HomeMy WebLinkAboutItem 7.2 AutoPulse Resuscitation
CITY CLERK
File # D~[YJ[Q]-~[]
AGENDA STATEMENT
CITY COUNCIL MEETING DATE: September 18, 2007
SUBJECT:
Report on the AutoPulse Resuscitation System
Report Prepared by David RochaDeputy Chief, Alameda County
Fire Department and Joni Pattillo, Assistant City Manager
ATTACHMENTS:
1. June 14, 2006 letter from Dr. James E. Pointer. Alameda County
Medical Director
2. July 12, 2007 letter from Dr. James E. Pointer, Alameda County
Medical Director
3. Article from the June 14,2006 Journal of American Medical
Association
~
1. Receive the Staff Report and authorize the removal of the
AutoPulse from use in the City of Dublin.
RECOMMENDATION:
~
2. Direct Staff to pursue credit for the two Dublin AutoPulse units
as part of purchase agreement for Zoll E series heart monitors.
The full amount of credit for the two units that were donated by
the Dublin/San Ramon Womens Club would be allocated
exclusively for the City of Dublin's portion ofthe Zoll E series
heart monitors.
FINANCIAL STATEMENT:
Adequate funds have been budgeted in the City's Fiscal Year 2007-
2008 budget for the estimated first year lease-purchase cost for
Dublin's allocated cost for the replacement heart
monitors/defibrillators.
DESCRIPTION:
Back2:round
The General Federation of Women's Club (GFWC) of Dublin/San Ramon Women's Club made a
generous donation to the City of Dublin by purchasing two AutoPulse Resuscitation Systems. The units
have been in use at Fire Station 16 and Fire Station 17. Until more data became available, it was
determined that the best course of action was not to pursue the acquisition of an AutoPulse device for
Station 18. The unit at Fire Station 16 has been utilized for over two years and the unit at Fire Station 17
COpy TO: General Federation of Women's Club (GFWC) of Dublin/San Ramon Women's Club
Page 1 of3
ITEM NO.
'71~
is just approaching two years of service. When the systems were accepted, it was Council's direction that
Staff study the AutoPulse System and return to Council with an evaluation of its findings.
AutoPulse Description and Findings
The AutoPulse serves as a resuscitation adjunct, mechanically circulating blood to the heart, brain, and
other vital organs during cardiac arrest treatment. Proponents say that because the AutoPulse applies
pressure more broadly across the chest cavity than manual CPR, it does a better job of increasing blood
flo,:\, throughout the body. Laboratory tests on animals, along with hospital based trail studies, have shown
that the machine pushes as much as one-third more blood through the heart than manual CPR.
Conceptually, the concept of AutoPulse appears to be sound. Unfortunately, there are serious concerns
being raised throughout the medical community regarding its efficacy and safety. As a result of
information from the California EMSA Medical Director that Zoll has voluntarily withdrawn sales in the
California market, the Alameda County EMSA Medical Director asked all County ALS providers not to
purchase or acquire further AutoPulse devices in June of2006.
A report that was completed in June 2006 in the Journal of the American Medical Association looked at
the use of AutoPulse for patients suffering out-of-hospital cardiac arrests. One study, conducted in nearby
San Francisco, found improved survival rates for patients who received automated Cardiopulmonary
Resuscitation (CPR) versus those who received manual CPR. The second study (Attachment 3), a large
randomized prospective AutoPulse Assisted Prehospital International Resuscitation (ASPIRE) trial,
conducted in the states of Washington, Ohio, and Pennsylvania, as well as two provinces in Canada,
arrived at a vastly different conclusion. This research found that patients who received the AutoPulse had
worse neurological outcomes and poorer survival rates than patients receiving manual CPR. The ASPIRE
researchers, in fact, stopped the study after early data showed that only 5.8% of the patients receiving
AutoPulse CPR survived to hospital discharge compared with 9.9% of the patients given manual CPR.
According to the City of Richmond, Virginia EMS Medical Director Joseph P. amato, an internationally
recognized CPR researcher, there is no clear cut reason for the vast differences in the studies. "I don't
think any of us understands all the reasons for the difference in outcomes," he says. "But it's probably how
the device is used, when in the CPR sequence it's used, the quality and intensity of the training, and the
quality improvement (Q1) provided to medics using the device."
Unfortunately, the Alameda County Fire Department (ACFD) cannot report that the device has had a
positive effect on the outcome of any patients in the City of Dublin. However, it would be inappropriate
to base this evaluation on the ACFD's experience with AutoPulse because the number of deployments of
the devices was too low to be considered a reasonable reflection or valid basis for conclusion. The
devices were only deployed five times. ACFD staff prefers to evaluate the device based on larger
statistical valid studies. Based on the information from the studies, ACFD would like to remove the
AutoPulse units from service in the City of Dublin and use this as an opportunity to offset the cost of
acquiring Zoll E series heart monitors.
There is an opportunity to participate in the registry project sponsored by Zoll, in order to continue to
'study' patient survival rates, to see if the device actually works as intended. Unfortunately, this move
puts ACFD on the sidelines for a period of time with no guarantee that the risks of using the device
outweigh the benefits. Since overall survival rates for cardiac arrests are already low, and because
variables in delivering patient care abound, it is difficult to say if the AutoPulse machines truly make a
difference.
Proposed Next Steps
In ACFD Staff's opinion, the first rule of medicine - do no harm to the patient - should prevail. To that
end, staff believes the AutoPulse units should be removed from service, opting to improve ACFD's ability
to provide acute cardiac care in other ways.
Currently, the ACFD is planning to replace its entire fleet of outdated Zoll 1600 heart monitors .with new
state-of-the-art Zoll E Series heart monitors. This purchase which the City of Dublin has budgeted for
fiscal 2007-2008 will result in a significant improvement in providing advanced life support and cardiac
care to all citizens served by the ACFD. Every fire engine and truck in the City of Dublin would carry a
Zoll E series heart monitor.
The Zoll E Series heart monitors will enable the department to comply with current Alameda County EMS
protocols, as well as the 2005 American Heart Association (AHA) Guidelines. With advanced cardiac
monitoring and diagnostic capabilities, it will enable our paramedics to rapidly identify those patients who
are suffering an acute heart attack and insure they are transported to the closest county-designated facility
with a cardiac catheterization lab. In addition to monitoring heart rate and rhythm, the E Series will
monitor blood pressure and pulse oximetry, as well as continuous exhaled C02 for patients who require
insertion of a breathing tube.
With the 2005 AHA Guidelines emphasizing the need for improved CPR for all cardiac arrest patients,
ZOLL's patented Real CPR Help technology, offered standard in the E Series, will enable paramedics to
continuously monitor the rate and depth of chest compressions, both at the scene and during transport to
the hospital. ZOLL CodeReview Software will also provide data capture of the event for QA analysis.
Additionally, the Alameda County Fire Department is participating in a promising trial study utilizing the
Res-Q-POD. The device, technically called an impedance threshold device, works by preventing
unnecessary air from entering the chest during CPR. Studies show that when air is prevented from going
into the lungs during the recoil phase of a chest compression, a greater vacuum is created in the thorax.
This negative pressure draws more blood into the heart, doubling the blood flow during CPR. The benefits
of this circulatory enhancer (recommended by the AHA) include increased blood flow to the heart and
brain during assisted ventilations, and increased opportunity for survival after acute cardiac arrest. The
Res-Q-POD works with all standard resuscitation equipment currently in use by the Alameda County Fire
Department. About the size of golf ball, this device serves as a strong compliment to existing medical gear
and treatment protocols.
During ACFD's negotiations with Zoll Medical regarding the purchase of 32 E Series heart monitors;
Staff secured a commitment for a credit of $20,000 trade-in value for Dublin's two existing AutoPulse
units.
RECOMMENDATION:
Receive the Staff Report on 1) AutoPulse and remove the AutoPulse from use in the City of Dublin.
2) Direct Staff to pursue credit for the two Dublin AutoPulse units as part of purchase agreement for Zoll
E series heart monitors. The full amount of the credit for the two units that were donated by the
Dublin/San Ramon Women's Club would be allocated exclusively for the City of Dublin's portion of the
Zoll E series heart monitors.
FROM :ACFD
FAX NO. :5106183445
Sep. 10 2007 04:00PM P5
Michael D. King, Director
James E. POInter, Medical Director
ALAMEDA COUNTY HEALTH CARE SERVICES AGENCY
PUBLIC HEALTH DEPARTMENT
EMERGENCY MEDICAL SERVICES AGENCY
1000 San Leandro Blvd. Suite 100 · San Leandro, CA 94577
(510) 618-2050 · Fax: (510) 618-2099
MMI!~((l~1I1Y
MEMORANDUM
DATE:
June 14,2006
TO:
All ALS Providers
FROM:
James E. Pointer, MJ)
cc:
Michael King, EMS Director
~lJB.JRCT:
Auto Pulse
I would like It) updatc you on the interaction between Zoll COl'}Xlrati(ln and the Stllte Elnergl:lney Medical Sc:rviccs Authority
regarding the Auto Pulse CPR device. According to Doctor Cesar Aresteiguicta, Zoll has voluntarily withdrawn sales ill the
California market To honor this commitment, I would a.,k that mme of yo II purehl1.'1c or acquire further ALlto Pulse devices
Wltilfurther notice.
The company has agrecd to ~m "Auto Pulse Registry" in which all CPR cases involving the Auto Pulse in C,l.Iifornia will he
tracked to insure safety, While this is not a. randomi7..cd c1inicaltri~ll. atlcast wc will have a repositol')' 01'811 Auto Pulse uses.
It will be a requirement for all providers in California to participate in the regislry.
There is llls(l a randomi:.red clinical trial that is currently being organized among a number of EMS center ill the United States.
Probably, no definitive decisions abuut tho AulO Pulse will be made until the results of that trial are known,
1 continue to have concerns regarding the safety and efficacy of this device. , appreciate YOUT cooperation with the EMS
Authority in insuring safety fur our cardiac HITest pati'lnls in Alameda CULlnly. please oontinuc:w submit Auto Pulse cases to
Alameda County eMS until tho state registry is in operation.
ATTACHMENT 1
FROM :ACFD
FAX NO. :5106183445
Sep. 10 2007 04:00PM P6
Michael D. King, Director
James E. Pointer. Medical Director
ALAMEDA COUNTY HEALTH CARE SERVICES AGENCY
PUBLIC HEALTH DEPARTMENT
EMERGENCY MEDICAL SERVICES AGENCY
1000 San Leandro Blvd. Suite 200 · San Leandro. CA 94577
(510) 618-2050 · Fax: (510) 618-2099
July 12. 2007
Chief Sheldon Gilben
Alameda County Fire Department
835 E. 14th Street
San Leandro, CA 94571
Dear Chief Gilbert:
You asked me to briefly comment on the AutoPulse and related technologies concerning the administration of
cardio pulmonary resuscitation (CPR).
The AutQPulse device, while it is FDA approVed. never under went rigorous testing prior to its introduction to the
market. In the last several years, three studies have been performed that have looked at, in some fashion, the
performance or the outcome and safety of the Auto~ulse. One of the studies was discontinued early because of
worst cardiac arrest survival for the device as compared with conventional CPR. Another study did show a
superiority of the AutoPulse, but the methodology was flawed In that some patients alSo received hypothermia
therapy that has been shown in other studies to improve survival. The third large study, showed no difference,
baSically, between the Auto~uls8 and conventional CPR. Perhaps, more important is the question of safety with
the device. Autopsy findings in three or more cases have .shown unexpected thoracic vertebral and posterior rlb
fractures directly attributable to the machine. For all of these reasons, the state of California Emergency Medical
Services Authority recommended discontinuation of this device. Even though now the device Is "legal" in
california, participation in a data registry Is requIred.
A related technology that shows tremendous promise is the Res-Q-POD. This respiratory impedance device
througn a decrease in intra thoracic pressure and subsequent improvement in coronary and neurologic perfUSion
has been shown to improve survival in cardiac arrest. We at the EMS agency are very pleased that Alameda
County Fire is one of the Initial test sites and a "center of excellence" for a demonstration project, approved by the
EMS commission, involving the Res-Q-POD.
I hope this information has been helpful; please contact me if I can be of further assistance.
Sincerely,
':JL r~:I:.
