Vascular access for hemodialysis (HD) patients is one of the most critical aspects of their overall care and well being, and an arteriovenous fistula (AVF) is the preferred option for this (Renal Association guidelines). Although there is a lower rate of infection with AVFs as opposed to alternatives such as prosthetic grafts, AVFs continue to have complications. These not only include minor problems such as bruising after needling, pain and discomfort during needling, and prolonged bleeding from the puncture site postdialysis but also encompass more serious problems such as aneurysm formation, stenosis, and high output cardiac failure. Rates of thrombosis varying from one per 1,000 fistula days with routine monitoring1 to an overall rate of 17% in upper-arm AVFs2 have been reported. In one study of radiocephalic AVFs without any preemptive intervention, the thrombosis rate was reported as 0.043 per patient-year.3 Significant stenosis (>50%) was found at angiography in 79 of 191 patients monitored over a 5-year period4 and an overall complication rate of 42%–44% has been reported.2,5 More significantly, data from the Dialysis Outcomes and Practice Patterns Study (DOPPS) study demonstrate that in the United Kingdom, 44% of hospitalizations for HD patients relate to vascular access,6 whereas in the United States, suboptimal vascular access is associated with increased mortality with a relative risk of 1.23.7 The method by which needles are inserted may have a significant effect on the incidence of some of these complications.
There are a number of different techniques for needling AVFs. “Buttonhole” or constant-site needling, first described by Twardowski in 1979,8 has been suggested as a technique that may lead to a reduction in many of these problems, particularly prolonged bleeding, bruising, and aneurysm formation in comparison with the traditional “rope-ladder” technique of varying the needling site as currently used in most renal units. The buttonhole technique involves repeated needle insertion in exactly the same site at exactly the same angle over several dialysis sessions until a track has been formed, after which a less sharp needle can be used to access the AVF via the preformed track. In contrast, using the traditional technique, needles are inserted a short distance above or below the previous needling site, progressively moving up or down the AVF until there is no further area suitable for needling when the site reverts to the original area and the process starts again.
Recently, a nonrandomized, prospective observational study9 suggested a reduction in interventions, principally angioplasty, a reduction in aneurysm formation, and a reduction in subcutaneous hematoma with buttonhole needling, although an increase in infection was noted. A further single-center Dutch study10 found a significant improvement in ease of cannulation with the buttonhole technique and a subsequent comment on this study by the originator of the technique, Twardowski11 supports buttonhole needling as superior to the rope-ladder technique and outlines some of the reasons why this method has not yet gained widespread acceptance.
We performed a randomized controlled trial comparing these two techniques. The primary outcome was the pain score for all patients; secondary outcomes included bleeding times after needle removal, maximum transverse diameter of AVF, and complication rate.
Patients and Methods
Patients were recruited from a renal unit of the district general hospital and its two satellite renal units. All patients (n = 83) dialyzing with an AVF in NHS Highland at the beginning of the study were considered for inclusion. Patients dialyzing with arteriovenous grafts (n = 17) or tunneled central venous catheters (n = 3) as well as three unable to give written informed consent and one with preexisting buttonholes were excluded. Of the 59 patients recruited (mean age 60 ± 30 years; 34 men and 25 women; 15 diabetic), three withdrew before randomization (two died and one transplanted). All patients recruited gave their written informed consent. Patients were randomized in blocks of four for logistic reasons, 28 to buttonhole and 28 to the control group. In the buttonhole group, nine patients were diabetic, mean age was 61 years, and mean fistula age was 28 months. In the control group, 10 patients were diabetic, mean age was 60 years, and mean fistula age was 25 months. There was no significant difference between the groups on these measures.
The control group continued to dialyze using 25-mm-long sharp 15 G needles. In the buttonhole group, new buttonhole tracks were created using 25-mm-long sharp 15 G needles before changing over to “blunt” needles, also 15 G 25-mm-long (Nipro, Zevantem, Belgium).
The tracks were created by one principal nurse, but logistically it was impossible for only one nurse to be present to insert all the sharp needles required to create the track, so one other nurse observed exact site and angle of the initial insertions and then took over duties for insertion of sharp needles when the principal nurse was unavailable. Not more than two nurses were required to create any single track.
T-sterile swabs soaked in povidone-iodine 10% (betadine) were used for skin disinfection, one for each buttonhole site. Arterial and venous buttonhole sites were cleansed with separate swabs in a circular motion from the scab outward and allowed to dry. Two swabs were then soaked with saline and placed over each scab; a separate needle was used for each site to remove the scabs.
