ORIGINAL CONTRIBUTION
Arterial Puncture Using Insulin Needle Is Less Painful Than With Standard Needle: A Randomized Crossover Study Irwani Ibrahim, MD, Ying Wei Yau, MD, Lizhen Ong, MD, Yiong Huak Chan, PhD, and Win Sen Kuan, MD
Abstract Objectives: Arterial punctures are important procedures performed by emergency physicians in the assessment of ill patients. However, arterial punctures are painful and can create anxiety and needle phobia in patients. The pain score of radial arterial punctures were compared between the insulin needle and the standard 23-gauge hypodermic needle. Methods: In a randomized controlled crossover design, healthy volunteers were recruited to undergo bilateral radial arterial punctures. They were assigned to receive either the insulin or the standard needle as the first puncture, using blocked randomization. The primary outcome was the pain score measured on a 100-mm visual analogue scale (VAS) for pain, and secondary outcomes were rate of hemolysis, mean potassium values, and procedural complications immediately and 24 hours postprocedure. Results: Fifty healthy volunteers were included in the study. The mean (standard deviation) VAS score in punctures with the insulin needle was lower than the standard needle (23 22 mm vs. 39 24 mm; mean difference = –15 mm; 95% confidence interval = –22 mm to –7 mm; p < 0.001). The rates of hemolysis and mean potassium value were greater in samples obtained using the insulin needle compared to the standard needle (31.3% vs. 11.6%, p = 0.035; and 4.6 0.7 mmol/L vs. 4.2 0.5 mmol/L, p = 0.002). Procedural complications were lower in punctures with the insulin needle both immediately postprocedure (0% vs. 24%; p < 0.001) and at 24 hours postprocedure (5.4% vs. 34.2%; p = 0.007). Conclusions: Arterial punctures using insulin needles cause less pain and fewer procedural complications compared to standard needles. However, due to the higher rate of hemolysis, its use should be limited to conditions that do not require a concurrent potassium value in the same blood sample. ACADEMIC EMERGENCY MEDICINE 2015;22:315–320 © 2015 by the Society for Academic Emergency Medicine
INTRODUCTION
R
ecent systematic reviews have showed that venous blood gas is comparable with arterial blood gas in pH and bicarbonate estimations.1,2 These findings support the increasing use of venous blood gas to replace arterial blood gas by emergency physicians.3,4 However, there is a poor correlation of pCO2 and pO2 between venous and arterial blood.1,2 Therefore, there is still a need for emergency physicians to perform radial arterial punctures to evaluate ill
patients where knowledge of pCO2 and pO2 are crucial to guide management. In current practice, these punctures are usually performed using hypodermic needles with sizes ranging from 22 to 25 gauge.5–7 Arterial punctures are painful and can create anxiety and phobia of needles in patients.8 The perception of pain is influenced by both individual and technical factors. Differences in pain perception were observed between different sexes, ethnic groups, and hand laterality.9–11 Needle characteristics such as small gauge, short length, and sharp tip have also been
From the Emergency Medicine Department and Department of Surgery (II, YWY, WSK), the Department of Laboratory Medicine (LO), and the Biostatistics Department (YHC), National University Health System, Singapore. Received July 8, 2014; revision received September 17, 2014; accepted September 22, 2014. Presented at the American College of Emergency Physicians Scientific Assembly, Seattle, WA, October 2013. The study was supported by the National University Health Systems Junior Pitch for Funds. The authors have no conflicts of interest to declare. Supervising Editor: Brian Hiestand, MD, MPH. Address for correspondence and reprints: Irwani Ibrahim, MD; e-mail: [email protected].
© 2015 by the Society for Academic Emergency Medicine doi: 10.1111/acem.12601
ISSN 1069-6563 PII ISSN 1069-6563583
315 315
316
Ibrahim et al. • ARTERIAL PUNCTURE USING INSULIN NEEDLE IS LESS PAINFUL
shown to mitigate pain experienced during punctures.12 A widely available insulin syringe (Terumo Corp., Shibuya-ku, Tokyo, Japan) has an attached 13-mm 29gauge needle that is smaller and shorter (by approximately 1 cm) than the typical hypodermic needle. The insulin needle appears to display some characteristics that can reduce pain. Because this syringe was intended to deliver insulin subcutaneously, there have been no studies to evaluate its suitability to draw blood for laboratory investigations. Previous studies addressing the issue of pain in arterial punctures have recommended the use of local anesthesia before arterial punctures.6,8,13 Similar recommendations have been made by professional bodies.14,15 However, except for anesthesiologists, the compliance rate has been low (2% vs. 60%).16 Some of the reasons cited for noncompliance were perception that local anesthesia is as painful as the arterial puncture itself, distortion of the skin architecture upon intradermal infiltration, and additional time needed to administer the local anesthesia.8,13 We propose the use of the smaller-gauged insulin needle for arterial blood gas sampling without prior administration of local anesthesia, to obviate the additional step and to overcome the barriers to noncompliance. This method can benefit patients and may potentially change practice should it be proven to be less painful. The study hypothesis was that arterial punctures using the insulin needle are less painful than the standard needle. Our primary objective was to compare the difference in the pain score from radial arterial punctures between the 29-gauge insulin needle and the standard 23-gauge hypodermic needle. The secondary outcomes were rate of hemolysis, mean potassium values, and procedural complications. METHODS Study Design This was a randomized controlled crossover trial. The study was approved by the local institutional review board (DSRB 2011/01799) and registered with ClinicalTrials.gov (NCT01996189). Study Setting and Population We conducted the study between May 2012 and March 2013 at the emergency department (ED) of the National University Health System in Singapore. Volunteers were recruited through invitation by e-mail and poster advertisements around the hospital. Adult volunteers age 21 years and above were included. The exclusion criteria were concurrent analgesic use in the preceding 24 hours, presence of painful conditions such as fractures and rheumatoid arthritis, and history of peripheral vascular disease. Study Protocol Randomization, Allocation Concealment, and Blinding. Block randomization was performed to determine whether each volunteer would first receive the arterial puncture in the right or left hand and the type of needle used for the first arterial puncture. The volunteer would then receive the second arterial puncture using the
