Resuscitation 83 (2012) 107 112
Contents lists available at SciVerse ScienceDirect
Resuscitation
jo u rn al hom epage : www.elsevier.com/locate/resuscitation
Experimental paper
Pharmacokinetics of intraosseous and central venous drug delivery during
cardiopulmonary resuscitation ,
Stephen L. Hoskinsa, Paulo do Nascimento Jr.a,b, Rodrigo M. Limaa,b, Jonathan M. Espana-Tenorioa,
George C. Kramera,"
a
Resuscitation Research Laboratory, Department of Anaesthesiology, University of Texas Medical Branch, 301 University Blvd, Galveston, TX 77555-0801, United States
b
Sao Paulo Medical school, Department of Anesthesiology, Unesp, Botucatu, SP, Brazil
a r t i c l e i n f o a b s t r a c t
Article history:
We compared the pharmacokinetics of intraosseous (IO) drug delivery via tibia or sternum, with central
Received 27 January 2011
venous (CV) drug delivery during cardiopulmonary resuscitation (CPR).
Received in revised form 20 July 2011
Methods: CPR of anesthetized KCl arrest swine was initiated 8 min post arrest. Evans blue and indocyanine
Accepted 26 July 2011
green, each were simultaneously injected as a bolus with adrenaline through IO sternal and tibial needles,
respectively, n = 7. In second group (n = 6) simultaneous IO sternal and IV central venous (CV) injections
were made.
Keywords:
Results: Peak arterial blood concentrations were achieved faster for sternal IO vs. tibial IO administration
Intraosseous
(53 Ä… 11 s vs. 107 Ä… 27 s, p = 0.03). Tibial IO dose delivered was 65% of sternal administration (p = 0.003).
Cardiopulmonary resuscitation
Time to peak blood concentration was similar for sternal IO and CV administration (97 Ä… 17 s vs. 70 Ä… 12 s,
CPR
respectively; p = 0.17) with total dose delivered of sternal being 86% of the dose delivered via CV (p = 0.22).
Pharmacokinetics
Tracers Conclusions: IO drug administrations via either the sternum or tibia were effective during CPR in anes-
Drug delivery
thetized swine. However, IO drug administration via the sternum was significantly faster and delivered
a larger dose.
© 2011 Elsevier Ireland Ltd. All rights reserved.
1. Introduction in the field was 2 min for first attempts and 5 min when further
attempts were required.6 The overall success rate to establish an
Survival from out-of-hospital cardiac arrest depends on a IV line in the field for medical emergencies is less than 75%.6 8
sequence of therapeutic interventions termed the  chain of sur- There remains a need for more rapid vascular accesses for drug
vival by the American Heart Association (AHA). This sequence delivery during CPR may be one way of improving survival. Intra-
includes rapid access to emergency medical care, cardiopul- venous access during cardiopulmonary resuscitation (CPR) can be
monary resuscitation (CPR), defibrillation, advanced care, and post difficult even for an experienced caregiver. Intraosseous vascular
resuscitation techniques such as hypothermia, percutaneous coro- (IO) access is an established rapid, safe, and effective alternative
nary interventions, and implantable cardioverter-defibrilators.1,2 for peripheral intravenous drug delivery.8,9 The American Heart
Unfortunately, survival rates after cardiac arrests are dismal Association and the European Resuscitation Council Guidelines for
(2.5 10.5%).3 5 More rapid vascular accesses for drug delivery dur- Pediatric Life Support recommend IO access via the tibia for pedi-
ing CPR may be one way of improving survival. atric patients.12,13 In the last 10 years, several large bore IO needles
Intravenous access during CPR can be difficult even for an for adult patients have become available that use IO access via the
experienced caregiver. In one study, the median time required sternum, tibia and humerus. These devices have been evaluated in
to establish an intravenous (IV) line by well-trained paramedics both patients and animals.8,10,11 Use of these devices provides rapid
access to the systemic circulation during normovolemia.7,8,10,14
However, the effectiveness of IO drug delivery via different anatom-
ical sites during CPR has been under evaluation.

A Spanish translated version of the abstract of this article appears as Appendix
We used a swine model of cardiac arrest to determine the phar-
in the final online version at doi:10.1016/j.resuscitation.2011.07.041.
macokinetics of IO delivery of a double dye tracer method during

Financial support: American Heart Association Texas Affiliate Grant-in-Aid
CPR using simultaneous IO injections in the sternum and tibia. We
#0455157Y.
" also compared the pharmacokinetics of tracer administration via
Corresponding author. Tel.: [phone]; fax: [phone].
the sternum vs. central venous IV administration.
