Content uploaded by Elisa Nurmi
Author content
All content in this area was uploaded by Elisa Nurmi on Feb 06, 2019
Content may be subject to copyright.
Intraosseous access in neonates and infants: risk of severe
complications –a case report
P. K. Suominen
1
, E. Nurmi
1
and K. Lauerma
2
1
Department of Anaesthesia and Intensive Care, Children’s Hospital, University of Helsinki and Helsinki University Hospital, Helsinki, Finland
2
HUS Medical Imaging Center, Radiology, Children’s Hospital, University of Helsinki and Helsinki University Hospital, Helsinki, Finland
Correspondence
P. K. Suominen, Department of Paediatric
Anaesthesia and Intensive Care, Children’s
Hospital, Helsinki University Hospital,
University of Helsinki, Stenb€
ackinkatu 11,
FI-00029 Helsinki, Finland
E-mail: pertti.suominen@hus.fi
Conflicts of interest
All the authors have no financial or other
relationships that might pose a conflict of
interest.
Submitted 20 July 2015; accepted 2 August
2015; submission 19 April 2015.
Citation
Suominen PK, Nurmi E, Lauerma K.
Intraosseous access in neonates and infants:
risk of severe complications –a case report.
Acta Anaesthesiologica Scandinavica 2015
doi: 10.1111/aas.12602
Gaining vascular access in a neonate during cardiopulmonary
resuscitation is crucial and challenging. Intraosseous (IO) access
can offer a fast and reliable method for achieving emergency
access for fluids and drugs when venous access fails in a critically
ill child. IO access can however result in rare, but serious adverse
events including compartment syndrome and amputation. We
describe a case resulting in leg amputation due to IO infusion in
a neonate after resuscitation and therapeutic hypothermia. We
compared 10 tibia X-rays in three age groups. The mean medul-
lary diameter of the proximal tibia at the recommended site for IO
access was 7 mm in neonate, 10 mm in 1- to 12-month-old
infants, and 12 mm in 3- to 4-year-old children. This provides a
narrow margin of safety for the correct positioning and the avoid-
ance of dislodgement of the IO needle. The correct position of the
IO needle should be confirmed by bone marrow aspiration and
fluid bolus. Unnecessary touching of the IO needle after fixing it
in place should be avoided by inserting a luer-lock catheter with
a three-way stop-cock for IO drug and fluid administration. Regu-
lar observation of the circulation and possible swelling of the
leg should be performed. The IO administration of inotropic infu-
sions should also be avoided after the initial resuscitation phase.
When treating with therapeutic hypothermia, it may be wise to
remove the IO needle much earlier than the currently recom-
mended 24 h because of the problems in peripheral circulation
and its monitoring.
The intraosseous (IO) route for fluid resuscita-
tion and administration of drugs was introduced
in 1922 and gained popularity during the fol-
lowing decades.
1
The use of IO access subse-
quently decreased with the advances in the
quality of peripheral and central IV catheters.
1
In recent decades, IO access has been estab-
lished as a method for gaining emergency access
for fluids and drugs when venous access fails in
a critically ill patient and its use is recom-
mended by international guidelines.
1–3
IO
access is recommended in cases for which three
attempts at peripheral venous access have failed
or when >90 s have been spent trying to secure
access.
2
The preferred site for IO access insertion
in neonates and children is the anteromedial
surface of the tibia 1–2 cm below the tibial
tuberosity, due to the easy accessibility of the
subcutaneous cortex.
1–4
The distal femur is
another recommended insertion IO site in neo-
nates and children.
1–4
A low overall complica-
tion rate of 1% has been reported.
1,3,4
Compartment syndrome is a rare, but recognized
complication, however.
1,3–9
We report a case
Acta Anaesthesiologica Scandinavica (2015)
ª2015 The Acta Anaesthesiologica Scandinavica Foundation. Published by John Wiley & Sons Ltd 1
CASE REPORT
resulting in leg amputation after IO infusion in
a neonate treated with therapeutic hypothermia
after resuscitation. A signed consent was
received from the parents for the case report.
Case report
A previously healthy 24-day-old boy with a cir-
culatory failure was transported to the emer-
gency room (ER) of a University Hospital. An
intravenous (IV) line was placed in the scalp
vein. Suddenly supraventricular tachycardia
(SVT) was converted to a brief episode of ven-
tricular tachycardia and finally pulseless electri-
cal activity. CPR was started immediately. Two
doses of adrenaline and Ringer acetate bolus
10 ml/kg was given before the IV line stopped
functioning. The extremities of the neonate were
cold and the peripheral veins were not visible.