James E. Pointer, MD
Medical Director
Cc; Michael King, EMS, Director
Dale Fanning, EMS, Acting Asst. Director
ATTACHMENT 2
FROM :ACFD
_ ORIGiNAL ffiNTllffiU'nON
FAX NO. :5106183445
Sep. 10 2007 04:01PM P8
Manual Chest Compression vs Use of an
Automated Chest Compression Device .
During Resuscitation Following
Out-ot-Hospital Cardiac Arrest
A Random.ized Trial
AI Hallstrom. PhD
Thomas 0. Uoa. M 1>, MPH
N" .._..._.
Mich..~ellt ~ayro, MD
James Clll'irtenson. MD
-" ....
Amly R. ~~~~~ MI>. ...
Vince N. Mosesso, JJ', M.IJ
..,..'......, ''''_''M
T.UiH Van OUingham, BSl':l_.....
Micllt'lo Oll\ul'kll, RN
Sarah POrlningtulI, !iN
.L~;IUl}' White, MS
Slophon Vahn, F.MT:~.
JanlllH Husar, RMT-I'
Maloy 1"~ Mon'is
UlC)flarll 1\. Cobb, MD
O 1JT-O~.-H(lSl'lTAL <':i\RDIAC
atrt:st claims h\\ndred:'l of
thOU:'liUld:<o (,r li,,~ llnnu-
ally in NMlh Mnl~{ica. Suc-
cessful resuscitation depends (,II H t:\l-
ordinatcd set of actions including early
cardiopulmonary resU,:;dtl.llinn (CPR).
I1igh-quality Cl'R may he. itllpMtllll1 ror
both cardiac and brain re.:'IU:'Icila-
don. T.' In animal invl'-,r;rigalifJl1~, fewer
intr.rnlptions of CPR before and after
dcJihrillaLion have improved cardiac
and neurological o\\tcomc:.'1,.7 Thl~ ur-
der of resuscitation interventions may
also h(~ in1portant, eg, s\\rviva1ll1ay he
improved byperformin~ CPR by lmer-
gel1cynledical se.rvices (P.MS) pcn;on-
nel prior to defihri1l~ljC)tI..9.?
See also pp ;'62' &lid 2661.
"." ,-
U20 lAMA, J~nc ! 1. ~O()(. Vul 2')'1, Nn. H (R~l'rintcd)
Conted High-quality cardiopulmonary resuscitation (CPR) may improve both car-
diac and brain resuscitation following cardiac arrest. Compared with manual chest com-
preSSion, 8.n automat~ load-distributing band (lOB) chest comprenion d~vite pro-
duces greater blOQd now to vi1-a1 organs and may improve resuscitation outcomes,
Objective To compare resuscitation outcomes following out-of-hospital cardiac; ar.
rest when an aulomated lDB-CPR device was added to standard p.mergency medical
services (EMS) care with manual CPR.
D8$"n, SettIng, and Patients Multicenter, randomized trial of par.lcnts E!X~ri-
enclng out-oI-hospltal cardiac arrest in the United States and Canada. The a priori pri-
mary population was patients with c:ardinc arrest that was presumed to be of cardiac
origin and that had occurred prior to the a.rrival of EMS pCl'!lonnel. Initial study enroll-
ment varied by site, ranging from late July to mid November 2004; all sites halted study
enrollment on Mar~h :11, ~005.
Intervelltlon Standard F.MS care for cardiac arrest with an LOB-CPR device (n =554)
or manual CPR (n::::517).
Main Outcome MeuunI. .The primary end point was survival to 4 hours after the
911 call. Sp.condary end points were survival to hospital discharge and neurological
status among survivors.
Results Following the first planned interim monitoring conducted by an independenl
data and safety monitoring board, study enrollment was terminated. No difference ex-
isted in the primary end point of survival t~ 4 hours between the manual CPR group
and the LDB.CPR group overall (N:-:1071; 29.5% vs 28.5%; P=.74) or among-the pri-
mary study population (n ::-767; 24.7% V5 26.4 %, respectively; p.. .62). However. among
the primary population. survival to hospital discharge was 9.9% in the manual CPR group
ilnd 5.8% in the LDB..CPR group (P= .06, adjusted for covarlates and clustering). A CP..
rebral performance category of 1 or 2 at hospital discharge was recorded In 7.5% of
pcltients in the milnual CPR group and in 3.1 % of the LDB-CPR group (P.. .006).
Conclusions Use of an aut.omated LOB-CPR device as implemented in this study WtU
associated with worse neurological outcomes and a trend toward worse survival than
manual CPR. Device design or implementation strategil'lS ~uire further evaluation.
Trial R.statratlon c:Iinicaltrials,gov Identifier: NCToo120965
lAMA. 2006:2!i5:262O-262B www.jo...U....1
AuIhot Afflllallons: Dep~rtmcnh ot BiDttatistics (Dr
Hallstrom and Mss Van Ottinl:ham and Mor~5} lInd
M_dilil1C! (0'5 RtIIl and Cobb and Ms Olsufka). Uni-
versity of Washington. Seattll!: Department of Forner.
ge'I~Y Merllclllc. Ohio State University. Columbus (Dr
Sayre and Ms V1Ihlte): Bl1tish Columbla Ambulttlll'l' Ser'
vice. Vancouver (Dr Christenson): Calgary I~"'er'
ganey MIldIOlI ~..tvlr:fl.. C':1,l1,.ary. IIIb<:rtn (Dr Anton
and Mr nhll); D~~.,.hT'ftnt "f Frnergcncy Medldrnl
(PI M01CS10 and Mr Husar). Schoo! at Medicine, uni-
versity of Pitbbur{:h. PitbburP1. Po; and St Paul's Hos-
pit;~. Vancouver. al1tish <::olumbta eMs !>eollinl,:!(.n).
Correspondi"fl Auth'l,; III Hallstrom. PhD. Depart-
rnmt of 1I10Jt.1tistrcs, University of Wamlnston, , '107
NF. 4';tn St. Suite 505. Seattle, WA 981015 (~phtflu
.washln&Luh.ft,lu).
@inoD AmP.lil:a1l MtlCli..al Assod.tlon. All ~hbl """,,",~d.
Duwnlouded fr\llIl www.julTlll.comon September 10,2007
ATTACHMENT 3
FROM :ACFD
Ob:ll:rvlltion~ 0 r re~H~lle per.~onnel
indkate that nlElintainlng consistent
comprCSS\OM is il difficult task.10 In the
lahorat.MY, trained paramedics pco-
vlde shallower and slowCl. COll:),pr('.$-
slons over time Without KWlking.lI,12
Chest compressions ortl~O. dQ oot
ac:hieve ~iddim: recommendations
wilh rl~.g~rd to depth, r.m:, and hands-
off time.13.1.'
Thl~ llt-sitl: 10 provide nptimal chest
(,:omprc~~..'iinn:,; led 1:l~ t.he development
l~f autnlnated mechanical chtst com.
pression devices. The AutoPulsc
Re~ll:;l:illltion Syiltl~m (ZOLl Circula-
tjnn, Sunnyvale, Calif) is a lllad-
distributing band (LDn) d'r(:ul'Ylf\~rl~lI-
tial chest compression dcvi(:l~ with IIn
electricallyal'.tuatcd c:omil rkl i n~ hand
00 ill:ihorl bll.c:khl~ard and has been ap-
proved by the US Food and Drug Ad-
minl!lltrat,ICill for use in am:,mpted re-
suscitation of cllrdif.\(,. /.Irres!. Tn pig
models and in-hOb-pital cardiac arI'CSI in
hl\manS, this LDB-CPR dl'.vkl~ pro-
duces grE:all:r blood now to the heart
and brain than manual C:PR hy trained
individuals or dle automated mechani-
cal piston CPR devicc.,~,'b Ani.nl,al ill-
vestigatiol1 has demOtlSUlllCd II grclllcr
likelihood of neurologically intact sm-
vival in prolonged vcntrkular fibrHla-
ti on carolac arrest with LDB-CPR. I 'f
In t.hill ~tudy. the Autol'\.\lse As-
sisted 1>rebm;pitallnl~rn!l1 ill till. I Rl"_sLls-
FAX NO. :5106183445
Sep. 10 2007 04:02PM P9
COMPARISON OJ! MANUAL CPR WITH COMPRf:SSION DEVICe
cilatim1 (ASI'IRli) trial, we compatcd
LOB-CPR wit.h mllm\al CPR lluring out-
of-hospital cardiac arrest, W(~ hYPolh-
l"_~ized that 4-hotu" survivil.1 would he
greater among paUcnts randomized to
LOB..CPR compared with tholle ran-
domized to manual CI'R. Secondary
out.c:omes were 5urvivllllo hospital. d is-
dlil.r~l': and nellTologiclll fl\n\;tion al
hospital discharge.
METHODS
Study Design
Th~ st.udy was COndl\ctl-d hi Calgary,
Alberla; Colmnbus, Ohio: !'luhurbll of
I'ltl!'lhurgh, Pa; Seattle, Wash; and Van-
couver, Blitish Col.ull,hia. Br.C::!l.lJSI~ or
ditlen:nct:S in lhl~ liMe c:our!llt. of t.t.h-
ICS rmew and approval amI EMS lnlin-
ing sl:h(~dules, inilh\l study enroll.,
ment varied by site, ran.g;ng from Illll~
July to mid November 2001. (ThULl: 1).
All sites halted study emollmellt on
March 31, 2005.
Cost and the im:l1tlVt':t1ience of c.ar-
tying the l.Dn-CP'R (h~vke t.o epis(,deO'l
at which it would u.ot hl~ a~~ignl\d dk-
tall'.J I.h(\ UO'le of duster randomiZation
with crossover, Clusters w\~n~ bllsed Or)
Iln F.MS iltation or group of Slalions and
crOS$ovr.T OCCUlTed at specified time in,,,
I.t':TVals (4 weeks to 2 months), The du:r
lcri,ng unit. wa!l based on a combina~
tion ~~f F.MS !:y!:tl".m operaUonal and
design considerations ll.l avnid 8T1;val
of hot,h reO'lponse teams assigned 10
manual CPR al1d to automated me-
chanical L!)~-CPR. Withi.n a givt~ sill'.,
half of the ch.\sters were Tllfldornized to
the control (manual CPR) and halfto
the intervention (T .DB-CrR) wuh sub.
sequent aLtemation bl':twl':en the inter..
vent.ioll and the control. Thl~ rolation
1.lt':1i(~d was chosen partly for conve-
nience of device transfer. tlll~ntil~rl~ I.l~m-
poral balam:\~ c.luring t,he trial, and to
avoid cycliCal phl!nomena.18
The .~UJdy wa!l C011ducted under the
rl~gulat.ion~ for emergency exception
from informed consent. which Tl':-
quirt': t.ac:h uS !lite to inform its com-
munity of the proposed trial and sCl~k
Its opinion.IQ,10 N otifieaLioll of indi-
viduals enmlled in the stt\dyis also rc-
quin:d, and the. c.llmmunily ITIlisl he in-
rlmnr.d of the final results. At each site,
it,duding lhl'\ coordinatin~ center, the
primary research review boar!') re!lpon-
stble for guidance, review, and OVl~~-
sight of human subjec,ts' prolct:lillll at)-
proved the study. In addition. rcsel\n:h
review boards at most receivi~ hospi.-
tals also nwil~w<,:d Il.lld approved the
smdy 1.0 alh~w smdy personnel to pro-
vl.de. limely llotifica.r.lon to the patient
or family.
Tlll~ study also convened an inde..
pendent data and safety monitoring
board consisting of :i paramedic, an
EMS phy~ician, a biostatistician, and a
!!ble 1',Clusters and Enroll.rnc::nt ~ Site::
HiKtorioalellrvival rate.
NoJlo181 (%)
VAIltriculsrfibrilletiol1 ~/25 (16) tV4::! (14.::>) :'\I/I~I (31) NA
All f.lllrliclpwlll; 5/11 5 (4:3J. NA . _ .... 40/299 (13) 53/852 (6.2)
Nu. of $lllCJy vehIC!~:1 . :'\4 AI_S 213 AL$ 4'1 BLS 43 BLS
No. of clll~e~. 311 VehicleS 1 Ao.~ole.$. ? AI'llIlH II Arll6ll
Cl1l8ter/rotatioll. matl!l (SD) . 1.0 (1 , 'I) $.0 (? ,0) ,,?:~,~1 (12.3) 1\;1.2 (9.6)
Hotati?~!"!~~1 1 mo 1 rTlQ 2 fY\(l l;'I I'll<
Stllrt dill'll (1111 in 2001) July 22 October 28 Sl3ptl;mb;!r g August 17
^hl~r~i:lliurl~: Al.S, adv~"~M i~-~'upport; "i;l.;, i:t~~ Ifa support: EMS, em(l'Aeney m~.nI !~wviuo!:: NA. nut rJvnilublo. .. .
