a b s t r a c t

Objective: To compare pain associated with venous catheterization after administration of topical ethyl chloride vs placebo among emergency department health care providers.

Methods: We conducted a randomized, double-blind, placebo-controlled, crossover trial among a convenience sample of health care provider volunteers in a tertiary care urban emergency department. We randomly allocat- ed subjects to initial treatment (ethyl chloride vs sterile water aerosol spray) and catheterization site (left or right antecubital fossa). After venous catheterization placement and discontinuation, subjects underwent a 5-minute washout period. All subjects then underwent venous catheterization in the contralateral antecubital fossa after administration of the alternative agent. We measured all outcomes after discontinuation of the second catheter. The primary outcome was difference in pain verbal Numeric rating scale score (0-10) between the 2 agents. Sec- ondary outcomes included preferred agent (binary) and future willingness to use agent on patients (5-point Likert scale).

Results: Thirty-eight health care providers were recruited; all completed the study. Median pain verbal numeric rating scale scores were 4 (interquartile range, 2-5) for placebo vs 2 (1-4) for ethyl chloride. The effect size for pain reduction with ethyl chloride compared with placebo was 2 (95% confidence interval, 0.5-2; P = .001). Most subjects (68.4%) preferred ethyl chloride to placebo. Five-point Likert scale scores measuring willingness to use preferred product on future patients were higher by 2 (95% confidence interval, 1-3) among subjects pre- ferring ethyl chloride vs placebo.

Conclusions: We found that topical ethyl chloride yields a greater reduction in pain associated with venous cath- eterization compared with topical placebo.

1. Introduction

Intravenous catheter insertion is a frequent cause of pain for patients [1,2]. Venous catheterization is a common procedure, with one study reporting approximately 15% of all emergency department (ED) patients undergoing intravenous access [3]. Interventions reducing the discomfort associated with intravenous cannulation may significantly alleviate the discomfort associated with ED visits. Several options exist to reduce pain associated with venous catheterization. Analgesic creams are one possibility but may take up to an hour to achieve clini- cally significant pain relief [4]. Intradermal anesthetics offer another alternative but are invasive and painful [5]. A third option is topical

? Funding: None.

?? Disclaimers: The views expressed herein are those of the authors and do not reflect the official policy or position of Brooke Army Medical Center, the US Army Medical Department, the US Army Office of the Surgeon General, the Department of the Army and Department of Defense, or the US Government.

* Corresponding author.

E-mail address: [email protected] (M.D. April).

skin refrigerant, or vapocoolant, which potentially provides fast-acting noninvasive analgesia.

Several studies that examine the use of topical skin refrigerants exist. The agents in these studies include fluorohydrocarbon [6] and alkane vapocoolants [7,8]. Prior studies of ethyl chloride yield conflicting results. One unblinded randomized study of patients undergoing intra- venous catheterization prior to elective surgery demonstrated no signif- icant analgesia with ethyl chloride vs no intervention [9]. Conversely, 3 unblinded randomized studies of patients undergoing venipuncture demonstrated superior analgesia with ethyl chloride vs no intervention [5,10,11]. An unblinded randomized crossover trial of hemodialysis patients demonstrated superior analgesia with Ethyl chloride spray vs placebo [12]. To our knowledge, there are no prior blinded studies com- paring ethyl chloride to placebo. A blinded study would provide a useful contribution to the literature. Given the subjectivity of pain, such an investigation would benefit from a randomized crossover trial design in which patients serve as their own controls [13].

This study seeks to determine the effectiveness of topical skin refrig-

erant compared with placebo in alleviating the pain associated with venipuncture. It is a randomized, double-blind, placebo-controlled, crossover trial of health care provider volunteers. Through double-

http://dx.doi.org/10.1016/j.ajem.2016.01.039 0735-6757/

blinding and a Crossover design, it seeks to achieve an optimal measure of the effect of ethyl chloride on subjective pain experienced due to venipuncture. Our hypothesis is that topical ethyl chloride spray will lead to a greater reduction in pain associated with intravenous cannula- tion when compared with placebo.

1. Methods
   1. Study design and setting

We conducted a single-center randomized, double-blind, placebo- controlled, crossover trial at an urban tertiary care hospital. The annual ED census is approximately 80 000 patients. The institutional review board approved the study. The ClinicalTrials.gov registration number is NCT02499965.