E-mail addresses: [e-mail], [e-mail] (G.C. Kramer).
0300-9572/$  see front matter © 2011 Elsevier Ireland Ltd. All rights reserved.
doi:10.1016/j.resuscitation.2011.07.041
108 S.L. Hoskins et al. / Resuscitation 83 (2012) 107 112
2. Methods Immediately following the injection of KCl the electrocardiogram
(EKG) displayed a typical ventricular fibrillation (VF) waveform.
2.1. Animal preparation Ventilator support was terminated at this time. Cardiac arrest was
followed by an 8-min period of untreated ventricular fibrillation.
The study protocol was approved by the University of Texas CPR was then initiated and delivered by a mechanical chest com-
Medical Branch s Institutional Animal Care and Use Committee pression device (Thumper® Michigan Instruments, Grand Rapids,
(IACUC). UTMB animal facilities are accredited by the American MI) at 100 compressions per min (without supplemental O2) and
Association for the Accreditation of Laboratory Animal Care. at duty cycle rate of 50%. A compression depth was set at 2-in. and
The experimental model was Yorkshire swine (25 35 kg). The chest compressions were delivered in an anterior/posterior posi-
night before the experiment food was withheld from the animals, tion centered on the sternal body. After 1-min of CPR pre-tracer
though they had free access to water. Pre sedation was induced arterial blood samples were taken. The volume of solution utilized
the day of the experiment by an intramuscular injection of telazol, was 1.5 ml followed by a 1.0 ml of saline flush.
ketamine, and xylazine. A 22 gauge peripheral intravenous catheter
was placed in the ear vein to deliver fluids and alpha chloralose. The 2.3. Tracers
animals were anesthetized for the surgical prep with 2 4% isoflu-
rane by facial mask and then intubated orotracheally using direct Evans blue (EB) (Sigma Aldrich, St. Louis, MO) 5.0 mg/ml, and
laryngoscopy. Animals were placed supine on a heating blanket to indocyanine green (ICG) (Alkorn, Buffalo Grove, IL) 2.5 mg/ml were
ć%
maintain body temperature between 38 and 39 C. Surgical areas used randomly in each site for the consecutive experiments as trac-
were scrubbed and covered with sterile surgical drapes. Mechani- ers to determine the relative arterial appearance times and dose
cal ventilation was established at a tidal volume of 15 20 ml/kg and delivered from the IO and central venous routes. Both ICG and EB
a ventilatory rate of 12 16 breaths/min to maintain end tidal car- dyes are inert and have no known biological activity. Each bolus of
bon dioxide between 30 and 40 mmHg. Thereafter, isoflurane was tracer contained 0.014 mg/kg of adrenaline (epinephrine). At 2-min
discontinued and anesthesia was maintained with an IV infusion of post CPR (0 time point) the tracers EB and ICG were co-administered
1% alpha chloralose via the catheter in the ear, administered as an simultaneously to the designated two paired sites in Protocol I
initial bolus of 50 mg/kg and sustained with a continuous infusion (sternal IO and tibial IO) and in Protocol II (central venous IV and
at 10 mg/kg/h. sternal IO). Rapid injection of the 2 3 ml of tracer solution was
The carotid artery was cannulated for arterial blood sampling immediately followed by a 1 ml flush to clear the needle. Arterial
via an incision of the right side of the neck. A central venous blood samples were taken every 10-s for 5 1/2 min and then at every
catheter was placed via the external jugular vein to provide dye 30-s for the remainder of the 8-min time period. After completion
tracer administration into the central venous circulation. Catheters of the study CPR was stopped and the animal was euthanized with
were placed into the aorta, via right femoral artery, and femoral a high dose of ketamine and KCl.
vein for acute monitoring and recording of mean arterial pres- Plasma tracer concentrations in arterial blood were determined
sures and drug delivery by sampling arterial blood, respectively. spectrophotometrically (Beckman Coulter DU 800 spectropho-
IO needles Jamshidi (Baxter, Deerfield, IL) or EZ-IO® (VidaCare, San tometer, Brea, CA) using absorbance wavelengths of 805 nm for ICG
Antonio, TX) were placed in the manubrium 5 cm caudal of the ster- and 620 nm for EB. Calibration standards of EB and ICG were pre-
nal notch, and at 3 cm distal of the tibial tuberosity, respectively. pared in plasma and used to calculate the concentrations of EB and
Correct placement was confirmed by cross section at necropsy. Lac- ICG from arterial blood samples. The area under the curve (AUC) of
tated Ringer s solution was administered at a rate of 15 ml/kg/h arterial tracer concentration divided by the tracer dose was used as
during surgery. Standard hemodynamics were monitored (Hewlett a measure of the drug delivered to the systemic circulation during
Packard, Andover, MA) throughout the experiments. Data were the first 8 min after drug injection (0 480 s). The ratio of the AUC
recorded via a multi channel analog-digital data acquisition pro- for both tracers was used as a measure of the relative drug delivery.
gram using PowerLab (AD Instruments, UK).
