Comparison of a Vacuum Splint Device to a

Rigid Backboard for Spinal Immobilization

DAVID Fl. JOHNSON, MD,'T MARK HAUSWALD, MD,‘|'

In thisstudy. comparison ol a vacuum splint device to a rigid backboardwas made with respect to comlort. speed ot application, and degree olimmobilization.The study was a prospective, nonblinded comparativestudy conducted at a statewide emergency medical services (EMS)training lac‘ ‘ and included a convenience sample ol emergencymedical technician (EMT) and paramedic students. The vacuum splintwas judged to be significantlymore comlortahle on a 10-point scale thanthe rigid backboard alter subiects had been lying on each device tor 30minutes (F < .001). it was also taster to apply: 131.6 2 2&3 secondsversus 154.6 : 22.2 seconds (P < .001). Various measures ol immobiliza-tion were siml at lot the two devices. The vacuum splint provided betterimmobilizationol the torso and less slippage on a gradual lateral tilt.Therigid backboardwith head blocks was slightly better at immobilizingthehead. Vacuum splints utter a significant improvement in colnlort over atraditionalbackboard lor thepatient with possible spinal injury. They canbe applied in reasonable time times and provide a similar degree ofimmobilizationwhen compared to a standard rigid backboard. (Am .1Emerg Med 1996;14:369-312. Copyright C 1595 by W.B. Saunders Com-F3"Yl

Prehospitzil spinal immobilizationhas become the stan-dard of care for transporting patients who may have tservicalspine injuries.‘-2 The usual procedure is to strap the patientonto a wooden backboard. securing the head with tape incombination with blanket rolls, sandbags. or a variety ofcommercial devices.” A rigid cervical collar is often usedand the torso is secured with straps. Although some refine-ments to this system have been made. the practice of using ahat. rigid surface as it splint for the spine has changed verylittle since the 19605.‘

Rigid backboards can cause significant head. neck. andback pain in normal volunteers} If similarpain develops intrauma patients, it may be difficult to differentiate pain as aresult of trauma from that caused by a cervical collar or thebackboard. Because emergency physiciansbase the decisionto obtain cervical spine radiographs. in part. on the presenceof neck pain. many patients may be getting unnecessaryradiological procedures.“ ’

Vacuum splints contain numerous pol styrene beadsencased in a flexibleouter shell. They are i itially soft and

From the ‘New Mexico EMS Academy. and the fDepartmerit olEmergency Medicine, University ol New Mexico School or Medicine.Albuquerque.

Mariuscnpt received June 29, 1995, returned August 5. 1995;revision received August 24, 1995, accepted September5. I995

Presented at the Society lor Academic Emergency Medicine,Washington Dc. May 22. 1994

Address reprint requests to or Johnson. New Mexico EMS Acad-emy. 2700 Yale SE. Albuquerque. NM 87106.

Key Words: Emergency medical services. prehospltal. spinalimmdtntizatton, trauma.

Copyright c 1995 oy w B. SaundersCompany0735-5757/96/1404-000555.00/0

CY STOCKHOFF MS, EMT-I’

malleable.but when air is removed they become rigid. Sincethese devices conforiti to the shape of the patient. it seemslikely that they would be more comtonable than othersplinting devic s.

We Compared ii vacuum splint mattress to ii standard.wooden backboard with respect to patient comfort. We alsocompared the two devices with regard to the adequacy ofimmobilizationand speed of application using human volun-teers.