IO access at the left proximal tibia was placed
by an anesthetist using a power-driven device
EZ-IO (Vidacare, San Antonio, TX, USA) and a
15 mm needle. Unfortunately, after a fluid bolus
of 5 ml, the calf of the neonate became swollen
and the IO needle was removed. Another
15 mm IO needle was placed into the left distal
femur. A return to spontaneous circulation was
achieved after 18 min of CPR. The femoral IO
access was functioning well until the baby
started moving his limbs, which caused the dis-
lodgement of the IO needle. Therefore, a third
(15 mm needle) IO access was placed into the
right tibia. Bone marrow was easily drawn into
a syringe and a fluid bolus was easily injected.
Furthermore, a good haemodynamic response to
adrenaline bolus was achieved. An arterial
catheter was subsequently placed into the right
femoral artery.
The neonate was transferred to a pediatric
intensive care unit, where he was intubated,
ventilated and an adrenaline-infusion was
administered via the IO needle. Therapeutic
hypothermia (32–34°C) was started and lasted
for 24 h. Several attempts to place a femoral
central line with ultrasound guidance failed.
Jugular or subclavian veins were not accessed
because of the risk of triggering a new episode
of SVT. Therefore, a continuous adrenaline-infu-
sion 0.15 lg/kg/min and noradrenaline-infusion
0.1 lg/kg/min were administered via the IO
needle instead of peripheral IV catheters.
After 24 h of therapeutic hypothermia, the
patient was slowly warmed over the next 12 h.
A serious complication of the right lower limb
was noticed 24 h after placement of the IO nee-
dle. The limb was cold and pale with a clear
demarcation line below the IO needle insertion
place was detected. The IO needle was removed
and anticoagulation with subcutaneous enoxa-
parin 0.5 mg/kg was started. Ultrasound-guided
central venous access was achieved in the right
femoral vein. Two days after the placement of
the IO needle epidermolysis of the skin of the
right distal limb was detected (Fig. 1). An
angiography and several fasciotomies of the
right limb muscles were performed in the opera-
tion room (OR). The circulation was thought to
be sufficient in the right limb. However, 5 days
after the placement of the IO needle during the
second operation all lower limb muscle com-
partments were found to be non-viable and an
amputation below the knee was performed. At
1 year of age the child had haemodynamically
recovered and betablocker medication has been
stopped uneventfully. The child has neurologi-
cally developed according to age and is able to
stand with the lower limb prosthesis and take
assisted steps.
Comparative medulla-diameter measurements
We compared 10 tibia X-rays in three age
groups which were: 1- to 28-day-old full-term
Fig. 1. Ischemia and epidermolysis of the right limb 2 days after the
insertion of the IO needle.
Acta Anaesthesiologica Scandinavica (2015)
2ª2015 The Acta Anaesthesiologica Scandinavica Foundation. Published by John Wiley & Sons Ltd
P. K. SUOMINEN ET AL.
neonates, 1- to 12-month-old infants, and 3- to
4-year-old children, archived in the database of
the X-ray department. Quantitative measure-
ments of the medullary diameter in antero-pos-
terior and lateral dimensions of these X-rays
were made at the recommended site of IO access
anteromedial surface of the tibia. Children with
syndromes or deformities in lower extremities
were excluded. The measurement site was 1 cm
and 2 cm below the proximal end of the tibia
for neonates and infants respectively, because, it
was not possible to visualize the tibial tuberos-
ity or epiphyseal plate in neonates and infants
in these X-rays (Fig. 2). The measurement site
in 3- to 4-year-old children was 1 cm below the
tibial tuberosity (Fig. 2). The results of the mea-
surements are shown in Table 1.
Discussion
Achieving venous or IO access is challenging
during neonatal resuscitation especially after the
first days after birth, when the readily accessible
umbilical vein is not available.
2,3
IO access can
offer a fast and safe method for gaining an emer-
gency access in a critically ill child.
1–4
The sinu-
soids and central venous canals in the
medullary cavity of long bones do not collapse
during hypovolemia or shock, unlike the
peripheral veins, which can be seen as a point
in favor of IO access.