2.4. Statistics
2.2. Protocol
Summary data are expressed as means Ä… standard error of the
Two protocols were employed with simultaneous injections; mean (SEM). To test for differences of appearance times a paired
both of them were terminal studies. Protocol I (sternum vs. Student s t-test was conducted. Correlation coefficients for the rela-
tibia) compared the pharmacokinetics of two different dye tracers tionship of mean arterial pressure (MAP) to appearance time were
administered intraosseously and simultaneously via the sternum calculated utilizing Sigma plot software (Systat Software Inc., Ver-
and the tibia, respectively. Protocol II (sternum vs. central venous sion 11, San Jose, CA). A two-sided alpha level of significance of
IV) compared the pharmacokinetics of IO administration of dye <0.05 was used for assessing statistical significance.
tracers via the sternum with a simultaneous administration via
central venous IV. A 60-min baseline time period was established 3. Results
after completion of instrumentation. Lactate and blood gas vari-
ables were monitored to ensure that the animals had sufficiently Data on appearance time and dose delivered for all individual
recovered from the surgical procedure and reached a physio- animals and groups are presented in figures and tables.
logic baseline before experimental data was collected. Heparin,
10,000 units was administered IV prior to the induction of cardiac 3.1. Appearance times
arrest. During low flow states such as cardiac arrest, blood sampling
can be difficult if the lines become clotted. Prior to the induction Fig. 1(A and C) and Table 1 display data for each experiment
of cardiac arrest, the animals were administered a ketamine bolus of appearance times calculated in seconds, between injection and
(30 mg/kg) to achieve a deeper anesthesia plane and avoid any dis- time to peak tracer concentration, in Protocol I sternal IO and tibial
tress during the cardiac arrest and resuscitation. IO injections (n = 7). Mean time to maximum concentration was
Cardiac arrest was induced by rapid IV administration of 10 ml 53 Ä… 11 s for the sternal injection compared to 107 Ä… 27 s the tibial
of saturated potassium chloride (KCl) (Hospira Inc., Lake Forest, IL) injection. The range was from 20 to 90 s and 40 to 240 s for the
solution via central venous catheter followed by a 10 ml saline flush. sternal and tibial routes, respectively (p = 0.03). Time to half (50%)
S.L. Hoskins et al. / Resuscitation 83 (2012) 107 112 109
Fig. 1. The two upper graphs show appearance times of tracers vs. time: Protocol-I (tibial IO vs. the sternal IO): appearance times of tracers tibia (Graph-A) vs. sternum
(Graph-C). Concentrations were normalized in this figure to the maximal concentration in order to better visualize time differences to peak concentration. The two lower
graphs show dose delivered to the arterial blood calculated as dose injected (mg) by aortic blood concentration ( g/ml) for the same protocol tibia (Graph-B) and sternum
(Graph-D).
maximum concentration was 22 Ä… 3 s using the sternal route and 3.2. Dose delivered
50 Ä… 8 s for the tibial route (p = 0.006).
Fig. 2(A and C) and Table 2 show the appearance times of trac- Dose delivered was determined by using an area under the curve
ers for Protocol II, sternal IO and central venous IV injections (n = 6). analysis (AUC) for aortic concentration divided by injected dose.
Mean peak time to the maximum tracer concentrations after simul- Fig. 1(B and D) and Table 3 show the doses of tracer delivery to
taneous injections, via IO and central vein were not significantly the aortic blood, for each animal of Protocol I, calculated as AUC.
different 97 Ä… 17 s and 70 Ä… 12 s, respectively (p = 0.17). Times for The ratio of the AUC between Protocol I (tibial IO vs. sternal IO) is
tracers to reach their 50% maximal concentrations were 36 Ä… 4 s a measure of the relative effectiveness of dose delivery via the two
for sternal IO and 30 Ä… 4 s for the central vein routes (p = 0.06). routes. The tibial IO route delivered less dose to the arterial blood or
Fig. 2. The two upper graphs show appearance times of tracers vs. time: Protocol-II (sternal IO vs. central venous IV): appearance times of tracers central venous (Graph-A)
vs. sternum (Graph-C). Concentrations were normalized in this figure to the maximal concentration in order to better visualize time differences to peak concentration. The
two lower graphs show dose delivered to the arterial blood calculated as dose injected (mg) by aortic blood concentration ( g/ml) for the same protocol central venous
(Graph-B) and sternum (Graph-D).
