Comparing activation and recoil forces generated by epinephrine ...

Letters to the Editor Comparing activation and recoil forces generated by epinephrine autoinjectors and their training devices To the Editor: Epinephrine is widely accepted as the initial treatment of choice for patients with anaphylaxis.1-3 Prefilled epinephrine autoinjectors (EAIs) remain the primary method for delivering epinephrine to patients in whom anaphylaxis develops in the out-of-hospital environment. Three EAIs that are currently available in the United States include EpiPen (Dey LP, Napa, Calif), Twinject, and Adrenaclick (Sciele, Division of Shionogi, Tokyo, Japan). Each of the EAIs also have needleless replicas that serve as epinephrine autoinjector training devices (EAITDs) that allow patients to simulate the administration of an actual EAI. The purpose of this pilot study was to measure both the activation force (AF) required to trigger each device (real EAIs and their respective EAITDs) and any reaction force (RF) or recoil that is generated immediately after activation of the devices into ballistics gelatin. We hypothesized that EAITDs do not mimic the mechanical forces associated with administration of an actual device. We then compared the differences in AF and RF between each brand of EAI, each EAI and its respective EAITD, and each brand of EAITD. A total of 30 EAIs and EAITDs (5 of each brand of EAI and their respective trainers) were evaluated. Only EAIs containing 0.3 mg of epinephrine per injection were used in this study; however, it should be noted that within a given brand of autoinjector, the mechanisms for activation are identical for both the 0.3- and 0.15-mg injectors. All devices were new and had never been activated before testing. This was done to eliminate the possibility of the EAITDs ‘‘wearing out’’ over time with repetitive use, which might have altered the measured AFs or RFs. We used Gelita (Sioux City, Iowa) ballistic gelatin, which has been extensively used in medical research to effectively simulate human soft tissue. The gelatin was prepared in a uniform fashion according to the manufacturer’s guidelines and was poured into plastic drinking cups (Solo, Urbana, Ill) and refrigerated at 3.38C. All cups were refrigerated until immediately before testing to ensure uniformity of gelatin. In addition, circular denim swatches of uniform shape and thickness were placed on top of the gelatin and secured with 4 straight pins along the periphery (see Fig E1 in this article’s Online Repository at www. jacionline.org). An Instron 5567 (Instron Corp, Norwood, Mass) universal testing machine fitted with a 500-Newton (N) load cell was used to measure the AF and RF of each device in a randomized fashion. The universal testing machine measures the force-displacement relationship of materials and objects through motion of a crosshead (see Fig E2 in this article’s Online Repository at www. jacionline.org). During testing, each EAI/EAITD was placed in a custom fixture (see Fig E3 in this article’s Online Repository at www.jacionline.org) that was aligned vertically and horizontally and secured with a clamp. The crosshead was then positioned several millimeters above the trigger end of the EAIs/ EAITDs and displaced downward at 25 mm/min or 0.42 mm/s. Data were collected at 100 Hz, and time, crosshead displacement, and crosshead force were recorded.