alternative needle in the contralateral hand. Allocation concealment was achieved using opaque sealed envelopes. The volunteers and the outcome assessors were blinded to the type of needle used by means of a curtained table specially constructed to perform the arterial punctures in this study. Needles and Arterial Punctures. In the intervention arm, the insulin syringe (Terumo, 1.0 mL with attached 29-gauge needle and 13 mm in length) was used. The comparator was the hypodermic needle (B. Braun Sterican 23-gauge needle and 25 mm in length) attached to a 3-mL syringe, henceforth known as the “standard needle.” Each volunteer received two arterial punctures performed by the same doctor approximately 5 minutes apart. Two doctors performed the arterial punctures (II and YWY). The arterial puncture procedure and technique are described as follows: The wrist was immobilized in hyperextension. After the radial artery was located, the injection area was prepared with alcohol swabs. The needle was then inserted into the radial artery at a 45-degree angle and allowed to flow spontaneously to collect 0.5 mL of blood. If gentle suction was needed to facilitate the collection, it was done slowly over 10 to 15 seconds to minimize hemolysis in the blood draw. At least 0.5 mL of blood was drawn, which was the minimum amount required to perform point-ofcare testing in clinical practice. Upon completion of the blood draw, the needle was withdrawn from the skin and the site was compressed for 3 to 5 minutes. For both techniques, the needles were removed from the syringe before the whole blood was emptied slowly over 10 to 15 seconds into a microcentrifuge tube. The microcentrifuge tube was capped and immediately transported to the central laboratory for analysis. Specifically for the insulin syringe, the attached needle was removed using the Becton Dickinson Hub Cutter (BD Worldwide) to minimize the risk of needlestick injuries. Outcomes Measures The primary outcome was the pain score measured on the visual analogue scale (VAS) score. The secondary outcomes were rate of hemolysis, mean potassium values, and procedural complications. VAS for Pain. Immediately following each puncture, each volunteer was asked to assess the pain that was caused by the puncture by placing a mark on a 100-mm VAS sheet. The VAS showed “no pain” at 0 mm and “worst pain” at 100 mm. The outcome assessor (a nurse, another physician, or research assistant) was blinded to the type of needles used. Degree of Hemolysis. At the central laboratory, the microcentrifuge tube was left to stand for at least 30 minutes to allow blood clotting to take place. Subsequently, the whole blood was centrifuged at 3000 9 g for 5 minutes to separate the serum and blood cells. The serum was used to analyze the hemolysis index and potassium value using Ortho Clinical Diagnostics Vitros analyzer (Johnson & Johnson). The hemolysis index has been previously validated against free hemoglobin con-
ACADEMIC EMERGENCY MEDICINE • March 2015, Vol. 22, No. 3 • www.aemj.org
centration.17 We defined the presence of hemolysis as a hemolysis index of more than 120 mg/dL. Potassium values were also measured, as this is the most likely analyte affected in a hemolyzed blood gas sample.18 Procedural Complications. Each volunteer was assessed for complications of excessive bleeding, swelling, and bruising immediately after the procedure by the physician who performed the arterial punctures. For delayed complications, a telephone call was made 24 hours later to assess for development or progression of complications, using a simple questionnaire (Data Supplement S1, available as supporting information in the online version of this paper). Data Analysis The sample size was calculated to detect at least 10-mm difference (with a conservative SD 25 mm) in the VAS with a power of 0.8 and a two-tailed alpha of 0.05 on a crossover design.8 A total sample size of 50 subjects was required. The data were analyzed on PASW Statistics for Windows, version 18.0. Intention-to-treat analysis was followed. To account for the crossover design, a mixed model was applied to compare the differences in VAS scores and potassium values across the insulin needle and standard needle groups, with subjects as random effects and adjusting for the order of needle (first or second order) and hand site (right or left hand that was punctured first). The McNemar test was used to analyze the differences in the rates of hemolysis and procedural complications. Statistical significance was set at p < 0.05. RESULTS Samples Fifty healthy volunteers were recruited. The characteristics of these volunteers are shown in Table 1. For the first puncture, 26 volunteers were allocated to the insulin needle and 24 to the standard needle (Figure 1). One patient withdrew after the first puncture with the standard needle, leaving a total of 49 arterial punctures with the insulin needle and 50 arterial punctures with the standard needle. In 48 of 49 punctures using the insulin needle, there was success in obtaining adequate volume
Table 1 Characteristics of Volunteers (N = 50) Characteristic Age (yr) Male sex Ethnic group Chinese Indian Malay Other Hand laterality Right-handedness Left-handedness Ambidextrous Previous arterial punctures
Number
Mean SD or %
50 22
32 8.5 44
31 12 3 4
62 24 6 8
47 2 1 0
94 4 2 0
317
of blood for analysis, compared to 43 of 50 punctures using the standard needle. The procedure was abandoned if an attempt was unsuccessful, due to increasing pain experienced by the volunteers. However, VAS scores were still obtained from these volunteers. Primary Outcome The mean VAS score was lower in punctures using the insulin needle compared to the standard needle (Table 2). There were no differences in the VAS scores with regard to the order of needles or the hand where the punctures were instituted first. Secondary Outcomes The rate of hemolysis was higher in blood samples obtained with the insulin needle (Table 3). The mean hemolysis indices for insulin and standard needles were 109 62 and 72 137 mg/dL, respectively. The mean potassium value was higher in samples obtained with the insulin needle (Table 2). None of the puncture sites made with the insulin needle developed complications upon completion of the procedures, unlike puncture sites made with the standard needle, where 24% developed bruising. No other complications were observed. At 24 hours, 12 volunteers were lost to follow-up. Of the remaining volunteers, 13 punctures by the standard needle developed complications (three swelling and 10 bruising) compared to two punctures by the insulin needle (one swelling and one bruising). DISCUSSION Our study demonstrated that radial arterial puncture using the insulin needle was less painful than with the standard needle. The controlled crossover design enabled each volunteer to act as his or her own control and obviated factors that may affect the pain score such as age, sex, skill of the doctor performing the procedure, order of needle insertion, and the first hand of arterial puncture (right or left). In addition, using healthy volunteers ensured that certain confounding factors such as concurrent analgesic, painful conditions, and underlying illnesses obtunding pain sensation such as hypercarbia were removed. Arterial punctures with the insulin needle were associated with very low rates of immediate and delayed complications compared to the standard needle. The smaller 29-gauge insulin needle was associated with decreased pain compared to the 23-gauge standard needle. A previous study showed that that the needle size significantly affects pain during insertion, where 63% of volunteers reported painful insertions using 23gauge needles and 52% using 30-gauge needles.19 However, this study involved subcutaneous punctures in the abdomen and thigh and was not specifically used for arterial blood sampling from the radial artery. Our study supports the observation that use of smallergauge needles was associated with less pain even if the needles were used for blood sampling, as evidenced by the 15-mm difference in pain scores. A recent, larger, randomized controlled trial compared two syringes, with outcomes similar to those of our study.20 The heparinized insulin syringe with attached 26-gauge needle
