sharp versus blunt dialysis needle use with ... - Wiley Online Library

ORIGINAL RESEARCH

SHARP VERSUS BLUNT DIALYSIS NEEDLE USE WITH BUTTONHOLE METHOD: OPEN RANDOMISED TRIAL Claudio Morselli1, Paolo Chiari2, Tiziana Aliberti1, Guglielmo Celli3, Silvia Catalani1, Ignazia Miale1, Vanessa Melandri1, Lorenza Bianchi1 1 ` Sanitaria Locale of Bologna, Bologna, Italy Dialysis Centre Ospedale Bellaria, Azienda Unita 2 Department of Medical and Surgical Sciences, University of Bologna, Bologna, Italy 3 Master’s Degree Course in Nursing Sciences, University of Bologna, Bologna, Italy

Claudio M., Chiari P., Tiziana A., Celli G., Silvia C., Ignazia M., Vanessa M., Lorenza B. (2015). Sharp versus blunt dialysis needle use with buttonhole method: open randomised trial. Journal of Renal Care 41(4), 213–221.

SUMMARY Background: Current protocols recommend the use of a blunt needle to access the arteriovenous fistula via a buttonhole. This study aims to demonstrate whether a sharp needle can be used at the same buttonhole site without causing complications. Goal: To measure and compare fistula cannulation failures between the use of blunt and sharp needles. Plan: Open-crossover randomised controlled trial. Participants: Adult out-patients who had provided consent and were on dialysis with a mature arteriovenous fistula and buttonhole cannulation. Outcome measures: Failed cannulation—difficulty in inserting the needle and the trampoline effect; incidence of complications, such as infection and haematomas; times to haemostasis; patients’ pain; and patients’ preferences. Results: Based on analysis of the data from the 35 patients enrolled, no significant differences were detected in failed cannulation of the fistulae between the use of a blunt needle and a sharp needle for the 335 venous accesses (p ¼ 0.071). However, a significant difference was detected for the 335 arterial accesses (p ¼ 0.001), in which the sharp needle was more effective. Significant differences were also detected in the difficulty of insertion and in the trampoline effect for both venous and arterial access (p < 0.05); the use of a sharp needle was more favourable. Conclusion: This study demonstrates an increased incidence of failed cannulation using a blunt needle compared with using a sharp needle, although this was not significant. In addition, the use of a sharp needle did not result in any increase in complications.

K E Y W O R D S Arteriovenous fistula Buttonhole technique Cannulation Haemodialysis

INTRODUCTION BIODATA Paolo Chiari is researcher in Nursing Science, Department of Medical and Surgical Science, University of Bologna, Director of Evidence-Based Nursing Study Centre of the Teaching Hospital of Bologna, Director of the Master in EBP and research methodology and clinical care at the University of Bologna, President of the Master’s Degree Course in Nursing and Midwifery Sciences, University of Bologna. CORRESPONDENCE

Paolo Chiari, Centro Studi EBN, Azienda Ospedaliero–Universitaria di Bologna, Via Albertoni 15, 40138 Bologna, Italy Tel.: þ39 051 636 1461 Fax: þ39 051 636 1375 Email: [email protected]

Effective haemodialysis (HD) treatment requires adequate functional vascular access. International guidelines suggest that an arteriovenous fistula (AVF) is the optimal choice for vascular access in patients requiring HD (Fluck & Kumwenda 2011). Of the available types of access, the AVF demonstrates the longest duration of use and has a low complication rate (Fluck & Kumwenda 2011). The Buttonhole (BH) technique, also known as the ‘constant site’ technique, in which the needle is persistently inserted into the same site on the skin to create a virtual tunnel through which the AVF can be accessed, was described for the first time in 1979 (Twardowski 1979). The tunnel maturation process is principally characterised by two phases. In the first phase, a sharp needle is used until a tunnel forms (this process usually takes longer than one month); the second phase abandons the sharp needle in favour of a blunt needle to avoid trauma to the vessel that might compromise the AVF and increase the number of complications (Twardowski 1979; Ball 2006; Pegoraro 2008).

© 2015 European Dialysis and Transplant Nurses Association/European Renal Care Association

Journal of Renal Care 2015

213

Claudio et al.

