Abstract
BACKGROUND: Arterial puncture is often painful for patients. The aim of this study was to compare use of local anesthesia as a eutectic mixture of 2 local anesthetics, lidocaine and prilocaine, versus placebo.
METHODS: We conducted a double-blind, randomized controlled trial. Subjects were eligible if arterial puncture was indicated. The primary outcome was an experienced pain > 2 on a numerical pain rating scale. As having had a previous experience of arterial puncture was expected to be predictive of the current response, we planned 3 comparisons between use of local anesthesia and placebo: in the whole sample, among subjects with a painful previous experience, and among subjects with a painless previous experience. Multiple testing was analyzed using the Bonferroni correction for the primary outcome. The secondary outcome was the numerical pain rating scale score itself. All analyses were performed on an intention-to-treat basis.
RESULTS: A total of 136 subjects were included in this study. The primary outcome occurred in 20.9% in the active arm versus 37.7% in the placebo arm in the whole sample (relative risk 0.55; 95% CI when adjusting for multiple testing ranged was 0.28–1.09, P = .10; 95% CI without adjustment was 0.32–0.97, P = .038). No significant heterogeneity in the study treatment effect was found when considering previous painful or painless arterial puncture (P = .70). The numerical pain rating scale score was 1.55 ± 2.03 in active group versus 2.09 ± 2.15 in the placebo group (P = .13).
CONCLUSIONS: We found that application of a eutectic mixture reduced the number of painful arterial punctures by 50% compared with placebo. However, this result was not statistically significant. (ClinicalTrials.gov registration NCT01964248.)
Introduction
Arterial blood gas (ABG) testing is used to analyze patients' hematosis and acid/base state. Arterial puncture for ABG is often challenging for nursing staff and painful for patients. However, few providers use local anesthesia during arterial puncture.1,2
An observational study was carried out in our unit from 2006 to 2010 to evaluate the pain intensity of arterial puncture for ABG and to determine pain predictors. According to this study, no criterion seemed to predict pain, and pain intensity was highly variable between subjects.
Topical anesthetics reversibly block nerve conduction near the site of administration by targeting free nerve endings in the dermis or mucosa, thereby producing temporary loss of sensation in a limited area. To prevent patient pain, anesthetics could be administrated topically or through infiltration. However, injections of local anesthetics are painful3,4 and can worsen needle anxiety or cause tissue edema.5
Different types of anesthetics and routes of administration have been evaluated with discordant results regarding effective pain relief.6-12 France et al3 compared the use of subcutaneous lidocaine and ethyl chloride versus no treatment on pain intensity during arterial puncture. Ethyl chloride was not found to reduce pain, and pain felt during lidocaine injection was similar to that of arterial puncture without anesthetic, suggesting a limited benefit of subcutaneous lidocaine.3 In another study, vapocoolant spray (ethyl chloride) did not reduce pain during arterial puncture.13 Bobbia et al14 compared ultrasound-guided arterial puncture versus conventional sampling. The authors reported that ultrasonography increased the number and the duration of the procedure but had no effect on pain.14 The gauge of the needle may be associated with the degree of pain during arterial puncture, but the results are controversial. Patout et al15 compared pain experienced during arterial punctures performed with 23 French or 25 French needles, but the authors reported that the needle size had no significant impact on pain felt during arterial puncture.15 Conversely, Ibrahim et al16 compared the standard 23 French needle with an insulin needle and reported that arterial puncture using insulin needles was less painful than using standard needles.
A eutectic mixture of lidocaine and prilocaine, local anesthetics of the amide group, applied to diffuse into the stratum corneum, epidermis, and dermis to reach the superficial nerve endings, is commonly used for percutaneous anesthesia of healthy skin during blood sampling.17 This anesthetic, topically administered, has the potential to decrease pain experienced during arterial puncture; however, the efficacy of this method has not been demonstrated yet.
Thus, we conducted a randomized controlled trial comparing local anesthesia with a eutectic mixture of 2 local anesthetics, lidocaine and prilocaine, versus placebo. The primary objective was to demonstrate that local anesthesia with a eutectic mixture of lidocaine and prilocaine applied before arterial puncture for ABG allows a greater reduction of frequency of painful arterial puncture compared to placebo. The secondary objectives were to compare the pain felt by subjects between both groups and in subgroup analyses (ie, subjects with unpleasant or painful memories of prior arterial puncture for ABG versus subjects without such memories).