Ibrahim et al. • ARTERIAL PUNCTURE USING INSULIN NEEDLE IS LESS PAINFUL
Enrollment
318
Assessed for eligibility (n = 51)
Excluded (n = 1) concurrent antipyretic
Allocated to insulin needle as 1st puncture (n = 26)
Allocated to standard needle as 1st puncture (n = 24)
•Received allocated intervention (n = 26 )
•Received allocated intervention (n = 24)
Insulin needle as 2nd puncture (n = 24)
Standard needle as 2nd puncture (n = 26)
•Received allocated intervention (n = 23) One withdrew after 1st puncture
•Received allocated intervention (n = 26)
Follow-up
Allocation
Randomized (n = 50)
Analysis
Secondary outcome (24-hr complications) Lost to follow-up (n = 12)
Analyzed (n = 50)
Figure 1. Flow diagram of the study design.
Table 2 Continuous Outcomes Needle Type Outcome Visual analogue scale (mm) Number Mean SD Median (IQR) Range Potassium values (mmol/L) Number Mean sd Median (IQR) Range
Insulin
Standard
49 23 22 15 (6 to 35) 1 to 80
50 39 24 40 (17 to 57) 2 to 99
48 4.6 0.7 4.5 (4.2 to 4.7) 4.0 to 8.5
43 4.2 0.5 4.2 (3.9 to 4.4) 3.5 to 6.2
Adjusted* Mean Difference (95% CI) and p-value
–15 (–22 to –7) p < 0.001
0.4 (0.2 to 0.7) p = 0.002
*Adjusted for order of needle and hand site (both p > 0.05). IQR = interquartile range.
was compared to the new safety-engineered nonheparinized syringe with attached 25-gauge needle. Likely due to the small difference in needle gauges, their results showed no difference in the VAS scores for pain between the two needles, which supports our findings that the difference in pain score may be related to the
difference in the needle gauge. The authors recruited all ED patients who required radial punctures and each patient received a single puncture. In contrast, our study recruited healthy volunteers, each receiving bilateral punctures with different needles. Hence, our study permitted a more “controlled” setting not influenced by
ACADEMIC EMERGENCY MEDICINE • March 2015, Vol. 22, No. 3 • www.aemj.org
319
Table 3 Categorical Outcomes Needle Type Outcome Hemolysis* Number Number of samples hemolyzed (%) Immediate complications Number Number of immediate complications (%) 24-hour complications Number Number of immediate complications (%)
Insulin
Standard
p value
48 15 (31.3)
43 5 (11.6)
0.035
49 0 (0.0)
50 12 (24.0)
120 mg/dL. (There are eight volunteers with unpaired needle comparators due to inadequate blood volume obtained for blood analysis: one using the insulin needle, seven using the standard needle). †Twelve subjects were lost to follow-up and one did not have insulin needle puncture.
patient factors and various medical conditions that may alter pain perception. The insulin needle may be an alternative to using local anesthesia prior to the standard needle in arterial punctures to decrease pain. A previous study found a significant difference in the VAS score of 13 mm (p < 0.0002) between punctures with and without local anesthesia.8 This difference was comparable to that in our study (15 mm). Another study comparing prior local anesthesia with 22-gauge needles against direct puncture using 25-gauge needles showed that punctures with local anesthesia were less painful.6 We postulate that the smaller needle used in our study would yield comparable pain scores with the group receiving local anesthesia first, but this study did not test this assumption. One needs to exercise prudence when using the insulin needle in arterial punctures to obtain a serum potassium value in the same blood draw, because potassium values could be falsely elevated by hemolysis of the blood sample. There was a higher rate of hemolysis from punctures using the insulin needle (31.3% vs. 11.6%) that corresponded to a higher mean potassium value (mean difference = 0.4 mmol/L), with wide confidence intervals (95% CI = 0.2 to 0.7 mmol/L). This could lead to significant clinical errors if the results were relied on disregarding the level of hemolysis. Therefore, the utility of the insulin needle for arterial punctures should be limited to conditions where only the arterial pCO2 and pO2 are required, for example, in patients with chronic obstructive pulmonary disease. Blood draw using a standard needle is recommended in circumstances where the potassium value needs to be determined urgently, such as in patients with renal failure. LIMITATIONS We opted to compare the insulin needle against the standard needle without the use of local analgesia as this was a more common practice among nonanesthesiologists.16 A three-arm trial (arm 1, standard needle; arm 2, insulin needle; and arm 3, local anesthesia prior to standard needle) would be have been ideal. However,
we were limited by the number of punctures that can be instituted in a single volunteer. This circumstance was in contrast to similar studies comparing methods of venipuncture or venous cannulation where more puncture sites were available. The doctors who performed the arterial punctures were not blinded to the type of needle used, which can contribute to bias in the outcomes observed. However, we took careful steps to ensure that the patient and the outcomes assessor were blinded during assessment of the primary outcome. There was a possibility that the rate of blood draw using the two needles was not equivalent and may have contributed to the differences in pain score. Blood draws using the smaller-gauge insulin needle may have subconsciously been slower compared to the standard needle and could have led to lower pain scores. For practical purposes, it was not possible to standardize the rate of blood draw. The amount of blood drawn to qualify for a successful attempt in our study was small (0.5 mL). In clinical practice, this amount is sufficient to perform the point-ofcare testing where only two to three drops of blood (200 to 300 lL) are required. This method may not be suitable in clinical areas where the blood samples are required to be transported to the laboratory for analysis. We did not assess if the blood sampled was actually arterial in origin. Although a blood gas estimation for all samples would have been ideal, we believe that inadvertent venous puncture was less likely in healthy volunteers with no history of peripheral vascular disease. There is a theoretical possibility that hemolysis in the blood samples can affect arterial blood gas analysis, thus limiting the use of insulin needle in arterial punctures. A previous study of hemolyzed venous blood showed approximately 5% differences in the pO2 and pCO2 values.18 However, the degree of hemolysis in the samples studied was greater than in our study (8,900 mg/dL vs. 109 mg/dL). Our study protocol required blood samples to stand for at least 30 minutes to assess for hemolysis that could have paradoxically increased the rate of hemolysis from both needle
320
Ibrahim et al. • ARTERIAL PUNCTURE USING INSULIN NEEDLE IS LESS PAINFUL
samples. In clinical practice, point-of-care testing is typically done on a whole blood analyzer shortly after sampling. Hence, the degree of hemolysis in real-life practice is likely to be less severe, allowing for the use of insulin needles in radial arterial punctures.