METHODSTwo methods oi spinal immobilization were compared with

respect to comfon, speed ofnpplicniion. and degree or immobiliza-tion. A vacuum splint device (Evdc U Splint. Hartwell Medical.Carlsbad. CA) was compared to a traditional wooden backboard

The first phase or the study evaluatecl the comfort and speed ofapplication of each device. A convenience sample of 30 studentswas recruited from four emergency medical technician (EMT)-Basic classes. None of the students had any prior emergencymedical services (EMS) field experience or knowledge or spinalimmobilizationtechniques. All students were given training in theuse or each device and allowed approximately 4 to 6 hours ofpractice to master each of the techniques in a practical ldb setting."Hie students worked in groups of four. with one student selected toact as the "pzitlcnti"

The vacuum splint was applied according to the manufacturer'srecommendations except that 2—lnt~h tape was used to secure thehead rather than the velcro strrips provided by the manufacturer.Previous experience with these straps indicated that they werecumbersome and did not adequately secure the head. After thesplint had been flattened and smoothed out. air was removed tomake it rigid in order to allow the subject to be logrolled onto thedevice. The air intake valve was then opened to allow air back intothe system to soften it. The torso was secured using the suppliedstraps. Air was then removed from the vacuum splint by onestudent using -.i foot pump supplied by the manufacturer.As ill!’ was

evacuated, a second student molded the splint to the head and neckas the third maintained manual stabilizationof the head. The headwas then secured with tape.

The wooden backbtxird was applied in a similar fashion with 3Students logrolling the subject onto the board. The torso was

secured unrig 2-inch nylon straps applied in a cross-the-heartfashion and in n parallel technique over the legs. This strappingtechnique was the most consistently taught method in local EMTcourses and was therefore used in the study. The head was securedwith it headblock (Headhugger. Fcmoewiishington. Inc.. Wilming-ton. OH) and limb tripe. Rigid cervical collars were not used witheither device during this portion of the study to eliminate discom-rort caused by the L’t)ll1lrIt~‘Ell,

An instructor timed each procedure and students were randomlyassigned tor placement on the vacuum splint or the woodenbackboard first. After lying on each device for 30 minutes thestudents rated comfort on a |0~p0int scale. A score or I representedthe cnmfon ofli Concrete slab and it] was equivalent to their bed at

369

370 AMERICAN JOURNAL OF EMERGENCY MEDICINE I Volume 14, Number 4 I July 1996

home. Each participant had the opportunity to evaluate bothdevices.

The second phase of the study compared immobilizationofsubjects while in each of the devices Thirty additional studentswere recruited from two EMT—Basic classes and one paramedicclass. Students from the paramedic class had had previous experi-ence with the 'gid backboard technique but not with the vacuum

splint. Individuals were immobilizedin the same manner describedabove for the timed trials. When immobilizationwas measured,however, each device was evaluated both with and without theapplication of a rigid cervical collar (Stifneck, Laerdal, LongBeach. CA).

Immobilizationwas measured by placingeach individual and thetest device on 2 platfomi which could be tilted in a lateral directionup to a maximum of 90 degrees from the floor (horizontal tilt).Theangle of tilt criuld be measured using a large protractor at the baseof the device (Figure 1). Students were instructed to verballyindicate Llte point at which they first noted slippage of any part oftheir body during a gradual lateral till. This angle was thenrecorded. The platform was then tilted to a full 90 degrees tomeasure lateral movement of the head. shoulders. and hips. Thisprocedure was chosen to simulate a maneuver commonly used byEMTs when a patient suddenly begins to vomit and is at risk foraspiration. To make these measurements, flexiblemetal arms wereattached to the tilt platform using magnets (Figure 2). The metalarms were initially placed in Contact with the nose, right shoulder,and right hip. Movement away from each flexible arm was

measured in centimeters after the subject was fully tilted to 90degrees.

Except where otherwise indicated, results are presented asmean : SD. The Watson-Vvilliams test was used to analyzedifferences between angles.“ All other data were analyzed usingStatCtrnplucs Plus version 6.0 statistical software (Marlugistics,lnc., Rockville. MD). The Mann-Whitney rank sum test was usedto analyze the ordinal scale cornfort'd:lta, and the t test was used forspeed of application. immobilizationwas analyzedwith analysisofvariance (ANOVA)using theTukey test tor multiple comparisons.A critical value of P < .03 was considered significant.

