Anaesthesia, 2002, 57, pages 140154

................................................................................................................................................................................................................................................

R E V I E W A RT I C L E

Anions and the anaesthetist

D. G. Maloney,1 I. R. Appadurai2 and R. S. Vaughan2
1 Specialist Registrar and 2 Consultant Anaesthetist, Department of Anaesthetics, University Hospital of Wales, Heath
Park, Cardiff CF14 4XW, UK
Summary

Anions are the negative components of most chemical structures and play many important
physiological and pharmacological roles that are of interest to the anaesthetist. Their relevance is
reviewed with a particular emphasis on the inorganic anions (halides, bicarbonate, phosphate and
sulphate) and the significance and limitations of the anion gap. Organic anions (albumin, lactate) are
also discussed, albeit briefly. The suitability of anions for their role in neurotransmission and
acid2base balance is outlined.
Keywords Anions: chloride; fluoride; bromide; phosphates; sulphates; albumin; lactate. Acid2base
equilibrium: anion gap.
.................................................................................................
Correspondence to: Dr D. G. Maloney
E-mail: declan_maloney@hotmail.com
Accepted: 9 June 2001

Anions have many important physiological and pharmacological functions and have, to a large extent, been
ignored in anaesthetic textbooks. They are invariably
linked to cations but can have pronounced effects in their
own right. Although their concentrations are measured in
the extracellular fluid compartment in clinical practice, the
ion shifts that occur across the cell membrane are not only
involved in acid2base and electrolyte balance, but may also
form the basis of the molecular mechanisms of anaesthesia.

and is normally present in the body, unlike the other

halogens. Because it is exchangeable with bicarbonate, it
has a role in maintaining a normal acid2base state,
normal renal tubular function and in the formation of
gastric acid. It has been the subject of review in the
context of critical care medicine [1] and has important
implications for the anaesthetist in the areas of fluid
resuscitation and acid2base balance.

Halogens
Halogenation of hydrocarbons and ethers decreases
flammability and increases anaesthetic potency. It also
increases cardiac sensitisation to catecholamines in the
order: F2 . C12 . Br2 . I2. Fluorination increases
the stability of adjacent halogen2carbon bonds but may
also increase convulsant properties, especially if the
compound has four or more carbon atoms or contains
other halogens.

Chloride intake
The daily requirement of 13 mmol.kg21 is absorbed
from the upper part of the small intestine by the Na1/
K1/C12 co-transport mechanism. The active absorption
of sodium and potassium creates an electronegative
gradient in the chyme along which chloride ions move.
However, in the ileum and large bowel, the enterocytes
have an anti-transport mechanism that exchanges chloride
ions for bicarbonate. The physiological significance of the
latter is uncertain but impairment of this transport
mechanism leads to congenital chloride diarrhoea, an
autosomal recessive disorder.

Chloride
Chloride is the main anionic constituent of extracellular
fluid (ECF) contributing < 100 mosm.l21 to its tonicity,

Chloride output
The greatest loss is from the stomach in gastric juices
(<1.5 mmol.kg21.day21). There is a further loss from the

Inorganic anions

140

q 2002 Blackwell Science Ltd

Anaesthesia, 2002, 57, pages 140154

D. G. Maloney et al.  Anions and the anaesthetist
................................................................................................................................................................................................................................................

gastrointestinal tract (<1 mmol.kg21.day21) in the form

of bile, pancreatic and intestinal secretions. Chloride is
secreted into the intestinal lumen via channels regulated
by protein kinases. One of these is activated by protein
kinase A and therefore by cyclic adenosine monophosphate (cAMP). Some bacteria, notably Vibrio cholerae,
produce a toxin that binds to the Gs protein, producing a
marked increase in cAMP, leading to profuse secretory
diarrhoea. Normally about 15 mmol.day21 are lost in
sweat.
In the renal tubule, chloride excretion varies with the
need for bicarbonate. Chloride is excreted with ammonium ions to eliminate hydrogen ions in exchange for
sodium. This occurs mainly in the proximal tubule and
can result in the production of acid urine even in the
presence of a metabolic alkalosis, as seen in pyloric
stenosis. In the ascending limb of the loop of Henle,
chloride is reabsorbed with sodium. Regulation of
chloride is passively related to sodium and inversely
related to plasma bicarbonate. Therefore, aldosterone
indirectly influences chloride concentrations.
Chloride shifts
As erythrocytes move through the peripheral tissues,
carbon dioxide is converted by the action of carbonic
anhydrase to bicarbonate. About 70% of the bicarbonate
produced will diffuse into the plasma and chloride shifts
into the cell to maintain electrochemical neutrality. The
reverse occurs when the blood reaches the lungs. These
exchanges are known as chloride shifts. In respiratory
disturbances, up to 30% of an acid load can be buffered by
similar chloride2bicarbonate ionic shifts between the
intracellular fluid (ICF) and extracellular fluid [2]. In
chronic respiratory acidosis, a compensatory metabolic
alkalosis occurs with a subsequent chloride shift from

the erythrocyte to the plasma in exchange for bicarbonate. In respiratory alkalosis, hydrogen moves from the
intracellular space to maintain electroneutrality, whereas
excess bicarbonate moves into the erythrocyte in
exchange for chloride. The causes of chloride flux are
given in Table 1.
Implications for fluid resuscitation and acid2base management.
The choice of fluids containing chloride ranges from
hypertonic, e.g. 1.8% NaCl, to hypotonic, e.g. 0.18%
NaCl, in dextrose saline. These are normally chosen for
their sodium content and osmolarity and therefore their
effect on expanding the ECF. They are usually unbuffered
and can have a profound effect on acid2base balance.
It is important to be aware that over-reliance on normal
saline as a resuscitation or maintenance fluid will result in
a hyperchloraemic acidosis [35]. This can have significant effects in the postoperative period [6] that may be
ascribed erroneously to other pathology [7]. It can be
corrected by the judicious use of sodium bicarbonate and
is preferable to allowing hyponatraemia to develop perioperatively, as may occur with transurethral resection of
the prostate (TURP syndrome). There may be an added
benefit to using hypertonic saline as a `preload' before
regional anaesthesia in such cases [8].
Just as an excess of chloride can cause a metabolic
acidosis, it is notable that metabolic alkalosis and
hypokalaemia are the hallmarks of chloride depletion.
The classic example is that of gastric outflow obstruction,
in particular pyloric stenosis [9]. Metabolic alkalosis can
be divided, for therapeutic purposes, into saline-resistant
and saline-responsive types. Urinary chloride concentration can be helpful in determining the aetiology of the
alkalosis provided no diuretic has been administered [1]. A
urinary chloride of , 10 mmol.l21 implies that the

Table 1 Causes of chloride flux.