10.
11. CONCLUSIONS Arterial punctures using the insulin needle cause less pain and fewer procedural complications compared to the standard needle. Reduction in pain may translate to better patient satisfaction and tolerance, particularly if repeated arterial punctures are required during the hospital stay. However, due to the higher rate of hemolysis, its use should be limited to conditions that do not require a concurrent potassium value in the same blood sample. The authors thank the staff of National University Hospital Emergency Medicine Department for their support, particularly Associate Professor Sunil Sethi, Clinical Associate Professor Malcolm Mahadevan and Dr. Seet Chong Meng for their advice in this study.
References 1. Kelly AM. Agreement between arterial and venous blood gases in emergency medical care: a systematic review. Hong Kong J Emerg Med 2013;20:166–71. 2. Byrne AL, Bennett M, Chatterji R, Symons R, Pace NL, Thomas PS. Peripheral venous and arterial blood gas analysis in adults: are they comparable? A systematic review and meta-analysis. Respirology 2014;19:168–75. 3. Malatesha G, Singh NK, Bharija A, Goel A. Comparison of arterial and venous pH, bicarbonate, PCO2 and PO2 in initial emergency department assessment. Emerg Med J 2007;24:569–71. 4. Ma OJ, Rush MD, Godfrey MM, Gaddis G. Arterial blood gas results rarely influence emergency physician management of patients with suspected diabetic ketoacidosis. Acad Emerg Med 2003;10:836–41. 5. Ak A, Ogun CO, Bayir A, Kayis SA, Koylu R. Prediction of arterial blood gas values from venous blood gas values in patients with acute exacerbation of chronic obstructive pulmonary disease. Tohoku J Exp Med 2006;210:285–90. 6. Giner J, Casan P, Sanchis J. Sampling arterial blood with a fine needle. Chest 1997;111:1474. 7. Dev SP, Hillmer MD, Ferri M. Arterial puncture for blood gas analysis. N Engl J Med 2011;364:E7–9. 8. Giner J, Casan P, Belda J, Gonzalez M, Miralda RM, Sanchis J. Pain during arterial puncture. Chest 1996;110:1443–5. 9. Riley JL, Robinson ME, Wise EA, Myers CD, Fillingim RB. Sex differences in the perception of noxious
12. 13.
14.
15.
16.
17.
18.
19.
20.
experimental stimuli: a meta-analysis. Pain 1998;74: 181–7. Rowell LN, Mechlin B, Ji E, Addamo M, Girdler SS. Asians differ from non-Hispanic whites in experimental pain sensitivity. Eur J Pain 2011;15:764–71. Pud D, Golan Y, Pesta R. Hand dominancy–a feature affecting sensitivity to pain. Neurosci Lett 2009;467:237–40. Gill HS, Prausnitz MR. Does needle size matter? J Diabet Sci Tech 2007;1:725–9. Lightowler JV, Elliott MW. Local anaesthetic infiltration prior to arterial puncture for blood gas analysis: a survey of current practice and a randomised double blind placebo controlled trial. J R Coll Phys Lond 1997;31:645–6. O’Driscoll BR, Howard LS, Davison AG; British Thoracic Society. BTS guideline for emergency oxygen use in adult patients. Thorax 2008;63:vi1–68. [No authors listed]. [Normative of arterial gas analysis. Working Group of the SEPAR for the practice of arterial gas analysis] [article in Spanish]. Arch Bronconeumol 1998;34:142–53. Sado DM, Deakin CD. Local anaesthesia for venous cannulation and arterial blood gas sampling: are doctors using it? J R Soc Med 2005;98:158–60. Khoo R, Wang C, Saw S, Sethi S. Automated Serum Indices Measurement on Vitros 5600R Integrated System. American Association for Clinical Chemistry Conference, Chicago, IL, July 2009 Lippi G, Fontana R, Avanzini P, Sandei F, Ippolito L. Influence of spurious hemolysis on blood gas analysis. Clin Chem Lab Med 2013;51:1651–4. Arendt-Nielsen L, Egekvist H, Bjerring P. Pain following controlled cutaneous insertion of needles with different diameters. Somatosens Mot Res 2006;23:37–43. Baskin SB, Oray NC, Yanturali S, Bayram B. The comparison of heparinized insulin syringes and safety-engineered blood gas syringes used in arterial blood gas sampling in the ED setting (randomized controlled study). Am J Emerg Med 2014;32:432–7.
Supporting Information The following supporting information is available in the online version of this paper: Data Supplement S1. ALLAY study: arterial puncture using insulin syringe is less painful than standard syringe, a randomised crossover trial.
