Intraosseous line placement for antidote injection by first responders and receivers wearing personal protective equipment

www.elsevier.com/locate/ajem

American Journal of Emergency Medicine (2011) 29, 373–381

Original Contribution

Intraosseous line placement for antidote injection byfirst responders and receivers wearing personalprotective equipment☆,☆☆,★

Stephen W. Borron MDa,g,⁎, Juan C. Arias MDa, Charles R. Bauer MDa,Thomas Philbeck PhDb, Patti Hass RN, MSNc, Wayne Lawson PA-Ca,Diana Montez RN, BSNb, Miguel Fernández MDa,Inkyung Jung PhDd, Donald J. Gordon PhD, MDe,f

aDivision of Emergency Medicine and the South Texas Poison Center, Department of Surgery,The University of Texas Health Science Center, San Antonio, TX 78229-3900, USAbVidaCare Corporation, San Antonio, TX 78216, USAcEmergency Center, University Hospital, San Antonio, TX 78229, USAdDepartment of Epidemiology and Biostatistics, The University of Texas Health Science Center,San Antonio, TX 78229-3900, USAeEmergency Health Sciences, School of Allied Health Sciences, The University of Texas Health Science Center,San Antonio, TX 78229-3900, USAfEmergency Medical Services, San Antonio Fire Department, Bluff Creek Tower, San Antonio, TX 78229, USAgDepartment of Emergency Medicine, Texas Tech University Health Science Center, El Paso, TX 79905, USA

Received 4 February 2009; revised 22 October 2009; accepted 24 October 2009

AbstractStudy objective: Early antidotal therapy may be lifesaving in hazardous materials victims. Intravenousline placement is difficult while wearing personal protective equipment (PPE). We assessed the abilityof protected, experienced first responders and limited-experience first receivers to place intraosseous(IO) lines for antidote administration.Methods: Six first responders donned 4 (A, B, C, and D) and 12 first receivers donned 2 (C and D)United States Environmental Protection Agency PPE levels in random order and then placed IO lines in1 of 4 anatomical sites in 12 anesthetized Spanish goats. Observers timed interventions until bolusinjection of isotonic sodium chloride solution.

☆ Presentation: This paper was presented at CBMTS VII, Spiez Laboratory, Spiez, Switzerland, on April 15, 2008, and at the American College ofEmergency Physicians Scientific Assembly on October 27, 2008.

☆☆ Funding: This study was funded by Merck Santé, Lyon, France, manufacturer of Cyanokit (hydroxocobalamin), and by VidaCare Corporation, SanAntonio, TX, manufacturer of the EZ-IO intraosseous needle system.

★ Conflict of interest: Dr Borron has received research funding, consulting fees, and speaker fees from Merck Santé and Dey, LP (which manufactures anddistributes hydroxocobalamin, respectively).

⁎ Corresponding author. Department of Emergency Medicine, Texas Tech University Health Science Center, El Paso, TX 79905, USA. Tel.: +1 210 5675762; fax: +1 210 579 6599.

E-mail address: [email protected] (S.W. Borron).

0735-6757/$ – see front matter © 2011 Elsevier Inc. All rights reserved.doi:10.1016/j.ajem.2009.10.009

374 S.W. Borron et al.

Results: First responders placed IO lines successfully in 100% of cases. The median (interquartilerange) times to completion (in seconds) were as follows: level A, 43.5 (23.0); B, 45.0 (29.0); C, 40.0(15.0); D, 30.0 (17.0). First receivers placed IO lines successfully in 91% of cases. The median(interquartile range) times to completion (in seconds) were as follows: level C, 42.0 (19.5); D, 37.0(11.0). There were no significant differences in time to completion among PPE levels (overall orpairwise) or between operator groups. Two (4%) of 48 line placements resulted in recognizedextravasation due to penetration of the opposite cortex. Infusions were completed successfully.Conclusion: Hazardous materials first responders and receivers can effectively place IO lines in a goatwhile wearing PPE. Intraosseous lines may facilitate earlier administration of antidotes in hazardousmaterials victims.© 2011 Elsevier Inc. All rights reserved.