Hypochloraemia (, 95 mmol.l21)

Hyperchloraemia (. 106 mmol.l21);

Gastrointestinal loss

Vomiting, e.g. pyloric stenosis

Nasogastric suction

Increased intake

Intravenous fluid infusions

Increased renal excretion

Diuretics

Decreased renal excretion

Uretero-ileostomy;
Renal tubular acidosis

Endocrine

Hyperaldosteronism
Syndrome of inappropiate
antidiuretic hormone secretion
(SIADH)
Cystic fibrosis

Endocrine
Ionic shift
Concentration effect
Artefactual

Diabetes insipidus
Respiratory alkalosis
Severe dehydrationSampling error
Laboratory error

Ionic shift

Chronic respiratory acidosis

Dilutional

Congestive cardiac failure

Infusion of hypotonic solutions

Artefactual

Sampling error
Laboratory error

q 2002 Blackwell Science Ltd

141

D. G. Maloney et al.  Anions and the anaesthetist

Anaesthesia, 2002, 57, pages 140154
................................................................................................................................................................................................................................................

patient should respond to saline, whereas a urinary

chloride of . 20 mmol.l21 suggests that a saline-resistant
metabolic alkalosis is present. The latter is caused mainly
by hyperadrenocorticism in which renal bicarbonate
excretion is stimulated and these patients remain alkalotic
despite chloride administration.
Fluoride
Fluorine has been known to decrease the boiling point,
increase the stability and generally decrease the toxicity of
anaesthetic vapours since 1932 [10]. Although it forms
stable bonds, it was not until after World War II that
advances in chemistry enabled the fluorination of organic
compounds to become a reality.
The problem with fluorination became apparent soon
after the introduction of methoxyflurane with the first
report of nephrotoxicity [11], followed shortly thereafter
by the conclusive dose2response study by Cousins &
Mazze [12]. An apparent renal threshold for nephrotoxicity was established at . 50 m mol.l21 and went unchallenged for . 20 years. It was even extrapolated to other
fluorinated volatile anaesthetics, a view that can no longer
be supported [13, 14]. It was found to be more common
in cases in which cytochrome P450 2EI induction had
occurred [15, 16] or in which there was pre-existing renal
disease [17]. The former was noted with chronic isoniazid
therapy, obesity [18, 19], untreated diabetes mellitus and
alcohol abuse [16]. It has not proven to be a problem with
children taking antiepileptic drugs [20].
As explanations were sought, the original concept of
MAC-hours gave way to that of the area under the
fluoride concentration curve (AUC) [21]. Methoxyflurane has a much larger AUC than expected, given that the
half-life of fluoride in blood is 90 min. It also has a much
larger AUC than any other volatile agent, including
sevoflurane. The difference was at first thought to be
related to both the greater lipid solubility and the degree
of metabolism of methoxyflurane. Sevoflurane undergoes
considerable metabolism to fluoride ions (35%) but its
lipid solubility is low and therefore its uptake by tissues
much less. However, studies have shown that even after
sevoflurane anaesthesia that produced fluoride ion concentrations . 50 m mol.l21 for prolonged periods,
nephrotoxicity did not occur [2225]. Similarly, prolonged anaesthesia [26] or sedation on the intensive care
unit (ICU) [2730] with isoflurane did not produce
clinically significant kidney damage. Enflurane, however,
has been observed to cause decreased urine concentrating
capacity at a fluoride concentration . 33.6 m mol.l21
[31]. It was subsequently proposed that there may be
additional or confounding factors in the aetiology of
nephrotoxicity that include intrarenal formation of
fluoride ions [32], formation of other toxic compounds,
142

e.g. hexafluoroisopropanol (HFIP) [33], or changes in

renal plasma flow. More recently, it has been observed that
maximum fluoride concentrations do not exceed
20 m mol.l21 in children undergoing sevoflurane anaesthesia and they are therefore unlikely to be at risk from
renal toxicity [34, 35]. Despite some early misgivings [33],
sevoflurane is now well established as the agent of choice
for inhalational induction in both adults and children.
It has been noted that at a urinary pH . 7.0 [36, 37],
mannitol, but not furosemide, increases fluoride excretion
[38], although this has not been put to specific therapeutic
use. Fluoride also stimulates periosteal and endosteal bone
formation, but as it replaces the hydroxyl ion in the
hydroxyapatite lattice, there may be a consequent loss of
strength and the potential for fracture. However, there
have not been any reports of fluorosis after long-term
exposure in operating theatre personnel outside the
setting of drug abuse [39].
Bromide
Locock introduced bromide in 1857 as a sedative
anticonvulsant. Its inclusion lends stability to molecular
structures, as exemplified by the non-depolarising neuromuscular blocking drug vecuronium bromide and the
inhalational anaesthetic halothane. The metabolism of
halothane leads to debromination and the bromide may
reach psychoactive concentrations (480 m g.ml2l) after
anaesthesia [40]. This has been observed in both adults
[4143] and children [44]. The extent of this debromination is not possible to predict [45]. The half-life of
bromide is 1214 days and its cumulative effects are an
important consideration with repeat anaesthesia, e.g. for
neuro-oncology treatments [44, 46]. Elevated concentrations have also been noted in anaesthetic [47] and
operating theatre personnel [48] when measured using
X-ray fluorescence spectrometry. However, psychoactive
side-effects have not been demonstrated in these groups.
Debromination has not been shown to be a problem
with neuromuscular blocking drugs, as the amounts used
are much smaller. Halides in high concentrations have
been shown to displace some basic drugs from albumin,
but this has not been demonstrated with bromide up to
concentrations of 4 mmol.l21 [49].
Bicarbonate
Bicarbonate has two main physiological functions. It
forms the main buffer and facilitates the carriage of
carbon dioxide in the blood (80% as bicarbonate). These
two functions are inextricably linked to the body's main
system for the regulation of pH. This relationship is
expressed in the Henderson2Hasselbach equation. The
modified Henderson equation does not contain negative
q 2002 Blackwell Science Ltd