.In 1,1 wluJ!,l.llllr ('Y~I,,", """"Il'X")\f 'YK'K'Jii.ill..,t..,,;,IIWl and pal'smsdlce typIctJly 818ff the 3M'" r'.l!)p';rIl1in!l' vel'iole end "'''I''''' 11/1 ,dl "''''''~I'''''''Y 11'1,.1.1.181 dillp8tCl_. In Q :.!-tIQr
~'?"'l' tho firl'llil1f' (Ill S) i.. r:rlmrx",ml <>1 ,"'..rg'.II:Y medlc:llled'lnlcllll18 who respond to alllJlT1~ncy rT1Ild~ dl~p<d,,'x"i. TIm """"nd Ii", (M.S) it. """,plJeed 01 pl1t'M'ledlca
",d I. ~'."ry",J Fe. "X,,,, ...iOUr. "u"diUu"" ;rdJdin~ CIIll'dIaO arreat. I yplcaliy ElLS p<;Ir~onnrll arriYfI nllll nt 1110 'il:mll ..III."J~h ,fir~..lc~l i" "n.., S;nll.lll8l...0U8.
tAl ,II'lY Iimr,l, holf tlllfllll;l~"ly Vnt\i"t'lf. _ 1ll;!IlyrllJil to l()Bd-dI81r1but1l'lg IJMd osrdloplllmol'll"Y rr"~J\"U1\~iQrl ')Qf"pr~l!''''('1'l ~1'Ic11ht\ ..,Itlill' h$11 1<. r\'\lll'\Ua1 oompreaelon.
*1JIti1"''''~''1 fIK""( ill .""lr.,,,I;II fillrUlall"" ,
A
... ,_!n~~14)
FMS syRI9rn
TIer'"
..,
B c 0 e T01aI
1.n.111) (n . ;101) (n .. 348) (n.g~ (N = 1071)
2 ~ ;:> + fire :2
-..,...------
Participant Enrollment by Site
~'100fi Amr.rlt:illl Mf>I!it:a1 A.....odalioll. All rlptl l~,
f)(>wI\I,~adcd from wlvW.j31l\8.ocm l.'I1 September 10. '2001
39/1~~ (27,m
N(\.r>l
8S ALa
4 VOhiclo IJr:?uPlngs
7,6 (6,1)
llwk
NovWlTlbQr 1 7
167
5\ ,
4,3 (~.~L__
(R\!prin1ild) JAMA, Jun. 14, :WU~-vo\ 2~~, No. 12 :1(;:111
FROM :ACFD
FAX NO. :5106183445
(;OMl'AR'ISON OF MANUAL CPR WlT~ c:OMl>RESSION DEVICE
Fi....... Flow of Participants in Trial
--'--'-'-
C=_ r14Ir,dOlrlZaUoo. _~_:>
,..----~ - ..... , - ----...----......
I Loe-cPR
1__ 70<1 Total oa~*
ManYa! CI'R
E')fli '~'nICIa=lt':!.
156I.1_f.XCI,1(lflt'jT
~~I\!l(ld~IRy
:~ tJrl!;IQr"t('Ir or Wnrd of SI"~lu
20 OIl NLoll'lesuaaltale ()O:1er
40 Or.nd nn I\rrivoIIMU,
(It'FlOnly
51 Trull.
lHlill'1~s;,"1l'''Y
Z5 NO:llUCtY VIInlCln or
r>er_M1a\!:l(>;!lli'
517 ~O_
H~ Nol'lllrlmaIY (~
57 Cildlac Amlat Allar EMI:i
I'nlval
12 Nc"..lIlIJiIII; I::liuluw
I~. .I\l:1V\ln<:'MI w.,~J"pn"
~l . HMn", !;h.ly
M\(ll...,,,,, !IrrM!ri
373 I'rlmaIV ()(lmparfe"n Eplaoon
1",;Iutltttllll r;;,',8lY AnOllyeia
"l
150 c- ~.nl,l<inrlt
;<() ^aO<J-.1Sy
n f'n>u"." Ul Wotu <II Slole
20 Do Not 1;(l\Il,,,QlhiltG (.>r<:IQr
40 Dead on f'/rMII WIth
("'JlI1 01'1\'
~8 'Tt8(lmA
~ RcxoulIl SUII;P'I'V
:>7 Nl? illlldy Vr.I1'o'" or
f"""",,nn'~ Il!Sr""'"
C'~(~1'1I9~(.~~'~:'-_.]
I (~) Nnnprlnu"v c......
~q1 (wilnr: I\m~f ^ttrr f:MA
An1v(\l
06 Nunc~rdlac ElIOIo(1Y
III lIdvoncod LWv 3\~)port
:-uO ~ Dwfurw Sludv
^,nIlululloo ^rrlvu<l
:.ul4 P'irl:"uy CW''lnsriKul. EJ,IiI!lULl&i
Irn','c:fl)d In Prln"IIY ^'''IIr.;l1i
cr'R indicates cardiopulmonary rVlUlcitation; EMS, AmllrAenr.y medlcal5ervlces; LOB, load"di'itril1u~n!l bi\lld,
'Cost and the Inconvenience of canying the LOB.CPR device ,,, ..pl~ude~ at which It would not.,. assigned
dictated the use of clu5ter randonllza~on with crossov~. Within ~ givll'l 'iiI.., hftll "r the dusters were I'$n-
domizCKl to thc.t:Untrol (marllJ~1 CPR) $11<1 "~II!n lhe Intervention (LOB,CPR) with SUblC\,ucnt 1I1lcrniltion b<1.
tween the Intervention and the control,
tSome pationt5 had marc thOl" 1 llllclu51()n.
dinkal inv~_~1igfoll(1r. A.q prt>_qpE:cifled in
Ih(~ .~Iudy rrul.(~col, the d~\la .md sk\fety
monitoring board was to revi.t:w safety
and interim progress when approxi-
mately one thi.rll and IWlIlhitds or pa-
tirnt~ had bl~cn clltf,ned.
Population
Adults with out-uf-ho!iphal cllrdillC ar-
rest who rc(:civc.d allempf,~d rt>.qu~clta,'
tiun bya paflld1"llllt\~EMS agency were
enrolled unless an exclusion (:riterioTl
waS pr~Ctlt (FIGlHtP.). ratienl~ treated
by EMS li\1bscqllcmly defl':rminl':d to
meet exclus10n crllcl'ia were exchlded
froln t.he analysi.....
A primary compal'ison populatiol1.
patients who were in cardilw arrest at
the time ofr.MS arrivalllnd wh\I.~(~ car-
diac llTTl"_ql wa~ considered to be of car-
diac origin, was chosen a priori as the
popt\lation most likely to benefit hom
(~he~t. compre~sion~. Cardiac etiology
Wll!l det('nnined hy t.he liite ~tudy co-
O1'dinat()r ()l' investigator based on the
26U JAMA.J"n.. 14. :!()Oti v"l ~Y" Nu, n (R*prlnted)
EMS report forms and h()~ital rec:ords,
Early in. tbe cnl'IJlIml~ll, !ltudy adhet-
(~m:(': (application of the LDB-CPR de-
vice ba.1ed on otlt~of-hospital report)
Wa.1 vel'}' low at site J) w.lr.n thl~ ad-
vanced life s\lpporl lInit arrived he-
fore the study unit. Arrival 01' an ad-
vanccU me support unit 90 seconds 01'
longer before the ~tudy Imit waf; addl':cl
as a sitc-!l-pcdfk (~.lCclU$ion from the: pli-
lllRry comparison populsUnn.
study Protocol and Intervention
the device used in Ihi~ ~tt\dy h, ap-
proved by the US Pood and Drug Ad-
Il:llnistrQ tion and i~ the size of a ~a1f
backhoard, weighs 15.8 kg, and opel'..
at(',!l for lon~er thall 60 minutes on
hattery power. To \lSe the device, the
patient's uppet Imdy clothing ill re-
tlluvl'.d, Ilnd the.l)atient is placed su"
pinl~ i)tll1,,-. hl1ckhosrd. An8~inch wide
lOB anchored behind the hackboard i~
wrapped dtclW1fctenttally arOlllld rhl':
patien.t's cht'.Sl arId is dOlll':d anfl':riOl'ly
Sep. 10 2007 04:02PM P10
with Velcro. The device automatically
adjullt~ the length of the belt so that il
tlts snugly across the chcst, Dcvlcl~-
reguhlted, re.petitive shortening of the
belt :>lluee7.~ lhe. thorllt~ic 1~t1Vity. ~ell"
l~r~tinl=: arlt~rlll'l t:i rculllUon. Belt length-
ening 'enahle.~ passive decompression
of the che~f" 'fhe device provides
compressions at a rate of 80/min (~un-
figured 91115 compressions with a 3-S1,.1;-
ond pause for ventilations or continu-
()Il!: che.st c01npressionl: without any
w,t\lilslory p:l.lI!le.
DurinK a l'un-in period ranging
from 0.7 to 2.8 months, EMS person-
nrJ. integrated the automated device
in In nUI-Or-hnspital care, Initial train-
ing of t:::MS penonncl jncl\ldl~d
hands-on .~kill practice l.\sing thc
cl~vlt.('. wilh II. mannequin and a video
prellentation with rationale for tht~
LDB~CPR device. Refreshcr training
waS not specified by design to best
rCI,IJt:atc rl\lIl-wo1'1d cllndttions and
W~\ll highly variable d\lTi1tg the study.
The protocol allowed 3 vpllvn.~ rm
Ih(~ resusclL:nlon intervention. ini-
tially all Slll'S eholle ol)tion 1, a "quick
lvvk (<6 seconds)" rhythm recordinll
followed by circulatory dfort with
eithl':T thl': I.DI:l-CPR device (:1.'1 soo~ as
relldy)'ot manual compressiolls hymll-
dllmi~ed assigl1m(~Il1. After approxi-
JJ'1atcly 2 mitlulcl1 ()f 2()() manual COIn"
rrj~~~iMII1, ,1 rhythm assessment was
perfonned.
Opti.on :2 was immediate CPR with
manual compressions regardlcSll tlf ran-
domitation until the tlrstshock assess-
ment. Site C, the only EMS with a com-
prehensive quality..improvenlent effort
to reduce paUSl~!; to (:h.c~t ~~(,Impn~);-
sions, changed its resusdtation iptt.~r-
ventioll to o11t.ioll two 110 days after
staning the study. The changC'. was
implemented after quality-improve-
nunt re,view identified prolonged I i rnt~
withOUI l:Otllpt(':ssion::; while deploy-
ing lhc LOB-CPR ()(~vl(:~.
Option 3 allowed analysis, and
shock if apprt1prlalc, bdQr(~ hl~glt\Uing
CPR. In aU cases, afte.1" rhythm aSSf.'.\iS-
mt>.nt. and shock if indicated, addi-
tional nect>_qllary compressions were to
hE'. l)erformed manually or with the
11;);2.0(16 I\m~I"..n M"t1i"..1 A....."'latlo... All rlpts ~~
n(>Wl)lc)a(lcd 'a'nm www.jalna,colll on September 10, 2007
FROM :ACFD
LDB-CPR de,vice according to ran-
domization. In all othe.r aspects, sitet'
followed (heir st.andard res\\scitatiott
prl)t.ocol until the patient was dec1an~d
dead or regained 5tablt~ Srlll1llUlelllls
circ\llation IInd WllS t,ram:ported to
and arrived allhe emergency depart'
men.t.
End Points
The primk\ry end point was ddincd 1:\9
survival with spontan~~oll$ drt~u]ation
... hou ~ aflt~r tI,e 911 call. This mea,
!'lure avoids inherent incoru;il;tencie5 In
site..ta..sUe variatioIlb in the. ddhlition
of Mtldmiltance to the hospital. n Scc-
ondary cnd points included di..~charge
from the hl)spital imd ce.re.bral perlor-
m.anCl~ t:lItl~gory !'lC:ore at discharge from
111l~ ll()spi lill that wa.c; obtained tram the
hospitall'ecords. ~1
Data Collection
1)ata were collected. from ro.MS (('.-
ports, defibrillator r~~wrd in,;(.~. !l sl udy
questionnaire, and hospi,l~l' rl~CMtis.l7.