Nevertheless, to our knowledge, the literature contains no studies that compare AVF complications resulting from the use of a sharp needle with those of a blunt needle. However, certain studies have been performed in which the traditional rope– ladder (RL) technique, which uses a sharp needle, has been compared with the BH technique.

LITERATURE REVIEW Complications were examined in six primary studies, of which three were randomised controlled trials (RCTs) (Struthers et al. 2010; Chow et al. 2011; MacRae et al. 2014b), two were observational (Van Loon et al. 2010; Pergolotti et al. 2011) and one was a ‘prepost’ study (Ludlow 2010). The observed complications included pain, infection of the injection site, the presence of haematomas at the site, the presence of aneurysms, bleeding, anxiety and fear of accessing the AVF, failed cannulation, insufficient flow to perform haemodialysis and the inability to proceed further with cannulation.

INFECTION AT THE SITE Not all of the studies were able to demonstrate a significant causal association between infection and the type of cannulation (Struthers et al. 2010; MacRae et al. 2014b). In particular, the studies by Chow et al. (2011) and MacRae et al. (2014b) showed an increased risk of infection using BH. However, in most studies, the follow-up period was too short to evaluate the occurrence of infection, even infections that are localised to the skin.

INSUFFICIENT FLOW/INABILITY TO PROCEED FURTHER USING THE NEEDLE No differences in dialysis session quality in terms of flow capacity of the access have been reported (Ludlow 2010). In one study, 21% of those using BH showed insufficient blood flow or inability to proceed following cannulation. For the control group, this event occurred in 27.6% of cases; however, this was not a significant difference (p > 0.05). Failed cannulation is defined as the need to use another needle for venous or arterial access. The study by Van Loon et al. (2010) demonstrated a greater risk of this event in the BH group (p < 0.0001), in which the average number of failures was 8.1, than in the RL group, in which the average number of failures was 3.7. However, when cannulating BH, the risk of failure might be due to the operator’s skills or abilities. In the centre at which their study was performed, a survey that concluded in spring 2012 showed a 41% rate of failed cannulation and replacement of the needle (a high incidence that was likely due to insufficient technical ability in using a blunt needle). A subsequent study performed in November 2012 showed a reduction of 30% in failed cannulations (Morselli et al. 2012). No other indications of failed cannulation followed by substitution of the blunt needle with a sharp needle have been reported in the literature. Indeed, many authors suggest generating a second access in these cases using the area or RL technique (Twardowski 1979; Ball 2006; MacRae et al. 2014).

HAEMATOMA OF THE ACCESS SITE, BLEEDING AND LOSS OF BLOOD AROUND THE NEEDLE The data regarding the development of haematomas are also contradictory. The study by Chow et al. (2011) showed an

PATIENTS’ PREFERENCES The BH technique has been shown to be widely accepted by both patients receiving haemodialysis and nurses who use this technique daily for AVF access (Struthers et al. 2011).

Shown below are the principal complications discussed in the above studies which serve as parameters for comparisons between cannulation using a sharp versus blunt needle. PAIN The evidence is contradictory. Some studies show an increased perception of pain during the access of the AVF in the group in which the BH technique was used (Van Loon et al. 2010), whereas other studies showed no differences between the groups (Chow et al. 2011; MacRae et al. 2014b) or even reduced pain in the BH group (Ludlow 2010; Pergolotti et al. 2011). In some studies, the use of local anaesthetics has been analysed as an indirect indicator of perceived pain (Struthers et al. 2010; Van Loon et al. 2010; Chow et al. 2011); this practice was used less frequently in the BH group.

214

increase in the incidence of haematomas in the BH group, whereas Ludlow (2010), Van Loon et al. (2010) and MacRae et al. (2014b) showed substantially contrasting results, in which the haematoma incidence in the BH group was lower than that in the control group. Bleeding, even expressed as the time to haemostasis, is statistically similar between the two groups according to the studies by Struthers et al. (2011) and Ludlow (2010). Alternatively, Pergolotti et al. (2011) demonstrated that bleeding was less frequent by 23%, in the BH group and that bleeding increases with age. No significant difference (p > 0.05) in loss of perivascular blood has been detected between these groups. In particular, the study by Ludlow (2010) detected 17 (17.9%) blood loss events in the RLR group compared with 2 (6.9%) in the BH group; this difference was not significant.