Study population

We recruited a convenience sample of ED health care providers to participate in the study. Eligible subjects included health care workers (residents, nurses, or medics) 18 years or older who routinely either order or perform venous catheterization on patients. Exclusion criteria included history of hypersensitivity to ethyl chloride, upper extremity amputation or neurologic deficits, or recent injury to or infection of the antecubital fossae. Study investigators confirmed eligibility by phys- ical examination. Participation was voluntary and investigators did not record any demographics or other identifying information.

We publicized the study for 1 month prior to data collection via department-wide e-mails and announcements at weekly grand rounds. On the day of the study, research assistants approached health care providers in the ED patient care areas and invited them to participate. Consent forms disclosed that the subjects would undergo experimental topical therapies prior to antecubital intravenous catheterization in

both arms. Subjects understood that they would undergo random allo- cation to initial receipt of their right vs left arm and 1 of 2 interventions and that by study end, both arms would undergo cannulation and they would receive both therapies. Subjects received no additional details regarding the topical treatments. Study investigators obtained written informed consent from all subjects. Research assistants (nurses or medics) blinded to the study interventions applied the topical treat- ments, inserted the intravenous catheters, solicited subject responses to study outcomes, and recorded all outcomes on data collection forms. We documented subject intervention and arm allocation in accordance with the CONSORT statement (Fig. 1) [14].

Study protocol

All study subjects underwent antecubital intravenous catheteriza- tion with both topical skin refrigerant and placebo. Research assistants (nurses or medics) not acting as study subjects performed all venous catheterizations using 20-gauge needles. A single research assistant performed both cannulations for each subject. We did not collect detailed information regarding each research assistant’s prior back- ground and experience. Similarly, we did not collect any data allowing us to link any subject to the research assistant administering his or her venous catheterizations.

When undergoing the study intervention, subjects received sprays from aerosol containers filled with ethyl chloride (Gebauer’s Ethyl Chloride; Gebauer Company, Cleveland, OH). When undergoing placebo administration, subjects received sprays from identical aerosol containers filled with sterile water (Nature’s Tears EyeMist; Bio-Logic Aqua Research Technologies Incorporated, Grants Pass, OR) stored at 36?F to better simulate a topical refrigerant sensation. Pharmacy personnel prepared all aerosol containers to obscure contents to blind study investigators, subjects, and research assistants to contents. Research assistants sprayed the antecubital fossa with the aerosol

Fig. 1. Flow diagram of patient enrollment, allocation, follow-up, and analysis.

container contents for 5 to 8 seconds. Assistants directed sprays away from both the patient’s and their own face to prevent odor detection. Venous catheterization was then performed within 15 to 30 seconds.

We used a simple randomization sequence for subject allocation. Subjects first underwent randomization to initial receipt of ethyl chloride or placebo (double blind). Subjects then underwent randomiza- tion to either left arm or right arm for initial therapy (unblinded). After a 5-minute washout period, subjects underwent topical application of the therapy not already received in the contralateral upper extremity antecubital fossa. Pharmacy personnel coordinated appropriate aerosol container delivery based on subject allocation. Study investigators, research assistants performing venipuncture, and subjects all remained blinded to subject allocation until study completion.

Measurements

After placement and removal of both venous catheters, investigators used hardcopy data collection forms to record all outcomes. Subjects answered questions regarding pain, product preference, and willing- ness to use those products on themselves or their patients in the future. Investigators double-entered all hardcopy forms into an Excel database (version 14; Microsoft, Redmond, WA).

The primary outcome was the difference in subject response to the question “rate the pain of intravenous catheterization” for ethyl chloride vs placebo. This rating comprised an 11-point ordinal pain verbal numeric rating scale (VNRS) as previously used in the ED setting [15]. This scale ranged from 0 on the left, reading “least possible pain,” to 10 on the right, reading “worst possible pain.” An additional item requested that subjects designate “which product was most useful in reducing the pain of intravenous catheter insertion?” Subjects next responded yes or no to the question “if you had to have intravenous catheterization in the future, would you want to have the method that you rated as superior performed on you?” Finally, subjects provided a response on a 5-point Likert scale to the question “when you order or place venous catheters in the future, how likely are you to use or recom- mend the method that you rated as superior?” This scale ranged from “very unlikely” on the left to “very likely” on the right.

Data analysis

Sample size calculations assumed that 11-point pain VNRS scores and distributions approximate those from Visual analog scale measurements [15]. Based on previous research, we powered our anal- ysis to detect a 2-point difference in 11-point pain scale [16,17]. We assumed a standard deviation of 3 points [16]. Given ? = .05, ? = 0.20, 2-sided testing, and each subject serving as his or her own control, our sample size estimate was 36.