St\\dypers011l1d reviewed the dcfibl"il-
lator's digital electrocardio~rll,r.llli(: re-
cording when availablc; otll~rwise de.t~-
trocardiographic paper strIps Wl~re
revitwed. In addition, lHI c)('pl~rlct\CI~d
arrhythmia resean~h nurse al Ihe: QQtll
cootdinttti.o,g (:\~ntl,".r reviewed initial
del~' rC>I~l\rd IClgraphic records. If that re.'
view resulted in a discrcpanry, the prin-
dpAI \nvl:SII~alnr,;, masked to treat.
men t. a...!lignment, categorized the initial
rhythm. Oat-II collected from the hos-
pital record',were primarily \lsed to
monitor for adverse effects of l:hl~SI.
compression and for ll$l~e)'tai.ning end
points.
Samp'. Size
For sample size, calculation, p[uicnt.~
were assumed to be indepcndl:nt, al-
though the troRiOVt~r wilhin d'Ul';tp.r'llo...
tentiallycollld he more eftlcient.2J The
1.-houl.' survival rat.e wa!'; assumed to be
the avcragc of ,Ill". bll,;e!ine admit.t.ance
and discharge survival raIl's. Based on
available data from participating siles,
we estimllted !';urvival to 4 hours (pri-
mary outcome) in the mamlal CPR
group to be 17.8%. Dasl,,~l on reported
FAX NO. :5106183445
Sep. 10 2007 04:03PM P11
C.OMPARISON OF MANUAI_ CI'Il wlTtr COMPRESSION DF.VICF.
improvements in restoration of spon,
IIUIMUS circulation ra'c:~ from II !<Omall
llhservatioiw study I)f ,he device,' a :~.5lj{,
rel:ltive improvt:"I\~nt was hypoth-
esized (it:, an inl,ervention group pri-
mal:Y llUII:ome rate of 24,O%).1~ Th\ls.
,111.'. $tudy required 1837 pMlilmts illlhe
primary COmpB11$on ~roup to :tchicve
II power 0(90% using a 2-sidt:d test with
a level of ,05 (ba:ll.'ll on sequential moni-
toring, 2 interim looks, .\Uclan O'Brien.,
Fleming type hllundsry). ':; The design
inl~luded the possibilityofinl.T.t.'.llf>lng tbl:
!lample size based on (hl~ oh!lel'Ve.d out,-
conles to that Pl1int,lO
$tatistlcal An.lysls
Compllri!;(l1\!l were made by intelltion-
to-ll'~lIt a!l!';ignment. Logistic regres-
sion wa!; applied using ge,neral1zed Lln-
ear mixed models wll.h the l't)hust
:;anuwkh p_c;tlmator of the variance to
compare the outconle oflndividual epi..
!;ode.~ between the 2 study groupS.27'JO
Models were adj\lsll~l for cowriat:p_" ]lre-
viously demonstraled to predict sur-
vivap1Jl at; well as duster (II sou~ce uf
nonindependmcl:) .
A single a pril)ri subgroUI) analysi!l tll"
the prin18ry population Willi !lpedfied
based on initial rhythrn (11.~Y"t,o]e, vell-
tlicular fibrillationlV~~lllricular tachy-
cardia, p u l.sdr.s.'1 e:leCl.rical activity). In
6.1% (471767) uf participant'l, electro-
cardiographic rhYlhm wall not avaH-
able an.d thl~ IIIJlllltllllt~d external defi-
brillator dic:ll\Ot advi5~ to shock. TheSE:
were assumed tl") he a~ystole or pulse"
le~ dectricl:\l activity. Three of the 47
cases were assigned thl: rllylhm ob-
served at the next clcct'rtlC~I'(liographi(:
analysis. In the remaining -14 cases, thc
initial rhythm was imputed ha~ed 1m [lIe-
tots that dil;criminall.-d signU1C'..I1ntly be-
tweell patienu with initial rhythm I)f
pul~eless electrical activity an.d llbj'Stole.
1'05t hOI: s\1bgrnup analyses evalu-
ated whcthl.T the itll1~,l'Vellf.ion effect dif-
tt.Ted by she or by thl'. time since the sile
began enrullihg patients. lnteractillllS
were te:>tell u:>ing llll t,mel'action tl!'.rm
between treatment group and the '~IJ-
variate of int.f'.rellt..
Amllyses were l~ondt\cted ming SPSS
ver!iion12.0 (SPSS, Chkago, 111) and R
~}O(l6 ^mClnCiOn Mr.dit.lll A~"<)Clatloll. All r1pts 1'C:ilCI'Vl:d.
nOwllll\adcd from www.jama.com (.Ill :Seplember 10. ZOO?
version 2.3 (R ....oundat.ion for Statisti-
cal Computing;' statislit.....ll 50I"tWSl"f'.. Un-
Less explicitLy stt.\led, P values are un,,,
lIdju:;I\'.d fM covsriates or clustering. Por
thl~ r~imary' and secondary end points.
p values we.re generally atijustl:tl; hy
protocol, the cc levc1 for the }lrimary end ,
point was !let at .05.
R!5UL TS
The data ~nd r;l(fel.Y !n\")nitoring board
mc,t on. Mal'l~h 11,200.5, and again on
March 2R, 2005, to review .the rl:sulu;
for 757 patients en rlllll~d Ihr\)ugh.Janu.,
ary 31, 2005, and recllll\lnended S\lS-
pension of enroUmen l unl it dillS for the
:,~ 141,stie.llts enrolled d\uing Pebrllltr.y
atlll March cl)uld be eval\\ated. Results
promptp.d additional data eoHee,tion.
indlldin~ estimates of chest compres-
sion duration in the first' minul.es of
the relSu$t~llalioll effort, drugs admin-
il;tered prior to the patient arriving at
the hospital, mode of in-hoopltal death,
and other details indiClttil1g lutl~, helln,
or cerebral damage. OnJu"c 27, 2()();,
the steering committee reviewed thcs('
expallded data and recomm(~nd(~d lhal
the trial be haltcd.
There were: 51 dusters, and tilt. ilV-
erage number of episodes pl~ du~tl:r flt".r
i'olation interval ranged from.l.8Iu 2'5
CTable: 1). The nl.Hnber vr Jllllient epi-
sodes enrolled at site:> varied from 120
to 391. Thel'e were a total of 1377 epi-
sodes, of which 373 in. the manual r.PR
group and 394 j,n ,I'll: 1.0B-CPR ~l'OUp
were eligible fo~ stuuy enrlllllllenl.
(Pigur.c).
Demographic features, cardiac ar-
l~_~r, circumsT~nCE:S, and treannCilt char-
IKlcrlSllcs Were jtenel'ally similar be-
tween the treatment groups. Among
primary cases, patients in the: LDn-
CPR group were more likely to re-
l:dVt'. epinr.phr'i11e (P= .0:,) aild bave
longer time intervals to fir.;t shock (fc It
patients found in vcntrir.\lblr fthrillationl
veIltricular lachycat'llla) (P=.OO1). tet.-
mination of ret'us(,',i tatlV('.I'.ITllrl (p =.() I),
and hospital transport (P=.O I)
(TABU 2). In the LDB-CPR grot1p.
th~~ dcvil:c waS spplif'.d durinjt the Te"
suscitation to 83.8<~, nf the primary
l~l\."e.", 73.,'5% ofllancardiac cause cascs,
(R..,.nrotlld) JAMA. JUII~ 14. lOO6-Vo! 2.95. No, i.l 2.21
FROM :ACFD
Sep. 10 2007 04:03PM P12
FAX NO. :5106183445
C.OMPARISON OF MANUAl. (]>R WITH COMPRHSSION DIiVlCE
!.~I* 1. ,Epi5ode Characteristics Ilr Primary Case Status and. T\'Qatrn.~nt Group.
Nonprl~ry c-
r--'
Manual CPR
(n ..144) ...._._
I'rlmary 0a8e
r--' .
Manu" CPR
(n . 373)
LD-..CPIIl
(n = 394)
LDB.oCPR
(n = 160)
!'\tRillS
l!nwltnllR~ ?~.urlkrlown
Witl1lllllled by bySlandfJr
... Wltnllfl8ed by i:MS .....
Cpn ~rf(lrTTlfl(:i by byatander _
f't Ihllo 10(,Ation
Age. mew! (SO), y_...
Mlln
"'ody typet
rhln
NOlmoJ
01>>'..0
Morbidly obe...'\e
Nul r~porlccl*
Rhylrltn
VFII'II1aellllW VT
~~.~!I11I~.\~c\l c\CUvlly
AllyRlolf! _..
Uncertain
Time from 91! 0611, mAfIfl (SD). min
Firftt vMiC1~
SMjy vehlde . . .
~~ life lu,.,~url yuhiele
EMS persunncl I)eI1ormfl(J CI"R
LD~-CI:'i'l COO1prO:>RRic;;;R
Total
Time Irom 91 I C1A11. ~n (SO). rnin _
Time frorn 911 c;)J! .~o Initial rhyfb~.I;lllRElRRIT1Anl. mUW'1 (Sl))~ mln
Timfl fmm Al1 cl1JI 10 flr:rt 3hock lor In~ial rhythm of VFNT,
mO<\n (SU), mlnS
Proportion oiiri';;. rir~t-i, mln on elflctrocllrdiogrsm ._..
with.comprlllllll~~!ll. rY19an (Sl)t~l?, ~1Io:>nta]
AdvarlcecJ ~1irwaY placed ...
rime from III 1 ('.all, mesn (SD). min
Inl~flnnUlllln!, inmed
Trrne from 911 c!1il, r'r1O;).r' (SD), mln
fpi!'1f!flhriM Rdrnini.a,tei-etm... ...
OOSll. mean (SO), ITl2
V~~OllrtWlin Ildmlnilltlll'lItJ_
fJo~, rTlIlf,In (SO). U
RlmrhnnRffl Rdministered
_..
II'IIr'l\vf!noLJs Rdmin~strali()r' or cJrip. n1WrI (131)), mFq
Died ill ~(;em:)
Tim~ 110.0.1 g.! I 0811, mRAn (SO). ';;;;;;j"" .
1"ran3parLOd 10 hospital . ...
Time from 911 0011 to ~rr'l", eMS bogan Iran8r'lott,
r~'\n (SD), mlnll
Hypoltllirrni~ 1t\9l'llpy
Prlortoho$pltslarr~1 ~(2.2).._.... 9_H~ :1(0.9) '7(1.8)
In m!lP1tII1 13 (9.0) 9 (6.6) 19 (!l.I) 23 (5.6)
^1,brwiatIOn&: erii: oardlopulmonwy r91l1._\l1Ion: ~M~, mmwy""cy Il'ledlcal 8eI\I1ceS: LOE;. lood'rli"rihuting hnn<1; NA, I ",llll,~~i'~e: VF. Yen1l1cular 1b!111a~on; 1(1', ~ntrinulnr
tachyeEll'dl..
'Vlluea !lI'8 expra88d 113 nl.mb~r lpnr..",hKJ") ,,,,In.., o1herwI8EIlncllcmed. All ccmparioorll.; p;.'.OS \'>X0fIP1 r...r rJl\ttain compsrllOI1s smOrl!l the prirnQry 'u".'" "" inolir.~tA<1.
1 ~ody type wu ..limited by Ihe PI1T'f,lflllJdIcu j"",1,irl!llh/l pAtu,"!.
.j.Po<: .001 (""~'l!l\)ted difference lleoauee device bell OOIIId nol 1",~,(Jmmodlte very !hin or ytry OO(\J)r, pl)!"ll}m'l,
fP- ,001.
lIP = .0:3.
'lI1"- .01.
219 (66.6)
I,(~' (44.4)
N^
.1.2,7, (32.2)
0911 7.~~L-
fi!i.!i (Ih.e)
_.. 252 (64,01_
IH~(51.5) .