110 S.L. Hoskins et al. / Resuscitation 83 (2012) 107 112
Table 1 Table 3
Appearance times in seconds from injection to maximum tracer concentrations and Dose delivered for tibial vs. sternal IO injections calculated as area under the curve
half (50%) maximal concentration. for aortic concentration g/ml divided by dose injected (mg) over 480 s after injec-
tion. The relative effectiveness of the two routes is shown as a ratio of the area under
Tibial IO vs. sternal IO injection
the curve (AUC), tibial IO divided by sternal IO.
Animal (n = 7) Peak concentration* 50%Peak concentrationż Relative dose delivered of tracers (Tibial IO vs. sternal IO injection AUC0 480 )
s
Sternum Tibia Sternum Tibia
Animal AUC* ( g s/ml) Ratio
86 80 110 36 57
Sternum Tibia Tibia/sternum
21 90 150 22 68
18 80 240 25 85 21 912 450 0.49
34 20 40 15 25 18 776 382 0.49
35 30 100 18 50 34 601 400 0.67
39 20 50 22 33 35 645 368 0.57
36 50 60 13 35 39 509 423 0.83
Mean 53 107 22 50 36 511 418 0.82
SEM 11 27 3 8 86 783 545 0.70
CI 30 75 55 158 16 27 34 65 Mean 677 427 0.65
SEM 57 22 0.05
CI, confidence interval (confidence level = 95%); SEM, standard error of the mean.
CI 564 789 383 470 0.6 0.7
*
p = 0.03  peak concentration  tibia vs. sternum.
ż
CI, confidence interval (confidence level = 95%); SEM, standard error of the mean.
p = 0.006  50% peak concentration  tibia vs. sternum.
*
p = 0.003  comparison between AUC0 480  tibia vs. sternum.
65 Ä… 5% as compared with the sternal route, mean AUC s difference
was statically significant (p = 0.003).
success rates for the pediatric patient population 18 65%.16,17 A
Fig. 2(B and D) and Table 4 show the actual values and ratio of the
prospective study of successful prehospital IV placement in 583
AUC between Protocol II (sternal IO vs. central venous IV). The ster-
patients showed that the success rate at first attempt was 74% (368
nal IO route was 86 Ä… 10% as effective as the central venous route
patients).6
in tracer delivery, although the mean AUCs were not significantly
Physicians have long sought alternate routes for the rapid
different (p = 0.22).
administration of drugs during cardiac emergencies, circulatory
shock, and low flow states. The endotracheal route is often used
4. Discussion as a convenient and rapid alternative for IV delivery of selected
drugs. However, efficacy of endotracheal delivery of drugs is
To the best of our knowledge the present study is the first to use controversial.18,19 The IO route provides access to systemic circu-
a double tracer technique to assess effectiveness of simultaneous lation via the bone marrow cavity which provides a noncollapsible
drug delivery, during CPR into two IO sites. delivery point into the central circulation for emergency infu-
Overall the study demonstrated that the intraosseous (IO) route sions and for drug delivery in the operation room setting.20
is an effective means of delivering drugs during CPR for tibia and Current American Heart Association guidelines and the Interna-
sternum IO sites. tional Resuscitation Council Guidelines recommend the IO route
Peripheral IV lines are the most commonly used routes for as first vascular access in pediatric emergencies such a cardiac
drug delivery by EMS personnel. An absence of venous blood flow arrest.13 21 For adult cardiac arrest IO is the first alternative when
and low pressure during cardiac arrest can lengthen the time to intravenous access is delayed or impossible.13,22 The success rate
obtain peripheral IV access and delay critically needed drug ther- when IO access is used is 81 100%8,10,11 and the time to establish
apy. Experienced medics can achieve IV access rapidly under ideal a IO line varies between 20 s and 1.5 min.8,10,23 The most common
conditions. However, prehospital conditions in the field transport adverse effect associated with IO infusion is extravasation and this
to hospital, and the skill levels of medics can vary widely. Clini- complication has been reported in 12% of patients.24 Compartment
cal studies have shown that peripheral IV access times can range syndrome, osteomyelitis, and tibial fracture are rare, but have been
from 2 to 49 min.6 8,15 The success rate for establishing periph- reported.9,24,25
eral IV access after cardiac arrest and difficult IV is variable and
ranges broadly between 30 and 75% in adult6 8 patients, with lower
Table 4
Dose delivered for sternal IO versus central venous IV injections calculated as area
Table 2 under the curve for aortic concentration g/ml divided by dose injected (mg) over
Appearance times in seconds from injection to maximum tracer concentrations and 480 seconds after injection. The relative effectiveness of the two routes is shown as
half (50%) maximal concentration. a ratio of the area under the curve (AUC), sternal IO divided by central venous IV.
