The Influence of High Spinal Anesthesia on Sensitivity to

Midazolam Sedation
Bruce Ben-David, MD*, Sonya Vaida, MDt, and Luis Gaitini, MDt
Departments of Anesthesia, *Herzliya-Haifa (Horev) Medical Center and tbnai Zion Medical Center, Haifa, Israel
We tested whether a high spinal anesthesia may alter
the susceptibility to the soporific effects of sedatives.
Twenty ASA grade I and II women undergoing elective
abdominal hysterectomy were randomly allocated into
two groups. Patients in Group I were given a subarach-
noid injection of 12 mg hyperbaric tetracaine and those
patients who after 10 min had a sensory level of T4-6
(10 patients) were included in the study. Ten additional
patients (Group II) received no spinal injection. Induc-
tion of anesthesia was performed on all patients by in-
jecting 1 mg of midazolam intravenously every 30 sun-
til the patient failed to respond to three repeated
commands to squeeze the anesthetists hand. This was
considered the induction dose or end-point for the pur-
poses of the study. Patients were then given a neuro-
muscular blocker, ventilated with oxygen, nitrous ox-
ide, and a volatile anesthetic, tracheally intubated, and
maintained under general anesthesia for the remainder
of the operation. The dose of midazolam administered
to the point of patient failure to respond to command
was 7.6 + 0.72 mg SEM for Group I and 14.7 + 1.16 mg
SEM for Group II, (P < 0.0001). These results support the
conclusion that patients having a high spinal anesthetic
are more sensitive to the sedative effects of midazolam.
(Anesth Analg 1995;81:525-8)
I
n 1988, Caplan et al. (1) published their landmark
retrospective review of closed insurance claims.
Their work revealed 14 cases of unexplained car-
diac arrest during spinal anesthesia. One of the pat-
terns common to their cases was that all patients had
received intraoperative sedation. The authors sug-
gested the possible importance of respiratory changes
produced by sedation in combination with the physi-
ologic changes of spinal anesthesia. In an accompany-
ing editorial Keats (2) speculated that the deafferenta-
tion caused by spinal anesthesia and the loss of
facilitatory proprioceptive input into the respiratory
center may alter the respiratory response to these
drugs. Insofar as these sensory inputs also have pro-
jections into the reticular activating system, the loss of
this neural traffic consequent to a spinal anesthetic
may also alter wakefulness and the sensitivity of the
central nervous system (CNS) to the soporific effects
of sedative drugs. This study undertook to examine
the question of whether patients are more sensitive to
the sedative effect of midazolam in the presence of a
high spinal anesthetic.
Accepted for publication April 12, 1995.
Address correspondence to Bruce Ben-David MD, Medical Direc-
tor, Herzliya-Haifa (Horev) Medical Center, 15 Horev St., Haifa,
Israel.
Methods
After protocol approval by our institutions human
studies committee, and after patient informed consent,
22 ASA grade I and II women who were to undergo
elective hysterectomy were prospectively entered into
the study. Exclusion criteria included age 70 yr or
more, use of any opioid or sedative medications in the
week prior to surgery, a history of alcohol or drug
abuse, history of any psychiatric or neurologic abnor-
mality, refusal of spinal anesthesia, or a history of
untoward reaction to benzodiazepine.
No patient received premeditation. The women
were randomly allocated to one of two groups. Group
I received a spinal anesthetic using 12 mg of hyper-
baric tetracaine. After 10 min, if the spinal block had
reached a sensory level to pinprick testing of T4- 6, the
patient was included in the protocol. If an inadequate
level was achieved, the patient was not included in the
study protocol. No patient had a level higher than T4,
and two were excluded because of sensory levels be-
low T6. An intravenous (IV) solution of lactated Ring-
ers solution (500-1000 mL) was infused to maintain
systolic blood pressure within 10% of its initial level.