Arterial Puncture Using Insulin Needle Is Less Painful Than With Standard Needle: A Randomized Crossover Study Irwani Ibrahim, MD, Ying Wei Yau, MD, Lizhen Ong, MD, Yiong Huak Chan, PhD, and Win Sen Kuan, MD
Abstract Objectives: Arterial punctures are important procedures performed by emergency physicians in the assessment of ill patients. However, arterial punctures are painful and can create anxiety and needle phobia in patients. The pain score of radial arterial punctures were compared between the insulin needle and the standard 23-gauge hypodermic needle. Methods: In a randomized controlled crossover design, healthy volunteers were recruited to undergo bilateral radial arterial punctures. They were assigned to receive either the insulin or the standard needle as the first puncture, using blocked randomization. The primary outcome was the pain score measured on a 100-mm visual analogue scale (VAS) for pain, and secondary outcomes were rate of hemolysis, mean potassium values, and procedural complications immediately and 24 hours postprocedure. Results: Fifty healthy volunteers were included in the study. The mean (standard deviation) VAS score in punctures with the insulin needle was lower than the standard needle (23 22 mm vs. 39 24 mm; mean difference = –15 mm; 95% confidence interval = –22 mm to –7 mm; p < 0.001). The rates of hemolysis and mean potassium value were greater in samples obtained using the insulin needle compared to the standard needle (31.3% vs. 11.6%, p = 0.035; and 4.6 0.7 mmol/L vs. 4.2 0.5 mmol/L, p = 0.002). Procedural complications were lower in punctures with the insulin needle both immediately postprocedure (0% vs. 24%; p < 0.001) and at 24 hours postprocedure (5.4% vs. 34.2%; p = 0.007). Conclusions: Arterial punctures using insulin needles cause less pain and fewer procedural complications compared to standard needles. However, due to the higher rate of hemolysis, its use should be limited to conditions that do not require a concurrent potassium value in the same blood sample. ACADEMIC EMERGENCY MEDICINE 2015;22:315–320 © 2015 by the Society for Academic Emergency Medicine
INTRODUCTION
R
ecent systematic reviews have showed that venous blood gas is comparable with arterial blood gas in pH and bicarbonate estimations.1,2 These findings support the increasing use of venous blood gas to replace arterial blood gas by emergency physicians.3,4 However, there is a poor correlation of pCO2 and pO2 between venous and arterial blood.1,2 Therefore, there is still a need for emergency physicians to perform radial arterial punctures to evaluate ill
patients where knowledge of pCO2 and pO2 are crucial to guide management. In current practice, these punctures are usually performed using hypodermic needles with sizes ranging from 22 to 25 gauge.5–7 Arterial punctures are painful and can create anxiety and phobia of needles in patients.8 The perception of pain is influenced by both individual and technical factors. Differences in pain perception were observed between different sexes, ethnic groups, and hand laterality.9–11 Needle characteristics such as small gauge, short length, and sharp tip have also been
From the Emergency Medicine Department and Department of Surgery (II, YWY, WSK), the Department of Laboratory Medicine (LO), and the Biostatistics Department (YHC), National University Health System, Singapore. Received July 8, 2014; revision received September 17, 2014; accepted September 22, 2014. Presented at the American College of Emergency Physicians Scientific Assembly, Seattle, WA, October 2013. The study was supported by the National University Health Systems Junior Pitch for Funds. The authors have no conflicts of interest to declare. Supervising Editor: Brian Hiestand, MD, MPH. Address for correspondence and reprints: Irwani Ibrahim, MD; e-mail: [email protected].
© 2015 by the Society for Academic Emergency Medicine doi: 10.1111/acem.12601
ISSN 1069-6563 PII ISSN 1069-6563583
315 315
316
Ibrahim et al. • ARTERIAL PUNCTURE USING INSULIN NEEDLE IS LESS PAINFUL
shown to mitigate pain experienced during punctures.12 A widely available insulin syringe (Terumo Corp., Shibuya-ku, Tokyo, Japan) has an attached 13-mm 29gauge needle that is smaller and shorter (by approximately 1 cm) than the typical hypodermic needle. The insulin needle appears to display some characteristics that can reduce pain. Because this syringe was intended to deliver insulin subcutaneously, there have been no studies to evaluate its suitability to draw blood for laboratory investigations. Previous studies addressing the issue of pain in arterial punctures have recommended the use of local anesthesia before arterial punctures.6,8,13 Similar recommendations have been made by professional bodies.14,15 However, except for anesthesiologists, the compliance rate has been low (2% vs. 60%).16 Some of the reasons cited for noncompliance were perception that local anesthesia is as painful as the arterial puncture itself, distortion of the skin architecture upon intradermal infiltration, and additional time needed to administer the local anesthesia.8,13 We propose the use of the smaller-gauged insulin needle for arterial blood gas sampling without prior administration of local anesthesia, to obviate the additional step and to overcome the barriers to noncompliance. This method can benefit patients and may potentially change practice should it be proven to be less painful. The study hypothesis was that arterial punctures using the insulin needle are less painful than the standard needle. Our primary objective was to compare the difference in the pain score from radial arterial punctures between the 29-gauge insulin needle and the standard 23-gauge hypodermic needle. The secondary outcomes were rate of hemolysis, mean potassium values, and procedural complications. METHODS Study Design This was a randomized controlled crossover trial. The study was approved by the local institutional review board (DSRB 2011/01799) and registered with ClinicalTrials.gov (NCT01996189). Study Setting and Population We conducted the study between May 2012 and March 2013 at the emergency department (ED) of the National University Health System in Singapore. Volunteers were recruited through invitation by e-mail and poster advertisements around the hospital. Adult volunteers age 21 years and above were included. The exclusion criteria were concurrent analgesic use in the preceding 24 hours, presence of painful conditions such as fractures and rheumatoid arthritis, and history of peripheral vascular disease. Study Protocol Randomization, Allocation Concealment, and Blinding. Block randomization was performed to determine whether each volunteer would first receive the arterial puncture in the right or left hand and the type of needle used for the first arterial puncture. The volunteer would then receive the second arterial puncture using the