1
Unfortunately, the mean
medullary diameter of the proximal tibia at the
recommended site for IO access is only 7 mm in
a neonates and 10 mm in infants (Table 1).
According to a previous study in newborns the
medullary diameter is only 2 mm at the mid-
point of the shaft of the tibia.
10
The length of
the tip of the smallest pediatric needle of the
EZ-IO device, used in the present case was
4 mm. The stylet and Y-shaped needle has to be
placed within the medullary cavity when insert-
ing the needle (Fig. 3). This makes the correct
placement of the IO needle in neonates and
infants and the avoidance of needle dislocation
extremely challenging.
Compartment syndrome is a rare complication
caused by extravasation of fluids and drugs.
Extravasation can be caused by incomplete IO
needle insertion or by penetration of the oppo-
site cortex of the tibia and by producing several
holes in the cortex when attempting IO access
and by needle dislocation after successful place-
ment.
7
In the present case, the first IO insertion
attempt failed but after their initially successful
placements the second and third IO needles
became dislocated during the treatment. Accord-
ing to previous reports semiautomatic devices
may reduce the IO needle insertion time and
the number of insertion attempts needed in
comparison to manual IO devices.
4
Semiauto-
matic devices provide also a more regular entry
site into the cortex, which may minimize the
risk of extravasation.
9
The major emphasis in
neonates and small children must be the correct
positioning of the IO needle and also the pre-
vention of dislodgement rather than the fast
insertion time. The needle is in the bone mar-
row when the needle enters the IO space and a
loss of resistance is detected. When using a
power-driven EZ-IO device in neonates, the
drill has to be stopped within the average mar-
gin of 3 mm so that the needle will remain in
the IO space and not penetrate the opposite cor-
tex. The manufactures of the IO devices such as
AB C
Fig. 2. X-rays of left proximal tibia in
antero-posterior dimension in (a) a neonate,
(b) an infant, and (c) a 3-year-old child.
Acta Anaesthesiologica Scandinavica (2015)
ª2015 The Acta Anaesthesiologica Scandinavica Foundation. Published by John Wiley & Sons Ltd 3
INTRAOSSEOUS ACCESS IN A NEONATE
the EZ-IO should consider providing different
sizes of IO needles that have been adjusted in
proportion to the patient’s size and that also cor-
responds to the bone marrow space instead of
providing a shorter version of the same caliber
of 15 gage adult needle for neonates and infants
(length 45 mm vs. 15 mm).
In hindsight, several predisposing and dis-
putable factors for this missed and grave
complication could be recognized. The main
issue was that the inotrope infusions were
administered at a slow rate via the IO needle
because the central line could not be placed.
Although, successful IO administration of con-
tinuous infusion of inotropes in an infant with
septic shock has been previously reported,
11
the
slow rate of inotrope infusions in the present
case led to the delayed recognition of signs of
extravasation. It can be speculated that in the
present case the partial dislocation of the
Y-shape tip of the IO needle was such that it
protruded outside of the tibial cortex, and this
may have caused a continuous small leak of the
inotropic infusion. In addition, vasoconstriction
of the lower extremity was further increased by
therapeutic hypothermia of 34°C, which delayed
the detection of poor circulation in the right
limb. The present case emphasizes to use the IO
access primarily for emergency purpose and the
importance of the use of ultrasound guidance
not only to achieve central venous but also
peripheral access.
13
Extravasation that caused compartment syn-
drome, which in the worst case scenario,
resulted in amputation of the leg has been
reported at least six times previously in case
reports,
1,3,4,12
but surprisingly it is not even
mentioned as a possible complication in some
review articles.
5–9
Almost all of the six previ-
ously reported cases of compartment syndromes
that lead to leg amputations have occurred in
neonates and infants, including one septic shock
survivor that had bilateral below knee amputa-
tions.
5–9
In addition to large amounts of fluids,
irritating solutions and drugs such as inotropes,
sodium bicarbonate, hypertonic fluids, and cal-
cium have been associated with these incidents
as in the present study.
5–7
It is highly likely that
the cases with severe compartment syndromes
are underreported, especially when taking into
consideration the most common scenarios of IO
needle placement such as pediatric cardiac arrest
and drowning, which have low survival rates.