1. Introduction

1.1. Background

Treatment of hazardous materials (hazmat) victims iscomplex. Victims are assumed to be contaminated unlessknown otherwise to protect the victim and rescuers.Contaminated victims often receive only basic medical care(spinal immobilization and airway opening) before undergo-ing decontamination. Hazmat team site entry, decontamina-tion station deployment, victim rescue, removal of clothing,and decontamination all take time. Patients exposed to highlytoxic and rapidly lethal compounds, such as cyanides, riskdeterioration or death before receiving lifesaving therapy,including antidotes.

Innovative approaches to hazmat victims have beenproposed. Protected rescuers might initiate antidotal treat-ment for victims even in the hot zone, at the discretion of theincident commander, before decontamination [1]. Althoughsuch an approach admittedly runs counter to current dogma,reality on the ground may demand consideration ofalternatives when faced with overwhelming conditions. Aterrorist chemical attack may involve dozens of victims. Itmay be impossible for a small number of protected rescuers toextricate a large number of cyanide-exposed victims in a timeframe permitting lives to be saved. Early administration of anantidote in the hot zone might permit self-extrication by somevictims or buy time for additional rescuers to arrive and donpersonal protective equipment (PPE, Fig. 1). This approachappears logical given the long decontamination delaysanticipated in chemical incidents, particularly with masscasualties. Specific antidote administration requires provi-sional characterization of the substance by source identifica-tion, environmental sampling, toxicant-specific warnings, orclinical toxidromes. Some antidotes require direct vascularaccess, including the recently approved cyanide antidotehydroxocobalamin. However, even if initial treatment isrelegated to the warm zone during decontamination, rescuersstill must wear chemical PPE, rendering intravenous (IV) lineplacement difficult [2-4]. This problem extends to thehospital, where hazmat receivers may have to decontaminatevictims who have undergone inadequate or no decontamina-tion before arrival [5,6].

Intraosseous (IO) puncture is an alternative to IV lineplacement in settings where time is critical and motor skillsare limited by PPE [4,7]. Intraosseous devices designedspecifically for emergency infusion have proven their valuein prehospital care [8-12], in emergency departments (EDs)[13-15], and in the wars in Iraq [16] and Afghanistan [17].Basic emergency medical technicians (EMTs), not normallypermitted to start IVs, may place IO lines [18] depending onlocal law, augmenting advanced life support capacity inmulticasualty situations. Multiple pharmacokinetic studieshave shown IO and IV administration routes to becomparable [19-22]. Emergency drugs normally adminis-tered IV may be given by the IO route [15].

The EZ-IO (VidaCare Corporation, San Antonio, TX)(Fig. 2) was developed in our institution. Its battery-poweredmechanical driver requires little force for needle placement,resulting in minimal damage to the underlying bone. Theergonomic design permits easy handling even while wearinggloves. The device is approved by the US Food and DrugAdministration for children and adults for any drug that maybe given intravenously during emergencies [23].

Successful human IO antidote administration has beenreported [24]. Intraosseous administration of anticonvulsantsin organophosphate-poisoned pigs has been studied [7]. TheParis Fire Brigade recently administered hydroxocobalaminIO to a child with smoke-related cyanide poisoning (personalcommunication, JL Fortin).

Few authors have studied IO placement while wearingPPE. Previous studies were not performed in live animals[3,4,6,25,26]. Experience in a mannequin model has shownthat the EZ-IO may be placed by operators in PPE faster thanIV line placement [4]. We studied IO placement by rescuerswearing PPE in a goat model, with subsequent administra-tion of hydroxocobalamin or isotonic sodium chloridesolution to determine whether IO antidote administrationmight be feasible in hazmat incident response involvingcyanides. Live animals were chosen for use in this study topermit the assessment of hemodynamic effects of antidotesadministered by the IO route. The administration ofhydroxocobalamin or isotonic sodium chloride solutionand subsequent hemodynamic monitoring are the subjectof another article (Am J Emerg Med, in press) and notdiscussed further here.

375Intraosseous line placement while wearing PPE

2. Materials and methods

This study was approved by the University of TexasHealth Science Center, San Antonio, Institutional AnimalCare Unit Committee (protocol 08027-71-01-B2). Allexperimentation was performed in compliance with existingUS Department of Agriculture regulations regarding thehumane treatment of animals. A veterinarian assessed the

Fig. 1 Personal protective equipment used in the study. Level A FRSP. LD FRCV. Levels of PPE as defined by the US Environmental Protection

health of the animals and supervised the administration ofanesthesia and experimentation.