Quick Look
Current Knowledge
Arterial puncture is painful. Application of a Eutectic mixture of lidocaine and prilocaine is commonly used for percutaneous anesthesia of healthy skin during sampling blood, without evidence of benefit.
What This Paper Contributes to Our Knowledge
Application of a Eutectic mixture of lidocaine and prilocaine allowed us to reduce the number of painful arterial punctures by half in comparison with placebo, without significant differences between groups. Our results provide an evidence-based answer to the question about the use of local anesthesia for pain reduction during arterial punctures.
Methods
Study population and ethic review
Between December 12, 2012, and October 2, Study population and ethic review 2015, patients hospitalized in or referred to the Pneumology Department were eligible if they required arterial puncture for ABG. Patients were not included if they met any of the following criteria: age < 18 y, inability to report a pain score, unfeasible radial arterial sampling, patients with pain > 0 on a numerical pain rating scale before the arterial puncture, known hypersensitivity to amide-bonded group local anesthetics or to any other component of the lidocaine/prilocaine cream, known porphyria, pregnancy, or the absence of written informed consent.
This single-center, double-blind (subjects, caregivers, investigators, and outcome assessors), parallel-group, randomized clinical trial was approved by the ethics board (CPP Ouest 6, n°2012-000489-39) in February 2012. Written informed consent was obtained from all subjects before randomization by physician investigators.
Randomization, Masking, and Interventions
At the time of inclusion, subjects were randomized via a computer-generated blind-fashion assignment sequence to receive either prilocaine/lidocaine cream (ie, the anesthetic group) or the placebo cream (ie, the placebo group). Nurses in the respiratory care unit administered all study treatments and performed all of the arterial punctures.
The active treatment evaluated was a lidocaine/prilocaine 5% cream marketed by Aguettant Laboratories (Lyon, France). The placebo, chosen for its identical texture, color, and smell to preserve blinding, was Excipial Hydrocreme (Spirig Pharma, Egerkingen, Switzerland). The tested treatment was prepared and blinded by the hospital pharmacy in identical single-dose tubes of 2 g. To obtain an anesthetic effect of at least 5 min, the cream was applied 2 h before the puncture at a radial artery perception site. The amount of cream to be applied was defined by the single-dose tube. The dose of cream was then covered with a transparent adhesive film. Other local treatments applied at the arterial puncture site were not allowed to avoid any risk of interactions with the evaluated treatments. Arterial puncture for ABG was performed using standardized 23 French needles in both groups.
Endpoints
The primary end point was an experienced pain > 2 (yes/no) on a numerical pain rating scale from 0 to 10. Pain intensity was measured on a numerical pain scale immediately before and just after the arterial puncture. The scale chosen was recommended by the institution's Pain Control Committee. No pain was represented as the 0 of the scale, with worst imaginable pain as the 10 of the scale. Because having had a previous experience of arterial puncture was expected to be predictive of the current response, we planned 3 comparisons between local anesthesia and placebo groups: in the whole sample, among subjects with a painful previous experience, and among subjects with a painless previous experience.
Before the arterial puncture, previous arterial puncture and related feelings (ie, the nurse asked subjects if they had ever had a previous arterial puncture and, if yes, their sensation during this previous puncture) were recorded, as were subjects’ apprehension, physical condition, sample cutaneous conditions of realization, and baseline characteristics. After the arterial puncture, procedural difficulty, success, and subjects’ treatment tolerance were collected. The secondary end point was the numerical pain rating scale score itself from 0 to 10.
Sample Size
The trial was designed to establish the superiority of local anesthesia over placebo to reduce the frequency of painful arterial puncture. We hypothesized a reduction of this frequency from 30% (with placebo cream) to 5% (with lidocaine and prilocaine cream). Two subgroup analysis were planned prior to randomization; the first was for subjects with unpleasant or painful memories of previous arterial punctures (group 1), and the second was for subjects without unpleasant or painful memories of previous arterial punctures (group 2). To maintain the overall probability of type 1 error at < 5%, each individual hypothesis (whole sample, group 1, and group 2) was performed at the nominal risk (α = 0.05/3). For an overall α-error of 5% and a 90% power to detect the expected difference between the anesthetic and placebo groups in the whole sample, the required sample size was 136, anticipating a maximum withdrawal rate of 10%.