Pain scores using a visual analogue scoring system and use of local anesthetic (LA) were assessed before randomization then at 6 months after introduction of the buttonhole technique. The time taken for bleeding to stop postdialysis was measured before randomization and at month 6. Fistula problems off treatment, including pain, bruising, and bleeding, were also recorded at these times. Patients from the buttonhole group indicated their preferred technique at month 6 and a questionnaire on nursing satisfaction with both techniques was completed at the end of trial. Fistulas from both groups were photographed and maximum transverse diameter measured prerandomization and at the end of trial. Other complications including thrombosis, infection, infiltration, and bleeding during dialysis were recorded from both groups.
Normally distributed data were analyzed using the nonpaired t test and skewed data using the Mann-Whitney U test using GraphPad Prism software. The study required 15 patients per group to have 85% power to detect a 50% reduction in pain scores based on preliminary data collected within the unit.
A mean of 13.7 (median, 11) sharp needle insertions were required to create a suitable track for consistent blunt needle insertion. One AVF always required sharp needle insertion throughout the length of the trial. There was no correlation between either the age of the AVF or the age of the patient and the number of sharp needles required to create a track.
The use of LA was reduced in the buttonhole group by month 6: nine of the 22 patients completing the trial reduced or stopped using any LA, six of these stopped completely, compared with one of 25 in the control group (χ2 9.5, p < 0.01) (Figure 1). Although pain scores were not reduced, the buttonhole group had a median pain score of three of 10 pretrial and 2.5 after 6 months and the traditional group had a median score of one at both points.
Overall AVFs in the buttonhole group did not increase in size: average increase was 1% ± 22% (mean ± SEM), whereas in the control group increased in size by 30% ± 7% (p < 0.01), equivalent to an absolute increase of 5 mm (Figure 2). In the buttonhole group, 11 decreased in size, one by 9 mm, compared with only three in the control group.
Blood flow through the dialyzer was 350 ml/min (mean) in the buttonhole group and 358 ml/min in the control group with no significant difference between groups either before the trial or after 6 months.
There was no difference in postdialysis bleeding times in either group. Bleeding times in the buttonhole group were 342 ± 36 seconds before the trial and 300 ± 26 seconds after 6 months. In the control group, bleeding time was 402 ± 33 seconds pretrial and 400 ± 31 seconds after 6 months.
Buttonhole needling was strongly preferred by patients in the buttonhole group and by nursing staff, with 21 of the remaining 22 patients expressing a preference for this technique. Among the nursing staff, 15 of the 23 preferred buttonhole needling, with seven expressing no preference and only one nurse preferring the traditional method of needling (Figure 3).
There were 11 reported episodes of bleeding from needle sites during dialysis in the buttonhole group (six of these before buttonhole tracks being created and occurred in only two of the 28 patients), compared with 17 episodes in the control group. There were 19 reported infiltrations (peripuncture subcutaneous hematoma) in the buttonhole group, eight of these before buttonhole tracks being created, seven during creation of the tracks, and four in established buttonhole tracks (three with sharp needles and one with a blunt needle). From the control group, 27 infiltrations were reported. Each group had one fistula thrombosis, and the buttonhole group had one fistula infection.
During the trial, six patients withdrew from the buttonhole group (one for no reason given, one because of pain, one had a thrombosed fistula, one transplanted, and two died). Withdrawals from the control group included two patients who died and one had the fistula replaced with a graft because of multiple aneurysms.
Using one main nurse to create the track with observers substituting as needed, it took 4 weeks of thrice weekly dialysis to create tracks. This was longer than that had been suggested before our trial, but it was neither related to the age of the fistulas that we were using nor to the age of the patient. Because this was almost the first introduction of the buttonhole technique into the unit, it is possible that there was a learning curve and the number of sharp needles used to create a track might reduce with greater experience.
Although their pain scores did not change, patients using the buttonhole technique use less LA. Our visual analogue pain tool registered low levels of pain in both groups, and it may be that this tool was insufficiently sensitive to detect changes in the level of discomfort that the patients felt on needle penetration. This is supported by the observations that patients strongly preferred the buttonhole technique, and that their requirement for LA dropped significantly. Similar observations were made by Verhallen et al.10 whose home HD patients were cannulating their AVFs themselves. They also found no overall reduction in pain scores, although their patients, similar to ours, claimed that pain had reduced markedly. Their conclusion was that patients had been over-optimistic in their initial assessment of their pain, and this may have been the case in our trial too.