2.1. Selection of participants

Volunteers served as operators in this study. Five EMTparamedics (EMT-Ps) from the San Antonio Fire Depart-ment EMSMedical Special Operations Unit and 1 physician,

evel B FRSP. Level C FRSP. Level C FRCV. Level D FRSP. LevelAgency.

Fig. 1 (continued).

376 S.W. Borron et al.

all with formal hazmat incident response operations training(29 CFR 1910.120), acted as first responders (FRSPs).Twelve ED physicians, physician assistants, and nurses actedas first receivers (FRCVs). The extent of training of thesereceivers in PPE and in the use of the EZ-IO variedconsiderably, consistent with the actual clinical situationin our hospital. All operators completed a questionnaire

Fig. 2 The EZ-IO power driver device and needles.

before participation to quantify their previous training andexperience in hazmat operations as well as the use of theIO device.

2.2. Study design

Spanish goats weighing between 15.5 and 35.1 kg (mean ±SD, 24.1 ± 6.4 kg) on the days of experimentation wereobtained for the study from 5R Farms, Floresville, TX. Goatswere chosen for the study because of prior experience withtheir use in health care provider training sessions forplacement of EZ-IO needles. The animals arrived in theanimal care unit no less than 4 days before experimentationand underwent veterinary examination, including dewormingand complete blood count. They were fed normal goat chowand water ad libitum until the evening before the procedure,when they were fasted. Water was allowed up untilexperimentation. Anesthesia was induced with ketamine10 mg/kg intramuscularly and xylazine 0.2 mg/kg intramus-cularly, then maintained with isoflurane 1.5% to 2.0% in100% oxygen via endotracheal tube. Mechanical ventilationwas maintained throughout the experiment. A carotid arterialline was placed for monitoring of vital signs. Sixteen goatswere used for the experiments and randomized to practice(n = 4), hydroxocobalamin (n = 6), or isotonic sodiumchloride solution (n = 6) groups. Animals were killed at theend of experimentation using sodium pentobarbital.

A complete list of PPE used is found in Table 1.Examples of the actual PPE used in this study are found inFig. 1. Designation of PPE was in accordance with USEnvironmental Protection Agency nomenclature [27]. LevelD for FRSPs consisted of their routine work uniform,

Table 1 Personalprotectiveequipmentwornduringexperimentation

Level Operator Chemicalprotective suits

Airwayprotection

Gloves

A FRSP Vapor-protectivefullyencapsulating a

SCBA e 2 Layers(nitrile i orlatex j, butyl k)

B FRSP Full, hoodedchemicalsplash suit b

SCBA e 2 Layers(nitrile orlatex, butyl)

C FRSP Full, hoodedchemicalsplash suit b

APR f 2 Layers (nitrileor latex, butyl)

C FRCV Chemicalsplash suit c

PAPR g 2 Layers (nitrileor latex, butyl)

D FRSP Routine workuniform witheye protection

N-95 h Nitrile or latexexaminationgloves

D FRCV Nonsterile gownand eyeprotection d

N-95 h Nitrile or latexexaminationgloves

a Tychem TK 648, Lakeland Industries, Ronkonkoma, NY.b Tyvek laminated with Saranex 23-P, Kappler CPF3, Guntersville, AL.c Tyvek laminated with Saranex 23-P, MarMac Mfg Co, McBee,

SC.d AMD Ritmed Isolation Gown Model A 8010, AMD-Ritmed, Inc,

Tonawanda, NY 14150.e Scott Air-Pak 75, Scott Health & Safety, Monroe, NC.f Scott-O-Vista Full Facepiece, Scott Health & Safety.g 3M Breathe-Easy, 3M Occupational and Environmental Safety

Division, St Paul, MN.h Gerson Isolair APR Model 2735 N95 Healthcare Particulate

Respirator, Louis M Gerson Co, Middleboro, MA.i Sempercare Nitrile PF,Model NIPFT105, Sempermed, Clearwater,

FL.j Diamond Grip PF, Model MF-300-L, Microflex Corporation,

Reno, NV.k Ansell Sol-Vex Model 37-155, Ansell Healthcare, Red Bank, NJ.