Statistical Analyses
Continuous variables were expressed as mean ± SD. Subject characteristics were compared between anesthetic/placebo groups using the Student t test or the Wilcoxon test when appropriate for quantitative data, and using the chi-square test or the Fisher exact test for qualitative data. Painful arterial punctures frequencies were compared between groups using a generalized linear model with binomial distribution and logarithmic link function. This allowed estimating relative risks with confidence intervals and testing the homogeneity of the relative risks across the 2 predefined subgroups thanks to an interaction term in the model. Adjusted P values for multiple comparisons concerning the primary outcome were calculated according to the Bonferroni method in which the P values are multiplied by the number of comparisons. Pain scores were compared between the anesthetic and placebo groups using the Student t test. No correction was used for secondary outcomes. The statistical analyses were performed using SAS 9.4 (SAS Institute, Cary, North Carolina).
Results
Subjects
Between December 2012 and October 2015, 136 subjects were included in the study and randomized to the anesthetic group (n = 67) or to the placebo group (n = 69) (Fig. 1). One subject received placebo cream instead of the evaluated treatment. Baseline characteristics of subjects are reported in Table 1. In the overall population, gender was mostly male (66.9%), mean body mass index was 29.18 ± 7.07 kg/m2, most subjects were retired (71.6%), most currently or previously worked in a manual trade (71%), and most of the subjects had a previous arterial puncture, and had no apprehension about the procedure. The mean time between cream application and arterial puncture was 136.85 ± 22.31 min.
Flow chart.
Baseline Subject Characteristics
Primary and Secondary Outcomes
Arterial puncture was reported as painful by 37.7% of subjects (26 of 69) in the placebo group and by 20.9% of subjects (14 of 67) in the anesthetic group with a nonsignificant between-group difference (relative risk 0.55 [95% CI 0.28–1.09], P = .10 using Bonferroni correction). The subgroup analysis (ie, subjects with unpleasant or painful memory for prior arterial puncture [group1] and subjects without unpleasant or painful memory for prior arterial puncture [group 2]) did not show significant heterogeneity of the relative risks (group 1: 43% vs 18%; relative risk 0.42; group 2: 29% vs 16%; relative risk 0.55, P value for interaction = .70) (Table 2).
Between-Group Comparison of Painful Arterial Puncture Frequency
We found no significant difference for pain score between groups in the whole sample (2.09 vs 1.55, mean difference 0.54 [95% CI –0.17 to 1.25]) or in the subgroup analysis (Table 3). No significant between-group differences were observed for the overall incidence of adverse events.
Difference in Pain Score Between Groups
Discussion
In this prospective, double-blind, randomized controlled trial, we found no significant difference in the reduction of the frequency of painful arterial puncture between groups (38% vs 21%, P = .10) or in the subgroup analysis (group 1: 43% vs 18%, P = .17; group 2: 29% vs 16%, P = .75).
To our knowledge, this is the first randomized, placebo-controlled trial that used the proportion of painful arterial punctures as the primary end point. Several reports used the decrease in pain intensity as a main end point to evaluate the effects of anesthetics.6,7,10-12,18-20 However, if decreasing pain intensity of this procedure is a clinically relevant issue for patients, the benefit might be incomplete as some will still feel pain during the puncture. The use of our end point was more accurate as the objective of using anesthesia was to ensure that the puncture was not painful at all for subjects. However, our hypothesis of a reduction in painful arterial puncture frequency from 30% to 5% was overestimated. Moreover, we expected a decrease in painful arterial puncture rates by 83%. Despite these issues, we observed that application of a eutectic mixture of 2 local anesthetics (ie, lidocaine and prilocaine) reduced the number of painful arterial punctures by 50% compared to placebo, which is clinically important for patients. The use of the Bonferroni correction to analyze multiple comparisons led to nonsignificant results. A hierarchical testing strategy, testing the whole sample first with a significance level of P < .05, would have been more appropriate and perhaps successful. however, the sample size would be lower and we do not think this testing strategy could affects the results.