Sternal IO vs. central venous IV injection Relative dose delivered of tracers (sternal IO vs. central venous IV
injection AUC0 480 s)
Animal (n = 6) Peak concentration 50%Peak concentration
Animal AUC g s/ml Ratio
Sternum IV Sternum IV
IV Sternum Sternum/IV
87 100 50 36.4 24
89 70 50 34 23 89 694 589 0.85
105 60 50 29 28 105 855 939 1.10
95 110 110 52 48 95 879 805 0.92
110 70 90 28 27 110 854 783 0.92
92 170 110 38 36 92 956 923 0.97
Mean 97 70 36 30 87 934 385 0.41
SEM 17 12 4 4 Mean 862 737 0.86
CI 64 129 45 94 28 42 22 37 SEM 38 87 0.10
CI 788 935 566 907 0.7 1.0
p = 0.17  peak concentration  sternum vs. central venous infusion.
p = 0.06  50% peak concentration  sternum vs. central venous infusion. p = 0.22  comparison between AUC0 480  sternum vs. central venous infusion.
CI, confidence interval (confidence level = 95%); SEM, standard error of the mean. CI, confidence interval (confidence level = 95%); SEM, standard error of the mean.
S.L. Hoskins et al. / Resuscitation 83 (2012) 107 112 111
Voelckel et al. showed that bone marrow blood flow was through the marrow, out of the injection site and into the venous
reduced by 70 80% after hemorrhage.26 During CPR the bone mar- circulation.
row flow is expected to be lower than in hemorrhagic shock. Sato The mean dose delivered via the tibial route was 65% and 53%
et al. and Del Guercio et al. showed in dogs and humans, respec- of the drug delivery via the sternum and central venous route,
tively that during CPR the cardiac output is only approximately respectively. However, even for the tibial route the half maxi-
20 30% of normal.27,28 In our study mean aortic appearance times mal concentrations were achieved in less than 1 min. Andropoulos
to the peak concentration of the tracer was 97 Ä… 17 s for the sternal et al. used HPLC analysis for the determination of tibial adrenaline
IO route which was not statistically significant (p = 0.17) compared delivery during CPR in lambs. The authors determined that the
to 70 Ä… 12 s for central venous route. Barsan et al. showed similar maximum arterial plasma adrenaline concentrations were similar
result in dogs with mean time to peak times for central venous between central venous and tibial IO delivery. However, they noted
infusion of 84 s with range between 53 and 100 s.29 Kuhn et al. reduced appearance time, after central venous administration com-
showed that the peak concentration of dye obtained with central pared to tibial IO injection after adrenaline injection.36
venous injection of indocyanine green during CPR in humans was Our measurements of appearance times and doses delivered,
at 30 s. However, only three patients were included on the study.30 coupled with an additional one or more minutes for establishing
Emerman et al. demonstrated in dogs that the interval of central a peripheral IV, suggest that even when using the slower tibial
venous injection to first appearance of the indocyanine green dur- IO route, one would effectively deliver drugs into the arterial cir-
ing CPR was 37 Ä… 17 s.31 Zuercher et al. showed mean time from culation during CPR in a shorter time than the time needed to
adrenaline injection to peak coronary perfusion of 60 Ä… 6 s when successfully start a peripheral IV. As such, the tibial IO route is
the drug was delivered via IO vs. 43 Ä… 4 after IV injection during both an efficacious and rapid means of delivering drug therapy dur-
CPR.32 These results are similar to our finding of time to the 50% ing CPR. The size of the saline bolus after the drug infusion may
peak concentration, i.e. central venous (30 s), sternal (22 s Protocol also have an important role on the time for maximum concentra-
I; 36 s Protocol II), and tibia (50 s). tion of the dye. If we had used a larger flush the effectiveness of
Some factors can affect the appearance times and the dose the IO tibial delivery may have increased. Wenzel et al. demon-
delivery in this study. One is that sternum is located closer to strated comparable vasopressin plasma level and hemodynamic
the central circulation when compared with the tibia location, variables when the drug was delivered both by the intravenous
which may facilitate the faster appearance of the drug on the and the tibial IO routes during CPR. However, the authors infused
systemic circulation when the drug is delivered into the ster- 20 ml of saline bolus compared with 1.0 ml used in the present
num. Second, there is a difference of blood perfusion between study.37
the two bones. It is likely that the sternum perfusion is better Based on the present data, we recommend that sternal IO route
than the tibia perfusion and this may facilitate the absorption of be considered as the first choice of drug delivery during CPR when
the drug to the systemic circulation. Gross et al. showed a wide IV access has not been established, and that the tibial IO route