Factorial ANOVAs were used to determine whether there were any differences in AFs and RFs between actual EAIs and EAITDs for the 3 different brands tested (see Table E1 in this article’s Online Repository at www.jacionline.org). Regarding AF, the ANOVA revealed a statistically significant interaction between the brand and type of injector (ie, trainer vs actual; P < .001). The trainer for the Adrenaclick and Twinject required less AF than their respective EAIs, but the EpiPen trainer actually had a higher mean AF than the EpiPen EAI. Post hoc comparisons were then performed to further describe the significant interaction detected by means of ANOVA. For actual EAIs, there were no statistically significant differences between brands with regard to AF (P 5 .175). However, there was a statistically significant difference in AF when comparing brands of EAITDs (P < .001). Post hoc comparisons of the EAITDs revealed that the EpiPen trainer required greater AF than the Adre_ .001 for each comparison). The naclick or Twinject trainers (P < Adrenaclick and Twinject trainers did not differ from one another in AF (P 5 .801). Regarding RF, ANOVA revealed a statistically significant interaction between the brand and type of injector (ie, training vs actual; P < .001). Although the EAITDs demonstrated lower RF than the actual EAIs for all 3 of the brands, the difference was much steeper for the EpiPen EAI, which had the highest RF recorded. Post hoc comparisons were again performed to explain the significant interaction that was detected. For real EAIs, there were statistically significant differences in RFs between the different brands (P < .001). Further analysis revealed that there was no difference between the Adrenaclick and Twinject EAIs with regard to RF (P 5 .151) and that the EpiPen EAI generated significantly more RF than the Adrenaclick (P < .001) or Twinject (P < .001) EAIs. For EAITDs tested, there were no statistically significant differences in RF between brands (P 5 .892). Our results indicate that there are no significant differences between the Adrenaclick, EpiPen, and Twinject EAIs in regard to how much force is required to activate each device, with a mean AF for all 3 brands of 24.4 N or approximately 5.46 pounds-force (lb-f; see Table I for the relationship between Newtons and pounds-force). These results are similar to the mean AF of the Epipen Junior EAI of 23.0 N that Schwirtz and Seeger4 reported in another pilot study that also used a mechanical tester and denim-covered ballistics gelatin. Although there were no differences between brands in terms of AF, it appears that the EpiPen EAI generated a much greater amount of RF when compared with the Adrenaclick and Twinject EAIs. The EpiPen EAI generated a mean RF of 42.2 N or approximately 9.49 lb-f compared with a mean of 15.4 N for the Adrenaclick and 18.9 N for the Twinject EAIs or 3.46 lb-f and 4.25 lb-f, respectively. With respect to EAITDs, our results revealed that the EpiPen EAITD required more AF (mean, 28.0 N or approximately 6.29 lb-f) than the EpiPen, Adrenaclick, or Twinject EAIs or EAITDs. All 3 brands of EAITDs appear to be similar in regard to RF generated. However, there are significant differences between EAIs and EAITDs with respect to the amount of RF, with the EAITDs generating much less RF than their respective EAIs. The greatest difference occurred between the EpiPen EAI and the EpiPen EAITD (42.2 vs 13.3 N, respectively, or approximately 1143

1144 LETTERS TO THE EDITOR

J ALLERGY CLIN IMMUNOL APRIL 2012

TABLE I. Description of Newtons and pounds-force 1 Newton (N) 5 the amount of force required to accelerate a mass of 1 kg at a rate of 1 m/s2 squared. Units are as follows: kg $ m/s2. Conversion factor: 1 N 5 0.22481 lb-f. Example: The difference between the nearly 10 lb of RF (9.45 lb-f) generated by an EpiPen EAI versus the almost 3 lb of RF (2.99 lb-f) generated by an EpiPen EAI trainer could be analogous to the difference in force being exerted on a subject’s hand while holding a 10-lb weight versus a 3-lb weight.

FIG 1. Example of graphic data collected from universal tester during pilot testing of a Twinject EAI. Negative forces represent compression. The AF was the crosshead force measured at the instant before injection of the needle into the gelatin. The RF measurement typically occurred between 100 and 400 ms after the device was activated.

9.49 vs 2.99 lb-f). Our results also indicate that the Adrenaclick and Twinject EAITDs underestimate how much AF is required to trigger their respective EAIs. There are some limitations to this study as well. Blinding of the investigator operating the universal testing device was challenging because of the different shapes and colors of each device. The unique holder that was designed for each brand of device was necessary, however, to ensure an upright and uniform position in the tester each time. An attempt was made to blind the name with a generic label covering the identity of each device; we could not cover up every difference between devices out of concern that we might add more weight or alter the natural function of each device. However, the investigator operating the universal tester was an engineer who had no previous knowledge about EAIs of any kind and had never handled or operated one before this study. There are differences in how each device is designed to be activated with EpiPen by using a ‘‘swing and firmly push’’ technique, as opposed to the Twinject and Adrenaclick recommendation to place against the thigh without any arm swing and then ‘‘press down hard.’’ Our mechanical testing device could not

simulate an arm swing, and it is unknown how including a variable such as arm swing might cause differences in the mechanical forces measured. Obviously, the use of ballistics gelatin as a tissue simulant has limitations in that it cannot perfectly mimic human tissue. The use of gelatin also mandated that we slowly compress the devices over a much longer time frame than would actually occur during administration of a real EAI (Fig 1). Using a faster compression rate would sometimes cause the entire device to puncture the gelatin, impeding free backward movement and therefore altering our recoil measurements. However, in the laboratory setting we believed that gelatin was preferred over animal/human tissue because of potential differences in the amount of subcutaneous tissue and firmness between samples. This pilot study confirmed our hypothesis that EAITDs do not simulate the mechanical forces of their respective EAIs. We also found differences between each brand of EAI and between each brand of EAITD. Although we could not determine whether the amount of RF generated by the devices resulted in any clinically significant ‘‘lift’’ upward, thereby reducing the likelihood of