Anaesthesia, 2002, 57, pages 140154

D. G. Maloney et al.  Anions and the anaesthetist
................................................................................................................................................................................................................................................

Table 2 Causes of bicarbonate flux.

Bicarbonate deficit

Bicarbonate excess

Gastrointestinal losses Small bowel, pancreatic, biliary fistulae Metabolic compensation

Diarrhoea
Bicarbonate administration
Cholestyramine
Renal losses

Renal tubular acidosis

Carbonic anhydrase inhibitors

Chronic vomiting with acid loss Gastric outlet obstruction

Chronic diuretic use
Metabolic alkalosis from potassium loss

logarithms and is consequently easier to understand and is

recommended to the reader.
Henderson2Hasselbach equation:
pH pK 1 log HCO2
3 /H2 CO3 :
Modified Henderson equation:
H1 :HCO2
3 kPco2 :H2 O:
The ratio of bicarbonate to carbonic acid is 20:1. Its pK
value is 6.1 and at first glance one would expect the
chemical buffering capacity at pH 7.4 to be poor.
However, it is the quantity of bicarbonate in plasma
(2428 mmol.l21) that is significant. Furthermore, as
blood becomes more acidotic, its pK value is approached
and the efficiency of the buffer system increases. One of
the most important features of the system is its
responsiveness. In the short-term, changes in pH can be
compensated for by adjustments of Paco2 achieved via the
lungs, and in the longer term by adjustments in plasma
bicarbonate achieved via the kidneys. It is important to
mimic this system and to increase the minute ventilation
appropriately when managing patients with a metabolic
acidosis. The ventilatory management of patients with
metabolic alkalosis is more difficult, as a normal minute
volume may reduce the patient's own compensatory
hypercapnia and initially worsen the alkalosis. Ideally,
ventilation should be avoided until the alkalosis has been
reversed. If this is not possible, ventilatory strategies that
allow Paco2 to increase, i.e. permissive hypercapnia,
should be considered. The main causes of changes in
plasma bicarbonate concentration are listed in Table 2.
Therapeutic uses of bicarbonate
Correction of metabolic acidosis
The first solutions of sodium bicarbonate (2.74%) were
formulated at the Westminster Hospital. It was first used
in an anaesthetic context in 1962 to treat the metabolic
acidosis that was thought at the time to cause `neostigmine resistant curarisation' [50]. Hunter had earlier
q 2002 Blackwell Science Ltd

Type II respiratory failure

Self-administration, e.g. treatment of dyspepsia
Iatrogenic, e.g. overcorrecting an acidosis

described this syndrome that caused high mortality in

patients with long-standing intestinal obstruction. The
therapy was a resounding success and marked a milestone
in the evolution of the therapeutic uses of bicarbonate.
The amount of bicarbonate, in mmols, necessary to
correct an acidosis can be calculated by the product of the
treatable volume (0.3  body weight in kg) and the base
deficit. This equates to the amount, in ml, of an 8.4%
bicarbonate solution (1 mmol.ml21). However, over treatment with bicarbonate can have many deleterious effects. It
may not only cause an initial increase in Paco2, worsening
intracellular acidosis, but it may also result in a rebound
metabolic alkalosis after correction of the metabolic acidosis.
Hypokalaemia may result from an intracellular shift of
potassium. The sodium load may cause severe hypernatraemia and increased osmolarity [51], which can result in fluid
overload in patients with heart failure, or intracranial
haemorrhage in neonates [52] and premature infants [53].
Bicarbonate can have adverse vasodilatory effects, especially
under anaesthesia [54], and it has been recommended that
invasive monitoring be used to guide administration [55].
Bicarbonate is best avoided in diabetic ketoacidosis, as it can
precipitate impaired oxygen delivery to the tissues by
unmasking the unopposed action of 2,3-diphosphoglycerate
(2,3-DPG) [56]. Bicarbonate administration causes a
decrease in ionised calcium as a result of increased binding
to albumin, and this may lead to tetany.
To avoid some of these complications, in particular the
increase in Paco2, other buffers have been investigated, e.g.
trihydroxymethylaminomethane (THAM), Carbicarb and
Tribonat. All these agents are anions. Guidelines for the
administration of THAM have been reviewed [57]. The
main advantages are regeneration of bicarbonate with a
lowering of Paco2, slow penetration of the intracellular space,
maintenance of buffering power in hypothermia and
effectiveness in a closed or semiclosed system. However, it
has been associated with severe thrombophlebitis and tissue
necrosis, alkalosis with hypoventilation and arterial vasodilatation, with a consequent decrease in coronary perfusion
pressure. Tribonat is a mixture of THAM, sodium

143

D. G. Maloney et al.  Anions and the anaesthetist

Anaesthesia, 2002, 57, pages 140154
................................................................................................................................................................................................................................................