181 (~8.5)
NA
13:' (:~h.4)
7Af.21.2)
66.2 (15.2)
2i.lS (85.7)
65 (~0.6)
:.16 (2<1,1;i)
59 (3tI.!J)
23 (14~~)
34 (21 ,32
&l.~fll:l.:;)
106 (6t!.:~L
:'1 (~i~.tl)
35 (25.0)
f;7 (39.6)
~lj(Il:l.1)
:jO (20.8) .
111.3(18.0)
9.'5(66.0)
33 (D.O)
H._133 (35.7)
84 (22:!?l
\l.(2.~)
1101 (30.6)
.~~(14.2)
1!i7(~9.B)
87(~?1)
IT (4.~'l).
n.l!~
2.2 (15.3)
45(31.9)
22 (15.:.1)
.~ ..(?:8)
50 (3U)
_ .1.1 pO.6)
!i5(~4,~)
:~!il ~4.4!
lj(::I.1.lJ
4:;!~~:!.I)
122 ("1,0)
79(~~
161 (1J.&...-
29 (7.~)
IIR ~1.R)
EI4 (25.~t._".
141.l (:;l.l:r)
. . 1? ~;~.;;.)
28 (17.5)
54 (33.8)
~l-l~:l)
9(5.6)
?~( 1 A.~J..
58 (40,3) .
62 (:;e,l)
( (4.Q)
5.1 t).i)
." 6.8 (3.2)
0.0(4.2)
7.&(2.7)
5.0(2,:.1)
e.7 (?7).
..~U(U)
7,9(2.0) .
6.9 (2.7)
8.1(M)
U.!:l (4~H~ .
14.8 (15.3)
._.5.8 (2.5)
7.!i(4~9.l.
. .... ....__~:! (-1.9)
12.2(8.5).;..
323 (82.0)
,n.9(4,5l
8.9 (3.0)
I Ul (fl. 1)
NA
NA
8.9 (2,91
9.7(3.1)
95 (59 .3)
. 1.407 (U,2)...__
12.8 (6.2)
In.G(9.9)
NA
NA
, 5:1 (150)
23.4 (2t1.1)
0.60 (ii:.ii2J [511
0.59 (0.21) [203.1
o.eo (0.90) [167)
0.A7 (0.25) 160J
.... 342 (SI'l.S)
'-1.9(6.3)
a4:'!(l;I!:l.t:l)
.!~.9(5.1=1)
283 (89.1)
3.9(2.5)
76 (19.:'9
.. ~i!:9 (0)
'!\lEl (30,0)
67.3 (44,7)
1 ~i.~ (:''53.8)
37.5 (11.7)
?Il 1 (r>A.~)
~14.!J (1'1.0)
:310 (!J3;n
111.4 (0.2)
31~J!!Ml
15.!5(Il:o.! .
:!t'lS (82.8)
4.0~.3)
.. .___ 52 (16.6)
40.0JO)
.. ..",,!j~ (39.7)
G6A (26.7)
130 (::34,9)
r~~:~JIO.7)
24<.1(65..1)
32.0 110.6)
'1;;':, (8:S.II) ..".. 1~J~6)
IQ..2 (13.8) , 18.1 (9.2)
122 (84:l) _.. 14:l (A9:j)
18.1 (1:.1.3) IA.8 (7.1)
112 (n8) 118 (7B.1)
S.?.!?.1) 3.9 (2.6)
12 (/J.:.I) ...... 1.6 (10.0)
::36,6 (11,3) 40.0 (0)
~ 1 (20.5) 40 (25,0)
70.3 (26.(J,l .....____.!?!:! e~l:l,a)
:l!:l ~!h.O) " .... ~g .(39..6)
~G.A (13.2) _~~~O (12.6)
10S (75.0) 111 (69.4)
36.7 (17.8) ~;5,5 (1?A.)
0100Cl AnlCric:1UI MecliA'"l (\,,,~llItlOII. All ~hts ftsel'VCd.
:16:14 .lAMA. Junl'. 14, !()()Il VII) 29';. Nil. 22 (Reprinted)
J.loW1lI~,..dcd !'rOlll ","ww.jams.com on SCpl:1'I11ber !O, ~007
FROM :ACFD
FAX NO. :5106183445
Sep. 10 2007 04:03PM P13
COMPARISON OJ' MANUAL CPR WITH COMPRESSION OF.VICF.
T~~..,_J. Out~?~~"b.r _~~_~_~.ent_~,~()~p Overall and by Rhythm Subgroup Among Primary Comparison Population-
PuI"1eCS
I!lectrlcal Actlvlty
,
I.,DB-C~III
(n :I 98)
21 (21.4.)
:;l8(:l8,6)
49 (60,0)
19 (1 !l.4)
~ rJ.1)
VF/Pulcoloss VT ,
r-o",' 'I
Mllnual CPR LDB-CP~ Manual CPA
, ,_,,,,jn ':' 119) __l~= 122) (n = 100)
~~,!rvivod, ~~.~ altor 911 ':all 49 141.21 ~, (4:'~~L....,._li! (~'~:~_
l)jed 01 GCan& 2.7 (2.2..7) 20 ('16.4) 30 (30,0)
Died in 1iI1111l'Qllrlt.y daparlrllilll1 .H (37.0) 49 (4Q,&... 44 (44.0)
rm,o in h(l(lpital ".,~,~7.fi) 3l'l~,5) 17(17.0)
Dlli!'ohllrgtld 81~ fmm hnRpit!\1 ~7 (~~.7) __..,,1.?J:!~,g) 9 (G.O)
CPC llCOrll
,I.. <::onsc1ous. and alert
2. CUrlllCioU3
3, Dependent
4. Unconscious
....,~._".~-"."........
6, Circulatory c;iIl~th 9:7 (77,3} lor. (Rfl.B) III (ll2.R) 95 (99.0) 15~ (Q9A) 171 (llll.3) 3..';6 (00.!l) 371 (Q4.9)
Abbr~: OPe, cerebml pri.'frnam.1G1 ot\t!i1tJorv: CPR;-;;~~~'tC)'wy n~I.4;t;nl1liul; I n~, r;~~b100t'ltIY tli1tllJ; vr:'~~~~~~ firR11~~lknl; yr. VWllt1r.lJl; I;H~~~~'i:\.
"Val.le8 are ellpI'IY8ed as number (percentage),
'I N"""(A,~;;li IJ~b\ WI1r1'l inc:'~l\J'lItIItl ror 5 RlJrviVtll'r.
~(i (19.~)
2(1.7)
2(1.7)
r)
.md 52.5% of cast's for which cardiac
arrest occuHt',d ,lifter T:MS arrival.
Amon~ rrimllry study path~'1I1,!;, the
mean (SD) time from 911 can t,(I ftr~t
U$~. of lhl~ ).nR-CPR device was 11.9
(45) mitl.UlI':!l with lI\ ml':c1ilolll of 10.9
minut\~s.
There w~q no signilkam djffl~l'l':n(:e
in s1.\rvivnll\t 4 hours after the l) 11 cnll
between the manual CPR group and the
I\\.uoml\r~d LDB-C.I'R 'groufl t'lverf'lll
(N.1<171; 2i)S/h Vl: 2't~%; P=.74) or
MroOl\g lbe primMry "'tmly pnpulil1i<l'n
Cn= 767; H.7%vs 26.4'J{,; P=.f'I'l). Sur.,
vivaltn hospital discharge wa.q Illwer in
lh\': I.PR-CPR group amnng p'rimary
t':p'i!lodes (.'i.li% ~ \1.\)% IP='c141: ad..
jUMI'.d f(lT c:t'lvl!.rial'l~:o> i1l1d c:lu:o>tC".ring.
p= .06), but :.;imil.ar amung lbl:.I1Ullpri-
mary cases (10.6% v:.; 11.9%; P=.72).
F.XdlllliTl~ 'i survivnr.~ with irH:mn-
pletE: neurological data, survival with
a cerebral performance categolY score
of 1 or 2 was recorded in 7.5% (281
371) of patients in the manual CPR
~oup com-pared with 3.1% (121391) in
tnc LDB-CPR ~nUl) (P=.006).
The survival effect of the LDB"CI'R
dl~vkl~ diffr.tl~d, hut IWI slp;nirl<:lllllly
(P.. .37), according to initial rhythm of
ventricular llbrillntion, pulseles.'l elec'
t,lical activity, or asystole. In contra.'lt
lO the venuicular fibrillation and pulse-
It':ss elec.trical activity sl\hgroup!l. out,
c:omp.s tre.~,Md better in thl'. I..DB-C1'){
grol.\l) in the asystole subgroup for
AsyBtale
I ,
Manual CPfil I.,IiFS-CPjq
(n = 154) (n = 174)
Ifij'IO.4) :iO (1/.2)
73 (4::.4l.... 86 (48.9)
etl (;42.9) 61 (ab.1)
1~ (Q.:.!L_,_~!!4:.~)..
1 (O.ll) 3(1.7)
All Primary O-t
I 1
MIInual C~1Il I.llI!~Pl
(n = 373) (n = 394)
92 (24.7) 104 (26.4)
.~~2J34.91... 133(~~~.
14 (41.3) 169 (40.4)
52 (13.9) 19 (20.1)
:.17(0,9) 2:.1(5.8)
fi (4.1L__~ (2,0)
5(4.1) _.:'.._..._
IS (5,0) 3(3.1)
() ._._ ",. "" .~ (2.gL___....
o
_....!...@.8)
o
..---
o
25(6.7)
3(0.8)
5(1,3)
2 (O.G)
6(1.5)
El(1,o)
7(1,0)
1 (0.3)
o
1 (1.0)
o
o
1 (0.6)
o
1 (M)..
I (0.0)
Table 4. Lo;istic Re~m55iOrl of SUrViva,1 to H05pital DischarJ;C*
Adjusted for ClUBterlng
Univsrisbfo P l\IIuttlvarillblo P
OR (96% Cllt Valuo OR 196% Cllt Valllo
Age pery ?E>7 .lO,~:O.QQ) .OO~ 0.98 (0.9S-0.99) .0 I
PEAtoVF 0.28(O.1~-0.55) ~:,OO1. 0,36(0.17-0,75) <.001
ASySlojotoVI O,0l:)(O.020.IS) <,001 0.09 (O,();:l.O,:!8) <.001
Wltno~L_______.,__,_ ._. S,(iI~e,l:IO-ICJ.:lO) __:'::.9~1 3.~~.lL~l)~~,.\lO) .\I~
SiLO C ~, 70 (2.1 0-1.l.51J) .. ,:::.l~?_~_ ::I. 70 (;>:~:7 :Oo.L ..-..:.~ .UO'I
Aeepons!! lima nf nrl!t vahiclll(ll,I!~' O. 72 (O./\O~?S6) < .001 0.70 (O.58-n.BS) 0::.001
f"uhllr. !~~~~.. _ . .... ....... ~~~I~t~-~_.::::90,~ ". 1.80 (0.9~ -3.40). ._. ._--E!
~r:l!3:CPR lr'l:lt\trnllrll \jI'Llup 0,57 (0.33-0.01)) ,045 0,56 (0,31 "1,00) ,06
IIbbreV8UOIl8: CI, conndencelnterlllli; CPR, cardioPUlmonary 1'93\15CItt\tion: WEt. 10m .tjkltrlb<ft.im bl1F1d: OR. odc:bJ.I'\l'
nll; ~. pullll!less eleclMcaI ectMtv; ',IF. ventricular llbr~letlon.
""1I~1,lt"," .....id..'.lf 1M" ""t Mi\l'lif....an! In \he mOI1l1l~: 11f1;,/Sllab!9! CJayIllmm all... Rl8/t (OH, 1.00 [ll'J% CI. O.Illl.l ,OOJ:
p - ,27 rudj\J<<Ind fo, nk,.I"n~"D: u~roru O<KJ,""lm, 20 (OR, 1.60 (lI6'll> CI, 0.06.2.60): P ,- .1 e l~dju.tod fOr
clua1Elrlll!lD; mQn (OR, 1.40 fP,5% r.1, n.7!l ~,701; p = ,23judjlJr.tlold ftx cluratllrifYJll; CPR Il'lrformlold by Q by.;tl1/1lJilr
(OH, ',10rv~%~. O.~;!';!,WlIP - ,~1 [w:fjUlt'edforCfuettril1!3l); I'JNponetdme of sdvE\nced lit....IPPortvel'lcIe (OR,
0,i1 fll:.'j!, cr. 0.11(-1 ,0:!1: P = ,1!) [o'JJ\l.'I1od for clY3lerlIl!lll, Inwmctl<Jr1ll CQrlalclnd: \l'ft\ment QI'Ovr;> x dlIy, from
QI1EIllWl.I'.. .81: treatment group X 9ItgC.,.... .12:\reatmentgroup >< rhythm(wttl1lmpuledrhythmvaluea).1' = .U.
tn,,, on~ 11Ighllr th8l'\ 1 III(f.... a hlgl1er Ukellhood (>f ~urvl\l81l1nd converaely. eg, !he <Xld8 01 AIlIV~ de.'>l'68Rtllly
0.08 If;r o,,~h yonr of ano, d,1(1rtl"'''' by O.3f. if r'>Ur1n in F't^ 11.."1,, vr:, nln.
4..hour survival (17.2% vs 10.4%) and
hospital dhichargc, (1. 7')h vs 0.6%)
(TABtJ~ 3).