This Slucly was approved by the Human Research and ReviewCommittee of the University of New Mexico. Informed consentwas obtained from each study participant.

FIGURE l. Measurement of the angle of till after slippage isinitially noted by the subject

FIGURE 2. Flexible metal arms used to measure movement ofthe head, shoulders, and hips after 2 90—degree tilt.

RESULTSThe vacuum splint device was found to be significantly

more comfortable (Figure 3). It was also faster to apply:131.6 1 24.3 seconds versus 154.6 : 22.2 seconds(P < .001).

The angle of initial slippage on gradual lateral tilt isshown in Table 1. Subjects in the vacuum splint could betilted significantly more than those on the wooden back»board before movement was noted. Application of a rigidcervical collardid not significantlychange these results.

Tables 2 and 3 compare each of the devices with andwithout a cervical collar with respect to movement of eachpart of the body after a full 90~degree tilt. Although therewas more movement of the head while in the vacuumsplint,more movement of the shoulders and hips was noted withthe backboard. The difference between movement at thehead and movement of the shoulders was less for thevacuum splint than for the backboard (without collar:0.8 : L4 cm v 71.5 t 1.7 cm, P < .001). Addition of Zlrigid cervical collardid not significantlychange this relation-shlp(0.6 : 1.5 cm v *l.8 : 1.5 cm. P < .001).

DISCUSSIONImmobilizationof patients withsuspected spine injuries is

a procedure commonly perfomaed by EMTS in the prehospi—tal setting. Procedures for accomplishingspinal immobiliza~tion are well described in prehospital care texts, but havechanged Very little since described by Farrington in 1968.":-‘Subsequent work has determined that immobilizationis bestaccomplished by placing tl1e patient on a rigid surface andsecuring the patient’s head with tape along with sandbags orsimilar devices.’ Rigid cervical collars alone do not signiti»cantly restrict movement of the head.’-9

Recently,Chan et al5 have shown that immobilizationon a

rigid backboard forjust 30 minutes can result in significantdiscomfort to the patient. Neck or backpain perceived by thepatient may prompt the emergency physician to order a

radiographic examination of the spine.‘ Therefore it ispossible that pain caused by a rigid backboard may result inunnecessary radiographs of the spine in trauma patients.

Rigid backboards have also been associated with the

JOHNSON ET AL I SPINAL lMMOB|L|ZAT|ON

12

to

371

TABLE2. Comparison ol Lateral Movement at Nose, Shoulders,and Hips in Each DeviceWhen Used Without a Rigid Cervical Collar

Movement (cm)Head Shoulders Hips

12

in

0 345575910B 1 2

FIGURE 3. Frequency of comfort ratings on a l0—point scale (1represents a concrete slab and 10 represents the comfort of thesubject's bed at home): (A) vacuum splint, median score = 5; (B)rigid backboard.median Score = 3 (Mzmn—Whi[ney test, P < .001).

development of pressure sores in spinal cordeinjured pa~rientsllv” Such patients are unable to take pressure off bonyprominences on their own. It is possible that a device that iseither padded or conforms to the shape of the patient mayreduce thesepotentiallydevastating complications.Vacuum splint devices have been available for several

years and offer the possibilityof improved comfort becausethey conform to the shape of the patient. Before thesedevices can be recommended for general use, they should beshown to immobilizeat least as well as a traditional rigidTABLE1. Angle ol Initial SlippageWith a Gradual Lateral Tilt

AngleN (degrees : so) P

Vacuumsplint withoutcollar so 29.2 : 7.0 }Backboardwithout collar so 19.2 2 4.3 <~°°‘