This case is presented to raise awareness of
the importance of the correct positioning and
the avoidance of dislodgement of the IO needle
in neonates and infants due to the small medul-
lary space of the neonate tibia. The correct posi-
tion of the IO needle should be confirmed by
bone marrow aspiration and fluid bolus. Unnec-
essary touching of the IO needle after fixing it
in place should be avoided by inserting a luer-
Fig. 3. The 15 mm pediatric needle set of the power-driven device
EZ-IO (Vidacare, San Antonio, TX, USA) with and without the stylet.
The distance between two thin solid lines is 1 mm.
Table 1 Medullary diameter of tibia measured from X-rays at
the preferred site of intraosseous access in neonates and
children (N=30).
Age
Antero-posterior
diameter (mm) Lateral diameter (mm)
<1 month 7.7 0,4 7.4 0.7
1–12 months 9.9 1.4 9.9 1.5
3–4 years 12.4 0.8 10.7 1.2
Values are mean SD.
Acta Anaesthesiologica Scandinavica (2015)
4ª2015 The Acta Anaesthesiologica Scandinavica Foundation. Published by John Wiley & Sons Ltd
P. K. SUOMINEN ET AL.
lock catheter with a three-way stop-cock for the
IO drug and fluid administration. Regular
observation of the circulation and possible swel-
ling of the leg should be performed. The IO
administration of inotropic infusions should
also be avoided after the initial resuscitation
phase. When treating with therapeutic
hypothermia, it may be wise to remove the IO
needle much earlier than the currently recom-
mended 24 h because of the problems in
peripheral circulation and its monitoring.
Because of the low incidence of severe emergen-
cies in neonates and infants, the experience of
IO access insertion remains sparse among physi-
cians. Therefore, regular simulation-based train-
ing with a neonatal mannikin with a real-size
medullary diameter is highly recommended.
References
1. Tobias JD, Ross AK. Intraosseus infusions: a
review for the anesthesiologists with focus on
pediatric use. Anest Analg 2010; 110: 391–401.
2. Biarent D, Bingham R, Eich C, Lopez-Herce J,
Maconochie I, Rodriquez-Nunez A, Rajka T,
Zindeman D. European Resuscitation Council
Guidelines for Resuscitation 2010 Section 6.
Paediatric life support. Resuscitation 2010; 81:
1364–88.
3. Engle WA. Intraosseous access for administration of
medications in neonates. Clin Perinatol 2006; 33:
161–8.
4. Olaussen A, Williams B. Intraosseus access in the
prehospital setting: literature review. Prehosp
Disaster Med 2011; 28: 717–8.
5. Taylor CC, Clarke NMP. Amputation and
intraosseus access in infants. BMJ 2011;
342: d2778.
6. Vidal R, Kissoon N, Gayle M. Compartment
syndrome following intraosseous infusion.
Pediatrics 1993; 91: 1201–2.
7. Launay F, Katschburian M, Bourelle S, Jouve JL,
Bollini G. Leg amputation after intraosseous
infusion in a 7-month-old infant: case report. J
Trauma 2003; 55: 788–90.
8. Oesterlie GE, Petersen KK, Knudsen L, Henriksen
TB. Crural amputation of intraosseous needle
insertion and calcium infusion. Pediatr Emerg Care
2014; 30: 413–4.
9. Dasgupta S, Playfor SD. Intraosseous fluid
resuscitation in meningococcal disease and lower
limb injury. Pediatr Rep 2010; 2: e5.
10. Rodriquez JI, Palacios J, Rodriquez S. Transverse
bone growth and cortical bone mass in the human
prenatal period. Biol Neonat 1992; 62: 23–31.
11. Berg RA. Emergency infusion of catecholamines
into bone marrow. Am J Dis Child 1984; 138:
810–1.
12. Luck RP, Haines C, Mull CC. Intraosseous access. J
Emerg Med 2010; 39: 468–75.
13. Egan G, Healy D, O’Neill H, Clarke-Moloney M,
Grace PA, Walsh SR. Ultrasound guidance for
difficult peripheral venous access: systematic
review and meta-analysis. Emerg Med J 2013; 30:
521–6.
Acta Anaesthesiologica Scandinavica (2015)
ª2015 The Acta Anaesthesiologica Scandinavica Foundation. Published by John Wiley & Sons Ltd 5
INTRAOSSEOUS ACCESS IN A NEONATE