Small doses of ephedrine were given IV as needed to
maintain arterial blood pressure if administration of
IV fluids alone did not maintain it. Group II patients
did not receive a spinal injection and received no IV
hydration prior to induction of anesthesia. All patients
01995 by the International Anesthesia Research Society
0003-2999/95/$5.00 Anesth Analg 1995;81:525-8 525
526 REGIONAL ANESTHESIA AND PAIN MANAGEMENT BEN-DAVID ET AL.
HIGH SPINAL ANESTHESIA AND MIDAZOLAM SEDATION
ANESTH ANALG
1995;81:525-8
had standard intraoperative monitoring initiated prior
to spinal block and induction. Prior to induction, pa-
tients had a face mask placed over the mouth and nose
and 100% oxygen was administered via the anesthesia
machine. The mask was held in place by the anesthe-
tist for the duration of induction.
Midazolam, 1 mg, was given into a freely running
IV line every 30 s. Ten seconds prior to each injection
patients were asked to squeeze the anesthetists hand,
which was left continually in the patients hand. The
anesthetist caring for the patient, while not blinded to
the presence or absence of a spinal anesthetic, was
blinded as to the purpose of the study, believing it to
be a study of the effect of spinal anesthesia on post-
operative pain and patient-controlled analgesia. No
response, as determined by the anesthetist, to three
promptly repeated and increasingly louder com-
mands was considered the end-point for the study and
no further midazolam was given. Patients were then
given a nondepolarizing neuromuscular blocker, ven-
tilated with oxygen, nitrous oxide, and a potent vola-
tile anesthetic, tracheally intubated, and maintained
under general anesthesia for the remainder of the
operation. Statistical analysis of the data was done
with the Mann-Whitney U-test and two-tailed un-
paired Students t-test using Statview 4.0 (Abacus
Concepts, Berkeley, CA).
Results
Both Groups I and II consisted of 10 patients each.
Originally, Group I had 12 patients, but two patients
were withdrawn from the study because of insuffi-
cient level of the spinal blockade (below T6). No sig-
nificant differences were observed between the groups
in terms of age or weight. For Groups I and II, the
average ages were, respectively, 52.4 ? 4.8 yr and 46.5
2 2.4 yr, and the average weights were, respectively,
67.2 ? 2.9 kg and 70.2 2 4.1 kg (mean + SEM). The
average level of spinal block to pinprick testing was
T4-7. Through the end of induction, the Group I pa-
tients received an average of 500 mL more lactated
Ringers solution than did the Group II patients. Mi-
dazolam dosing to the point of the patients failure to
respond to a command to handgrip were 7.6 + 0.72
mg for Group I and 14.7 f 1.16 mg for Group II. The
difference was significantly different with P < 0.0001.
Discussion
It has long been noted that patients given high spinal
anesthesia frequently fall asleep, and that as the level
of spinal blockade becomes higher, drowsiness be-
comes more frequent and pronounced (3). Greene (4)
found that patients with high spinal anesthesia require
only very low concentrations of inhalational drugs to
maintain unconsciousness. Surprisingly, however,
there has been little attention given this phenomenon.
Even in the wake of the review by Caplan et al. (1) on
unexplained cardiac arrests during spinal anesthesia
and the association of these cases with the use of
sedative drugs, the focus of research efforts seems to
have been on the respiratory system and not on wake-
fulness per se.
Gauthier et al. (5) compared spinal anesthesia (av-
erage sensory level T6), midazolam sedation (4.4 mg
IV average dose), and the combination (same spinal
level and midazolam dose) as to their effects on rest-
ing ventilation. Whereas an increased respiratory
drive was the primary effect of spinal anesthesia alone
in the group receiving both spinal anesthesia and mi-
dazolam sedation, the decreases in tidal volume,
minute ventilation, and mean inspiratory flow rate
and the increase in rib cage contribution to breathing
were significantly more than expected based on the
predicted sum of the individual interventions. The
nature of this superadditive interaction may be, as
Keats (2) proposed, a result of the deafferentation and
the loss of facilitatory neural input produced by the
spinal anesthesia.
The conclusion reached in this study is that high
spinal anesthesia markedly sensitizes to the sedative
effect of midazolam. We speculate that the primary
mechanism of this action is a deafferentation phenom-
enon and that the loss of facilitatory input to the
reticular activating system renders it more susceptible
to the sedative actions of a drug such as midazolam.