All analyses were intention-to-treat and performed using SPSS (Version 22; IBM, Armonk, NY). For the primary outcome, we com- pared ordinal pain VNRS scores between treatments using Wilcoxon Signed Rank Testing. For secondary outcomes, we calculated median VNRS scores for placebo and ethyl chloride stratified by initial treatment and arm allocation. We also compared ordinal pain VNRS scores for all first cannulation placements vs all second cannulation placements regardless of treatment or arm using Wilcoxon signed rank testing. Next, we compared willingness to have product used on self for future venous catheterizations (binary outcome) between subjects preferring ethyl chloride and subjects preferring placebo using Fisher exact test. Finally, we compared willingness to use product on future patients undergoing venous catheterization (ordinal 5-point Likert scale) between subjects preferring ethyl chloride and subjects preferring placebo using Wilcoxon rank sum testing. We calculated effect sizes for all comparisons with 95% confidence intervals. We generated confi- dence intervals for all ordinal data comparisons using Hodges-Lehmann estimators.

1. Results

Thirty-eight health care providers who routinely either place or order venous catheterizations on patients volunteered and underwent screening for study inclusion by the study investigators (Fig. 1). All sub- jects were eligible and included in the study. All subjects were active duty military personnel. All enrolled subjects underwent randomization to initial intervention and initial arm for venous catheterization. Twenty-one subjects first received topical ethyl chloride (8 allocated to undergoing initial catheterization on the left side and the remaining 13 to the right side). The remaining 17 subjects first received placebo (11 allocated to undergoing initial catheterization on the left side and the remaining 6 to the right side). Research assistants reported that all cannulation procedures were successful on first attempt.

Regarding primary outcome, pain VNRS scores were significantly

lower with administration of ethyl chloride compared with placebo (Fig. 2). Median pain scores were 4 (interquartile range, 2-5) for topical placebo vs 2 (interquartile range, 1-4) for topical ethyl chloride (Fig. 3). The effect size for pain reduction with ethyl chloride was 2 (95% confi- dence interval, 0.5-2, P = .001). Median pain VNRS scores for placebo and ethyl chloride were comparable between patients receiving placebo first vs ethyl chloride first. Similarly, median pain VNRS scores for place- bo and ethyl chloride were comparable between patients undergoing left arm cannulation first vs right arm cannulation first (Table 1). Medi- an pain score for all first cannulation placements was similar to that for all second cannulation attempts, with a difference of 1 (95% confidence interval, 0-2).

Most subjects reported that topical ethyl chloride provided superior venipuncture pain relief compared with placebo: 68.4% (95% confidence interval, 51.4%-82.5%). Willingness to have product used on self for future venous catheterizations (binary outcome) was comparable be- tween subjects preferring topical ethyl chloride vs placebo (Table 2). Willingness to use product on future subjects undergoing venous catheterization (per ordinal 5-point Likert scale) was greater among subjects preferring topical ethyl chloride vs placebo (Table 2).

All research subjects finished the study and completed the survey instrument. There were no adverse events.

1. Discussion

Venous catheterization is a common procedure for patients in the ED [2,3]. This procedure frequently results in patient discomfort [1]. Although prior research suggests that agents exist which might mitigate this discomfort, these agents frequently go unused. Topical skin refrig- erants offer an option likely to be attractive to both patients and providers because they are noninvasive and fast acting. Ethyl chloride’s mechanism of action relates to its role as a skin refrigerant. Ethyl chlo- ride is a gaseous substance that achieves this effect by abstracting heat as it evaporates after application to the skin [18]. The resulting skin refrigeration blocks sensory nerve conduction, achieving anesthesia [19]. This randomized double-blind, placebo-controlled, crossover trial among health care providers demonstrated decreased ordinal pain VNRS scores with venous catheterization after application of topical ethyl chloride as compared with topical placebo.

This study is unique in several respects. First, to our knowledge, it is the only double-blind crossover study of topical ethyl chloride use for subjects undergoing venous catheterization. Second, it is the first study of ethyl chloride conducted on health care providers. That said, previous studies of health care providers found that Alternative agents (eg, lidocaine cream and subcutaneous lidocaine) [16] reduce venipunc- ture pain.