bicarbonate, phosphate and acetate and has been proposed as

an effective buffer without the side-effects of its parent
compounds [58].
Emergency treatment of hyperkalaemia
Bicarbonate administration results in the intracellular
redistribution of potassium ions. This may restore ECG
morphology and occasionally lead to the return of the
cardiac output after cardiac arrest [59].
Decreasing latency of local anaesthetic blocks
There are numerous studies that show that the onset [60
64] and duration of action [65, 66] of local anaesthetic
blocks can be increased by alkalinisation of local anaesthetics. The proposed mechanism is that the alkaline pH
promotes the local anaesthetic to remain in the unionised
state, in which form it crosses the neural membrane.
Others have suggested that the effect is limited to solutions
containing epinephrine [67], albeit in vitro.
Forced alkaline diuresis for removal of acidic drugs or toxins
Myoglobinuria is a feature of rhabdomyolysis and cast
formation may cause acute renal failure. Myoglobin
remains soluble if the urinary pH is kept . 6.5. Ron
et al. first described a protocol for alkalinizing the urine
with bicarbonate and maintaining a diuresis [68]. Forced
alkaline diuresis has also been used as an adjunct in the
treatment of overdosage with tricyclic antidepressants but
is no longer recommended.
Phosphate
Phosphorus comprises < 1% of total body weight in the
form of phosphate ion. It is the most abundant anion in
the intracellular compartment. The majority (85%) is
stored in bone as hydroxyapatite crystals, with the rest
being stored in the soft tissues (14%) and blood (1%).
Approximately 1 g of phosphorus is ingested daily. This
is well in excess of the dietary needs as only < 70%
(700 mg) is absorbed, mainly through the small intestine.
Phosphorus absorption from the intestine is remarkably
efficient and hypophosphataemia secondary to deficient
absorption is rare in the absence of antacids. There are
both passive and active mechanisms, the latter being more
active when dietary intake is low, with absorption
approaching 90%. The kidneys normally excrete
700 mg.day21 by filtering 6 g and reabsorbing 5.3 g,
thus maintaining a balance between intake and output.
Phosphorus absorption occurs under the influence of
vitamin D, whereas phosphate excretion is primarily
influenced by parathyroid hormone.
Phosphate exists in the body in both organic and inorganic
forms. Most intracellular phosphate is organic, taking the
form of adenosine triphosphate (ATP), nucleic acids,
144

phospholipids and phosphoproteins. Extracellular phosphate

exists mainly in the inorganic form. This is the freely diffusible
form and is the fraction that is measured in plasma. In contrast
to calcium, only 12% of phosphate is bound to albumin and is
thus not affected by changes in serum albumin.
The functions of phosphate in the body have been
outlined in a recent review [69]. The main function is to
store and release energy via compounds such as ATP that
contain high-energy phosphate bonds. Phosphate is an
integral part of the structure of proteins, lipids and bone.
It is an intracellular and renal tubular buffer with a pKa of
6.8. Phosphate achieves maximum buffering power when
it is concentrated in the renal tubules where the pH is
closer to its pKa, although the ammonium buffer system
has a greater buffering capacity. It is the most important
intracellular buffer and is essential for the intermediary
metabolism of carbohydrate [70]. The activity of
phosphofructokinase, the rate-limiting enzyme in glycolysis, is increased by alkalosis. This causes increased
cellular uptake and hypophosphataemia. Therefore, over
ventilation of patients' lungs during surgery may not only
remove the stimulus to breathe (co2) but may also
potentially impair the ability to breathe.
Phosphate is a component of 2,3-DPG, which is
crucial to tissue oxygen delivery [71], and of the
intracellular messengers, cAMP and cyclic guanosine
monophosphate (cGMP). It is necessary for the optimal
functioning of both the immune system and the
coagulation cascade. The process of opsonisation and
phagocytosis by leukocytes, and many of the steps in the
coagulation cascade require ATP.
Regulation of phosphate in the ECF
The normal concentration of phosphate in the ECF is
maintained between 0.8 and 1.0 mmol.l21 by two
mechanisms.
Overflow mechanism. Phosphate filtered through the
glomerulus is normally reabsorbed through the proximal
tubules. Renal tubules have a transport maximum of
0.1 mmol.min21 and when more phosphate is presented,
the excess is excreted and is normally < 1015% of the
filtered load. This reabsorption from the proximal tubule
is dependent on parallel sodium absorption and therefore
diuretics that act on this part of the nephron will cause
phosphaturia.
Role of the parathyroid glands. The parathyroid glands,
which have a major role in the regulation of calcium ions,
can decrease phosphate ion concentration in two different
ways. Parathyroid hormone decreases the transport
maximum for phosphate in the renal tubules, thereby
increasing the proportion of phosphate that is lost in the
q 2002 Blackwell Science Ltd

Anaesthesia, 2002, 57, pages 140154

D. G. Maloney et al.  Anions and the anaesthetist
................................................................................................................................................................................................................................................

urine. It also promotes bone reabsorption with a loss of

phosphate from the ECF.

thought to occur only in patients with pre-existing

myopathies such as that which occurs in chronic alcoholics.

Causes of hypophosphataemia [69, 72]

1 Decreased intake and/or absorption: malnutrition,
aluminium antacids.
2 Increased urinary excretion: carbonic anhydrase
inhibitors, diuretics, the osmotic effect of glycosuria,
glucocorticoids.
3 Internal redistribution: alkalosis (metabolic and
respiratory), total parenteral nutrition and catecholamine
release which cause hyperglycaemia, resulting in an
intracellular shift of phosphate.
4 Haemodilution: hypophosphataemia has long been
recognised as a consequence of intravenous fluid therapy
during surgery. However, as the 2,3-DPG concentration
[71] and postoperative muscle function [73] are unaffected, it is of limited significance in fit patients with
normal cellular stores and does not require treatment.
However, established hypophosphataemia has been shown
to cause respiratory muscle weakness in the general
inpatient population [74]. It may even result in respiratory
failure on its own [75, 76] or in combination with other
causes [77, 78], such as pneumonia. There is a need to
recognise those subgroups of patients who are at risk from
chronic phosphate deficiency [79], as it is likely to
contribute to postoperative hypoventilation. Patients with
chronic obstructive airways disease may have phosphate
depletion from a combination of malnutrition, steroid and
diuretic use [80]. The severe alcoholic also suffers from
malnutrition. In uncontrolled diabetes, phosphate excretion and redistribution are increased [76] and hypophosphataemia may be made worse by insulin administration.
The cardiovascular effects of hypophosphataemia
include ventricular arrhythmias and, in chronic states, a
reversible congestive cardiomyopathy. Erythrocyte and
leukocyte dysfunction may ensue due to ATP depletion.
Haemolytic anaemia may occur because of increased red
cell fragility and 2,3-DPG depletion may result in impaired
oxygen delivery. Neurological features may be pronounced, usually developing over 710 days after hyperalimentation or refeeding has commenced, and consist of
agitation, confusion and hyperventilation that can progress
to seizures, coma and death. Skeletal demineralisation
occurs after chronic deficiency, especially where there is
increased bone turnover and this may mimic osteomalacia.
Metabolic acidosis occurs infrequently considering phosphate's renal buffering capacity and deficiency also impairs
ammonium production. Carbonate ions are also released
during demineralisation and buffer the hydrogen ions.
Conditions in which hydroxyapatite cannot be mobilised,
e.g. vitamin D deficiency and severe magnesium deficiency, may result in metabolic acidosis. Rhabdomyolysis is