377Intraosseous line placement while wearing PPE

augmented by eye protection, examination gloves, and anN-95 mask. For FRCVs, universal precautions gear wasworn (nonsterile gown, eye protection, examination gloves,and an N-95 mask). Level C for FRSPs was a chemicalsplash protective suit with full face mask air-purifyingcartridge respirator (APR) and 2 layers of gloves. Firstreceivers wore a chemical splash protective suit andpowered air-purifying respirator (PAPR) with 2 layers ofgloves. Level B was worn only by FRSPs and consisted of afull, hooded chemical splash suit and external self-containedbreathing apparatus (SCBA). Two layers of gloves wereworn. Level A was worn only by FRSPs and consisted of avapor-protective fully encapsulating chemical suit withSCBA. Two layers of gloves were worn. All participantswore rubber boots.

The EZ-IO was used for all IO placements. Adult andpediatric human needles were variably used for the purpose(“Discussion”). Hydroxocobalamin (Cyanokit) was manu-factured by Merck Santé (Lyon, France) and supplied byDey, LP (Napa, CA).

2.3. Interventions

All operators were given brief (~10 minutes) instructionson the use of the EZ-IO. Some, but not all, had priorexperience with the device (“Results”). They were thenallowed 1 hour to practice using the device, in duty clothingor PPE at their option, on plastic simulated bones designedfor the purpose and on an anesthetized goat designatedfor practice. Most operators chose to practice in streetclothing wearing only examination gloves and/or butylrubber overgloves.

Operators were randomized to a particular experimentalanimal and to order of level of PPE worn for each attemptbefore arrival. First responders sequentially donned each of4 levels of PPE (Fig. 1), placing an IO needle in 1 of 4preassigned nonrandomized anatomical locations in the goat(right or left forelimb, right or left hindlimb). First receiversdonned 2 levels of PPE sequentially in random order, placingan IO needle in 2 sites randomly chosen from the anatomicallocations listed above. In sum, 6 FRSPs operated on 6 animals(1 each); and 12 FRCVs operated on 6 animals (1 animalshared by 2 operators).

2.4. Methods of measurement

Operators were timed by an observer from the momentthey touched either the EZ-IO needle cover or the driverthroughout needle placement and bone marrow aspiration(where possible) and until they completed injection ofa 5-mL bolus of isotonic sodium chloride solution.Successful completion was noted by aspiration of bonemarrow and/or facile injection of 5 mL of isotonic sodiumchloride solution. If the operator or the observer haddoubts about the placement, a second attempt was per-mitted, with the timer continuing nonstop until the secondneedle was placed, whether or not it was successful. Thefinal IO anatomical site placement in each animal, whichwas to be used for the subsequent study of anti-dotal administration and hemodynamics, was verifiedby fluoroscopy.

2.5. Primary data analysis

All statistical analyses were performed using SASVersion 9.1.3 for Windows software (SAS, Cary, NC).Serial simple randomization was used for determination oforder of PPE, operator to animal assignment, and animal toantidote assignment.

Because of the small sample size and skewed data, time tocompletion was summarized as median and interquartilerange (IQR) for each PPE level within operator group. Timeto completion was contrasted using the Kruskal-Wallis testamong PPE levels and using the Wilcoxon 2-sample testbetween operator groups. All statistical testing was 2-sidedwith a significance level of 5%.

378 S.W. Borron et al.

3. Results

The results of the operator experience survey are foundin Table 2. As a requisite for medical special operationsunit participation, the FRSPs had all completed one ormore hazmat incident response operations level trainingcourses and generally had greater experience in wearingPPE. Some had also attended an Advanced Hazmat LifeSupport (http://www.ahls.org) or other hazmat courses.Hands-on experience in actual hazmat incidents waslimited, however, even for FRSPs. Most had both previoustraining and practical experience using the EZ-IO. Firstreceivers, on the other hand, rarely had formal hazmattraining or experience apart from having attended Ad-vanced Hazmat Life Support training, which offers nohands-on practice with PPE or IO placement. Most hadprevious training but very little patient care experience withthe EZ-IO. Four FRCVs had never donned PPE or used anEZ-IO before the day of the experiment.