Our finding of superior venipuncture pain relief with ethyl chloride is broadly consistent with most previous randomized studies of this agent. Two unblinded randomized studies of patients undergoing venipuncture demonstrated superior analgesia with ethyl chloride vs no intervention [5,11]. An unblinded randomized crossover trial of

Fig. 2. Subject pain VNRS scores associated with venipuncture. The horizontal axis separates data for each individual subject. The vertical axis represents the patient-reported pain VNRS score (0-10). Each pair of bars represents data for a single patient. The black bars represent pain associated with venipuncture after administration of placebo. The gray bars represent pain associated with venipuncture after administration of sterile water.

hemodialysis patients demonstrated superior analgesia with ethyl chlo- ride spray vs placebo [12]. One unblinded randomized study of patients undergoing intravenous catheterization in the ED [10] demonstrated that the decrease in pain 100-mm VAS with ethyl chloride vs no inter- vention was statistically significant but did not meet their threshold

for clinical significance (13 mm) [17]. A single unblinded randomized trial of preoperative patients with 5 intervention arms (topical ethyl chloride, subcutaneous lidocaine, EMLA cream, placebo, and no inter- vention) demonstrated no efficacy with topical ethyl chloride [9]. This alternative finding may relate to lack of blinding and differences in the

Fig. 3. Boxplot of subject reported ordinal pain scale scores from 1 to 10 associated with venous catheterization. The midlines represent the medians. The bottom and top box edges represent the 25th and 75th percentiles, respectively. The bottom and top “whiskers” represent the minimum and maximum values, respectively. Subjects reported significantly lower pain with venous catheterization after topical ethyl chloride than topical placebo (P = .001 by Wilcoxon signed rank testing).

Table 1

Median pain scores compared between venous catheterization with placebo vs ethyl chloride

research suggests that the experience of pain is unlikely to vary signifi- cantly based on age or sex [20]. Although previous research has exam- ined provider perceptions of their patients’ pain levels, [21] it is less

Study population Placebo (IQR)

Ethyl chloride (IQR)

Effect size (95% CI)

clear whether providers’ perceptions of their own pain might vary compared with those of nonproviders. Furthermore, recruitment of a

All participants (38) 4 (2-5) 2 (1-4) 2 (0.5-2)

Subjects receiving placebo first (17) 4 (3-5) 3 (2-4) 1 (0-2)

convenience sample of research study volunteers may also have affect- ed generalizability; we have no data to compare the characteristics of

Subjects receiving ethyl chloride

first (21)

Subjects undergoing cannulation of left arm first (19)

Subjects undergoing cannulation

of right arm first (19)

4 (1.5-6) 2 (1-2.5) 2 (1-3)

4 (2-5) 2 (2-4) 2 (0.5-3)

4 (2-6) 2 (1-4) 2 (0-2)

our study participants to those of other health care providers in our ED who would also have been eligible for study participation.

Beyond subjeCT characteristics, our study may also have limited generalizability given the procedural protocol. All subjects specifically underwent antecubital venous catheterization using 20-gauge needles

Abbreviations: CI, confidence interval; IQR, interquartile range.

study setting, subject population, providers placing the catheters, needle size, venipuncture locations, and techniques for topical anesthe- sia application.

Comparison of the effect size estimated by this study to that estimat- ed by previous research is difficult given that prior studies of topical ethyl chloride used 100-mm VAS measurements [5,9-12]. However, data exists suggesting a significant correlation between pain measure- ments using ordinal VNRS scores vs continuous VAS in the ED setting [15]. Visual analog scale pain scores in these studies for placebo ranged from 23.5 [11] to 38 [5], whereas scores for topical ethyl chloride ranged from 1 [11] to 26 [9]. Excluding the single study demonstrating no efficacy with ethyl chloride in reducing venipuncture pain, [9] the differences in pain VAS between patients allocated to ethyl chloride and placebo were very similar across studies, ranging from 19.4 [12] to 22.5 [11]. This difference appears generally consistent with the 2-point median difference in ordinal pain scales between venipuncture pain with ethyl chloride and that with placebo detected by our study.

Our study provides some insight into the willingness of health providers to incorporate use of ethyl chloride into their clinical practice. Providers preferring topical ethyl chloride to placebo reported a greater willingness to use their preferred agent on their patients in the future as measured by a 5-point Likert scale. In contrast, we detected no differ- ence in provider binary willingness to have their preferred agent used on themselves among those preferring ethyl chloride vs those prefer- ring placebo. It is possible that we detected no difference because the subjects considered the pain associated with this relatively minor procedure to be so insignificant that they felt ambivalent about the use of effective analgesic options. However, previous research among health care providers comparing non-ethyl chloride Topical anesthetics found that health care providers predictably have increased willingness to use agents on their patients, which they reported as more efficacious on themselves [16]. There was a trend to increased willingness to have the preferred agent used on self among those preferring ethyl chloride (96.2% vs 81.8%). We believe that the lack of statistically significant difference in this proportion simply reflects the fact that we did not adequately power the study to examine this outcome.