alternative needle in the contralateral hand. Allocation concealment was achieved using opaque sealed envelopes. The volunteers and the outcome assessors were blinded to the type of needle used by means of a curtained table specially constructed to perform the arterial punctures in this study. Needles and Arterial Punctures. In the intervention arm, the insulin syringe (Terumo, 1.0 mL with attached 29-gauge needle and 13 mm in length) was used. The comparator was the hypodermic needle (B. Braun Sterican 23-gauge needle and 25 mm in length) attached to a 3-mL syringe, henceforth known as the “standard needle.” Each volunteer received two arterial punctures performed by the same doctor approximately 5 minutes apart. Two doctors performed the arterial punctures (II and YWY). The arterial puncture procedure and technique are described as follows: The wrist was immobilized in hyperextension. After the radial artery was located, the injection area was prepared with alcohol swabs. The needle was then inserted into the radial artery at a 45-degree angle and allowed to flow spontaneously to collect 0.5 mL of blood. If gentle suction was needed to facilitate the collection, it was done slowly over 10 to 15 seconds to minimize hemolysis in the blood draw. At least 0.5 mL of blood was drawn, which was the minimum amount required to perform point-ofcare testing in clinical practice. Upon completion of the blood draw, the needle was withdrawn from the skin and the site was compressed for 3 to 5 minutes. For both techniques, the needles were removed from the syringe before the whole blood was emptied slowly over 10 to 15 seconds into a microcentrifuge tube. The microcentrifuge tube was capped and immediately transported to the central laboratory for analysis. Specifically for the insulin syringe, the attached needle was removed using the Becton Dickinson Hub Cutter (BD Worldwide) to minimize the risk of needlestick injuries. Outcomes Measures The primary outcome was the pain score measured on the visual analogue scale (VAS) score. The secondary outcomes were rate of hemolysis, mean potassium values, and procedural complications. VAS for Pain. Immediately following each puncture, each volunteer was asked to assess the pain that was caused by the puncture by placing a mark on a 100-mm VAS sheet. The VAS showed “no pain” at 0 mm and “worst pain” at 100 mm. The outcome assessor (a nurse, another physician, or research assistant) was blinded to the type of needles used. Degree of Hemolysis. At the central laboratory, the microcentrifuge tube was left to stand for at least 30 minutes to allow blood clotting to take place. Subsequently, the whole blood was centrifuged at 3000 9 g for 5 minutes to separate the serum and blood cells. The serum was used to analyze the hemolysis index and potassium value using Ortho Clinical Diagnostics Vitros analyzer (Johnson & Johnson). The hemolysis index has been previously validated against free hemoglobin con-
ACADEMIC EMERGENCY MEDICINE • March 2015, Vol. 22, No. 3 • www.aemj.org
centration.17 We defined the presence of hemolysis as a hemolysis index of more than 120 mg/dL. Potassium values were also measured, as this is the most likely analyte affected in a hemolyzed blood gas sample.18 Procedural Complications. Each volunteer was assessed for complications of excessive bleeding, swelling, and bruising immediately after the procedure by the physician who performed the arterial punctures. For delayed complications, a telephone call was made 24 hours later to assess for development or progression of complications, using a simple questionnaire (Data Supplement S1, available as supporting information in the online version of this paper). Data Analysis The sample size was calculated to detect at least 10-mm difference (with a conservative SD 25 mm) in the VAS with a power of 0.8 and a two-tailed alpha of 0.05 on a crossover design.8 A total sample size of 50 subjects was required. The data were analyzed on PASW Statistics for Windows, version 18.0. Intention-to-treat analysis was followed. To account for the crossover design, a mixed model was applied to compare the differences in VAS scores and potassium values across the insulin needle and standard needle groups, with subjects as random effects and adjusting for the order of needle (first or second order) and hand site (right or left hand that was punctured first). The McNemar test was used to analyze the differences in the rates of hemolysis and procedural complications. Statistical significance was set at p < 0.05. RESULTS Samples Fifty healthy volunteers were recruited. The characteristics of these volunteers are shown in Table 1. For the first puncture, 26 volunteers were allocated to the insulin needle and 24 to the standard needle (Figure 1). One patient withdrew after the first puncture with the standard needle, leaving a total of 49 arterial punctures with the insulin needle and 50 arterial punctures with the standard needle. In 48 of 49 punctures using the insulin needle, there was success in obtaining adequate volume
Table 1 Characteristics of Volunteers (N = 50) Characteristic Age (yr) Male sex Ethnic group Chinese Indian Malay Other Hand laterality Right-handedness Left-handedness Ambidextrous Previous arterial punctures
Number
Mean SD or %
50 22
32 8.5 44
31 12 3 4
62 24 6 8
47 2 1 0
94 4 2 0
317
of blood for analysis, compared to 43 of 50 punctures using the standard needle. The procedure was abandoned if an attempt was unsuccessful, due to increasing pain experienced by the volunteers. However, VAS scores were still obtained from these volunteers. Primary Outcome The mean VAS score was lower in punctures using the insulin needle compared to the standard needle (Table 2). There were no differences in the VAS scores with regard to the order of needles or the hand where the punctures were instituted first. Secondary Outcomes The rate of hemolysis was higher in blood samples obtained with the insulin needle (Table 3). The mean hemolysis indices for insulin and standard needles were 109 62 and 72 137 mg/dL, respectively. The mean potassium value was higher in samples obtained with the insulin needle (Table 2). None of the puncture sites made with the insulin needle developed complications upon completion of the procedures, unlike puncture sites made with the standard needle, where 24% developed bruising. No other complications were observed. At 24 hours, 12 volunteers were lost to follow-up. Of the remaining volunteers, 13 punctures by the standard needle developed complications (three swelling and 10 bruising) compared to two punctures by the insulin needle (one swelling and one bruising). DISCUSSION Our study demonstrated that radial arterial puncture using the insulin needle was less painful than with the standard needle. The controlled crossover design enabled each volunteer to act as his or her own control and obviated factors that may affect the pain score such as age, sex, skill of the doctor performing the procedure, order of needle insertion, and the first hand of arterial puncture (right or left). In addition, using healthy volunteers ensured that certain confounding factors such as concurrent analgesic, painful conditions, and underlying illnesses obtunding pain sensation such as hypercarbia were removed. Arterial punctures with the insulin needle were associated with very low rates of immediate and delayed complications compared to the standard needle. The smaller 29-gauge insulin needle was associated with decreased pain compared to the 23-gauge standard needle. A previous study showed that that the needle size significantly affects pain during insertion, where 63% of volunteers reported painful insertions using 23gauge needles and 52% using 30-gauge needles.19 However, this study involved subcutaneous punctures in the abdomen and thigh and was not specifically used for arterial blood sampling from the radial artery. Our study supports the observation that use of smallergauge needles was associated with less pain even if the needles were used for blood sampling, as evidenced by the 15-mm difference in pain scores. A recent, larger, randomized controlled trial compared two syringes, with outcomes similar to those of our study.20 The heparinized insulin syringe with attached 26-gauge needle