Times to completion for placement of the EZ-IO areshown in Table 3, and unsuccessful needle placements areannotated. Two operators (1 FRSP and 1 FRCV) required 2attempts in 1 level of protection each (level D and level C,respectively). Two placement failures resulting in completepenetration of the opposite cortex (2 FRCVs) were detectedby extravasation of contrast material into the soft tissuessurrounding the hindlimb under cinefluoroscopy. There were

Table 2 Operator prior experience questionnaire on use of PPE and

Operator Professionaleducation level

Hazmattraining

Hazmat drills(if yes, No.of incidents)

Hazm(if yeincide

FRSP 1 EMT-P A, B, C 6-10 0FRSP 2 EMT-P A, B, C 6-10 1-5FRSP 3 EMT-P A 1-5 0FRSP 4 EMT-P A 1-5 0FRSP 5 EMT-P A 1-5 1-5FRSP 6 MD A, B 6-10 1-5FRCV 1 MD B 1-5 0FRCV 2 MD None 0 0FRCV 3 MD None 0 1-5FRCV 4 PA-C None 0 0FRCV 5 MD None 0 0FRCV 6 MD B, C 6-10 1-5FRCV 7 PA-C None 0 0FRCV 8 RN C 1-5 0FRCV 9 MD A 6-10 1-5FRCV 10 MD B 1-5 0FRCV 11 RN C N10 1-5FRCV 12 PA-C None 0 0

MD indicates doctor of medicine; RN, registered nurse; PA-C, certified physiciHazmat training:

A. 40-hour minimum hazmat operations level courseB. 16-hour Advanced Hazmat Life Support Provider CourseC. Other: Advanced Disaster Life Support, Hazmat for Healthcare, National

no significant differences between PPE levels (overall orpairwise) or between FRSP and FRCV groups in times tocompletion, although completion times in level B werenumerically longer than those in the other levels and times tocompletion among FRSPs were numerically shorter thanthose among FRCVs.

Operators generally described the procedures as easy.Some operators commented that the butyl overglovesmade available to them were poorly fitting and that theynormally would have chosen better-fitting gloves. Visibil-ity did not appear to pose significant problems, although 1FRSP operator complained of fogging inside his levelC APR.

4. Discussion

Intraosseous line placement was readily accomplished inthis study even while wearing the highest levels of PPE,suggesting that earlier medical intervention may be possiblein the chemically contaminated patient. The ability to rapidlyadminister a lifesaving antidote, drugs for advanced cardiaclife support, and fluids for blood pressure support before orduring decontamination could theoretically improve patientoutcomes in chemical mass casualty or even single-casualtyincidents. In comparison with previously published data onIV line placement while wearing PPE, our data suggest that

the EZ-IO

at incidentss, no. ofnts)

EZ-IO practice(if yes, no. of trials)

EZ-IO experience(if yes, no. of patients)

N10 N101-5 6-101-5 1-51-5 1-51-5 6-101-5 00 00 01-5 01-5 00 06-10 00 0N10 6-106-10 6-106-10 01-5 00 0

an's assistant.

Fire Academy courses, military courses, etc.

Table 3 Successful IO placement and times to completion

Operator Time in seconds to completion/order of PPE wear/anatomical location

Level A Level B Level C Level D

FRSP 1 30 1 RF 35 2 LF 23 3 RH 21 4 LHFRSP 2 46 4 RF 70 3 LF 31 2 RH 26 1 LHFRSP 3 58 3 RF 41 2 LF 53 1 RH 31 4 LHFRSP 4 41 2 RF 43 3 LF 36 1 RH 29 4 LHFRSP 5 58 1 RF 116 4 a LF 46 3 a RH 43 2 LHFRSP 6 35 4 RF 47 1 LF 44 2 RH 70 b 3 LHMedian FRSPs 43.5 45.0 40.0 30.0IQR FRSPs 23.0 29.0 15.0 17.0FRCV 1 35 1 RF 31 2 LFFRCV 2 42 2 RH 46 1 RFFRCV 3 39 1 RF 68 2 RH d

FRCV 4 42 2 RH 27 1 LFFRCV 5 67 1 RF 41 2 LFFRCV 6 86bc 2 LH 23 1 RHFRCV 7 34 c 2 LH 35 1 RHFRCV 8 47 1 LF 40 2 LHFRCV 9 62 1 RF 37 2 LHFRCV 10 35 2 LH d 78 1 LFFRCV 11 50 1 RF 34 2 LHFRCV 12 38 2 RH 37 1 LFMedian FRCVs 42.0 37.0IQR FRCVs 19.5 11.0

No significant differences in time to completion were found among PPE levels within operator group. No significant differences in time to completion werefound between FRCV vs FRSP groups for PPE levels C and D. Levels of PPE: see text. RF indicates right forelimb; LF, left forelimb; RH, right hindlimb;LH, left hindlimb.

a Randomization protocol violation (PPE donned out of order).b Two attempts required.c Unsuccessful.d Randomization protocol violation (wrong anatomical site chosen).