This study had several limitations. We cannot establish the general- izability of our results to the population of ED patients undergoing intra- venous catheterization. We did not collect demographic data on our subjects, although all participants were active duty military health care providers suggesting a predominantly young and healthy popula- tion. Indeed, the 100% first pass Cannulation success observed in our study reflects this fact. Although this population likely differs from the population of ED patients undergoing venous catheterization, previous

only. Aerosol sprays all lasted 5 to 8 seconds with a 15- to 30-second delay prior to intravenous placement. Analgesia may vary from our findings outside of a research setting given likely real-world variability in venipuncture location, needle size, and aerosol application time.

We cannot prove effective blinding as we did not survey subjects after the protocol to ascertain their perceptions on study arm allocation. Our study population of health care workers in particular may have detected differences between the topical sensation and odor of ethyl chloride and those of sterile water. We sought to maintain blinding through obscuring aerosol containers, cooling the sterile water prepara- tion, and directing treatment sprays away from patients’ and research assistants’ faces to avoid olfactory detection.

Our choice of primary outcome measure may complicate the com- parison of our results with those of other studies. Prior ED studies of pain associated with venous catheterization have used ordinal pain scales [16]. However, previous studies examining ethyl chloride [5,9-12] or other topical skin refrigerants used 100-mm VASs [6-8]. Nevertheless, comparisons with these other study results may be meaningful given data suggesting a high correlation between pain mea- surements using ordinal scales vs continuous VASs [15].

Our choice of outcome measure is also limited in that it is subjective. Advantages associated with this choice include an outcome important to patients and low likelihood of measurement bias. However, this study does not capture many alternative objective measures potentially important to patients and providers. Examples include anxiety and time to successful intravenous venous catheter placement.

Finally, we did not collect details regarding the prior experience and backgrounds of the research assistants performing venous catheteriza- tions in our study. These data could have a material impact on the pain experienced during venipuncture. One might expect that those individuals undergoing cannulation by experienced nurses or medics would experience less pain compared with those undergoing cannula- tion by less experienced providers. We believe that our crossover design and assignment of a single research assistant to each subject likely control for any effect research assistant venipuncture experience had on subject pain scores.

The literature would benefit from additional investigation into the use of topical ethyl chloride in the ED setting. Our study focused on health care workers instead given our desire for a crossover design. Now that our findings provide further evidence of the Analgesic efficacy of this agent, future studies should focus on ED patients. In addition to measuring pain, such research should examine additional objective outcomes to include numbers of venipuncture attempts required to achieve cannulation, cost, anxiety, time to intravenous catheter place- ment, bleeding, and patient satisfaction. Objective outcomes over longer-time horizons (eg, 2 weeks after cannulation) to include delayed

Table 2

Outcomes compared between subjects reporting superior pain relief with ethyl chloride vs placebo

Variables	Placebo better (12)	Ethyl chloride better (26)	Effect size (95% CI)
Desire for preferred agent to be used on self in future, % (95% CI)	81.8 (48.2-97.7)	96.2 (80.4-99.9)	14.3 (-5.6 to 34.2)
Median willingness to use preferred agents on patients, 1-5 (IQR)	2 (1.25-3.75)	4 (4-5)	2 (1 to 3)

Abbreviations: CI, confidence interval; IQR, interquartile range.

bleeding, phlebitis, and cellulitis would be additional useful data. Such studies would ideally use post hoc surveys to confirm effective blinding. Studying a range of ED patients, needle sizes, and venipuncture loca- tions would greatly improve the generalizability of these future studies. Future studies should also consider new and Novel methods to improve blinding subjects to treatment allocation. Such methods might include literal blindfolding given that ethyl chloride can visibly evaporate and blanch the skin upon application. In addition, investiga- tors should consider use of aromatherapy or nasal plugs to mask olfac- tory detection of ethyl chloride. Finally, future studies would benefit from subject surveys at study completion to ascertain the effectiveness

of blinding efforts.

This randomized trial suggests that topical ethyl chloride provides efficacious pain relief among health care workers undergoing venous catheterization. Although this study is the first double-blind, placebo- controlled study of this agent, our results are consistent with much of the existing literature suggesting that skin refrigerants effectively reduce pain associated with this procedure. This agent appears rapid in onset and safe with no adverse events reported. Emergency depart- ment providers should consider incorporation of this agent into their Treatment protocols including venous catheterization or phlebotomy. Given the frequency of this procedure in the ED setting, topical ethyl chloride has the potential to significantly decrease pain and improve satisfaction among ED patients.