Ibrahim et al. • ARTERIAL PUNCTURE USING INSULIN NEEDLE IS LESS PAINFUL
Enrollment
318
Assessed for eligibility (n = 51)
Excluded (n = 1) concurrent antipyretic
Allocated to insulin needle as 1st puncture (n = 26)
Allocated to standard needle as 1st puncture (n = 24)
•Received allocated intervention (n = 26 )
•Received allocated intervention (n = 24)
Insulin needle as 2nd puncture (n = 24)
Standard needle as 2nd puncture (n = 26)
•Received allocated intervention (n = 23) One withdrew after 1st puncture
•Received allocated intervention (n = 26)
Follow-up
Allocation
Randomized (n = 50)
Analysis
Secondary outcome (24-hr complications) Lost to follow-up (n = 12)
Analyzed (n = 50)
Figure 1. Flow diagram of the study design.
Table 2 Continuous Outcomes Needle Type Outcome Visual analogue scale (mm) Number Mean SD Median (IQR) Range Potassium values (mmol/L) Number Mean sd Median (IQR) Range
Insulin
Standard
49 23 22 15 (6 to 35) 1 to 80
50 39 24 40 (17 to 57) 2 to 99
48 4.6 0.7 4.5 (4.2 to 4.7) 4.0 to 8.5
43 4.2 0.5 4.2 (3.9 to 4.4) 3.5 to 6.2
Adjusted* Mean Difference (95% CI) and p-value
–15 (–22 to –7) p < 0.001
0.4 (0.2 to 0.7) p = 0.002
*Adjusted for order of needle and hand site (both p > 0.05). IQR = interquartile range.
was compared to the new safety-engineered nonheparinized syringe with attached 25-gauge needle. Likely due to the small difference in needle gauges, their results showed no difference in the VAS scores for pain between the two needles, which supports our findings that the difference in pain score may be related to the
difference in the needle gauge. The authors recruited all ED patients who required radial punctures and each patient received a single puncture. In contrast, our study recruited healthy volunteers, each receiving bilateral punctures with different needles. Hence, our study permitted a more “controlled” setting not influenced by
ACADEMIC EMERGENCY MEDICINE • March 2015, Vol. 22, No. 3 • www.aemj.org
319
Table 3 Categorical Outcomes Needle Type Outcome Hemolysis* Number Number of samples hemolyzed (%) Immediate complications Number Number of immediate complications (%) 24-hour complications Number Number of immediate complications (%)
Insulin
Standard
p value
48 15 (31.3)
43 5 (11.6)
0.035
49 0 (0.0)
50 12 (24.0)
120 mg/dL. (There are eight volunteers with unpaired needle comparators due to inadequate blood volume obtained for blood analysis: one using the insulin needle, seven using the standard needle). †Twelve subjects were lost to follow-up and one did not have insulin needle puncture.
patient factors and various medical conditions that may alter pain perception. The insulin needle may be an alternative to using local anesthesia prior to the standard needle in arterial punctures to decrease pain. A previous study found a significant difference in the VAS score of 13 mm (p < 0.0002) between punctures with and without local anesthesia.8 This difference was comparable to that in our study (15 mm). Another study comparing prior local anesthesia with 22-gauge needles against direct puncture using 25-gauge needles showed that punctures with local anesthesia were less painful.6 We postulate that the smaller needle used in our study would yield comparable pain scores with the group receiving local anesthesia first, but this study did not test this assumption. One needs to exercise prudence when using the insulin needle in arterial punctures to obtain a serum potassium value in the same blood draw, because potassium values could be falsely elevated by hemolysis of the blood sample. There was a higher rate of hemolysis from punctures using the insulin needle (31.3% vs. 11.6%) that corresponded to a higher mean potassium value (mean difference = 0.4 mmol/L), with wide confidence intervals (95% CI = 0.2 to 0.7 mmol/L). This could lead to significant clinical errors if the results were relied on disregarding the level of hemolysis. Therefore, the utility of the insulin needle for arterial punctures should be limited to conditions where only the arterial pCO2 and pO2 are required, for example, in patients with chronic obstructive pulmonary disease. Blood draw using a standard needle is recommended in circumstances where the potassium value needs to be determined urgently, such as in patients with renal failure. LIMITATIONS We opted to compare the insulin needle against the standard needle without the use of local analgesia as this was a more common practice among nonanesthesiologists.16 A three-arm trial (arm 1, standard needle; arm 2, insulin needle; and arm 3, local anesthesia prior to standard needle) would be have been ideal. However,
we were limited by the number of punctures that can be instituted in a single volunteer. This circumstance was in contrast to similar studies comparing methods of venipuncture or venous cannulation where more puncture sites were available. The doctors who performed the arterial punctures were not blinded to the type of needle used, which can contribute to bias in the outcomes observed. However, we took careful steps to ensure that the patient and the outcomes assessor were blinded during assessment of the primary outcome. There was a possibility that the rate of blood draw using the two needles was not equivalent and may have contributed to the differences in pain score. Blood draws using the smaller-gauge insulin needle may have subconsciously been slower compared to the standard needle and could have led to lower pain scores. For practical purposes, it was not possible to standardize the rate of blood draw. The amount of blood drawn to qualify for a successful attempt in our study was small (0.5 mL). In clinical practice, this amount is sufficient to perform the point-ofcare testing where only two to three drops of blood (200 to 300 lL) are required. This method may not be suitable in clinical areas where the blood samples are required to be transported to the laboratory for analysis. We did not assess if the blood sampled was actually arterial in origin. Although a blood gas estimation for all samples would have been ideal, we believe that inadvertent venous puncture was less likely in healthy volunteers with no history of peripheral vascular disease. There is a theoretical possibility that hemolysis in the blood samples can affect arterial blood gas analysis, thus limiting the use of insulin needle in arterial punctures. A previous study of hemolyzed venous blood showed approximately 5% differences in the pO2 and pCO2 values.18 However, the degree of hemolysis in the samples studied was greater than in our study (8,900 mg/dL vs. 109 mg/dL). Our study protocol required blood samples to stand for at least 30 minutes to assess for hemolysis that could have paradoxically increased the rate of hemolysis from both needle
320
Ibrahim et al. • ARTERIAL PUNCTURE USING INSULIN NEEDLE IS LESS PAINFUL
samples. In clinical practice, point-of-care testing is typically done on a whole blood analyzer shortly after sampling. Hence, the degree of hemolysis in real-life practice is likely to be less severe, allowing for the use of insulin needles in radial arterial punctures.