Several lines of evidence support this theory.
Afferentation theory predicts that drugs or maneu-
vers that stimulate sensory input, such as that from
muscle stretch receptors, will produce cerebral stimu-
lation. It is known that sleep-deprived subjects who
increase motor activity are less likely to fall asleep
than subjects in whom muscle quiescence is encour-
aged (6). In anesthesia, the administration of depolar-
izing neuromuscular blocking drugs increases muscle
afferent activity (7) and induces electroencephalogram
(EEG) arousal (8). Lanier et al. (9) used a dog model
with tracheal stimulation to demonstrate that active
muscle movement in lightly anesthetized subjects oc-
curred together with cerebral stimulation manifested
by increased EEG activation, cerebral vasodilation,
and an increase in cerebral blood flow greater than
that required to meet metabolic demands. Paralysis
with pancuronium abolished the movement induced
by tracheal stimulation and, at the same time, attenu-
ated the cerebral blood flow, cerebral vascular resis-
tance, and EEG responses.
If an increase in muscle afferent activity stimulates
the CNS, then the loss of resting muscle afferent input
(to say nothing of other sensory input) should reduce
the level of cerebral stimulation. Forbes et al. (10) used
ANESTH ANALG
1995;81:525%3
REGIONAL ANESTHESIA AND PAIN MANAGEMENT BEN-DAVID ET AL. 527
HIGH SPINAL ANESTHESIA AND MIDAZOLAM SEDATION
an isolated limb technique to demonstrate a reduction
of minimum alveolar anesthetic concentration for hal-
othane from 0.77% to 0.55% in patients who had re-
ceived 0.1 mg/kg pancuronium. Schwartz et al. (11)
examined the same phenomenon using EEG analysis
and found that increasing doses of pancuronium re-
sulted in increasing amounts of EEG isoelectricity for
a given level of inspired concentration of isoflurane.
The increased duration of EEG isoelectricity resulting
from the administration of pancuronium signifies that
the effect of isoflurane on the brain is enhanced by the
neuromuscular block. Moreover, this effect was re-
versed with the administration of neostigmine.
Certainly, high spinal anesthesia will cause a pro-
nounced reduction of muscle afferent activity, as well
as a loss of any other sensory input from the anesthe-
tized portion of the body. It therefore seems reason-
able to anticipate that high spinal anesthesia, by virtue
of its reduction of this stimulatory neural traffic, will
have a significant effect on wakefulness and the
brains sensitivity to sedative drugs.
Besides deafferentation, there are at least two other
potential explanations for the observed combined ef-
fects of sedation and high spinal anesthesia as well as
the drowsiness seen with high spinal anesthesia alone.
One possible explanation is that hypotension second-
ary to the spinal anesthesia decreases cerebral blood
flow with resultant somnolence. However, in this
study our attention to the maintenance of blood pres-
sure by the administration of fluids and ephedrine
makes cerebral hypoperfusion an unlikely factor. The
second potential explanation is that of a direct effect of
the local anesthetic on the brainstem. It was demon-
strated many years ago that the direct application of
procaine to the brainstem of experimental animals
caused loss of consciousness without respiratory or
vascular collapse (12). It is certainly possible that dur-
ing high spinal anesthesia low concentrations of local
anesthetic do extend through the cerebrospinal fluid
to reach medullary level. This factor may play a con-
tributory role.
Although we can only speculate as to the mecha-
nism, the clinical significance of the findings of this
study are clear-the sedative potency of midazolam is
increased in the presence of high spinal anesthesia. In
spite of the lack of measured blood levels of midazo-
lam the results are suggestive of a prominent pharma-
codynamic effect. The additional fluid loading given
to the spinal group would have been expected to
increase blood volume, and therefore dilute the blood
level of the midazolam induction dose. This should
have increased the midazolam dose required in the
spinal group and cannot be expected to have ac-
counted for the difference between the groups. Pers-
son et al. (13) compared midazolam pharmacokinetics
in patients having lumbar epidural anesthesia to a
level of T4-6 with patients receiving total IV anesthe-
sia with midazolam and alfentanil. They found no
difference between the groups in the pharmacokinet-
its of midazolam. Their data indicated that there is
both a larger volume of distribution of the central
compartment and a larger apparent volume of distri-
bution in the epidural group.
