British Journal of Anaesthesia 84 (1): 59–62 (2000)

Hyperbaric bupivacaine for spinal anaesthesia in 7–18 yr old

children: comparison of bupivacaine 5 mg ml–1 in 0.9% and 8%
glucose solutions
H. Kokki* and H. Hendolin

Department of Anaesthesiology and Intensive Care, Kuopio University Hospital,

PO Box 1777, FIN-70211 Kuopio, Finland
*Corresponding author

We have compared two hyperbaric bupivacaine solutions for spinal anaesthesia in 7–18-yr-old
school-aged children in a double-blind, randomized, parallel group, prospective study. Children
were premedicated with diazepam orally. Half of the patients were sedated with either
midazolam or thiopental. After lumbar puncture with a 27-gauge spinal needle, bupivacaine
5 mg ml–1 in either 0.9% or 8% glucose was injected in a dose of 0.3 mg kg–1. Maximum
cephalad spread and regression of block were tested by transcutaneous electrical stimulation.
Success rate, spread and duration of sensory block were similar in both groups. The highest
median level of sensory block was T4 (10–90th percentiles T1–T7) in the 0.9% glucose group
and T4 (T1–T5) in the 8% glucose group. Time to two segment regression of block was 83
(50–143) min in the 0.9% glucose and 85 (53–150) min in the 8% glucose group. The incidence
of adverse effects was similar. Six children were given etilefrin to treat hypotension and six
atropine for bradycardia. Nausea was associated with a high level of block. Shivering was
detected in 16 children.
Br J Anaesth 2000; 84: 59–62
Keywords: anaesthetic techniques, subarachnoid; anaesthetic local, bupivacaine; children;
anaesthesia, paediatric; physics, baricity; complications, shivering; vomiting, nausea
Accepted for publication: June 23, 1999

Spinal anaesthesia is popular in both small children1 and
elderly people but less attention has been paid to school-
age children. In Kuopio University Hospital, spinal anaes-
thesia has been used in 5000 paediatric patients over the
past 10 yr. Spinal anaesthesia produces rapid onset, profound
and uniformly distributed analgesia with good neuromuscu-
lar block. The amide local anaesthetics (bupivacaine and
lidocaine) are used regularly, and spinal anaesthesia allows
the use of a small dose with a low risk of systemic toxicity.
Baricity (weight of anaesthetic solution in relation to
weight of cerebrospinal fluid (CSF)) is one of the most
important factors claimed to influence distribution of local
anaesthetic solutions in CSF.2 Solutions administered most
frequently are hyperbaric3 as they produce more predictable
block in both adults4 and children.5 Spinal anaesthetic
solutions are rendered hyperbaric by adding glucose. How-
ever, commercially marketed solutions contain 8% glucose
giving a density far in excess of that required to render
them hyperbaric.
Studies in adults have found that addition of a small
amount of glucose to increase the baricity of bupivacaine
solution just into the hyperbaric range improved predictabil-
ity of spinal block.4 6 Different hyperbaric bupivacaine
solutions have not been compared previously in children.
In this study, we have evaluated the role of baricity in
block characteristics. Two different hyperbaric bupivacaine
solutions (0.9% glucose and 8% glucose) were compared.
Patients and methods
The study was approved by our Ethics Committee and was
conducted in accordance with the Declaration of Helsinki.
Both children and parents gave informed consent. We
studied 107 children, ASA I–II, aged 7–18 yr, undergoing
day-case surgery below the umbilicus. Patients with a
known contraindication to spinal puncture, such as increased
intracranial pressure, haemorrhagic diathesis or infection at
the puncture site, were excluded. Children with a neuro-
logical disorder, or allergy to bupivacaine or other local
anaesthetics were also excluded. Data were collected
between December 1997 and February 1999 and some of
the results of these patients have already been published.7
We used a double-blind, randomized, parallel group,
prospective study design. Patients were allocated randomly
(computer-generated) to receive spinal anaesthesia with

