, which is identical to the results in the present study. The same failure rate, 0.07 failures per year was found by Bermond et al.6 in a study of 25 patients followed for 467 months.
Three groups have reported a lower rate of events and failures with fistulae used with daily hemodialysis. Louis and colleagues found no failure of fistulae in 10 patients followed for 350 patient months. 9 Ting reported only 0.13 events per year and no failures. However his patients were fewer, eight compared with the present fourteen, and follow-up was much shorter, 7 vs. 21 patient years. 3 Similar low rates of failure and adverse events were reported by Quintaliani et al.7 None of 24 patients lost a native fistula during a mean follow up of 3.8 years, and the event rate was 0.02 events per year.
Thus, in the patients in the present study, vascular access survival and problems requiring intervention were generally similar to those reported by others for daily hemodialysis
Comparison with Blood Accesses in Regular Dialysis
Access survival and the frequency of access problems that the present authors and others have reported with daily hemodialysis are no higher than those reported for conventional three times weekly hemodialysis. In 1969, DePalma et al., 8 in the first report of chronic daily hemodialysis, used previous shunt failures as an indication for daily hemodialysis. He speculated that the decreased ultrafiltration volume needed with daily treatments would minimize episodes of hypotension and decreased shunt blood flow. Although no details appear in his report, the number of episodes of shunt clotting decreased with daily hemodialysis (DePalma, personal communication, 1998).
The first detailed comparison of blood access with regular and daily hemodialysis was published in 1975 when Louis et al.9 reported upon 10 patients on daily hemodialysis followed for up to 90 months (mean 35 months). No fistula was lost in this series, while there were 0.24 failures per year in the matched control group. 9 Using an unselected control population of patients on regular dialysis, Mastrangelo et al.10 found an 80% 6 year cumulative access survival with daily hemodialysis vs. 67% for regular hemodialysis. Quintaliani et al.7 reported 9.8 vs. 2.2 events per year in their comparison of 124 patients on regular hemodialysis with 24 patients on daily short hemodialysis. In an unadjusted comparison, Bermond et al.6 found that access problems on daily hemodialysis were 30–75% less than in patients on regular hemodialysis.
There are two studies using a crossover format where access complications were studied during a period of regular hemodialysis and then compared with the fate of the same accesses after the patients were changed to daily hemodialysis. Ting’s 30 patients had more problems during a 2 year period of regular dialysis than during 3 years of daily hemodialysis, 0.96 vs. 0.90 problems per year. 3 Similarly, Woods et al.5 reported 0.28 events per year in 72 patients during 606 months of observation on regular hemodialysis compared with only 0.05 events per year during 2,864 months of daily hemodialysis. This improvement with daily hemodialysis was highly statistically significant.
The present authors did not have a direct comparison with either a matched control group or to previous access data in our patients. However, the results we found are better than those in two recent reports on blood access in patients on regular dialysis. Saran and colleagues 13 reported a 70% 1 year survival of grafts in 1,944 patients and a 78% 1 year survival with fistulae in 900 patients, and Rayner et al.14 reported a 1 year fistula survival of 87% in 3,674 patients. The present authors had no failures in patients with fistulae during the first year on daily dialysis and an 80% function rate at 1 year in the patients with grafts.
So far there have been no reports of a worse outcome with any type of access in patients on daily hemodialysis compared with regular hemodialysis, and most have found a better outcome with daily hemodialysis. 11,12 Possible reasons for this include the more frequent heparinization of the access, better hemodynamics with less hypotension during and following dialysis, better hemostasis with less hematoma formation around the access, and less hypercoagulability or vasculopathy secondary to lower blood levels of homocysteine in patients on daily hemodialysis. 9
One criticism of all these studies is that patients were selected and those with frequent vascular access problems are unlikely to be selected for daily hemodialysis. Whereas this is probably true, it does not detract from the general conclusion reached by all investigators that daily dialysis in a patient with well functioning vascular access will have no adverse consequences for the access, and in all likelihood, access complications will decrease. This criticism is also invalidated by the studies that show better access function in patients when they move from regular three times weekly dialysis to daily hemodialysis.
This study was part of an FDA mandated clinical trial of the Aksys Ltd. Personal Hemodialysis system PHD. Aksys Ltd. financed the study. The authors thank the patients who participated in this study and willingly and accurately kept very detailed and time consuming records of their access and their treatments. They also thank the nurses and technicians at the DCI dialysis unit at the University of Missouri, as well as the RCG dialysis unit at the University of Mississippi and at the Home Hemodialysis Training Program of the Northwest Kidney Centers for their excellent work.
1. Lockridge R, Swafford A, Spencer M: Internal jugular catheter use in nightly home hemodialysis. Perit Dial Int 19: (Suppl.) 89, 1999.
2. Pierratos A, Goluboff D, Ouwendyk M: Vascular access for nocturnal hemodialysis (NHD) (Abstract). J Am Soc Nephrol 10: 215, 1999.
3. Ting G: Blood access outcomes associated with short daily hemodialysis. Hemodial Int 4: 42–46, 2000.
4. Vos P, Kooistra M:Daily Hemodialysis in the Netherlands
. Charlotte, NC: 5th International Symposium on Home Hemodialysis, 1999.
5. Woods J, Friedrich K, Port M, et al: Clinical and biochemical correlates of starting “daily” hemodialysis. Kidney Int 55: 2467–2476, 1999.
6. Bermond F, Quaglia M, Burdese M, et al: Survival of vascular access during daily dialysis: a comparison with other schedules in home hemodialysis and in limited care center. Nephrol Dial Transplant 17 (Suppl. 1): 283, 2002.
7. Quintaliani G, Buoncristiani U, Fagugli R, et al: Survival of vascular access during daily and three times a week hemodialysis. Clin Nephrol 53: 372–377, 2000.
8. DePalma JR, Pecker EA, Maxwell MH: A new automatic coil dialyzer system for “daily” dialysis. Proc Eur Dial Transplant Assoc 6: 26–34, 1969.
9. Louis B, Patel TG, Pinedo A, Snyder D, Gorfein P: Clinical experience with long-term 5 days-a-week hemodialysis. Proc Clin Dial Transplant Forum 5: 58–60, 1975.
10. Mastrangelo F, Alfonso L, Napoli M, DeBlasi V, Russo F, Patruno P: Dialysis with increased frequency of sessions (Lecce dialysis). Nephrol Dial Transplant 13: 139–147, 1998.
11. Twardowski Z: Blood access complications and longevity with frequent (daily) hemodialysis and with routine hemodialysis. Semin Dial 12: 451–454, 1999.
12. Twardowski Z: Primary fistula complications and longevity are better in daily (frequent) hemodialysis than in routine hemodialysis. Home Hemodialysis Today 1: 1, 1999.
13. Saran R, Dykstra DM, Wolfe RA, Gillespie B, Held PJ, Young EW: Association between vascular access failure and the use of specific drugs: the Dialysis Outcomes and Practice Patterns Study (DOPPS). Am J Kidney Disease 40: 1255–1263, 2002.
14. Rayner HC, Pisoni RL, Gillespie BW, et al: Creation, cannulation and survival of arteriovenous fistulae: Data from the Dialysis Outcomes and Practice Patterns Study. Kidney Int 63: 323–330, 2003.