It was not the intention of this study to examine
induction dose per se, but rather to use a common
end-point as a point of comparison of sedative po-
tency. Still, the doses of midazolam used in this study
are within the range of induction doses reported in the
literature. The wide range of induction doses reported,
0.1-0.4 mg/kg (14), is explained by several influenc-
ing factors. Induction dose is very dependent on the
speed of injection, and the different rates of adminis-
tration used in the various dose seeking studies
accounts for much of the variability. Slower adminis-
tration allows time for distribution of the drug, so that
the CNS is presented with a lower concentration of
drug. In addition, slower administration may allow
time for the development of acute tolerance. Other
factors that influence the induction dose include pre-
medication (opioids in particular), age (the elderly
require less drug), gender (males are more sensitive),
and ASA physical status (ASA grades III and IV re-
quire less drug). In this study, we tried to control for
all of these factors by limiting the subject population
to women patients of comparable age and weight and
of only ASA grades I and II, with no patient being
given a premeditation. Importantly, we adhered
closely to the administration rate set out in our proto-
col to eliminate this as a possible explanation for a
difference between groups. This study focuses on a
single end-point as a measure of sedation. It does not
address ensuing respiratory or hemodynamic re-
sponses, such as to laryngoscopy and intubation.
Therefore, the doses administered for the purpose of
this study do not necessarily represent true induction
doses of midazolam, and certainly not of midazolam
delivered as a bolus. Nevertheless, our conclusions do
reflect real differences between the two study groups.
The finding that high spinal anesthesia sensitizes to
the sedative action of midazolam is not likely to
change clinical practice which is typically that of titra-
tion to effect. It may, however, increase ones sense of
caution in sedating the patient with a high spinal
anesthetic. Midazolam has a depressant effect on the
respiratory responses to hypercarbia and hypoxia
causing a decrease in the slope of the hypercapnic
ventilatory response (5,15,16) and a decrease in the
slope of the hypoxic ventilatory response curve (17).
Midazolam at a dose of 0.1 mg/kg IV produces a
marked increase in upper airway resistance (18) at
least partially due to relaxation of the genioglossus
muscle (19) and a relapsing of the tongue posteriorly
528 REGIONAL ANESTHESIA AND PAIN MANAGEMENT BEN-DAVID ET AL. ANESTH ANALG
HIGH SPINAL ANESTHESIA AND MIDAZOLAM SEDATION 1995:81:5254
into the oropharynx. By increasing the sensitivity to
midazolam a high spinal anesthetic might indeed af-
fect the frequency and severity of these untoward
airway and respiratory effects. This may or may not be
in direct proportion to the effect of high spinal anes-
thesia on wakefulness and the sensitivity to sedatives.
Midazolam 0.1 mg/kg IV produces a sixfold increase
in total pulmonary resistance and a compensatory
load response characterized by an increase in inspira-
tory intercostal and expiratory abdominal muscle ac-
tivities with a reduction of diaphragmatic contribution
(20). A high spinal anesthetic might potentially inter-
fere in this compensatory ventilatory response by pro-
ducing a motor block of intercostal and abdominal
muscles.
Certainly this study could have been done differ-
ently using a given level of desaturation as an end-
point in place of handgrip to command. Such a study
would have addressed more closely the central clinical
question suggested by the foregoing respiratory data:
Does the combined use of high spinal anesthesia and
midazolam sedation increase the risk of hypoventila-
tion and hypoxemia. 7 However, that was not the
question we were asking in this study. In this work we
did not observe any incidents of desaturation below
Sao, of 85%, but we were repeatedly stimulating the
patient by calling her name and asking her to squeeze
the anesthetists hand. This is not analogous to an
intraoperative situation where the patient is left
largely unstimulated. Furthermore, our patients were
breathing 100% oxygen through a face mask, which
might have been arousing to the patient and would
have provided a buffer against hypoxia secondary to
apnea or hypoventilation.