Vacuumsplint with collar so 30.2 : 5.5Backboardwithcollar so 20.2 : 7.3 i <'°°‘

VacuumSplint 4.3 : 1,6 3.5 : 1.6 1.3 : 0.9P <.001 NS <.O01Backboard 2.6 : 1.1 4.1 : 1.7 3.4 : 1.5

backboard. In addition, they should be easy to use and mustbe applied in reasonable time frames so that prompt patienttransport is not delayed.Our study showed that a vacuum splint device is signifi~

cantly more comfortable thana rigid. wooden backboard. Inaddition, this device can be applied faster than a backboardusing accepted strapping techniques. The differences notedin speed of application were not, however, clinicallysignifi-cant. Althoughothertechniques for strapping the patient to a

rigid backboardmay have resulted in faster times, it seemsfair to conclude that application of the vacuum splint doesnot take significantly longer. Because of design differencesin the two devices, it was not possible to control forstrapping technique. We therefore chose a technique forstrapping a patient to the rigid backboard which was

commonly taught in local EMT courses.We measured several parameters of immobilizationfor

both devices. When subjected to a gradual lateral tilt,slippage of the subject was noted at a much smallerangle onthe rigid backboard thanon the vacuumsplint. When the twodevices were compared using a full 90-degree tilt, smalldifferences in immobilizationwere noted. Specifically, thevacuum splint was superior at immobilizationof the torso,whereas the backboard provided better immobilizationofthe head. It was interesting to note that the difference inmovementbetween thehead and shoulders was smallerwiththe vacuum splint than with the rigid backboard. Thissuggests that theremight be smallershear forces exerted onthe neck when a subject on the vacuum splint is subjected toan extreme lateral tilt.The larger degree of movement of the head while

immobilizedin the vacuum splint was believedby some ofthe study participants to be attributable to the design of thedevice. There did not seem to be enough padding around thehead and neck to allow tight molding in these areas. It maybe useful to alter thedesign of the vacuum splint so that it ispartially segmented to allow separate molding around thehead and torso. Some of the larger subjects believedthat thedevice needed to be wider to accommodatethem.Previous investigators have studied various immobiliza-

tion devices by taking goniometricor radiographicmeasure-TAELE 3. Comparison of Lateral Movement at Nose, Shoulders,and Hips in Each DeviceWhen Used With a Fiigid Cervical Collar

Movement (cm)Head Shoulders Hips

VacuumSplint 4.0 2 1.4 3.4 : 1.7 1.4 1* 1.0P < .001 < .003 < .001Backboard 2.3 1 1.1 4.6 :1,E 3.4 t 1.6

372 AMEFHCAN JOURNAL OF EMERGENCY MEDICINE I Volume 14, Number 4 I July 1996

merits while the subject is actively exerting pressure againstthe study device in various planes.3i9i'1” Interpretation ofthese studies is difiicult because the amount of pressureexerted by individuals of differing strengths is not standard-ized. These studies also did not measure movement of thetorso.’-9-‘3-*3 Movement of the torso relative to the headwould be expected to produce shearing forces on the neckand could be an important concem.We chose to use a 90-degree tiltwhen evaluating immobi-

lization because this is a procedure that is often employedwhen a patient with an unsecured airway begins to vomit.Our study design did not permit measurement of flexion orextension of the neck. This may be why we were unable toshow a benefit from a rigid cervical collar. In one previousstudy, the cervical collar, when used with tape and sandbags,prevented only extension of the neck but did not signifi-cantly reduce movement in otherdirections}Although the differences in movement of various parts of

the body were often statistically significant, the clinicalimportance is probably minimal. It is likely that there issubstantiallymore movement of the spine during extricationand placement of the patient on the immobilizationdevice.Significant movement of the lumbar spine. for example, hasbeen noted while using the logrollingmaneuver.”Regardless of the methodology used. this study suffers

from the same drawbacks of previous efforts to investigatespinal immobilization.The fact that normal volunteers withno spinal injuries were used limits our conclusions. Clearly.a large clinical study evaluating the effectiveness of spinalimmobilizationin preventing further injury in patients withspinal injuries is needed, Given the relative infrequency ofthese injuries and the difficulty in collecting accurateprehospital data. such a study may never be done. In theabsence of such a study, new methods for immobilizationcan only be compared to time-honored devices such as therigid backboard. If new devices can achievesimilar degreesof immobilizationwith more comfort for the patient. theymay represent an improvement in patient care.It should also be noted that this study was conducted in a