© The Board of Management and Trustees of the British Journal of Anaesthesia 2000
 Kokki and Hendolin
either bupivacaine 5 mg ml–1 in 8% glucose (Marcain
Spinal, Astra, Södertelje, Sweden) (bupivacaine–8% glucose
group) or bupivacaine 5 mg ml–1 in 0.9% glucose (bupiva-
caine–0.9% glucose group). The dose of 0.3 mg kg–1 was
based on our previous study of spinal anaesthesia in
children.5 Commercial ampoules of hyperbaric bupivacaine
5 mg ml–1 (in 8% glucose) and isobaric bupivacaine
5 mg ml–1 (in normal saline) with erased drug names were
prepared and coded in the hospital pharmacy. Their identical
appearance allowed blinding of the study. Local anaesthetic
(4 ml) from a coded ampoule was mixed with 0.5 ml of
hyperbaric bupivacaine. If the coded ampoule contained
hyperbaric bupivacaine, mixing resulted in bupivacaine in
8% glucose; if the coded ampoule contained isobaric
bupivacaine, this resulted in bupivacaine in 0.9% glucose.
The baricity of bupivacaine in 8% glucose was 1.025
and that of bupivacaine in 0.9% glucose 1.005 at room
temperature.
Children were not allowed solid food for 6 h but clear
fluids were allowed up to 2 h before induction of anaesthesia.
Each child was premedicated with diazepam 0.5 mg kg–1
orally, up to a maximum dose of 10 mg. EMLA cream was
used at the lumbar and venepuncture sites. An infusion of
0.9% saline was given at a rate of 5–10 ml kg–1 h–1.
Ketoprofen 2 mg kg–1 i.v. or rectal paracetamol 40 mg kg–1,
or both, were given for pre-emptive pain treatment.
Intraoperative monitoring consisted of non-invasive arter-
ial pressure measurements every 5 min, continuous ECG,
ventilatory frequency, peripheral arterial oxygen saturation
and end-tidal carbon dioxide concentration using a nasal
adapter. Appropriate treatment was given if systolic arterial
pressure or heart rate decreased to less than 75% of baseline.
All adverse effects were recorded.
Midazolam 0.03–0.05 mg kg–1 or thiopental 1–3 mg kg–
1 i.v. were used for sedation in children who felt uncomfort-
able after premedication or during operation, or expressed
a wish. Sedation was used in 53 children (thiopental in 30
and midazolam in 23 children) with similar distributions in
both groups. All sedated children were administered oxygen
and monitored closely by the anaesthetist or anaesthetic
nurse.
Lumbar puncture was performed in the lateral decubitus
position using a midline approach at the L3–4 or L4–5
interspace. Standard 27-gauge, 90-mm long spinal needles
(Yale, Becton Dickinson, Meylan, Spain or Pencan, B Braun,
Mellsungen, Germany) were used for spinal puncture. The
bevel and orifice of the needle were facing laterally, parallel
to the longitudinal dural fibres. Correct placement was
verified by free aspiration of CSF. After injection of local
anaesthetic, free aspiration of CSF was verified again and
the child was placed in the supine, horizontal position.
During spinal puncture the following variables were
recorded: interspace used; number of skin punctures; need
for redirection of needle; degree of difficulty of CSF
aspiration (1⫽easy, 2⫽difficult, 3⫽impossible); and time
to complete the block.
60
If there were signs of inadequate spread or duration of
spinal anaesthesia in relation to the operation performed,
fentanyl 1 µg kg–1 i.v. was given for supplementary anal-
gesia. The cases were recorded.
An electric stimulator (Microstim Plus, Neuro Techno-
logy, Houston, TX, USA) was used to evaluate the width
of the analgesic area, 15 and 30 min after injection of the
anaesthetic. The metal electrodes of the device were moved
in a rostral direction along the surface of the trunk using
continuous high-frequency stimulation (50 Hz, pulse width
0.2 ms, amplitude 30–60 mA) until the child reported mild
pain, indicating the upper border of the analgesic area.
Amplitude was adjusted to the lowest level so that the
response could just be elicited. The procedure was repeated
until the level of the first painful segment was confirmed.
Motor block was assessed using a modified Bromage scale
recording the child’s ability to flex the ankle, knee and hip
(0⫽no motor block, 3⫽complete motor block of the legs
and feet).
After operation, children were transferred to the post-
anaesthesia care unit (PACU) for continuous monitoring of
vital signs and regression of block. Regression of sensory
block by two segments was tested every 5 min and the
time recorded. If the child was in pain, fentanyl 1 µg kg–1
i.v. was given and the time recorded.
Patients were discharged when they were awake, able to
walk unaided, had stable vital signs for at least 1 h, had no
pain or only mild pain, had no nausea, retching or vomiting,
and were able to tolerate clear fluids. Time to discharge
was measured from spinal puncture to actual discharge
from the hospital/PACU. After operation (at home) children
were prescribed ibuprofen tablets 40 mg kg–1 day–1 in four
divided doses. Parents were encouraged to give ibuprofen
on the first night and first postoperative day, even if the
child did not seem to be in pain. Later parents were told to
give analgesics on an ‘as required’ basis. Follow-up of the
children at home was recorded using a diary.
Power analysis suggested that 48 children were required
in each group for a 90% chance at the 0.05 level of
significance of detecting a 10% difference in success rate
between groups. Categorical data were tested using the chi-
square test. For continuous data the Mann–Whitney test
was used. Association between independent variables was
tested by linear regression analysis. Results are presented
as median (10–90th percentiles), number (%) of cases or
the square of the correlation coefficient (r2) with 95%
confidence intervals (CI), as appropriate. The level of
statistical significance was set as P⬍0.05.
Results
Patient data (Table 1) and the characteristics of spinal
puncture were comparable between groups (Table 2). Ortho-
paedic surgery was performed in 84 children; 23 underwent
urological surgery, herniotomy, orchidopexy or circum-
cision.
 Spinal anaesthesia in children