379Intraosseous line placement while wearing PPE

IO lines may be placed more rapidly and with a greatersuccess rate than IV lines [2,3]. A recent direct comparisonof IO and IV placement in mannequins while wearing PPEbears this out [4]. However, we did not directly compare IOto IV placement. Given the many variables in our study, interms of personnel involved and their training, applicabilityof previous comparisons between IV and IO to our study isstrictly speculative.

Like IV placement, IO needle placement is not alwayssuccessful. We documented 2 cases of placement failure(4.2% of 48 placements). Placement failures have beenreported in the range of 0% to 28% in clinical trials wearingtypical work uniforms, with the EZ-IO generally associatedwith few placement failures [9,12,14,28]. Davidoff et al [9]reported a 97% success rate in their clinical trial of 250patients with the EZ-IO. Ben-Abraham et al [25] reportedfailure rates in the range of 20% in a group of rescuerswearing PPE and using the bone injection gun to attempt IOplacement in turkey bones ex vivo. Vardi et al [26] morerecently reported an 89% success rate with the bone injectiongun device in simulations involving 64 physicians in PPE.We attributed the failures in our study to overzealous use offorce (driving the needle hub to the skin) and use of humanadult needles in a small animal model. Both operators who

failed were FRCVs, one with no previous experiencewearing PPE or using the EZ-IO, the other having wornPPE numerous times and having performed IO placement inanimals, but never in humans. Interestingly, there were nofailures among FRSPs, who in general had both greaterexperience in PPE wear and in clinical use of the EZ-IO.Careful choice of needles and placement as well as increasedexperience should decrease the incidence of this complica-tion. In a review of IO use, Buck et al [15] reported thatextravasation, the most common adverse effect associatedwith IO use, has been reported in 12% of patients. Two casesof compartment syndrome have been reported [29].

The concept of delivery of antidotes in the hot zone runscounter to most current training and practice, and such anidea will no doubt be met with skepticism by some.However, experience suggests that in chemical mass casualtysituations, such as the Tokyo sarin incident, rescuer safetymay fall victim to the priority of saving lives [30]. Given asituation where the number of victims far exceeds thenumber of protected rescuers, a limited number of optionsexist. Byers et al [1] point out that in drills in the UnitedKingdom involving hazmat that used the standard approachof rescue from the hot zone, decontamination, and thentreatment, the time for victims to receive treatment was as

380 S.W. Borron et al.

long as 2½ hours. It seems to us that rapid administration ofan antidote by protected FRSPs in the hot zone might buytime for the arrival of additional appropriately protectedrescuers to begin the arduous task of extrication anddecontamination. Such an approach seems far safer thanhaving unprotected rescuers attempt extrication, as was thecase for most of the sarin victims [30]. Until such time that allemergency responders are trained and equipped for hazmatentry, difficult choices will continue to be made in chemicalmass casualty incidents.

We chose to study the IO route of administration as apotential alternative means of delivery of hydroxocobala-min, which normally is administered by the IV route. Themajor perceived disadvantage of hydroxocobalamin in thesetting of multiple chemical casualties has been therequirement for direct vascular access, which until nowcalled for IV placement. With the advent of spring-loadedsyringe pumps that require no electrical power [31], an IOinfusion of hydroxocobalamin could theoretically beinitiated in seconds; and the rescuer could move on tothe next victim. Although this infusion device has not yetbeen tested for this particular application, further evalua-tions are planned.

There exists a theoretical risk of introducing foreignmatter (chemical contamination or clothing) if the IO needleis introduced through potentially contaminated clothing.Removal of clothing and local skin decontamination beforeIO injection would further reduce the risk, but wouldincrease the time necessary to complete the procedure.Additional studies may shed light on this potential concern.