heterogeneity of bone blood flow comparing hematopoietic can- is also justified as second choice. The practical choices of which
cellous bones (red marrow) such as sternum, rib, ilium, and femur route to use in adults also depend on which IO devices are avail-
epiphysis (24 ml min-1 100 g-1) vs. nonhematopoietic bones (yel- able. There are currently 6 adult IO devices allowed for marketing
low marrow) such as tibia and mandible (2 ml min-1 100 g-1). The by the Food and Drug Administration (FDA). This includes two IO
authors also described a significant decrease in blood flow and devices for adult sternal access (FAST1 (Pyng Medical Corp., Rich-
an increase in vascular resistance in bone during hemorrhagic mond, BC, Canada) and Sternal EZ-IO (Vidacare Corp., San Antonio,
hypotension.33 TX)) and four IO devices for tibial access (SurFast (Cook Criti-
A key point during the CPR maneuvers is the quality of the cal Care, Bloomington, IN), Jamishidi (Baxter Allegiance, McGraw
chest compressions. To give effective chest compression is impor- Park), Bone Injection Gun (B.I.G., Waismed, Houston, TX), EZ-IO
tant that the rescuers or the devices used to perform the CPR push (Vidacare Corp., San Antonio, TX)).9,38,39 In pediatric patients, stan-
hard (e"5 cm) and fast (e"100/min).22 The chest should be allowed to dard butterfly needle, spinal needle, and pediatric versions of adult
recoil freely after each compression. Besides, approximately equal IO needles can be used. Most recently the humerus has been sug-
compressions and relaxation times should be used and interrup- gested as a route for IO delivery. Further work will be required to
tions in chest compressions should be minimized. If these chest assess the relative success of this route vs. the sternal and the tibial
compressions are not effective all the circulatory blood flow can be route.
affected including the bone marrow flow.22,34 Any anatomic dif- There are limitations to our study. First, swine are not humans
ference between the animals or any other factor that impair the and conclusive extrapolation to human patient responses cannot
dynamic of the chest compressions might result in differences in be made. The shape of the pig thorax is different from the human
cardiac output during this period, which might consequently delay thorax. In pigs, the ventricles are positioned in the center of the tho-
the appearance time of tracers on the systemic circulation. racic cavity, surrounded by lung tissues on all sides. In humans, the
The dose delivered of tracer via the IO route was similar to that right ventricle is positioned just under the sternum. This anatomic
delivered by central venous route. The sternal IO route delivered difference makes it more difficult to get a compression effect on
86% of the tracer to the aorta compared with central vein drug deliv- the heart of pigs. Chest compressions in pigs increase intratho-
ery. However, in one animal, the ratio between sternum/central racic pressure (thoracic pump mechanism), which in turns affects
venous infusions was 0.41 (Table 4). When we exclude this outlier the heart. In humans we have not only the thoracic pump effect
data point from the analysis, the resultant sternum dose delivered but also the direct heart pump mechanism affecting the heart by
via the route was 95% that of the central venous. The effectiveness of chest compression.34 Moreover, we did not measure the plasma
the IO sternal route for drug delivery during CPR may be due to one concentrations of adrenaline. We used dye tracers as a surrogate
or more factors. The red bone marrow of the sternum could pro- of drug delivery in place of the biologically active drug. However,
vide sufficient blood flow for rapid delivery of drugs to the great measurement of adrenaline would preclude comparison of simulta-
veins. Further, chest compressions may facilitate the drug egress neous injections. The significant variability of cardiac output during
out of the marrow and into the vasculature.35 Alternatively, the CPR results in an animal to animal variability of time to peak con-
IO delivery of tracer may be independent of marrow blood flow. centration and dose delivered; while simultaneous 2 tracer paired
It may be that a 1.5 ml bolus of tracer followed by the 1 ml flush studies provides for great precision for comparing differences. Fur-
used in our study is sufficient volume to advance most of the tracer ther, high background levels of endogenous adrenaline during CPR
112 S.L. Hoskins et al. / Resuscitation 83 (2012) 107 112
make precise assessment exogenous drug epinephrine impossible. 13. Pediatric basic and advanced life support 2010 International Consensus on Car-
diopulmonary Resuscitation and Emergency Cardiovascular Care Science with
Our study suggests that either bone marrow blood flow or the vol-
Treatment Recommendations Part-10. Resuscitation 2010;81:e213 59.