Table 1 Patient characteristics (median (10–90th percentiles) or number Table 4 Characteristics of sensory block in the two groups (median (10–90th
of patients) percentiles)). Times are after spinal puncture

Bupivacaine in 0.9% Bupivacaine in 8% Bupivacaine in 0.9% Bupivacaine in 8%

glucose (n⍧55) glucose (n⍧52) glucose (n⍧55) glucose (n⍧52)

Sex (M/F) 27/28 24/28 Height of sensory block T4 (T1–T7) T4 (T1–T5)

Weight (kg) 52 (27–65) 48 (30–64) (dermatome)
Height (cm) 156 (128–176) 157 (135–172) Regression of block by two 83 (50–143) 85 (53–150)
Age (yr) 13 (8–16) 13 (8–16) segments (min)
ASA I/II 44/11 38/14 Regression of block to T7 90 (59–161) 103 (61–180)
(min)
Regression of block to T10 112 (63–178) 115 (66–180)
Table 2 Characteristics of the spinal puncture in the two groups (median (min)
(10–90th percentiles) or number of patients) Time to first dose of rescue 210 (88–337) 181 (120–279)
analgesic (min)
Bupivacaine in Bupivacaine in Time to discharge from 237 (147–352) 240 (180–330)
0.9% glucose 8% glucose hospital (min)
(n⍧55) (n⍧52)