4.1. Limitations

Our study suffers from several limitations. First, our studywas performed in a controlled laboratory environment.Environmental factors such as light, temperature, humidity,victim clothing, responder fatigue, heat stress, and anxietywere not accounted for in the study. Any extrapolation to thehazmat scene must take these variables into account. Becauseof the complexity of the operations, randomization protocolviolations occurred twice, one in terms of PPE order and onein anatomical site order (requiring the subsequent operator tochange sites). Given the limited differences observed insuccess rates and completion times, we do not believe thatthese violations substantially altered the results. Differentlevels of experience in PPE and IO use, multiple IO sites, andorder of PPE use were considered important and included inthe study design. However, because of the limited number ofgoats used, these factors were not taken into account in theanalysis. Because meaningful power calculations were notpossible, the sample sizes we used were based on economicand logistical constraints.

Although the goat appears to be an acceptable model forIO needle placement, it does not precisely mimic human IOplacement because of differences in anatomy and bonedensity. Another limitation was the lack of an IV control

group due to the desire to limit animal numbers. Further-more, we chose to document direct evidence of correct IOplacement only for the fourth (final) IO placement in eachanimal because of the logistical difficulty in performingrepeated fluoroscopy in our operating suite. It is thus possiblethat other unrecognized IO placement failures occurred,although documented successful aspiration of bone marrowand easy injection of isotonic sodium chloride solution (theusual measures for clinical evidence of successful place-ment) diminish the likelihood of placement failure.

5. Conclusion

Hazardous materials FRSPs and FRCVs can rapidlyand reliably place IO needles in a goat model whilewearing PPE appropriate to their roles. The data suggestthat IO administration of antidotes such as hydroxocoba-lamin by hazmat responders is feasible where deemed safeand appropriate.

Acknowledgments

Particular thanks go to Dr Larry Miller of VidaCare forlogistical and equipment assistance. The study would nothave been possible without the generous support of the SanAntonio Fire Department EMS Medical Special OperationsUnit and, in particular, Donald J Gordon, PhD, MD, FormerMedical Director, and Terry Eaton, EMT-P, who recruitedthe medics and assisted with PPE. We would like to thankparamedics Brian D Worley, Robert W Dugie, EmmettGuzman, and Michael Pierce and 2 other medics whopreferred to remain unnamed, who acted as FRSPs, and DrClaudio Zeballos, Dr Ricardo Hernandez, Jr, Dr GeorgeLayton, and PA-C Terry Henrie, who (in addition to variousauthors) acted as FRCVs in the study. We appreciate thecareful review of the biostatistics in our article by Dr JoelMichalek. The authors deeply regret the loss of one of ourcollaborators in this study, Dr Michael Sanchez. Hisenthusiastic support of our research and of the bettermentof emergency medicine will be sorely missed.

References

[1] Byers M, Russell M, Lockey DJ. Clinical care in the “Hot Zone”.Emerg Med J 2008;25(2):108-12.

[2] Berkenstadt H, Arad M, Nahtomi O, et al. The effect of a chemicalprotective ensemble on intravenous line insertion by emergencymedical technicians. Mil Med 1999;164(10):737-9.

[3] MacDonald RD, LeBlanc V, McArthur B, et al. Performance ofresuscitation skills by paramedic personnel in chemical protectivesuits. Prehosp Emerg Care 2006;10(2):254-9.

[4] Suyama J, Knutsen CC, Northington WE, et al. IO versus IV accesswhile wearing personal protective equipment in a hazmat scenario.Prehosp Emerg Care 2007;11(4):467-72.

381Intraosseous line placement while wearing PPE

[5] Garner A, Laurence H, Lee A. Practicality of performing medicalprocedures in chemical protective ensembles. Emerg Med Australas2004;16(2):108-13.

[6] Udayasiri R, Knott J, Mc DTD, et al. Emergency department staff caneffectively resuscitate in level C personal protective equipment. EmergMed Australas 2007;19(2):113-21.

[7] Eisenkraft A, Gilat E, Chapman S, et al. Efficacy of the bone injectiongun in the treatment of organophosphate poisoning. Biopharm DrugDispos 2007;28(3):145-50.

[8] Macnab A, Christenson J, Findlay J, et al. A new system for sternalintraosseous infusion in adults. Prehosp Emerg Care 2000;4(2):173-7.

[9] Davidoff J, Fowler R, Gordon D, et al. Clinical evaluation of a novelintraosseous device for adults: prospective, 250-patient, multi-centertrial. Jems 2005;30((10):suppl):20-3.

[10] Gillum L, Kovar J. Powered intraosseous access in the prehospitalsetting: MCHD EMS puts the EZ-IO to the test. JEMS 2005;30((10):suppl):24-5.