very controlled setting. Clearly, a field study would behelpful in evaluating both devices with respect to otherfactors that may affect the usefulness of each device. Suchfactorsmight include use in extrication,durability,and easeof application in a variety of settings.Although the vacuum splint compared favorably to the

backboard in thisstudy, it has some drawbackswhich shouldbe mentioned. First, it is more expensive than a woodenbackboard. This is, however, a piece of equipment that maybe used frequently and is much less expensive than manyother pieces of prehospital equipment that are used onlyrarely. Second. it does not fit into the standard long, narrowbackboard compartment of an ambulance. Alternate meth-ods for storing the device, however, should not be a majorproblem.

One final point should be made with respect to vacuumsplint devices. Even with all of the air removed from thedevice. they are not rigid enough to serve as an extricationdevice in some situations. Indeed, traditional rigid back~boards serve two functions in prehospital care. In addition toimmobilization,they function as large spatulas to lift andmove patients. If vacuum splints are to gain widespreadacceptance, techniques for extrication of patients onto thesedevices must be developed and evaluated.

CONCLUSIONSWith respect to the variables evaluated in this study, the

vacuum splint seems to be a reasonable alternative totraditional spinal immobilizationusing a rigid backboard. Inaddition to being significantlymore comfortable. immobili-zation and speed of application are similar to a traditionalrigid backboard.The addition of a rigid cervical collar addedlittle to immobilizationon either the vacuum splint or therigid backboard in this study.

REFERENCESt. Spine management skills In Mcswain NE (ed). Prehospital

TraumaLife Support (ed 2). Akron, OH, Educational Direction, inc.1990, pp 2252562. Central nervous system. In Caroline NL (ed): Emergency Care

in the Streets (ed 3). Boston, MA, Little, Brown, 1937, PD 3453743 Fodolsky S, Earafl LJ, Simon HR, at al: Ellicacy oi cervical

spine immobilizationmethods.J Trauma19833234614654. FarringtonJD: Extricationoi victims-surgicalprinciples, 4 Trauma

l95S:E:493—5l25 Chan D. Goldberg R, Tascone A, et al: The effect of spinal

immobilization on healthy volunteers. Ann Emerg Med 199423:48-516. Baciiulis BL, Long WE, Hynes GD, et al: Clinical indications lot

cervical spine radiographs in the traumatized patient. Am J Surgl987.153:473~4777. FlingenbeigBJ, FlsherAK. Urdaneta LF, et al: Rational ordering

oi cervical spine radiographs following trauma. Ann Emeig Medl98E;l7:792-7963. Circular distributions: Hypothesis testing. In ZarJH (ed): Blasts-

listical Analysis (ed 2). EnglewoodClifls, NJ, Prentice Hall, 1934, pp44041599. Cline JR. Scheidel E. Bigsby F. A comparison or methods of

Cervical immobilizationused in patient extrication and transport. JTrauma1985252649-65310 Linares HA, MawsonAR. Suarez E, et alzitssociationbetween

pressure sores and immobilization in tile immediate post-iniuryperiod Orthopedics 19877102571-573

V1, Mawson AR, Biundo JJ, Neville P, et al Fllsk lactors for earlyoccurring pressure ulcers following spinal cord injury. Am J Phys MedRehabil 198816712342712. Flosen PB. Mcswain NE, Arata M, et al: Comparison of two

new collars. Ann Emerg Med1992;21:11E9—119513. Graziano AF. Scheidel EA. Cline JR, et al. A radlographic

comparison ol ptehospital cervical immobilization methods. AnnEmerg Med I9B7Il5:1127-113114. McGuire RA. Neville S. Green EA, et al: Spinal instabilityand

the log-rolling maneuver. J Trauma1987271525-531