Treatment of hypophosphataemia
The mainstay of treatment is supplementation or replacement. Hypophosphataemia has been shown to be a poor
prognostic factor in patients requiring intensive care [81].
Although an increase in single indices such as the P50 for
oxygen dissociation have been demonstrated [82], good
studies showing an improved clinical outcome with
phosphate therapy are, as yet, lacking. Various treatment
protocols have been described. Vannatta et al. [83]
concluded that it is safe to infuse 0.32 mmol.kg21 of
phosphate over a 12-h period. If that failed to increase the
serum phosphate by . 0.2 mmol.l21 then a larger dose of
0.5 mmol.kg21 was given over the following 12 h.
Another regimen described by Rosen et al. [84] used
much larger doses without complications. It involved
giving sodium phosphate 15 mmol (potassium phosphate
if the plasma potassium was , 3.5 mmol.l21) in 100 ml
over 2 h. If the serum phosphate concentration remains
, 0.65 mmol.l21, a repeat dose was recommended.

q 2002 Blackwell Science Ltd

Causes of hyperphosphataemia
1 Decreased renal excretion, increased intestinal reabsorption.
2 Exogenous sources: parenteral administration, poisoning [85].
3 Release after cell damage. This has been shown to
occur commonly after tumour lysis [86] but not after
abdominal aneurysm repair [87].
4 Internal redistribution: acidosis, reduced insulin
concentration.
5 Spurious: thrombocytosis, hyperlipidaemia.
No direct symptoms can be ascribed to hyperphosphataemia. It may cause arrhythmias, hypotension and
hypocalcaemia with associated tetany, renal failure, and
ectopic calcification if high concentrations are maintained
for long periods.
Treatment of hyperphosphataemia
Hyperphosphataemia in concentrations of up to three
times the normal are well tolerated in the short-term.
Therapy should be directed at the cause wherever
possible. This will involve the discontinuation of any
exogenous sources of phosphate. Phosphate binding
antacids given orally, e.g. aluminium hydroxide, may be
used to decrease the amount absorbed from the gastrointestinal tract. An intracellular shift of phosphate can be
achieved by the administration of hypertonic dextrose
solutions while maintaining a normal pH. Ensuring
adequate hydration will promote excretion by the
kidneys. When cell death or tumour lysis is the cause of
145

D. G. Maloney et al.  Anions and the anaesthetist

Anaesthesia, 2002, 57, pages 140154
................................................................................................................................................................................................................................................

hyperphosphataemia, further measures to increase renal

excretion, such as the use of diuretics, may be considered.
Haemodialysis and haemofiltration are both extremely
effective at phosphate removal and would be indicated in
renal insufficiency or where other treatment measures
have failed.
Sulphate
The dietary requirements for sulphur are met by the
breakdown of the amino acids methionine and cysteine.
Sulphur is excreted as the inorganic anion, sulphate. The
disulphide bond is crucial to the three-dimensional
structure and hence the function of proteins. Sulphate
readily forms conjugates with drugs and protein structures.
It is a major biotransformation pathway for phenolic drugs
in humans, particularly in neonates [88]. Propofol, for
example, has both glucuronidated and sulphated conjugates, e.g. 4-quinol sulphate. This pathway is a saturatable
system and it is thought that the depletion of inorganic
sulphate by exogenous substances can affect the metabolism
of neurotransmitters [88]. It can combine with haemoglobin to form sulph-haemoglobin, and cyanosis will occur if
the concentration exceeds 0.5 g.dl21 [89]. The latter is
more likely with the breakdown of phenacetin and
sulphonamides that have numerous sulphur atoms in their
structure. Sulphates may also form conjugates with
polysaccharides and proteoglycans on cell surfaces. These
have been postulated to form the trigger for anaphylaxis
[90]. Sulphates may accumulate up to 10-fold during renal
failure and may cause a metabolic acidosis with an increased
anion gap that responds readily to haemodialysis [91]. It has
also been suggested that sulphates may play a role in the
osteodystrophy of renal failure [92].
Anion gap
The principle of electroneutrality requires that the total
positive electrical charge in body fluids equals the total
negative electrical charge [93]. However, clinical laboratories only measure the most common electrolyte profiles,