In subgroup analysis, the results were similar, although there is a lack of data for subjects with unpleasant or painful memory of previous arterial puncture and for subjects without unpleasant or painful memory, including subjects without arterial puncture experience prior to this study. Indeed, 23% of subjects did not have a prior arterial puncture, so they could not answer the question about their experience of a prior arterial puncture, which led to a smaller sample size and thus a loss of statistical power.
Consistent with other studies, the pain felt by subjects during the arterial puncture, as measured using numerical pain rating, was not significantly different between groups. Aaron et al9 compared the effectiveness of topical tetracaine versus placebo gel prior to arterial puncture on pain measured with a visual analog scale. The authors concluded that tetracaine gel did not significantly decrease pain after arterial puncture in comparison with placebo gel.9 Similarly, Tran et al,7 in a randomized placebo-controlled trial, reported that the topical application of 4% amethocaine gel for 30 min was not effective in reducing the pain associated with arterial puncture compared to a placebo gel. In contrast, Youn et al11 reported significant benefit of topical anesthesia (ie, a eutectic mixture of lidocaine and prilocaine) versus placebo on radial pain and sympathetic response during transradial coronary angiography. For the protocol, Youn et al11 applied the anesthetic cream or placebo cream and, after 1–3 h, they injected lidocaine and then inserted the introducer sheath. Pain was compared between groups during lidocaine infiltration and during introducer sheath insertion. Youn et al11 noted a significant difference showing decreased pain in the anesthetic group (using a visual analog pain scale) during lidocaine infiltration but not during introducer sheath insertion. A plausible explanation is that lidocaine infiltration is more superficial than introducer sheath insertion; thus, the anesthetic cream appears to be more effective for superficial insertion in comparison with deeper insertion.11 This may explain the ineffectiveness of the anesthetic cream on pain during arterial puncture, which requires relatively deep insertion. Different authors have studied the effectiveness of lidocaine infiltration versus no infiltration on pain during arterial puncture, but the results are discordant.6,18,19 Two reports compared the effectiveness of lidocaine infiltration versus lidocaine and prilocaine cream, but these results were also discordant.12,20 In a randomized controlled trial, Haynes10 reported that use of cryoanalgesia (ie, an ice bag) for 3 min prior to arterial puncture significantly decreased pain in comparison with no cryoanalgesia.
Our study has some limitations. First, using a hierarchical testing strategy to test the whole sample first with a significance level of P = .05 and then testing subgroups with statistical differences would have been more appropriate However, the sample size would have been smaller, and we do not think this testing strategy would affect the results. Second, our hypothesis of a reduction in painful arterial puncture frequency from 30% to 5% was overestimated. Lastly, nurses’ experience and the length of time required for the puncture were not collected, which could be a confounding bias.
The strengths of our study are the use of a double-blind randomized design, a predefined and standardized protocol for local treatment application, and the use of predefined objective end points. Finally, to our knowledge, our study is the first randomized controlled trial to use the proportion of painful arterial puncture as the primary end point to evaluate the effects of local anesthetics.
Conclusions
We observed that application of a eutectic mixture of 2 local anesthetics, lidocaine and prilocaine, decreased the proportion of painful arterial punctures by 50% compared with placebo. However, this result is not statistically significant due to multiple testing and is a lower benefit than expected. We found no significant difference for pain scores between groups. Future studies should evaluate the best prevention of pain during arterial puncture to increase quality of care for this common procedure, which remains painful for many patients. Attention should also be given to identify predictive factors of painful arterial puncture to focus preventive efforts on the most sensitive patients.
Footnotes
- Correspondence: Marc Beaumont PT PhD, EA 3878 (GETBO), CIC INSERM 1412, Pulmonary Rehabilitation Unit, Morlaix Hospital Centre, 29600 Morlaix, France. E-mail: mbeaumont{at}ch-morlaix.fr
The authors have disclosed no conflicts of interest.
- Copyright © 2021 by Daedalus Enterprises