AVFs accessed using the buttonhole technique are less likely to enlarge and more likely to reduce in diameter. Some aneurysms reduced significantly in size. Overall, although buttonhole AVFs did not change, there was a significant difference compared with the control group in which AVFs increased in size during the 6 months of the study. This suggests that the traditional technique with repeated puncturing of small areas inflicts traumatic weakening on the vessel wall allowing arterial pressure to cause aneurysm formation.
Bleeding times were neither significantly different between the two groups nor there was a reduction in bleeding times within the buttonhole group. Buttonhole tracks might be thought to lead to longer bleeding times because of the presence of a persistent track or conversely to shorter bleeding times because of the compression of this shallow-angled track by the arterial pressure in the AVF creating a valve effect. In practice, there seems to be little or no difference according to needling technique and bleeding time presumably relates more to hemostatic factors and pressure in the AVF.
The technique is preferred by both patients and nursing staff. Although this was a new technique to the nursing staff on the renal unit, they quickly became adept at developing the tracks and accessing them once developed. The only nurse who preferred rope-ladder needling enjoyed the challenge of putting needles into difficult AVFs and felt that buttonhole needling made cannulation too easy.
The incidence of complications with the buttonhole technique is low. The overall incidence of infiltrations was not different between the groups, but it was notable that once a track was established there was a reduced incidence of this complication. There was one thrombosis in each of the groups and one infection in the buttonhole group during the trial, which was localized and resolved with antibiotic therapy. This trial was not powered to demonstrate whether there is an increased risk of infection with buttonhole needling. Further trials to clarify the absolute risk of infection with buttonhole needling and to investigate whether there is a reduced incidence of stenosis and thrombosis requiring intervention would be of great benefit.
We have demonstrated that buttonhole needling which reduces enlargement of AVFs is preferred by both nurses and patients and reduces the need for LA. At present, it seems sensible to recommend buttonhole needling for patients with significant aneurysm formation, those with discomfort or pain on needling, and those AVFs that have a limited area that would be suitable for needle insertion.
Supported by an unrestricted grant from Nipro that supported conference expenses (to J.S. and S.L.).
1. Lok CE, Bhola C, Croxford R, Richardson RM: Reducing vascular access morbidity: A comparative trial of two vascular access monitoring strategies. Nephrol Dial Transplant
18: 1174–1180, 2003.
2. Fitzgerald JT, Schanzer A, McVicar JP, et al
: Upper Arm Arteriovenous Fistula Versus Forearm Looped Arteriovenous Graft for Hemodialysis Access: A Comparative Analysis. Ann Vasc Surg
19: 843–850, 2005.
3. Basile C, Ruggieri G, Vernaglione L, et al
: The natural history of autogenous radio-cephalic wrist arteriovenous fistulas of haemodialysis patients: A prospective observational study. Nephrol Dial Transplant
19: 1231–1236, 2004.
4. Tessitore N, Lipari G, Poli A, et al
: Can blood flow surveillance and pre-emptive repair of subclinical stenosis prolong the useful life of arteriovenous fistulae? A randomized controlled study. Nephrol Dial Transplant
19: 2325–2333, 2004.
5. Hossny A: Brachiobasilic arteriovenous fistula: Different surgical techniques and their effects on fistula patency and dialysis-related complications. J Vasc Surg
37: 821–826, 2003.
6. Rayner HC, Pisoni RL, Bommer J, et al
: Mortality and hospitalization in haemodialysis patients in five European countries: Results from the Dialysis Outcomes and Practice Patterns Study (DOPPS). Nephrol Dial Transplant
19: 108–120, 2004.
7. Port FK, Pisoni RL, Bragg-Gresham JL, et al
: DOPPS estimates of patient life years attributable to modifiable hemodialysis practices in the United States. Blood Purif
22: 175–180, 2004.
8. Twardowski ZJ: Different sites versus constant sites of needle insertion into arteriovenous fistulas for treatment by repeated dialysis. Nephrol Dial Transplant
8: 978–980, 1979.
9. Van Loon MM, Goovaerts T, Kessels AGH, et al
: Buttonhole needling of haemodialysis arteriovenous fistulas results in less complications and less interventions compared to the rope-ladder technique. Nephrol Dial Transplant
25: 225–230, 2010.
10. Verhallen AM, Kooistra MP, van Jaarsveld BC: Cannulating in haemodialysis: Rope-ladder or buttonhole technique? Nephrol Dial Transplant
22: 2601–2604, 2007.
11. Twardowski ZJ: Utility of the buttonhole cannulation method for haemodialysis patients with arteriovenous fistulas. Nat Clin Pract Nephrol
3: 648–649, 2007.