i.e. sodium, potassium, chloride and bicarbonate. This

results in an apparent imbalance on the anion side of the
equation that is known as the anion gap.
Anion gap Na1 1 K1 2 Cl2 1 HCO2
3
The normal mean (SD) anion gap is 12 (4) mmol.l21,
although newer methods of assaying chloride have in
some series led to a lower range, and it is best to be guided
by the local laboratory. The factors that may increase or
decrease the anion gap are summarised in Table 3.
Albumin is strongly polyanionic at physiological pH
and is largely responsible for the magnitude of the anion
gap. Therefore, changes in its concentration and dissociation will lead to a corresponding change in the anion gap.
Gabow et al. [94] proposed that the anion gap is decreased
by a factor of 2.5 mmol.l21 for every 1 g.dl21 decrease in
albumin, and this has been validated scientifically [95].
However, it has not yet been shown that hypoalbuminaemia can mask a significant anion gap acidosis [96]. As
sodium is the largest portion of the measured cations,
changes in its concentration can have a profound effect on
the anion gap. Cations that are not included in the
equation, such as magnesium and calcium, normally do
not have an effect on the anion gap unless the disturbance
in their concentration is severe. Halide salts can cause
interference with most chloride assays, giving falsely high
values.
Significance of the anion gap
An increased anion gap in a patient is a poor prognostic
indicator [97] and should prompt consideration of
intensive care or high-dependency care. The anion gap
may also help diagnose the cause of a metabolic acidosis.
To counter an effective decrease in bicarbonate, there
must be a corresponding increase in another anion or a
decrease in cation concentration. If the anion gap and the
cation concentrations remain normal, the balancing anion
must be chloride, as this is the only other anion that is

Table 3 Changes in the anion gap.

Increases in the anion gap
Increased unmeasured anions

Decreases in the anion gap

Increased measured cations

Lactate, ketone bodies, drugs,

e.g. salicylates, albumin or
organic poisons
Phosphates or sulphates
Hypernatraemia

Decreased unmeasured cations

Alkalaemia
Laboratory error

Underfilled blood sample bottles

146

Decreased unmeasured anions

Decreased measured cations
Increased unmeasured cations

Albumin
Hyponatraemia

Metabolic acidosis
Laboratory error

q 2002 Blackwell Science Ltd

Anaesthesia, 2002, 57, pages 140154

D. G. Maloney et al.  Anions and the anaesthetist
................................................................................................................................................................................................................................................

measured, and the cause of the acidosis is likely to be

hyperchloraemia [3, 98].
However, if the anion gap is increased, then the
balancing anion must be one that is unmeasured. If urea,
glucose and electrolyte measurements are normal,
increased anions from renal failure, e.g. phosphate or
sulphate, or diabetes (3-hydroxybutyrate) are unlikely.
Other osmotically active substances must be present and it
is worth noting from the outset that if the anion gap is
. 25 mmol.l21, an organic cause is likely [94]. A lactate
level can be easily measured in most hospitals and a
toxicology screen normally excludes the more common
poisons such as alcohol and salicylates. If there is a high
index of suspicion for abuse of substances such as
methanol or ethylene glycol, then it is worth calculating
the osmolal gap (calculated 2 measured osmolalities).
Early treatment on this basis before poison levels were
available has been shown to be highly effective [99, 100].
The anion gap can also be useful in distinguishing the
primary imbalance in mixed acid2base disorders. The
simplest example is that of diabetic ketoacidosis with
alkalaemia [101]. In this condition, the metabolic acidosis
caused by the ketone bodies is countered by the metabolic
alkalosis caused by persistent vomiting. Arterial blood
gases may not be helpful if both conditions are of equal
severity. An increased anion gap demonstrates the underlying aetiology.
Limitations of the anion gap
It is important to be aware of the limitations [102] and
pitfalls in interpretation [96] of the anion gap.
1 Insensitivity: it has been found that the anion gap is
relatively poor at identifying small changes in ion
concentrations, especially lactate [103], even when they
would be considered clinically significant.
2 Competing factors: if two clinical conditions known
to cause opposing effects on the anion gap were to
coexist, the diagnostic usefulness of the anion gap may be
diminished.
3 Slow onset metabolic acidosis: when a metabolic
acidosis is of sufficiently slow onset and renal function
remains normal, the patient may be able to excrete
enough anions so that the anion gap remains normal. This
may explain why some cases of diabetic ketoacidosis may
present with a normal anion gap [104].
4 Mixed acidosis: occasionally, a metabolic acidosis can
be caused by a combination of conditions that have
opposing effects on the anion gap, potentially masking
one another. It is possible to resolve this diagnostic
conundrum through the calculation of the change in the
anion gap divided by the change in the bicarbonate
concentration (DGap/DHCO32) [105]. However, it
assumes that each mmol of acid will decrease the serum
q 2002 Blackwell Science Ltd

bicarbonate by a similar amount, but this will only hold

true when the proton and its conjugate base have the same
volume of distribution. This is certainly not the case in
lactic acidosis (lactate has a smaller volume of distribution
than the protons) or severe acidosis [96, 106].
5 Unidentifiable anions: it has been found in some
series that the offending anion contributing to the anion
gap cannot be identified despite special effort [94].
6 Laboratory error: an error in the measurement of
any of the cations or anions in the equation will be
translated into an error in the anion gap.
Strong ion difference
The limitations of the anion gap have led some researchers
to adopt a physicochemical approach using the strong ion
difference or gap (SID) [107]. This approach takes into
account not only the principle of electroneutrality but also
the law of conservation of mass. In order to understand the
body's defence of pH, it is necessary to understand the
regulation of three independent variables: SID, Pco2 and
ATOT (total weak acid concentration) [108]. The advantage
of this approach is that it takes account of the smaller cation
concentrations, plasma albumin and phosphate concentrations. It helps explain `dilutional acidosis' and how
hyperlactataemia can exist without acidaemia. It is also
more helpful than the anion gap in establishing the
presence of unidentified anions in the critically ill and in
mixed acid2base disorders [109]. It has been explored in
both adult and paediatric intensive care settings [110, 111].
However, it has not been introduced without controversy
and as yet has not received wide acceptance [112, 113]. It
does provide a different perspective on acid2base balance
and merits further study and debate.
Organic anions

The commonest organic anions in the body are albumin

and lactate and their physiological importance is discussed
here for completeness.
Albumin
Albumin is a medium-sized plasma protein with a
molecular mass of 69 000 Da. It is a polyvalent anion
and can readily accept further negative charge. Its effect
on the anion gap has already been discussed. The normal
half-life of albumin in plasma is between 18 and 20 days,
but this is not the case with exogenously administered
albumin, which has a shorter half-life. Although it is a
medium-sized molecule, it has a transcapillary escape rate
(TER) of 5%.h21. Ninety per cent of the extravascular
albumin returns to the circulation and the other 10%
undergoes catabolism.
147

D. G. Maloney et al.  Anions and the anaesthetist

Anaesthesia, 2002, 57, pages 140154
................................................................................................................................................................................................................................................