Acknowledgments

The authors acknowledge the contributions of Dr John Ward, PhD, to the statistical analysis presented in this manuscript.

References

1. Cummings EA, Reid GJ, Finley GA, McGrath PJ, Ritchie JA. Prevalence and source of pain in pediatric inpatients. Pain 1996;68:25-31.
2. Singer AJ, Richman PB, Kowalska A, Thode Jr HC. Comparison of patient and practi- tioner assessments of pain from commonly performed emergency department pro- cedures. Ann Emerg Med 1999;33:652-8.
3. Limm EI, Fang X, Dendle C, Stuart RL, Egerton Warburton D. Half of all peripheral in- travenous lines in an Australian tertiary emergency department are unused: pain with no gain? Ann Emerg Med 2013;62:521-5.
4. Hallen B, Olsson GL, Uppfeldt A. Pain-free venipuncture. Effect of timing of applica- tion of local anaesthetic cream. Anaesthesia 1984;39:969-72.
5. Armstrong P, Young C, McKeown D. Ethyl chloride and venipuncture pain: a com- parison with intradermal lidocaine. Can J Anaesth 1990;37:656-8.
6. Hartstein BH, Barry JD. Mitigation of pain during intravenous catheter placement using a topical skin coolant in the emergency department. Emerg Med J 2008;25: 257-61.
7. Hijazi R, Taylor D, Richardson J. Effect of topical alkane Vapocoolant spray on pain with intravenous cannulation in patients in emergency departments: randomised double blind placebo controlled trial. BMJ 2009;338:b215.
8. Page DE, Taylor DM. Vapocoolant spray vs subcutaneous lidocaine injection for re- ducing the pain of intravenous cannulation: a randomized, controlled, clinical trial. Br J Anaesth 2010;105:519-25.
9. Biro P, Meier T, Cummins AS. Comparison of topical anaesthesia methods for venous cannulation in adults. Eur J Pain 1997;1:37-42.
10. Robinson PA, Carr S, Pearson S, Frampton C. Lignocaine is a better analgesic than ei- ther Ethyl chloride or Nitrous oxide for peripheral intravenous cannulation. Emerg Med Australas 2007;19:427-32.
11. Selby IR, Bowles BJ. Analgesia for venous cannulation: a comparison of EMLA (5 mi- nutes application), lignocaine, Ethyl chloride, and nothing. J R Soc Med 1995;88: 264-7.
12. Celik G, Ozbek O, Yilmaz M, Duman I, Ozbek S, Apiliogullari S. Vapocoolant spray vs lidocaine/prilocaine cream for reducing the pain of venipuncture in hemodialysis patients: a randomized, placebo-controlled, crossover study. Int J Med Sci 2011;8: 623-7.
13. Sibbald B, Roberts C. Understanding controlled trials. Crossover trials. BMJ 1998; 316:1719.
14. Schulz KF, Altman DG, Moher D. CONSORT Group. CONSORT 2010 statement: up- dated guidelines for reporting parallel group randomized trials. Ann Intern Med 2010;152:726-32.
15. Bijur PE, Latimer CT, Gallagher EJ. Validation of a verbally administered numerical rating scale of acute pain for use in the emergency department. Acad Emerg Med 2003;10:390-2.
16. McNaughton C, Zhou C, Robert L, Storrow A, Kennedy R. A randomized, crossover comparison of injected buffered lidocaine, lidocaine cream, and no analgesia for pe- ripheral intravenous cannula insertion. Ann Emerg Med 2009;54:214-20.
17. Todd KH, Funk JP. The minimum clinically important difference in physician- assigned visual analog pain scores. Acad Emerg Med 1996;3:142-6.
18. Bierman W. therapeutic use of cold. JAMA 1955;157:1189-92.
19. Travell J. Factors affecting pain of injection. JAMA 1955;158:368-71.
20. Kelly AM. Does the clinically significant difference in visual analog scale pain scores vary with gender, age, or cause of pain? Acad Emerg Med 1998;5: 1086-90.
21. Marquie L, Raufaste E, Lauque D, Marine C, Ecoiffier M, Sorum P. Pain rating by pa- tients and physicians: evidence of systematic pain miscalibration. Pain 2003;102: 289-96.