Interspace used for spinal puncture Table 5 Perioperative adverse effects of the spinal anaesthesia in the two
L3–4 20 17 groups (number (%)).*One patient vomited
L4–5 35 35
Repeated skin punctures 4 6 Bupivacaine in 0.9% Bupivacaine in 8%
Redirection of needle 13 13 glucose (n⍧55) glucose (n⍧52)
CSF aspiration easy/difficult 50/5 49/3
Appearance of CSF at the hub (s) 2 (1–4) 2 (1–4) Hypotension 1 (2) 5 (10)
Time to complete (s) 40 (30–85) 40 (30–114) Bradycardia 3 (5) 3 (6)
Nausea 11 (20)* 10 (19)
Shivering 7 (13) 9 (17)
Table 3 Success rate of spinal anaesthesia in the two groups (number
(%)).*Fentanyl was administered for one child after an initially complete
block after 55 min; **fentanyl was administered after an initially complete
block after 115 min for one child; ***another was one of the failures etilefrin i.v. (Effortil, Boehringer Ingelheim, Ingelheim am
Rhein, Germany) for hypotension. Six children were given
Bupivacaine in 0.9% Bupivacaine in 8%
atropine i.v. to treat bradycardia. Twenty-one (20%) children
glucose (n⍧55) glucose (n⍧52)
had nausea, but only one child vomited once. Nausea was
Sensory block mild (median 3 on a 10-cm VAS) and did not require
Complete 52 (95) 51 (98)
treatment. Only two (7%) children suffered nausea when
Fentanyl i.v. 3 (5)* 1 (2)**
Motor block
the height of the block was less than T4. Sixteen (15%)
Complete 53 (96) 52 (100) children experienced shivering in the PACU.
Some movement 2 (4)*** Children were discharged after a median time of 237 min
in the bupivacaine–0.9% glucose group and after 240 min
in the bupivacaine–8% glucose group (Table 4). In both
The success rate of spinal block was high in both groups
groups there was a similar positive correlation between the
with no differences between groups (Table 3). Four children
age of the child and time to discharge from hospital: r2⫽
required supplementation with fentanyl 1 µg kg–1, three
0.22, time⫽97⫹1.0⫻(age in months), CI for b⫽0.47–1.5)
in the bupivacaine–0.9% glucose group and one in the
in the bupivacaine–0.9% glucose group; and r2⫽ 0.18,
bupivacaine–8% glucose group.
time⫽97⫹0.97⫻(age in months), CI for b⫽0.36–1.6 in
There was a similar variation in cephalad spread of
bupivacaine–8% glucose group.
sensory block in both groups (Table 4). Maximum extent
Post-dural puncture headache occurred in three children
of block was independent of age, weight, height or interspace
in the bupivacaine–0.9% glucose group and in one child in
used for spinal puncture.
the bupivacaine–8% glucose group. All four had mild
Regression of block was similar in both groups. Regres-
symptoms which resolved in 1 or 2 days.
sion of sensory block by two segments correlated with the
age of the child in the bupivacaine–8% glucose group (r2⫽
0.21, time⫽1.7⫹0.6⫻(age in months); CI for b⫽0.20– Discussion
0.98), but not in the bupivacaine–0.9% glucose group. We have compared the anaesthetic effects of bupivacaine
Rescue analgesics were administered to 42 children in 5 mg ml–1 in 0.9% and in 8% glucose solutions. Thus
the PACU: 25 (49%) in the bupivacaine–8% glucose group baricity was the main factor responsible for differences in
and 17 (31%) in the bupivacaine–0.9% glucose group. Time spinal block. In agreement with the findings of Wildsmith
to the first dose of analgesic was similar in both groups and co-workers,4 6 observed in adults, a slightly hyperbaric
(Table 4). bupivacaine solution produced a predictable sensory block
There were no differences between groups in the incid- but the spread of the block did not show the narrow range
ence of adverse effects (Table 5). Six children were given observed in adults. In this study, the success rate of spinal