Albumin has a fundamental role in the transport of

hormones and ions, and forms a reservoir or buffer
preventing wide variations in serum concentrations.
Competition between drugs does occur and is increased
during hypoalbuminaemia. The effects are more marked
for drugs that are highly protein bound, such as propofol,
than for midazolam or thiopental.
Albumin is thought to be responsible for 7580% of
the colloid osmotic pressure of human plasma. It was
considered that measures to maintain colloid osmotic
pressure in the critically ill could make a difference in
outcome. However, a systematic review [114] reported an
increase in mortality in critically ill patients who had
received albumin. This sparked a vigorous debate with the
methodology in the study being severely criticised [115
117]. Two recent reviews by Boldt [118] and Nicholson
[119] dealt with this vexed issue and are strongly
recommended to the reader.
Albumin is a scavenger of free radicals. It also exhibits
anticoagulant effects in its own right by inhibiting platelet
aggregation and enhancing the inhibition of factor Xa by
antithrombin III. Despite its multiple functions, it is not
crucial for life, as cases of analbuminaemia occur rarely.
These patients display only moderate oedema and are able
to handle drugs. It is thought that other plasma proteins
increase and take on the role of albumin [120].
Lactate
Lactic acidosis is one of the commonest causes of metabolic
acidosis in medical practice. Lactic acid is a weak acid
(pKa 3.8) and is therefore almost completely dissociated
at physiological pH. Lactate is a carbohydrate anion and the
normal concentration in blood is 12 mmol.l21. Lactate

may be produced or consumed by all cells, but skeletal

muscle, the gut, the brain and erythrocytes are the main
producers (1020 mmol.kg21.day21 in adults).
Lactate is an intermediary in carbohydrate metabolism.
Glucose is catabolised in a series of steps to pyruvate by
the Embden-Meyerhof pathway. Pyruvate in non-gluconeogenic organs has four potential fates: conversion to
acetyl coenzyme A by pyruvate dehydrogenase; transamination by alanine aminotransferase with glutamate to
alanine and a-ketoglutarate; carboxylation to oxaloacetate
or malate; reduction by nicotinamide adenine dinucleotide (NAD)H to lactate and NAD1 catalysed by lactate
dehydrogenase. Lactate can be used as a source of energy
but the process is very inefficient and it is normally
transported from the peripheral tissues back to the liver
(Cori cycle) where it is removed from the blood by
gluconeogenesis. Up to 70% of lactate is metabolised in
this way. The other 30% is metabolised by oxidation in the
liver and kidneys.
Lactate concentration in the cell is regulated by the
pyruvate concentration, the redox state of the cell and
hydrogen ion concentration in the cytosol. One step in
the Embden-Meyerhof pathway is essentially irreversible
and is catalysed by phosphofructokinase. In metabolic
acidosis, this enzyme is inhibited and lactate production is
decreased, whereas in metabolic alkalosis, phosphofructokinase is stimulated and lactate production is increased.
Causes of lactic acidosis;
Lactic acidosis was originally classified into two categories,
Type A and Type B. In Type A lactic acidosis, there is an
imbalance between oxygen supply and demand, and lactate
is generated by anaerobic metabolism. In Type B lactic

Table 4 Causes of lactic acidosis.

Type A

Type B

Decreased oxygen supply

Severe hypoxia
Shock
Severe anaemia
Congestive cardiac failure

Congenital errors of metabolism

Glucose 6-phosphate dehydrogenase

deficiency
Fructose 1,6 diphosphate deficiency
Type 1 glycogen storage disease

Excessive oxygen demand

Shivering
Hyperpyrexia
Strenuous exercise
Seizures

Acquired conditions

Drugs: phenformin, ethanol, methanol,

paracetamol, salicylates,
catecholamines
Diabetic ketoacidosis (decreased
pyruvate dehydrogenase activity)
Septicaemia
Neoplastic disease
Parenteral nutrition
(excess fructose/sorbitol)
Vitamin deficiency (thiamine, biotin);
Liver failure (metabolises 50%
of daily lactate production)
Renal failure (metabolises 2530%
daily lactate production)

148

q 2002 Blackwell Science Ltd

Anaesthesia, 2002, 57, pages 140154

D. G. Maloney et al.  Anions and the anaesthetist
................................................................................................................................................................................................................................................

acidosis, tissue hypoxia is irrelevant and abnormalities in

pyruvate metabolism are probably the cause. The causes of
lactic acidosis are listed in Table 4.
Significance of lactic acidosis
Although lactate per se is not detrimental and high levels
are found during severe exercise in healthy adults, it is
associated with a poor prognosis in critically ill patients
[121]. It is the trend rather than any single measurement
that is associated with a high mortality [122]. However,
therapies that have been aimed directly at lowering serum
lactate have failed to produce a significant impact on this
mortality rate [123]. Lactate has been cited as a marker of
tissue hypoxia for over 30 years [124]. This is now being
questioned because it is recognised that there are multiple
mechanisms through which it may increase in the
critically ill [125127]. It is thought that it is the inability
of non-survivors to increase oxygen consumption that is
at fault [128], and lactate remains high in this group as a
result. The exact origin of lactic acidosis, in particular the
anatomical site of lactate production in sepsis, has yet to
be elucidated. The lung is thought to be a significant
producer of lactate in acute lung injury [129].
It has been shown that the understanding of the
purpose and metabolism of lactate in Hartmann's solution
is lacking [130]. It was first added to an intravenous fluid
preparation because it was non-irritant and acted as a
source of bicarbonate and glucose after lactate metabolism. Therefore, it would be prudent to restrict the
volume of Hartmann's solution used in the fluid management of both septic and diabetic patients.
It is reassuring to note that anaesthesia does not affect
serum lactate despite the associated haemodynamic
effects, presumably because an adequate oxygen supply
is maintained throughout. This is the case for both general
[131] and regional anaesthesia [132].
Treatment of lactic acidosis
Although in patients with serum lactate concentrations
. 5 mmol.l21, mortality exceeds 51% at 3 days after ICU
admission [121], therapy that successfully addresses the
cause can result in a successful outcome despite the
absolute lactate level. The acute increases in serum lactate
that are seen peri-operatively, after major vascular surgery
or when tourniquets are used, do not result in increased
mortality, as the aetiology is usually self-limiting [133].
Oxygen delivery must be optimised, which may require
artificial ventilation, measurement of cardiac index and
the rational use of inotropes. Aiming for supranormal
haemodynamic targets has not always been successful in
changing outcomes [128].
q 2002 Blackwell Science Ltd