SHARP VERSUS BLUNT DIALYSIS NEEDLE USE WITH BUTTONHOLE METHOD: OPEN RANDOMISED TRIAL

AIMS OF THE STUDY In current protocols, the use of a blunt needle is recommended to access the fistula via a BH. This study aims to demonstrate whether a sharp needle can be used at the same BH site without causing complications. Our primary goal was to measure the incidence of cannulation failure using a blunt needle compared with using a sharp needle. Our secondary goals were to measure the incidence of difficulty of insertion and of the trampoline (incorrect placement of needle) effect; to compare the extent of pain at the moment of insertion and during dialysis in the case of acute events between the two groups; to measure the incidence of infection at the site and of haematomas, including the assessment of infiltration and oedema which may or may not impede the use of the access; and to evaluate bleeding after removal of the needle and the incidence of blood loss around the needle during the session. We also aimed to understand patients’ preferences concerning the use of these two types of needles and to measure the variability of the needle sizes.

MATERIALS AND METHODS This study was approved by the Independent Ethical Committee of the ASL (Local Health Company Unit) of Bologna, Italy. The patients signed consent forms and were informed that they could withdraw from the study at any time. STUDY PLAN This was an open randomised study that included active, crossover and monocentric controls and patients receiving haemodialysis treatment using the BH method. ASSIGNMENT OF TREATMENT After the patients had affirmed their consent to participate in this study, they were randomly assigned at each subsequent dialysis session to have their AVF cannulated, using either a sharp needle or a blunt needle. Each patient served as his or her own control due to the repeated random accesses. Simple randomisation lists were generated by the secretarial office of the clinic for each patient to avoid an imbalance between patients, as each patient underwent approximately 12 dialysis sessions per month. The randomised sequences were concealed in sealed, non-transparent envelopes. The choice of which needle to use was provided to each dialysis station via the sealed envelopes, which were used in numerical order at each dialysis session for each patient. It was not possible to assess the results

using blinding methodology. Each dialysis session constituted a complete ‘case’. ELIGIBILITY OF THE PARTICIPANT The study population consisted of out-patients undergoing haemodialysis treatment via an AVF using the BH method who attended a dialysis centre in Emilia-Romagna, Italy. The enrolled patients were adults undergoing haemodialysis treatment using the BH technique who provided written consent to participate in the study. Exclusion criteria were as follows: patients who were undergoing haemodialysis treatment using the BH method during a training phase; patients who were unable to understand this study or provide consent. TREATMENT DURING THE STUDY The BH technique is characterised by repeated insertion of the needle at the same location to generate a virtual tunnel through which the AVF is accessed. Once the site matured, we used, according to randomisation, a blunt (experimental) or sharp (control) needle to access the vessel. All measurements were performed during dialysis treatment, and the follow-up was concluded at the end of dialysis treatment. The needle insertion procedure was identical for the two groups, and is shown in Table S1. OUTCOME MEASURES The primary outcome was evaluated by measuring the incidence of failed cannulation of the AVF. Failed cannulation was defined as the failure of the needle to locate or reach the vessel. The secondary outcomes were as follows: - the incidence of difficulty of insertion (defined as difficulty with accessing the tunnel, although ultimately reaching the vessel); - the incidence of the trampoline effect (defined as the inappropriate placement of the needle that was inserted into the vessel, although ultimately reaching the vessel); - pain at the moment of insertion and during dialysis in the case of acute events according to the Numerical Rating Scale (NRS). This scale is composed of a rating scale from 0 to 10 in which 0 represents the lack of pain and 10 represents maximal pain (McCaffery et al. 1989; Ferreira-Valente et al. 2011);



215

Claudio et al.