J ALLERGY CLIN IMMUNOL VOLUME 129, NUMBER 4

epinephrine being delivered intramuscularly, the concern is that it might. Needle lengths of EAIs have already been cited as potentially inadequate to reliably deliver epinephrine to the muscle bed.5,6 Any amount of unexpected recoil that occurs while using a real EAI during an emergency might further reduce the likelihood of successful intramuscular administration.7 On the basis of this pilot study, it might be prudent for practitioners to inform patients that there are indeed differences with regard to how much force is required to activate different brands of EAIs, as well as differences in recoil generated. This might be of particular importance for those patients or providers who have been using one brand of EAI exclusively and then switch to a different brand of EAI. Encouraging patients to practice with real expired EAIs on tissue simulants, such as an orange, might also be beneficial. Regardless of the brand of EAI used, providers should instruct patients to firmly grasp the device and to continually depress the EAI into the thigh after activation occurs. This compression might also help displace subcutaneous fat and reduce the distance to muscle in some patients, potentially increasing the likelihood of intramuscular administration of epinephrine.5,6,8 We thank Aaron J. Bonham, MS, for performing statistical analyses on this project. We also thank Gelita (Sioux City, Iowa) for the donation of researchgrade ballistics gelatin and for guidance on the preparation and storage of gelatin for our research study. Ryan C. Jacobsen, MDa Trent M. Guess, PhDb A. Wesley Burks, MDc From athe Department of Emergency Medicine at Truman Medical Center, University of Missouri–Kansas City School of Medicine, Division of Emergency Medical Services at Children’s Mercy Hospitals and Clinics, Kansas City, Mo; bthe Department of Mechanical Engineering, University of Missouri–Kansas City, Kansas City, Mo; and cthe Division of Pediatric Allergy and Immunology, Duke University Medical Center, Durham, NC. E-mail: [email protected]. Supported by a grant from Emergency Physicians Foundation, Kansas City, Mo. Disclosure of potential conflict of interest: R. C. Jacobsen receives research support from the Emergency Physicians Foundation. A. W. Burks is a minority stockholder in Allertein and MastCell, Inc; is on the advisory board for Dannon Co Probiotics; has consultant arrangements with Exploramed Development, Intelliject, McNeil Nutritionals, Merck & Co, Novartis, Pfizer, Portola Pharmaceuticals, and Schering-Plough; is on the Expert Panel for Nutricia; receives research support from the National Institutes of Health, the Food Allergy and Anaphylaxis Network, the Food Allergy Initiative, the National Peanut Board, SHS, and the Wallace Research Foundation; has provided legal consultation or expert witness testimony in cases related to food allergy; is on the Medical Board of Directors for the Food Allergy and Anaphylaxis Network; is a Dermatological Allergy Committee Member for the American College of Allergy, Asthma & Immunology; is a member of the National Institutes of Health Hypersensitivity, Autoimmune, and Immune-mediated Diseases Study Section; is a member of the Reviewer Board for the Journal of Allergy and Clinical Immunology; and is member of the food advisory committee for the US Food and Drug Administration. T. M. Guess declares that he has no relevant conflicts of interest.

REFERENCES 1. Simons KJ, Simons FE. Epinephrine and its use in anaphylaxis: current issues. Curr Opin Allergy Clin Immunol 2010;10:354-61. 2. Lieberman P, Nicklas RA, Oppenheimer J, Kemp SF, Lang DM, Bernstein DI, et al. The diagnosis and management of anaphylaxis practice parameter: 2010 update. J Allergy Clin Immunol 2010;126:477-80, e1-42. 3. Kemp SF, Lockey RF, Simons FE. Epinephrine: the drug of choice for anaphylaxis. A statement of the World Allergy Organization. Allergy 2008;63:1061-70. 4. Schwirtz A, Seeger H. Are adrenaline autoinjectors fit for purpose? A pilot study of the mechanical and injection performance characteristics of a cartridge-versus a syringe-based autoinjector. J Asthma Allergy 2010;3:159-67. 5. Song TT, Nelson MR, Chang JH, Engler RJ, Chowdhury BA. Adequacy of the epinephrine autoinjector needle length in delivering epinephrine to the intramuscular tissues. Ann Allergy Asthma Immunol 2005;94:539-42.

LETTERS TO THE EDITOR 1145

6. Stecher D, Bulloch B, Sales J, Schaefer C, Keahey L. Epinephrine auto-injectors: is needle length adequate for delivery of epinephrine intramuscularly? Pediatrics 2009; 124:65-70. 7. Edwards A. Why and how to use real EpiPens to practice administration. Available at: http://www.associatedcontent.com/article/1590602/inadequacies_of_epipen_ trainers_and.html. Accessed January 26, 2011. 8. Frew AJ. What are the ‘‘ideal’’ features of an adrenaline (epinephrine) auto-injector in the treatment of anaphylaxis? Allergy 2011;66:15-24.