10.
11. CONCLUSIONS Arterial punctures using the insulin needle cause less pain and fewer procedural complications compared to the standard needle. Reduction in pain may translate to better patient satisfaction and tolerance, particularly if repeated arterial punctures are required during the hospital stay. However, due to the higher rate of hemolysis, its use should be limited to conditions that do not require a concurrent potassium value in the same blood sample. The authors thank the staff of National University Hospital Emergency Medicine Department for their support, particularly Associate Professor Sunil Sethi, Clinical Associate Professor Malcolm Mahadevan and Dr. Seet Chong Meng for their advice in this study.
References 1. Kelly AM. Agreement between arterial and venous blood gases in emergency medical care: a systematic review. Hong Kong J Emerg Med 2013;20:166–71. 2. Byrne AL, Bennett M, Chatterji R, Symons R, Pace NL, Thomas PS. Peripheral venous and arterial blood gas analysis in adults: are they comparable? A systematic review and meta-analysis. Respirology 2014;19:168–75. 3. Malatesha G, Singh NK, Bharija A, Goel A. Comparison of arterial and venous pH, bicarbonate, PCO2 and PO2 in initial emergency department assessment. Emerg Med J 2007;24:569–71. 4. Ma OJ, Rush MD, Godfrey MM, Gaddis G. Arterial blood gas results rarely influence emergency physician management of patients with suspected diabetic ketoacidosis. Acad Emerg Med 2003;10:836–41. 5. Ak A, Ogun CO, Bayir A, Kayis SA, Koylu R. Prediction of arterial blood gas values from venous blood gas values in patients with acute exacerbation of chronic obstructive pulmonary disease. Tohoku J Exp Med 2006;210:285–90. 6. Giner J, Casan P, Sanchis J. Sampling arterial blood with a fine needle. Chest 1997;111:1474. 7. Dev SP, Hillmer MD, Ferri M. Arterial puncture for blood gas analysis. N Engl J Med 2011;364:E7–9. 8. Giner J, Casan P, Belda J, Gonzalez M, Miralda RM, Sanchis J. Pain during arterial puncture. Chest 1996;110:1443–5. 9. Riley JL, Robinson ME, Wise EA, Myers CD, Fillingim RB. Sex differences in the perception of noxious
12. 13.
14.
15.
16.
17.
18.
19.
20.
experimental stimuli: a meta-analysis. Pain 1998;74: 181–7. Rowell LN, Mechlin B, Ji E, Addamo M, Girdler SS. Asians differ from non-Hispanic whites in experimental pain sensitivity. Eur J Pain 2011;15:764–71. Pud D, Golan Y, Pesta R. Hand dominancy–a feature affecting sensitivity to pain. Neurosci Lett 2009;467:237–40. Gill HS, Prausnitz MR. Does needle size matter? J Diabet Sci Tech 2007;1:725–9. Lightowler JV, Elliott MW. Local anaesthetic infiltration prior to arterial puncture for blood gas analysis: a survey of current practice and a randomised double blind placebo controlled trial. J R Coll Phys Lond 1997;31:645–6. O’Driscoll BR, Howard LS, Davison AG; British Thoracic Society. BTS guideline for emergency oxygen use in adult patients. Thorax 2008;63:vi1–68. [No authors listed]. [Normative of arterial gas analysis. Working Group of the SEPAR for the practice of arterial gas analysis] [article in Spanish]. Arch Bronconeumol 1998;34:142–53. Sado DM, Deakin CD. Local anaesthesia for venous cannulation and arterial blood gas sampling: are doctors using it? J R Soc Med 2005;98:158–60. Khoo R, Wang C, Saw S, Sethi S. Automated Serum Indices Measurement on Vitros 5600R Integrated System. American Association for Clinical Chemistry Conference, Chicago, IL, July 2009 Lippi G, Fontana R, Avanzini P, Sandei F, Ippolito L. Influence of spurious hemolysis on blood gas analysis. Clin Chem Lab Med 2013;51:1651–4. Arendt-Nielsen L, Egekvist H, Bjerring P. Pain following controlled cutaneous insertion of needles with different diameters. Somatosens Mot Res 2006;23:37–43. Baskin SB, Oray NC, Yanturali S, Bayram B. The comparison of heparinized insulin syringes and safety-engineered blood gas syringes used in arterial blood gas sampling in the ED setting (randomized controlled study). Am J Emerg Med 2014;32:432–7.
Supporting Information The following supporting information is available in the online version of this paper: Data Supplement S1. ALLAY study: arterial puncture using insulin syringe is less painful than standard syringe, a randomised crossover trial.