Thl~ results of tl:tl~ lo~i"'l it: r'1~grl~si.on
analysis of hospilal SllryivaJ. (or
the pdmary cOJl~,pad~ClII t::.t~es an~
shown ill TABLE 4. Traditional risk
factors for unSUC(:l~SSr 1.11 f',~susd tal ic.m
(older age, l.\1lWitllesse.d collapse,
longer response time, l10npublic loca..
tion, and burial rhythm of asyst(lJ.e or
pulse1ess electrical activo'll}') were con-.
firmed. The associatton het,ween hos-
pital survival and ttcatmenlgroup did
not differ over time from the start of
the study at l'.a(~h sitl~ (P-.R4 for lrlh~r'-
action, adjusted for other covariatt's
and clustering).
A'Ii expl~ctcd from hi'litorical rates,
"'\Jrviv:.tl WllS sigl,irit:lInlly 1:)(~I'I&,T in sill~
C cl1m.pared wilh olher sites (Table ...
.~nd T<\IIJ,F 5). HQWl'V~;r, the aSSllda-
tion between survival alld treatment
14t<lUP llid rlUI difrl.:t siwlirkanl.ly a1
site C compared with the other sites
(P =.12 for interaction. adjusted for
othl~r covartal\~li and dU!:iH~rinK;
Table i). [lolh before Qnd after the
Decemher 2R protocol chan~t:, EMS
pelsonnelllt $it~ C hlld hi~nel' proto-
col compliance and \1se.d th~. I.DH-C1'R
device earlier in the resusctt&\tive effort
C200fi AIIt...icIlR Mot.."..., ~..cl*Lu'll. All rlgh... ft_~<td.
(kepr1nted) JAMA. ,IUlle 11, :lOO6-V,,' 29~. N<>. 22 :'1525
Downloaded from www.jUlnu.L\Qnl Oil Sepl~'lllblll' 10, 2007
FROM :ACFD
C:::OMP^R150N OF MANUAl. CPR WITH COMPRES.'iIO!'ll DEVICE
thll11 F.MS t>er~onnd at the orner sit\"$
(Tllbl~ 5).
^ PO$I hoc multtvariabl.e 1Ill.11IYt>i",
wl.ich focused on patknl~ who were
t.reated relatively qui.ckly afl'~r thl".ircar.
dhl.c arrest (witnessed primary c:al>es
found in vcmrh:ular rihril1ation or
pulsdess t::lcctricaJ.ll~~t'ivity), indicated
that as first vchJ.cl\.~ t'Cl'lpon!le time shon-
elu~d, patients in I hI': manual el'R gt'Ol.lp
Wt~rt~ increasingly 1110re likdy to .~lIt-
vi"e t.O hOlll1ital discharge compllrt~d
with pat.ient.~ in the LDB-CPR grvup
(interactton P=.06), AI !l.ll minU1.ell of.
rl"!lpon!le ttme, the model iHdil:llh~d the
lrcl:ltmcm ~l'OupS would have the Sllml~
SUl'Vival. . .
Mode of death in th\.~ hosr'l1Jl.1 was
similar between I h(~ I.reatment groups.
Approxi.m.rltr.ly~'5'J{, died within 48
hours from 9. pre!lumed cardiac cause.
FAX NO. :5106183445
COMMENT
In this lr.ial comparing manual CPR
with aULOmatl"d LD8-CPR, inlcdm. (l~-
sults prompled early tennin!lllon ~ r(~t':-
ommcndcd hy the data and :;afcly tk1(I.tl.i-
taring buard. Although4-h\.,1.u :;1.1rvival
W!l:; :;imtlar hetween trel\lrn\~r11 grour'"
among primary cardiac ~m.'.st epi-
:;odc$, hO!lpital discharge. survival WII:;
lower in t.he LDB.CPR group (:5.8% v$
9.9%) as was survivlll with intact neu"
rological sUltuS.
Evidel1cc iI1llicaleS that incre:a.sed
blood flow (iurillg erR should trans-
late to a highe:r. lik('.1ihood of SUCCCSll-
rul resuscitation,~ Th~ I.DB-C.PR de-
vice evaluated i.n Int., !ltudy prOdl.1CeS
grutcr drc;u lal ion I,han manual CPR in
ani.mal ml)dels of cardia\.: arT'\~"t.I~,17 Tn
ob$l~rvai:ional h\.lman studi~~ nr 111t~. d~-
vice, most but not aU hlVl':sl igator.q have
Sep. 10 2007 04:04PM P14
indicated greater likelihood of return
of spontaneQu:,; cil'(lllation compared
with historical (;\.1111r\.)I$, with I stl.ldy
demonstrating b~ll ~r llurvival to hos-
pit.al discharge. ~.,:lJ.:l4 The rl'_';ults of the
current. randomiZed study wt':re not ex-
pected and thcf\~ i~ II\') nhvtmls expla-
nation.
One potential explarlluj\')n is that
patients in the manual CPR ~up bell.'
l.~fitH~d from a H."lWdlOme etl'cct.!!Illt:h
that manual CPR quality i.nitilllly
exc:.cc.deu ~tandard practice.'~ Con-
v~r'l'Idy I there cOl.lld have been It "1(~lIrrl-
ing I.~ul've" for use of the device with
performance cxp\.~Clcd to improve over
time. However, during the last 2
monlh~, .qUrViVlll to hospital dischl\tg\.~
for primary cases was 8.l % for mallual
C1'R and 5.0% for LDn-CPR, nnclinftll
similar to those frl.ltrllh~ init.ial months
Table.~~. comp~~~on of Site C With Other 4. Sitel; for ~rirnilt)' Cases'
other Sites
WI\TI~~~~...e3rc'll,c llrreST
A<JC. mCl.'ln ($1,)). y
VFNT
_........_--~" ~..
TIme from 911 cllll. mean (SD), rnin
First EMS yilt,i!;IQ
A<ly;lf1CO(l life SUPJXlrt VAnlnlA
FiI'llT llhOt'.k forVFNT
T..rrninution OII'(l~I.I!\CltMlve effort
- ,""
T rsI1aported tu hoopi tal
LDS-CPR txlnlpT'lolY91onG tlllomP\O<I
.Rem ~I'nher 28 (Option 1 ror sitQ OJ I
Aftfll' December 28 (Option 2 'for lIil" elll __
Time from 911 cwll to LD~',CPR
COmpre~!lI011~., I1'lflfln (SO). mln
8eforll [)l;K;tInll;,of 211 (Option 1 for sit'" C)~
AilOr Ocoomber;>8 (Oflr!~~:? f~r 1l~'C)j1
$urv~ .0:4 n afrflr 911 CRII .
DIll<'.hsrged e1NII from hospilal.
C:PC ~core ...
1. Conscio~IS and we~ ' 4(4.5) 5. (1.5) ] 11 (16.~) ...,,1 (U;) J
2, ConSCIO~I$. . $ (1.0) ...~ 0 ?J~.~!L
:J, c:.~.~l;lI1t I (O.~L 3 (0.9) .OUO~ 4 (6.9) . ...~ (6.2)
4, lJncnnllCious 1 (0.3) 1 (0.3) 1 (1.5) 0
6, Circulatory d\lCrth 278 (99.8) 313 (96.0) 49 (75..1) 5!! (U92)
AI ~1I'~3tl<lnR: CI"O. oerebr81 per1ormElll<"..e oategory; GPI'l, C";;'~~"'ll'''n.lI'Y ~~~ItaUon: "MS. .rmr!)~ncY medl<:lC\! ~;";il",": I.nn, k,....j-<i>\lrihllil~ 'b~Jld: ";"'. V9n1I1ouJar fIbI1lle.
liutl; Yr, y.l...,ncular taonycaJ'(lIQ. .,.
'ViIlUti are expreaaed u numb9r (p9~~.) \I~~"''' "I.~...r..w.o. illdloated.
I" values are unadlusted. Thoo mod. o<.>r1sisted.Jf u w"ii""lc~ (.)cll'.Rtllnt) plU!l8lte plUs tre!\tmel'll !l'mifl r~.l" ,Olll x 1".t1mllllt group (IdenUIy 01' log-linear link) UIII...~ Qttlll'Wi>;r' "ul-
<:AI""j, Nu h'~~01\ terms wer;, significant.
*O"tic., 1 ct,r..."d;lr. rhythm recording ioiowed by crcularory ~ wilh niloollro LIJ~"CIJl.\ device or menual CPA.
gThe. mor.tel oonl:;p;lnrl ur " 'f:lriaflli. (cclIlstant) piUS lite P1U8 dale Ibtfor. t1I1d n!l"r Il<Ulllrr.K,r 28) pllJ8 a1te x dIlt8 (before aM aflllr DeC'-"l,b", ?Il).
IIOpton 2 defined Ill.' l'nr'.KII.,Ic\ CPR wt1I1 mAlllJAI OOI11pr-'ons l'e\I9r*~ tJ[ r;IIKk>rllimMrl Until tne Irst 81'1OCK _'rrYlm,
~\')I'C seem 'I and ~ 1'3 e, 4.!Ind 5,
252" JAMA,J'''''' 14, 20()(''''''YIII ~~~. Nil. U (Roprlntod)
I
Manual CPA
(n = .~07)
...~lil (49.9.)
I:lti.l,l(lli..q
97 (31 .8). ...
- '----'
LDB-CPR P Value
(n .. 32.9L__ f?~~~!.
152 (4~:.2) ,13
US.7 (15,6) .57
101 (31,4) .76
5.9(2.2)
Z:? (4.4)
9.8(3.2)
.34.7(10.;.1)
31.4(10.5)
6,7(;>.1)
~.Q (~.8)
12.0(6.6)
:lH.:> (10.6)
~.~(10.4\...._.
"';.001
.22
.$:l
.08
150 (02.4) 1
118 (00.3) .
66 (??1)
20 (8.5)_
, lotI (4,3) 'J'
13.1 (5.0)
8.?J24.9)
16 (4.H)
.006
-=:Jllll
24 fJtiA)
17 (25.8)
9.U(2.1) ]
10.9 (3,1)
;)2 (a.'M)
7(10.0)
.55
'(14.__
...;001!l
Site C
P value
for Treatment
G~_
.26
.87
.87
I
Manual OltFl
(rl=66)
:;~~ (41i.1i)
67.7 (IS.H) .
22 (33.3)
'1
LDB-OPR
(n ..65)
:?S (35/1)
..Ho.5 (16.0)
.?1 (32.3)
.no'l
4.01(1.2)
6.4 (0.4)
9.6(2.8)
32.1 (11.3)
36.Q(10.2)
5.0 (l.tI)
a.lit-HI)
10.6 (2.7)
35.4 (14,;>)
39.1 ('~.6)
.93
.88
.05
.m
,09
.Ol~
?7 (Q3.1) ]
. ~~.Jl'.7 .2)
.7\:l!l
.009
.000 11
4)2006 AmericAn M~jc"ll\~~I""lallclll. All r1pu rE8Cl'\l'Cd.
Downloaded f100m www.jamu.<...l11l<.1I1SCJll..lIIh~1. 10,2007
FROM :ACFD
FAX NO. :5106183445
of the: study (11.7% and 8.0%, re!lpec..
tivdy).