ume of injectate, or both, are sufficient for tracer delivery through
14. Von Hoff DD, Kuhn JG, Burris HA, Miller LJ. Does intraosseous equal intravenous?
the emissary veins to the superior vena cava. We studied young pigs A pharmacokinetic study. Am J Emerg Med 2008;26:31 8.
15. Stein J, George B, River G, Hebig A, McDermott D. Ultrasonograpically guided
with healthy hearts and peripheral vessels, while clinical ventric-
peripheral intravenous cannulation in emergency department patients with
ular fibrillation occurs largely in older patients with some amount
difficult intravenous access: a randomized trial. Ann Emerg Med 2009;54:33 40.
of peripheral artery disease. The pig is the most often used animal
16. Doninger SJ, Ishimine P, Fox JC, Kanegaye JT. Randomized controlled trial
of ultrasound-guided peripheral intravenous catheter placement versus tra-
model of cardiac arrest and CPR.26,37 Finally, data on tibial IO injec-
ditional techniques in difficult-access pediatric patients. Pediatr Emerg Care
tions in swine with their short legs may not be comparable to that
2009;25:154 9.
of adult humans with longer legs farther from the heart. Blood flow
17. Brunette D, Fischer R. Intravascular access in pediatric cardiac arrest. Am J Emerg
Med 1988;6:577 9.
in the leg and bone marrow cavities below the diaphragm could be
18. Orlowski JP, Gallagher JM, Porembka DT. Endotracheal epinephrine is unreliable.
less in humans than in pigs during CPR.
Resuscitation 1990;19:103 13.
19. Caen AR, Reis A, Bhutta A. Vascular access and drug therapy in pediatric resus-
citation. Ped Clin N Am 2005;55:909 27.
5. Conclusions
20. Joseph G, Tobias JD. The use of intraosseous infusions in the operating room. J
Clin Anesth 2008;20:469 73.
Both tibial and sternal IO routes are an effective means of deliv-
21. Pediatric advanced cardiovascular life support: 2010 American Heart Associa-
ering life saving drugs during CPR. Dye tracers delivered via tibial tion guidelines for cardiopulmonary resuscitation and emergency cardiovascu-
lar care Part 14. Circulation 2010;122:S876 908.
IO or sternal IO routes of anesthetized swine reached maximal con-
22. Adult advanced cardiovascular life support: 2010 American Heart Association
centrations in the arterial blood during CPR in less than 2 min with
guidelines for cardiopulmonary resuscitation and emergency cardiovascular
both, a faster and a greater dose delivered using the sternum route care Part 8. Circulation 2010;122:S729 67.
23. Lamhaut L, Dagron C, Aprotesei R, et al. Comparison intravenous and
than with the tibial route. Sternal IO and central venous routes are
intraosseous access by pre-hospital medical emergency personnel with and
not different considering pharmacokinetics of tracers during CPR
without CBRN protective equipment. Resuscitation 2010;81:65 8.
in swine. 24. Fiorito BA, Mirza F, Doran TM, et al. Intraosseous access in the setting of pediatric
critical care transport. Pediatr Crit Care Med 2005;6:50 3.
25. Rosetti VA, Thompson BM, Miller J, Mateer JR, Aprahamian C. Intraosseous infu-
Conflict of interest
sion: an alternative route for pediatric intravascular access. Ann Emerg Med
1985;14:885 8.
26. Voelckel W, Lurie K, McKnite S, et al. Comparison of epinephrine with vaso-
Dr. Kramer is an inventor on patents for intraosseous technolo-
pressin on bone marrow blood flow in an animal model of hypovolemic shock
gies and a compensated consultant to Vidacare 2007 2010.
and subsequent cardiac arrest. Crit Care Med 2001;29:1578 92.
27. Sato S, Okubo N, Satsumae T, et al. Arteriovenous differences in PCO2 and cardiac
output during CPR in the dog. Resuscitation 1994;27:255 9.
References
28. Del Guercio LMR, Coomaraswany R, State D. Cardiac output and other hemody-
namic variables during external massage in man. N Engl J Med 1963;269:1398.