In conclusion, we find that high spinal anesthesia
markedly increases the sensitivity to the sedative ef-
fects of midazolam. This factor should be taken into
consideration when sedating the patient under spinal
anesthesia. The known risks of the interaction of spi-
nal anesthesia and midazolam sedation on respiratory
function in conjunction with this finding of altered
CNS sensitivity in the presence of high spinal anes-
thesia should give us a due sense of caution in the use
of midazolam sedation for patients under spinal
anesthesia.
References
1. Caplan RA, Ward RJ, Posner K, Cheney FW. Unexpected car-
diac arrest during spinal anesthesia: a closed claims analysis of
predisposing factors. Anesthesiology 1988;68:5-11.
6. Kleitman N. Sleep and wakefulness. Chicago: The University of
Chicago Press, 1963:195-7.
7. Lanie; WL, Iaizzo PA, Milde JH. Cerebral function and muscle
afferent activity following intravenous succinylcholine in dogs
anesthetized with halothane: the effects of pretreatment with a
defasciculating dose of pancuronium. Anesthesiology 1989;71:
87-95.
8. Lanier WL, Milde JH, Michenfelder JD. Cerebral stimulation
following succinylcholine in dogs. Anesthesiology 1986;64:
551-9.
9. Lanier WL, Iaizzo PA, Milde JH, Sharbrough FW. The cerebral
and systemic effects of movement in response to a noxious
stimulus in lightly anesthetized dogs. Anesthesiology 1994;80:
392-401.
10.
11.
12.
13.
14.
15.
16.
Forbes AR, Cohen NH, Eger EI II. Pancuronium reduces halo-
thane requirement in man. Anesth Analg 1979;58:497-9.
Schwartz AE, Navedo AT, Berman MF. Pancuronium increases
the duration of electroencephalogram burst suppression in dogs
anesthetized with isoflurane. Anesthesiology 1992;77:686-90.
Kostar H, Kasman LP. Spinal anesthesia for the head, neck, and
thorax; its relation to respiratory paralysis. Surg Gynecol Obstet
1929;49:617.
Persson MI, Nilsson A, Hartvig I. Relation of sedation and
amnesia to plasma concentrations of midazolam in surgical
patients. Clin Pharmacol Ther 1988;43:324-31.
Reves JG, Fragen RJ, Vinik HR, Greenblatt DJ. Midazolam:
pharmacology and uses. Anesthesiology 1985;62:310-34.
Forster A, Morel C, Bachmann M, Gemperle M. Respiratory
depressant effects of different doses of midazolam and lack of
reversal with naloxone: a double-blind randomized study.
Anesth Analg 1983;62:920-4.
Forster A, Gardaz J-P, Suter PM, Gemperle M. Respiratory
depression by midazolam and diazepam. Anesthesiology 1980;
53:494-7.
17.
18.
19.
20.
Alexander CM, Gross JB. Sedative doses of midazolam depress
hypoxic ventilatory responses in humans. Anesth Analg 1988;
67~377-82.
Montravers I, Dureuil B, Desmonts JM. Effects of intravenous
midazolam on upper airway resistances. Br J Anaesth 1992;68:
27-31.
Leiter JC, Knuth SL, Krol RC, Bartlett D. The effects of diazepam
on genioglossal muscle activity in normal human subjects. Am
Rev Respir Dis 1985;132:216-9.
Molliex S, Dureuil B, Montravers I, Desmonts J-M. Effects of
midazolam on respiratory muscles in humans. Anesth Analg
1993:77:592-7.
2. Keats AS. Anesthesia mortality-a new mechanism. Anesthesi-
ology 1988;68:2-4.
3. Greene NM. Physiology of spinal anesthesia. 3rd ed. Baltimore:
Williams & Wilkins, 1981:45-54.
4. Greene NM. Hypotensive spinal anesthesia. Surg Gynecol Ob-
stet 1952;95:331.
5. Gauthier RA, Dyck B, Chung F, et al. Respiratory interaction
after spinal anesthesia and sedation with midazolam. Anesthe-
siology 1992;77:909-14.