Another possible ex.planation Cor
Ih\~ nUlt:omr.s is .hat c1~J'lI(lym(T1I liml':
fm' the LDB-CPR device w~s prn-
longed, Mean. time to first shock ill
prirlllolry t:loISI'.S wi1h initial rhythm
of ventricular.f1hril1atioll occurred
2.1 minutes later in the LOB-CI'R
group, While de.Vice de.ploymem tittle
Willi not mcaliureu dircl:lly, site C
MPl)li(!tl thl! dcvll~t! t!l\r.lit~I' l\1").cl mllre
frequently than the other sites and
yc.i !:'howed gr('~ter relative haurd for
the int.ervention (Tahle 5).
Another implementation-based
\':xplanarton i.q enrolhnr.nt bias, "F.nthu-
!liasm for the aUlomall~cJ r.DR-CPR
device could have m.otivllted F.MS per-
~\(mnd tll em!)11 plltiETlts wl'll usually
Wllllld havc been. dt!l:larcd dead on
arrival. Thill may have. ol~cl1rrt'!d it'! 11
few cases because 2\ more rrimllr)'
patients were enrolle.d in the LOB-CPR
grO\lP compared with the manual CPR
grotlp. HoweVer, almost all Long-tellu
stlrvivors were among patients whOSE:
initial rhythm was ventricular flbrtlla"
ti.on, pulsc.lt'SS ven.tri.(.'ular ta(~hy(.'al'dia,
or pulsr.ll"....~ der.l.ri<:lll ;,J.l~livitYI llnd rllr
whom enrolhnenl llnd ballclil1.l! l~hl:lr-
al~h'.ristiL:~ we'rl'. l:OItlI.Jarllbll~ bclwcl:.D
the 2 study groups. Moreover, the
adverse intervention relation.'lhip was
seen among patient'! presenting wil.h
ventrtcular' fihlillation, a gmul' that
would l'out.inely receive resuscit.ation
and Cor whotn c.nrolltncnL bias was
\lnlikdy,
Othcr potcmiltl explanaLions for our
findinp;.'l may hl!. rell1ted t.o thp. dirp.ct
physiological effecr~ of rhe autnmau.d
device. Me:dications administe:n:d with
superior blood !low gem:rate:d hy the
device might. exceed rherllpeutic thresh-
oldsllnd insteilcJ bl:'. toxic. However, Wt',
IIrc unawan~ of l~idcncr. for such an
r:rrl'.d. An addi1illnllll:nn.~il"~ratil.ln
;1:l.Volvcs the. manner in which blood
now is g;l':nl~rlltl~d (ie, HO I~OItl prl,'.sSlt1m;!
luin with the LDB,..CPR device vs
manual CPR rates of ===100 compres-
sll1nMnin), There may be an as-yc.t un-
mapped relationship between time,
flow, and reperfusion injury when early
Sep. 10 2007 04:04PM P15
C()MPARI~ON OJ; MANUAl.. (:I'R WITH (:OMI'RE5SI{JN n:F.VKE
low blood flow may gt:Ilcrlltc less n:pe:r..
fusion injury. )6.)7 Thel'c i~ also the PO!!-
si.bility lhilt dll:'.st compres!lim1l'; by the
l.DB-CPR dl~"ke may cause dirl~l:'
physical damage to thc cardiopulmo-
nary system, although nwiew of hos-
pill..!. rl~l:ord.~ 1.11 1Il1lnltor rOT 1\c.Iv\'.rst~
evell t.~ did not overtly identify this pOB-
si hili ty. )~
In additiotl, there is a I in 40 t~hau(;e
thaI lht~ atlvc~!:'e !:'urviv~l OUL\':ll[nt~
wuld h~ve (lC".I~urrp.d lHlder the mdlllY-
pothcsll; of no trcatmenl dfctl, In this
regard. thl~ possi.bility or Ilnl~tp~al risk
in the grm\ps randmnized at !lite c:
should be considered. lhat. site ac-
C"ounted for 40% of the stlrvivors, and
survival ill it~ manual CPR gt'\mJl was
SUbS1l1l1tial1y gl'l~atl".r than in previous
}'C.aT.s.
Thl~ effect of J .nD-CPR ~~OUlpres-
silm may have difft'!red ele.pendlng nil .
Ihl'. pre.~ent.inf( rhythm or time fnltYll:Ql-
lapse to resuscitat.ion effort. l'at.ietlts
with asystole, potentially most consis..
tent with \lUtrcattxl al,1l kmgt!\' arrest
d\lratk1n, appeared to ben.eIit from the
LDB..CPR compression whereas those
wi.th Vt~ntri.cutaT fibril.Jll1ion or pulse-
ll".~s dedrical activi1y llppellred lu ex-'
pt~ri.t~Ilr.t'! harm. In a !.JuS! hnc muhi.vari-
ilble l\nalysi.~ l)r wil.tll'::>.sed primary 1;1l:lC5
found in veutricular fibrillation or
puL~e1ess electrical activity, shorter re...
gponse times favored the manual CI'R
group, while the model indicated the
treat.ment ~roups WOt\1d have the same
~urvival when the respOllse time
n:achl',d 6.6 minutes (P,;, .06 (or inter-
action). To StHtlC t~XII~nt. thi.'l (in(llng
may be interpreted as consistent with
Ot.hCl' re:port.~ of oh~e:rvational human
snldie:s that have: evaluatt::d .this LDB.'
CPR device.21 These relationships and
th\':ir undl'.rlying mec:hanism~ rl:'.quire
additional i.nve.sligatlun,
Just as 'PtloradhcrcllC(~ cli1lllt~~ the oh-
ser....ed effect 11f a beneficial treatment.,
it also dilutes the effect of II nam1htl
I rl)llll1'\(~1\l. Thus, thl! Hhse-rved diffl~l"
et1ces betwern site C and the other sites
are compatible with the ovl-'ralL 1n1pres-
sion that this lm.plemenTaTion of m.e-
chanical CPR with the LDB-CPR de.
vice may be hanuhll. The.! difftTcnces
QZOOI'I Amer;".n Meclil;lllJ A"'.Of.i;clion, All ~Jghl. re!leJ:Vecl.
Oownloadcd from www.jAmn.c<11n on Septemher 10,2007
arc also compatiblc with the concept.
that t.he magnitude nr harm Ul;.\Y <lc::-
pl.'.ncl on. thc. eapabilitlc'\; of the EMS
sySl.em.
This study has se:veralltmil.ation!l.
The LDB-CPR device was imple-
Illenll~cJ III vllrinus slal/,(~~ ur rt!su!;dIS-
tion, a flexibility desis,tned to mini-
mize CPR interruptions. A protocol
requirhlg device implementation at a
~11l('lkula~ poinL llfl:arc might prothl\,:C
difft'!rl~nl results. F(It ('JCllmple, device
application in apparently late stages of
anest (the asystole s\lbgroup) ap-
peared tl"l he modestly hendidlll. Al-
though each sHe conducted a run-in
phase with the device, more lmenslve
traill j'ng or 1II longl~r fun-i t1 phase may
havp. prodl1~~ed different rl~ults. The
study cvah,ated the proportion of time:
with CPR during the fint 5 mimltcs of
EMS l'l':$Ust:italil)t1, bllt tlit! not c..'Vl1hl-
1111, Ihl~ "quality" llf mamllll CPR (ie,
1'I1tl'., dl'.]'lth, r~cl)il) or how mllnUl'lIl'ltld
LDB..CPR compression differed later Lll
the course of reS\lsd.tation. Because of
adverse trends in safety outcomes, the
study was te:mlinated j)lil1r to cmn-
plete el1l'Olhnmt. Although stopping the
study ror !:'tatisli<.:lI.1. futility was not part
of lht~ prl~pcdfied monitoring plau, the
l:lllldil itlll.al powl'.r to tklee! the hy-
pothesized difference in the primary
o\ltcome was only 0.55 at the time of
!ltudy l.el'milllJ.lion.
CONCLUSION
. As implemented in this study, du use
of an automated LDD-CPR device for
rl'_ql1Sdtalioll from I"IU1-nr-hospilllll:llr-
diac arrest appeared to result in lower
!lurvival and worst: l\t.urI1lngh~1l1 out-
COUles than traditional manual erR.
Device: desi~n and imple:me:ntatinn
strategies may need further preclini-
cal twaluation.
The results of this snldy underscore
the cmnplexily of re:'iu.~citali()n rrom
OtH-of-hospital cardiac an'est. l"urther
. rl~"I!aTl:h is T.Cq1liTl!d to nnde.rstand
the interaction of manual 01' assisted
chest compressions with other as"
peets of resuscltlHi.on such as the
phase of the arrest,'N drug choice and
dOSl\ timi.n.~ of d('.flbtillati.on, and trt~t-
(Rtlprinmdl)AMA,,June 14, l.00r>-v,,129'i. N". U 2ti17
FROM :ACFD
FAX NO. :5106183445
COMPARISON OJ-' MANUAL CPR WITH COMPRF."S10N l>liVICE
mt.l1ls such ;.IS hypot.hermia and coro-
nary rcper(usion.
AutIlorContrlbutlom; Dr H~II.I...rn Io~d 1,,11.~u:t!s~ 1:0
~II <,fthe data In the study and t.,kcs r~ponsibilityft.lI
the irrtcf,rity of !he c1~ta anti the accuracy of the data
analysis.
.~rlldy ':O~':I!rJI tmd design: Hal/stroln, ilia., Sayre,
Chrim-nson. Anton, M(I';~ln, V~" OtUnllhftlll, Cobb.
lI~qf/I~)t/(m of datA: Rea, Sayre. Christenson, Ant<m,
M_sso, Olsufkn, Pennlnl.(tl'n, Whit<!, Valin, 11l/!ar.
c:ohb.
Analysi~ arKllrrlerpret8tl.m of ~t,,: Hall~rom, Rea.
Sayre. Christenson. Anton, MO!Jt'~'J(I, V'1n Offinr,hur..,
Mo,rl'. Cobb.
Dr.llirlH "I III. manuscript: Hallstrom, k... Sayre,
Critical revision lJt Ihe mQn/J~.rlpl fill' ImporfJlnt 1/)-
Is/llCtual content: Hallstrom, Rea, SilY~, ehri,len""r.,
Ani on. MowM, Van otrlnl!ham, OIsuflc;., Pemlnglon.
Whit.., Vnhn, H"'ft., Mllrri.. Cnbb.
Statistical analysis; Hallstrom, Mmri..
REFERENCES
1, stielll, NichQI Co, willi, c;, eUr. Ilealth-related qual.
ity of liN is bettor tor cardiac arr<:st $\lrviVQ~ who r~-
celved citizen cardIopulmonary resuscitation.
C:iff.ullltlll/l. 2003:108~19!9-19~~.
2. Ewy GA. C:"rcliuL",..btlll 'esU1~lt<l~"Il: the nt!W car-
diupulmnnnry N".Iu."dl:&l:iuFI. C:;I'al/4'tJ(.)I). 2005:1'11 ~
~1114'~14~,
3. Oallaj!hef EJ, Lombardi G, Gennis P. Effoctivonen
of bystander cardioPlllmonary ~su)Qtil\tiun ftnd sur-
vival fullowinl,!: vuh'(-h"'pil;ll O\rdiac atTft.4t. lAMA.
lSl96;274:1922.1925.
4. Niemann fT. Cardlopubnonary I'f!$llitljt;,\ticm, N Fngl
J Mad, 1992:!21:10i5-1OS0.
$. V,, T. wr.il MH, TllIIIl W, et al. Adverse oulcomes
of In""...'pted ,....,,,,,,li"I,,,,,,I',"...,,. IIt"lll& nut,)-
mated dofibrillation. Cirwll1livrr. ~OO';10f5:~158- ~n.
6. Korn 1<8. Hllwir; RW, Berf, RA, S.,ndm" ^R, r;wy GA.
I mportanee of continuous chest compressions durinr,
",,,dk.pul"',,nary ll!I\lJ$Cltiltlon: Improved outA:Omedur-
ing ulmulated sincle lay-rescuer stl:nario, Cir'Clllation,
2002:105:645-649.
7, SteenS, ~jaoQ, PWreL, Pa~kevlclU5A,Sjoberj!T.