- the incidence of infection at the site according to the Visual Infusion Phlebitis (VIP) score, which was developed by the Intravenous Nurses Society (Gallant & Schultz 2006); - the incidence of haematoma at the site based on an assessment of infiltrations and oedema which do or do not impede access according to clinical judgment;

For analysis of the primary and secondary end-points, we used Pearson’s x2 test for non-parametric variables and Student’s ttest for continuous variables, particularly for the comparison between the means of the two groups. For both methods, the x2 test and Student’s t-test, the results were considered to be significant at a p-value of less than 0.05 based on a 95% confidence interval.

RESULTS - the incidence of bleeding after removal of the needle based on a calculation of the increase or decrease in the patient’s clotting time to haemostasis; - the incidence of loss of blood around the needle during the session; - the patient’s preference regarding the use of these two types of needles according to a VAS in which a horizontal line was marked at the two extremes as ‘negative perception of the action’ on the left and ‘positive perception’ on the right. These data were subsequently transformed into a numerical scale from 0 to 10, similar to the NRS (Wewers & Lowe 1990); and

Of the 60 eligible patients, 35 participated in this study. Data collection began on 26 November, 2013 and concluded on 19 December, 2013. A flow-chart describing this study is presented in Figure 1. The group of enrolled subjects consisted of 35 patients; their characteristics are shown in Table 1. This study subdivided the cannulation of the participants’ AVF into venous and arterial accesses, followed by randomisation according to the type of needle, either sharp or blunt. VENOUS ACCESS A total of 335 venous accesses were analysed. Of these, 169 were randomly assigned to using a blunt needle, and 166 were randomly assigned to using a sharp needle.

- the variation in needle size (15 G vs. 16 G).

SAMPLE SIZE To demonstrate a difference of 0.15 between the experimental and control treatments, as required for comparison to the literature, based on a significance level of 5% (two-tailed) and a power of 80%, it was estimated that at least 76 accesses per arm were needed. This sample size was calculated using Statistics Calculators v. 3.0 software.

216

Of the 169 accesses performed using a blunt needle, 12 ultimately used a sharp needle because of failed cannulation due to the following reasons: pain upon insertion of the blunt needle into the vein, the patient’s refusal to use the blunt needle, the trampoline effect, or failure to reach the vessel. Of the 166 accesses performed using a sharp needle, 4 ultimately used a blunt needle due to the patient’s refusal to use a sharp needle or pain caused by cannulation.

The overall number of participants was calculated, considering that approximately 30 patients who attended our Centre were realistically capable of adhering to the study requirements and providing informed consent. If these 30 patients were to undergo 3 dialysis sessions per week, then 90 venous accesses and 90 arterial accesses, for a total of 180 accesses, would be performed per week.

As shown in Table 2, for venous access, no significant differences in failed cannulation, perivascular blood loss, bleeding time or infection at the site were detected. However, significant differences in the difficulty of insertion, at p ¼ 0.009 (22 vs. 8), and in the trampoline effect, at p ¼ 0.019 (8 vs. 1), were detected between the use of a blunt needle and a sharp needle, respectively.

STATISTICAL METHODS The exploratory analyses were performed using descriptive statistics. The data were presented based on intention-to-treat.

No case of haematoma was observed. The 16 G needle size was used in 97 and 98% of the cases for the blunt needle and the sharp needle, respectively.



SHARP VERSUS BLUNT DIALYSIS NEEDLE USE WITH BUTTONHOLE METHOD: OPEN RANDOMISED TRIAL

Figure 1: Patient flow during the study.

ARTERIAL ACCESS Regarding arterial access, a total of 335 accesses were analysed. Of these, 169 were randomly assigned to using a blunt needle, and 166 were assigned to using a sharp needle. Of the 169 accesses performed using a blunt needle, 2 ultimately used a sharp needle because of failed cannulation due to the patient’s refusal to use a blunt needle.

Of the 166 accesses performed using a sharp needle, 4 ultimately used a blunt needle due to the patient’s refusal to use a sharp needle or pain caused by cannulation. In Table 3, regarding arterial access, significant differences in failed cannulation, at p ¼ 0.001 (10 vs. 0), in difficulty with insertion, at p ¼ 0.028 (13 vs. 4), and in the trampoline effect, at p ¼ 0.001 (11 vs. 0), were detected between the use of a blunt



217

Claudio et al.

Subjects N ¼ 35 (%)

Variables Sex Male Female Age (years)