61
 Kokki and Hendolin
anaesthesia with bupivacaine in 8% glucose was high as
only one child required fentanyl and sedation to complete
surgery. In our previous study with hyperbaric bupivacaine
in children, a similar high success rate was verified compared
with isobaric bupivacaine.5
We measured the analgesic area using transcutaneous
electrical stimulation (TES). TES has been found to be
equivalent to surgical incision;8 it is a reproducible stimulus
and is better than pinprick for assessment of height and
duration of anaesthesia and analgesia in children.5 However,
it is important to adjust the amplitude to the lowest possible
level to avoid unnecessary distress.
The results of our study confirm our earlier findings5 that
in school-age children, bupivacaine 0.3 mg kg–1 produced
median sensory block to T4 with a median duration of 80–
85 min. Duration of the block was adequate. However, the
maximum extent of block was unnecessarily high as one-
fifth of children had mild nausea during surgery. This
warrants further study of a smaller dose of bupivacaine
than 0.3 mg kg–1.
Hypotension and bradycardia are normal physiological
responses during spinal anaesthesia because of sympathetic
fibre block.9 In small children, cardiovascular stability
during spinal anaesthesia is good5 10 in contrast with school-
age children (as observed in this study). Furthermore, in
adults, spinal block with extensive spread impairs central
thermoregulatory control.11 Fifteen percent of children in
our study experienced shivering during recovery of spinal
block. Thus avoidance of extended block is also essential
to counteract hypothermia.
In summary, we have demonstrated that the use of
bupivacaine containing 0.9% glucose produced a similar
spinal anaesthesia in school-age children as bupivacaine in
8% glucose. Spinal anaesthesia was associated with a high
success rate and relatively low incidence of adverse effects.
62
References
1 Abajian JC, Mellish RW, Browne AF, Perkins FM, Lambert DH,
Mazuzan JE. Spinal anesthesia for surgery in the high-risk infant.
Anesth Analg 1984; 63: 359–62
2 Bridenbaugh PO, Green NM. Spinal (subaracnoid) neural blockade.
In: Cousins MJ, Bridenbaugh PO, eds. Neural Blockade in Clinical
Anesthesia and Management of Pain. Philadelphia: JB Lippincott,
1988; 213–51
3 Sethna NF, Berde CB. Paediatric regional anesthesia. In: Gregory
GA, ed. Paediatric Anesthesia. New York: Churchill Livingstone,
1994; 281–317
4 Bannister J, McClure JH, Wildsmith JA. Effect of glucose
concentration on the intrathecal spread of 0.5% bupivacaine. Br J
Anaesth 1990; 64: 232–4
5 Kokki H, Tuovinen K, Hendolin H. Spinal anaesthesia for paediatric
day-case surgery: a double-blind, randomized, parallel group,
prospective comparison of isobaric and hyperbaric bupivacaine.
Br J Anaesth 1998; 81: 502–6
6 Sanderson P, Read J, Littlewood DG, McKeown D, Wildsmith JA.
Interaction between baricity (glucose concentration) and other
factors influencing intrathecal drug spread. Br J Anaesth 1994; 73:
744–6
7 Kokki H, Heikkinen M, Turunen M, Vanamok, Hendolin H. Needle
design does not affect the success rate of spinal anaesthesia or
the incidence of postpuncture complications in children. Acta
Anaesth Scand 1999, in press
8 Petersen-Felix S, Zbinden AM, Fischer M, Thomson DA. Isoflurane
minimum alveolar concentration decreases during anesthesia and
surgery. Anesthesiology 1993; 79: 959–65
9 Covino BG, Scott DB, Lambert DH. Handbook of Spinal Anaesthesia
and Analgesia. Fribourg, Switzerland: Mediglobe SA, 1994
10 Saint-Maurice C. Spinal block. In: Saint-Maurice C, Steinberg
OS, eds. Regional Anaesthesia in Children. Fribourg, Switzerland:
Mediglobe SA, 1990; 119–25
11 Leslie K, Sessler DI. Reduction in the shivering threshold is
proportional to spinal block height. Anesthesiology 1996; 84:
1327–31