1. Cummins RO, Ornato JP, Thies WH, Pepe PE. Improving survival from sudden
29. Barsan WG, Levy RC, Weir H. Lidocaina levels during CPR. Ann Emerg Med
cardiac arrest: the  chain of survival concept. Circulation 1991;85:1832 47.
1981;10:73 8.
2. Nichol G, Aufderheide TP, Eigel B, et al. Regional systems care for out-of-hospital
30. Kuhn GJ, White BC, Swetneam RE, et al. Peripheral vs central circulation times
cardiac arrest. Circulation 2010;121:709 29.
during CPR: a pilot study. Ann Emerg Med 1981;10:417 9.
3. Hollenberg J, Bang A, Lindqvist J, et al. Difference in survival rates after out-of-
31. Emerman CL, Pinchak AC, Hagen JF, Hancock DE. Dye circulation times during
hospital cardiac arrest between the two largest cities in Sweden: a matter of
cardiac arrest. Resuscitation 1990;19:53 60.
time? J Intern Med 2005;257:247 54.
32. Zuercher M, Kern KB, Indik JH, et al. Epinephrine improves 24-hour survival
4. Rea TD, Cook AJ, Stiell IG, et al. Predicting survival after out-of-hospital cardiac
in a swine model of prolonged ventricular fibrillation demonstratins that early
arrest: role of Utstein data elements. Ann Emerg Med 2010;55:249 57.
intraosseous is superior to delayed intravenous administration. Anesth Analg
5. Olasveengen TM, Sunde K, Brunborg C, Thowsen J, Steen PA, Wik L. Intra-
2011;112:884 90.
venous drug administration during out-of-hospital cardiac arrest. JAMA
33. Gross PM, Heistad DD, Marcus ML. Neurohumoral regulation of blood flow to
2009;302:2222 9.
bones and marrow. Am J Physiol 1979;237:h440 8.
6. Lapostolle F, Catineau J, Garrigue B, et al. Prospective evaluation of peripheral
34. Liao Q, Sjoberg T, Paskevicius A, Wolfart B, Steen S. Manual versus mechanical
venous access difficulty in emergency care. Intensive Care Med 2007;33:1452 7.
cardiopulmonary resuscitation. An experimental study in pigs. BMC Cardiovasc
7. Constantino T, Parikh A, Satz WA, Fojtik JP. Ultrasonography-guided periph-
Disord 2010;10:53.
eral intravenous access versus traditional approaches in patients with difficult
35. Warren DW, Kissoan N, Mattar A, Morrissey G, Gravelle D, Rieder M. Pharma-
intravenous access. Ann Emerg Med 2005;46:456 61.
cokinetics from multiple intraosseous and peripheral intravenous site injections
8. Paxton JH, Knuth TE, Klausner HA. Proximal humerus intraosseous infusion: a
in normovolemic and hypovolemic pigs. Crit Care Med 1994;22:838 43.
preferred emergency venous access. J Trauma 2009;67:606 11.
36. Andropoulos DB, Soifer SJ, Schreiber MD. Plasma epinephrine concentrations
9. Buck ML, Wiggins BS, Sesler JM. Intraosseous drug administration in chil-
after intraosseous and central venous injection during cardiopulmonary resus-
dren and adults during cardiopulmonary resuscitation. Ann Pharmacother
citation in the lamb. J Pediatr 1990;116:312 5.
2007;41:1679 86.
37. Wenzel V, Lindner KH, Augenstein S, et al. Intraosseous vasopressin improves
10. Ong MEH, Chan YH, Oh JJ, Ngo ASY. An observational, prospective study com-
coronary perfusion pressure rapidly during cardiopulmonary resuscitation in
paring tibial and humeral intraosseous access using the EZ-IO. Am J Emerg Med
pigs. Crit Care Med 1999;27:1565 9.
2009;27:8 15.
38. Calkins MD, Fitzgerald G, Bentley TB, Burris D. Intraosseous infusion devices: a
11. Shavit I, Hoffmann Y, Galbraith R, Waisman Y. Comparison of two mechanical
comparison for potential use in special operations. J Trauma-Inj Infect Crit Care
intraosseous infusion devices: a pilot, randomized crossover trial. Resuscitation
2000;48:1068 74.
2009;80:1029 33.
39. Tobias JD, Ross AK. Intraosseous infusions: a review for the anesthesiologist with
12. Pediatric advanced life support: 2010 American Heart Association guideline for
focus on pediatric use. Anesth Analg 2010;110:391 401.
cardiopulmonary resuscitation and emergency cardiovascular care. Pediatrics
2010;126:e1361 99.