Available online November 10, 2011. doi:10.1016/j.jaci.2011.10.007

Predetermined challenge eligibility and cessation criteria for oral food challenges in the HealthNuts population-based study of infants To the Editor: Oral food challenges (OFCs) remain the gold standard for diagnosing food allergy.1 However, to date, most studies describing the use of OFCs for the purposes of diagnosing food allergy have been retrospective clinical audits that have not clearly described crucial methodological characteristics, such as eligibility for challenge and exclusion and inclusion criteria.2,3 Study outcomes from these retrospective audits are difficult to generalize to other clinic or population cohorts because undeclared physician and parent selection bias is likely to significantly affect the inclusion or exclusion of, for instance, a child with a history of anaphylaxis. Furthermore, the validity of challenge cessation criteria cannot be formally assessed if they have not been prospectively developed and applied using objective and reproducible allergic signs. To date, there are no standard cessation criteria for the definition of a positive OFC result. Differences in food challenge cessation criteria across different studies and different centers will hinder the ability to (1) compare food allergy prevalence estimates between studies, (2) identify risk factors for the development of food allergy (because phenotypes might vary across different study cohorts), and (3) assess the success of various treatment strategies (including oral immunotherapy). Using a clear definition of eligibility criteria (with prospective decisions regarding inclusion/exclusion of those with a history of previous reactions, including anaphylaxis) and predetermined cessation criteria, we describe outcomes from more than 1000 OFCs in 12-month-old population-recruited infants that will help to inform future standardization of food challenges. The study methods have been described in detail elsewhere.4,5 Briefly, all participating infants recruited from a population-based sample (n 5 4457; response rate, 73%; mean age, 12.7 months; SD, 0.8 months) underwent skin prick tests (SPTs) to peanut, _1 mm) on egg, and sesame. Those with a detectable wheal (> SPTs (21%) underwent a hospital-based OFC to peanut butter, raw egg, or tahini paste. A subset of those with positive results to raw egg underwent a challenge to baked egg in the form of a muffin (dose equivalent to one sixth of an egg). We prospectively developed minimum objective criteria for defining a positive food challenge result and hence stopping a challenge in infants based on data from published studies6 and expert peer consultation. Infants were considered to have food allergy (and thus not offered a food challenge) if a parent reported a recent history of an objective allergic reaction (see below for the criteria used to define an objective reaction within the HealthNuts study) occurring within 2 hours of ingestion of the food and the infant had a

1145.e1 LETTERS TO THE EDITOR

FIG E1. Denim patch overlying ballistics gelatin.



FIG E2. Universal tester used to measure AFs and RFs. The crosshead (thin black arrow) displaces downward, causing compression at the end of the autoinjector into the gelatin and ultimately triggering the device. The EAI (thick white arrow) rests on a denim-covered cup of gelatin.

LETTERS TO THE EDITOR 1145.e2

1145.e3 LETTERS TO THE EDITOR

FIG E3. Custom fixtures were created to hold each unique brand of the different EAIs/EAITDs during testing. Each fixture consisted of a 40 mm 3 40 mm 3 50 mm block with an opening in the center that was slightly larger than the outside shape of the EAIs/EAITDs, such that they were vertically aligned for testing but also that their motion was not impeded by the fixtures. The fixtures were constructed from ABS plastic by using a Dimension BST 1200es series 3D printer (Eden Prairie, Minn).


LETTERS TO THE EDITOR 1145.e4


TABLE E1. Descriptive statistics for EAIs and EAITDs Actual Injector

EpiPen Activation Recoil Adrenaclick Activation Recoil Twinject Activation Recoil

Trainer

No.

Mean

SD

95% CI

No.

Mean

SD

95% CI

5 5

23.4 N 42.2 N

1.9 4.0

21.7-25.1 38.7-45.7

5 5

28.0 N 13.3 N

5.8 3.2

22.9-33.1 10.5-16.1

5 5

22.6 N 15.4 N

5.7 4.3

17.6-27.6 11.6-19.2

5 5

16.0 N 13.1 N

0.6 1.4

15.5-16.5 11.9-14.3

5 5

27.1 N 18.9 N

2.6 1.9

24.8-29.4 17.2-20.6

5 5

16.5 N 13.7 N

0.9 1.0

15.7-17.3 12.8-14.6