[11] Fowler R, Gallagher JV, Isaacs SM, et al. The role of intraosseousvascular access in the out-of-hospital environment (resource documentto NAEMSP position statement). Prehosp Emerg Care 2007;11(1):63-6.

[12] Frascone RJ, Jensen JP, Kaye K, et al. Consecutive field trials usingtwo different intraosseous devices. Prehosp Emerg Care 2007;11(2):164-71.

[13] Nijssen-Jordan C. Emergency department utilization and success ratesfor intraosseous infusion in pediatric resuscitations. CJEM 2000;2(1):10-4.

[14] Fiorito BA, Mirza F, Doran TM, et al. Intraosseous access in the settingof pediatric critical care transport. Pediatr Crit CareMed 2005;6(1):50-3.

[15] Buck ML, Wiggins BS, Sesler JM. Intraosseous drug administration inchildren and adults during cardiopulmonary resuscitation. AnnPharmacother 2007;41(10):1679-86.

[16] Timboe HL, Bruttig SP, Ruemmler MW. Adult IO in the combat zone:the past, present and future use of intraosseous infusion by the U.S.military. Jems 2005;30((10):suppl):27-8.

[17] Cooper BR, Mahoney PF, Hodgetts TJ, et al. Intra-osseous access(EZ-IO) for resuscitation: UK military combat experience. J R ArmyMed Corps 2007;153(4):314-6.

[18] Guyette FX, Rittenberger JC, Platt T, et al. Feasibility of basicemergency medical technicians to perform selected advanced lifesupport interventions. Prehosp Emerg Care 2006;10(4):518-21.

[19] Prete MR, Hannan Jr CJ, Burkle Jr FM. Plasma atropine concentra-tions via intravenous, endotracheal, and intraosseous administration.Am J Emerg Med 1987;5(2):101-4.

[20] Cameron JL, Fontanarosa PB, Passalaqua AM. A comparative study ofperipheral to central circulation delivery times between intraosseousand intravenous injection using a radionuclide technique in normovo-lemic and hypovolemic canines. J Emerg Med 1989;7(2):123-7.

[21] Orlowski JP, Porembka DT, Gallagher JM, et al. Comparison study ofintraosseous, central intravenous, and peripheral intravenous infusionsof emergency drugs. Am J Dis Child 1990;144(1):112-7.

[22] Sapien R, Stein H, Padbury JF, et al. Intraosseous versus intravenousepinephrine infusions in lambs: pharmacokinetics and pharmacody-namics. Pediatr Emerg Care 1992;8(4):179-83.

[23] VidaCare emergency department data sheet. Available at: http://www.vidacare.com/download_file.php?file=DataEmergencyLR.pdf.Accessed December 27, 2008.

[24] Herman MI, Chyka PA, Butler AY, et al. Methylene blue byintraosseous infusion for methemoglobinemia. Ann Emerg Med Jan1999;33(1):111-3.

[25] Ben-Abraham R, Gur I, Vater Y, et al. Intraosseous emergency accessby physicians wearing full protective gear. Acad Emerg Med 2003;10(12):1407-10.

[26] Vardi A, Berkenstadt H, Levin I, et al. Intraosseous vascular access inthe treatment of chemical warfare casualties assessed by advancedsimulation: proposed alteration of treatment protocol. Anesth Analg2004;98(6):1753-8table of contents.

[27] Personal protective equipment. U.S. Environmental Protection Agency.Available at: http://www.epa.gov/emergencies/content/hazsubs/equip.htm. Accessed February 4, 2009.

[28] Waisman M, Waisman D. Bone marrow infusion in adults. J Trauma1997;42(2):288-93.

[29] Ribeiro JA, Price CT, Knapp Jr DR. Compartment syndrome of thelower extremity after intraosseous infusion of fluid. A report of twocases. J Bone Joint Surg Am 1993;75(3):430-3.

[30] Tokuda Y, Kikuchi M, Takahashi O, et al. Prehospital management ofsarin nerve gas terrorism in urban settings: 10 years of progress afterthe Tokyo subway sarin attack. Resuscitation 2006;68(2):193-202.

[31] Springfusors non-electronic infusion. Allied Medical. Available at:http://www.alliedmedical.com.au/clinical/springfusor/?c=3. AccessedMarch 20, 2008, 2008.