Tbe critical importance of minimal delay botweerl
chfit compressions and subsequent deflbrmatlon: I
Ilmtm...dyllnl'rlic: f!HI'I;lnlll\on. IleS/J<t:ltJltJon, 2lJO'.i :58:
249.258.
a. Cobb LA, Fahrenbrud1 CE, Walsh TR. at II. l"flu..
ence of cardlopulmonlllY re,uscltation prior to defi.
hrlllaaon In pa.tients with out-of-bospltal ventrleulilr
fibrillation. lAMA. 1999;281:1182-1188,
,. Wlk L, Han5en TB, Fylllng F. et ai, Delayin,
defibrillation to give basi~ (;llr'diopulmQnary r",usc;j-
. taaon to patients with out-of-hospital v9ntricu'ar
flbrillaliol\; a rahcklllllZed t.t1I1I, JAMA. 2003;2119:1389-
1395.
10. Hu~eyln 'I'S, Matthews Al, wills P. O'Neill VM.
I ,"pr(lVin" th\! ef1",~;v""l!!'" ..I cunllnuou. clt"'*<l t....l!!ol
comprc"ions: an e~pl'mlf~ry .Iudy. P,,,,';,m;ilJlUlltl.
2002:54:57-62.
1" Ochoa FJ, P.afrlillle-C(lrn~r~ t, LI.(A v, SIIt'llleIlUI
I. The effea of rescuer fatigue on Ihe quality of chest
L'\lmpre~$ions. Res'J~rJt.."Jrm. 1998:!'/:1'I!I-152.
12. Hightower D. Thomas SH. S1(>(1" C:I(, I)"n" 1<,
Mar~Il.IA, DfI<;~y In quality of closed-chest compres-
sions overtime, Ann Emerij Med, 199'1;~6;300-
303.
13. Alx:l/a liS, ^lvarad.).IP, My~rebu$t II, eLal. Qual-
Ity of cardiopulmonary resuscitatiQn during in-
h"'pltal cardiac arrest. lAMA. 2005:293:305-310.
14. Wok I, I(ramer-jllhAn.len J, Myklebust H, lit al.
2628 JAMA, .I11ll. L". 20M-Vol 295. No. 22 (P.eprintc<l)
ObWned funding: HaUstrum.
AdmJnistrativl, ledlniclll, or ,"alr.f;.,1 SlJpfJ(Ilt: Hall$!I1llll,
Van Oltitlghftln.
Study s~.rvlslon: Hallstrom. 1~0i\, Sayre, Christenson,
Anton, Mt.~IW, VA" OlD"I\Mm, OIwfka, PeMlngton,
white, Vahn, HUSilr. Cobb.
F1"anclal Disclosures: Or MOle$W reported recelv-
in~ ~ulpm..mt and supplie, for the AED dinical mal
from trn: ZOll CorpQra1i.""
Fundlnct5upport Thi, ~\Udy wa~ 'flon&Ored by kevl-
v.lnt l"..IlI'p<Jratlon, now part of the ZOll Corporation.
Role of &he Sponsor: The funding organizalion pro-
vided the automated d""ic..$, asslsted with training,
nnd ftlL'ended some of the inve~tiSlltor ni~etinl,!:S, but
had m, 1'<11... in Ihe design and cond~ct of the study:
eollectinn, rnftrhllletI1~nt, analysis. and lriterpretatlon
of the data; and pl'Opanrn...n ')r "pproval oftM ",;IIIU-
,cllp!. Tlle sponsor was provided a 1-month period
to lWicw and wmmcnl \l!'l+,~ manuscrfpt pdor1:o .(ub-
mission. The invcstir;ato" wc"" "ml..r n.. "hliltntilln
10 Incorporate any sum input.
Sep. 10 2007 04:05PM P16
Aekn_ledpnent The data and safety IIIUI.ltot1n./:
board C<ln~ of' ,ilWrenr.e llmwn, EMT .p, Depart.
ment of Emerpncy Medicine. Upstabl Medical Vni-
v"",",lty, Syr";""', NY; OIlVld tioff, Jr, MD. l'hD. Wake
Forest UnlvcMi1y Sd",,,1 ..t M",dlcln~, Wlnston-
Salem. NC: Jool SinJ:Cr, PIlD. $1: P'llll', 1-I000pltal, Vall-
<XIuvar, BrIIJ5I1 Columbia; and Ten!nce Valenzl.lel.~ MD,
Uniwrsity of Arizuna Cole~e Ilf Medlr.lne, Tuc.'IOn. each
d.lJl and lafely monltorln/: board member ~ived
a fixed i1mount of 1n11l"1lY ($.'000) Independent of how
m~ny tlmes the board would nead to meet. We ae.
knowlodgc tho dcdicat'ld eff01't5 of lhe foDnwlnll erner-
Jl:ency medical servIce <ll:enciC'l and their pel'3onnet;
British Columbia ^mbul./ltJ! SP.rVK'I!, IllIIdWln IiMS. Cal-
gary EMS, Eutern Arca Pre hospital Servlce~,
McKeesport Ambulance ((eseue Sorvlcc. Mcdicall
R_ T ....,m South AlltllolltY, Noltlowe5t ~MS. Ross-
W"t View EMS Authority. Seattle FiI'G OOpllrtmrmt,
Ttl-<:mllnKJ.lily south EMS, Columbus Division of Fil'l!
EMS, LIpper tirtin~on Firr. Divisi..". oml W",thing-
ton Fil'l! and EMS.
Quality of cardiopulmonary rcsus~itatlt\n durin8
out-,lf-ho.spltal cardiac arrest. lAMA. 2005:29'3:299-
304,
15, Tlmmorman S, Cardoso I.F, R~mlrll~ lA, Im-
proved homodynamics with a nuvel d1e.t t;Qmp","-
,Ion devieo durin" trcatll'lorrt <If i,,-h...pH'11 ~art!i."
arrest Prehosp Emorg Cil~, 200J;7;11i2.
16. Hillpedn HI{, Paradis NA, OmatoJP.lmproved he-
modynamics with a novel chest compression device
during a portlno model of cardiae arreRt. c"//'(Ulilt/(llj,
2002:104>:'~8,
17. Ikeno F, I<aneda H, Hongo Y,ct 011. ^ugmenta-
liWl of Ii..,,,! perfusion by a novel compression de-
vice increasos nourologicully intllet ~urvlv~1 in ~ pul-
eln" model at prolonged cardia.c arre~t. RCSII5Citation.
~lIO();15B: 1 09-118.
18. ~lIlllstrom AP. !'aradls NA. ~ro-randomiz.tion and
r.t~-I'lln.l/""illlu()t.ln emergency medicall'l!S8al'Ch: new
name! and risorous eriterift 11II1IId mcthc.>d~.
. RC$usel/ltllClrl. 2005;65:65-69.
151. ~rQtedion of Human Subjects, 4!S CFR ~46.116
(1997).
20. protection of Human Sublects. 21 CFR '50.29
(1979),
21. HerlltzJ, Ekstrom L. Wennerblom B, Axelnon A,
~,ml.( 1\, I-lClIIl1hAr/l S. Prollir\O!;lS amongsurvlvon; of
prehospitnl."...lj,": ,,,,.,,,t. Ilrm Emer!: Med. 1995:25:
58-63.
22. lacobs I. Nadkarni V, R..hr I, 1!t.1. C.'l'rJi~': arre_t
al'ld cardlopulmtmary ~u$Cltatlon outcome reports:
updatraml ~implification of the Vl$t\\i"lflfl'lplatcs for
rllN"",:it<ltlan re/llstrles: a statement for healthcare pro-
kssional~ m>m ~ l~"k (11";""( lile Intematiult"1 Liai-
son Commitltc on ResuscitativlI (At...,r;(~11 HCllrt A,-
wcialilln, tUropL'!tln ke~u5dtatlon Council. Australian
Resuscitation Council. Now Zealand RIl$lI~cilation
c:uundl. HNI't and Slroke Foundation of Canada, In-
tl:rAmerlcan Heatt Foundation. Resuscitation Coun-
cil~ of SlJuthem Afri<:n). Cllr.ll/llt/O/l. 200~ ;11 0:3385-
3397.
23, Connelly LB. DalaneinR the nunlber and size
of sites; un OCQnom;e approach II:> thlt optIlll;ll df!<llllt't
<'If clUAer samples, ControlClin Trials, '-(1(13;24:544..
559.
24. Casner M, Andol'5en D. l:illll~ SM. Th~ imp<<cl Q"
a new C:PR ~N.\iM. tlltvlClt Oil rate of return of sponta-
neous circulation in out.of.ho~I'ill "",rlji~" arr~sl, Pro-
/"'~P r.mel'gc.afe. 2005:S1:61-67.
25. E<<~l V~f/i;rm .ll1: .~/)/lw~,." for tl1/! DeS/CIl, ~/mu-
/(JtJon BIId IntJYim Monitorillt:o( "'!lx;hl~ (:II~iwl Trklls
U\l" M;J1Iu:tl. C:ilrnbridlle, M.iSS; cyteI software Corp:
2oo~.
26, Chen VH, D~Mcts Dl. Lan !l:K. In~reasinl:l th~
~amplll ~Izll When the unbl/nded interim result is
pl'llmi.inlt, Sllll MerJ. 2004;23:1023-1038.
27. Ven.hl"" WN. Ripley BO. Modern Applied 1fB-
IMil.. WiI), .~, 4110 ...1. N<'!w VorIc, NY: SpringeI' flub-
lishing Co; 2002.
28, RlJtcor$ WH, Regt'l!!lrJ(ll'l .~nd~rd error5 In clus-
tered samples. ~tata Tech Bull Rep. 1993;3:88-94.
29, Willi.ams RL. A note on robust variance estima-
~"" I", du."'r-,",rr~I"led data. Biometrics. 2000;!l6:
645-646.
liD. Wooldridge JM. Econometric Analysis of Cross
Sec.(iorl and Pan/11 O;tI8, l"~mbrld;.e, Ma~~: Mil "relS:
2002,
31, Vukmir R8. SUMVal and QuleQmt' from 1'"''''''-
pltaJ cardiac i1rrest.I1tIp:/lwww.i5!lub.com/ostialindex
.llhll?~mIFiI~P" th"j(lu,'n8IMlj rdm/voI4n.1/survlval
.)(1111. Accessibility verffied May 15, 2006.
32, Rea TO. Ei541nberll MS, Bcdcer U, Murray JA,
loIe~mp. T. Tflmporallrend, In sudden cardiac arrest
a 25-year emc:re:cncy medical ~ervices I"''''p'''~iv~.
CIrculatIon. 2003;107:2780-2785.
33. LairetJ, teeM. Acompi1I'i$011 Qfstiln<fard manual
c:ardlopulll1Onary resuscitation versus the AutoPulse
mec~I1t1k:a1 clll'dlopulmonary resuscitation device. Alln
fmers Med. 2005:46:S11~.
14. OI'I~llI J. AI. r'l,nerllenc.y m<!dlcal ""rvlce~ ~ys-
tem resuscitation stratelY: using AutoPulseThl as the
primary method for dclivcrin~ che~t wmp"".ilJn~ im-
proves survival fronl out.of.hospltal cardiac arrest (late-
b'litIIkUIA L!i.1 ptll,el.tatlon). I'resented at: AMerican
I' !I9I' I. AI",,:i.l:lon Re:lU5clta~on Science Symposium;
Novomber11-12, 2005; Dallas, Tex.
'Ii. ~mpl)p,1I JP, Maxey VA, Wat~on WA. Haw-
thomeeffcct; im~N~afl~~ ~hO"f'it;d ,..,...arch.llrm
F.mr:re; Moo, 1m,~6.6'Hl5. .
36. Anderson T. Vanden Hoek TL Pn!conditioning and
tI1e Qxidanru uf sudd~1'1 .IMI.h. OJfY ()fI/rl c.rn. c.are.
2003:9:194-1911.
117_ Black'tolle E, Morrtson M, koth MB, H2S In.
du~~s a suspt'O(lea .nlmal~iot't-lIke ~tate III mice.
Science. 2005:308:518. .
lB. Hart AI', A7.ar Vl, Hart 1<1{. stephens BG, Au-
topsy artifact created by the P.cvivarrt AutuPulge ~..
.lu'dtiltJon device. J forensIC SCI. :ZOO~:~O:164-168.
39. Wei:m,ldt MI., Beck~r LB. Rexu>citaUUfl after car-
diac a~st: a 3-phlL'lc timc,sQ",mv.. mild..!, !IIMA
2002j2H8~3()35-30311.
1:11006 Aomeric:1Ul Medical Assoc:lllllOll, AU rlghra ~Id.
Downloaded ftom www,jllmll.~'v'n 'Ill ~tlllt\'lmh..,. 10,2007