The use of bicarbonate has not been shown to

influence outcome in septic patients despite improvements in pH. Its use cannot be recommended at present
[126], although it is necessary as the buffer in lactate-free
haemodialysate. Some success has been reported in the
treatment of combined renal failure and lactic acidosis
with the use of continuous haemofiltration using lactate
free dialysate [134]. This treatment does not mask lactate
overproduction [135] and, in the face of sustained
hyperlactataemia, an underlying diagnosis of fulminant
hepatic failure or gangrenous bowel should be considered.
Dichloroacetate stimulates pyruvate dehydrogenase,
thereby reducing lactate with an increase in bicarbonate.
It has been studied extensively but its use does not
decrease mortality and has not found much support [121].
Anions and ion channel receptors

The reversible nature of electrostatic bonds is central to

the proper functioning of ion channels whereby they are
able to mediate fast transmission throughout the central
nervous system (CNS). The intracellular influx of cations
into nervous tissue causes depolarisation and excitation,
whereas anions cause hyperpolarisation and inhibition.
Anions may be either the activating ligands, e.g. gammaaminobutyric acid (GABA) and glutamate, or the ions
that pass through the channels, e.g. chloride and
bicarbonate.
Gamma-aminobutyric acid is the principal fast inhibitory neurotransmitter in the CNS. It acts on the GABA
receptor that belongs to a family of ligand-gated ion
channels that include the acetylcholine receptor, glycine
and 5HT3 receptors. There are three recognised receptor
subtypes, GABA-A, GABA-B and GABA-C. It is the
GABA-A receptor subtype that is pivotal to the action of
anaesthetic, anxiolytic and anticonvulsant drugs. The
structure of the GABA-A receptor is reviewed regularly as
more is learned about its diversity [136138].
Activation of the GABA-A receptor increases the
chloride conductance of the neuron. This chloride influx
hyperpolarises the neuron and prevents an excitatory
postsynaptic potential from being transmitted. The action
of GABA can be enhanced by any of the following:
potentiating the binding of GABA to its own receptor;
increasing the frequency and/or duration of opening of
the ionophore; acting directly on the channel itself in high
concentrations. Barbiturates and neuro-active steroids
[139] have been shown to act via all three mechanisms.
Benzodiazepines, in contrast, increase the frequency of
channel opening but do not act directly on it. Their
enhancement of the affinity of GABA for its receptor is
also limited and their clinical effects are dependent on
receptor occupancy. Propofol increases GABA receptor
149

D. G. Maloney et al.  Anions and the anaesthetist

Anaesthesia, 2002, 57, pages 140154
................................................................................................................................................................................................................................................

binding in a concentration-dependent manner and acts

synergistically with benzodiazepines [140]. This, combined with absence of reversal with flumazenil [141],
suggests that propofol acts at a separate site. Volatile
anaesthetic agents also have been shown to act through
GABA potentiation, again at different sites from intravenous agents [142, 143]. Evidence is increasingly in favour
of there being a protein site of action for general
anaesthetics [144].
There are three major subtypes of glutamate receptors:
N-methyl d-aspartate (NMDA), dl a-amino-3-hydroxy5-methylisoxazole-4-propionate (AMPA) and kainate.
The NMDA receptor is of particular importance for
anaesthetists because it is the known site of action of
ketamine [145]. It is involved in long-term potentiation
that is widely regarded as an experimental analogue for
the mechanism of learning and memory in the brain
[146]. This has implications for the control not only of
chronic pain syndromes [147], such as phantom limb pain
[148], but also for preventing awareness during surgery
under general anaesthesia. The associated calcium influx is
thought to be the principal mediator of neuronal cell
death during anoxia or stroke [149]. This has led to the
development of specific antagonists such as dizocilpine
(MK-801), with the aim of neuroprotection.
Conclusions

Anions play many important physiological and pharmacological roles that are of interest to the anaesthetist. Their
ability to accept a proton makes them the logical choice
for buffer systems, e.g. bicarbonate, urate, glutamate and
phosphate. They provide the power behind cellular
functions (phosphate). They are also involved in gas
exchange (bicarbonate) and transport functions, e.g.
glucose2lactate and albumin. The halogen group is
responsible for making the working environment safer
by decreasing the explosive risk of volatile agents. Because
of the reversible and surmountable nature of the ionic
bond, they are the perfect choice for the receptoragonist
interactions that are now thought to underpin the cellular
mechanisms of anaesthesia.
Acknowledgments

We thank Miss Kelly Barber and Mrs Margaret Johnson

for their help with the preparation of this manuscript.
References
1 Koch SM, Taylor RW. Chloride ion in intensive care
medicine. Critical Care Medicine 1992; 20: 22740.
150

2 Willatts SM. Normal acid base balance. In: Willatts SM, ed.
Lecture Notes on Fluid and Electrolyte Balance. Oxford:
Blackwell Science, 1987: 91104.
3 Miller LR, Waters JH, Provost C. Mechanism of hyperchloraemic metabolic acidosis. Anesthesiology 1996; 84:
4823.
4 Prough DS, Bidani A. Hyperchloraemic metabolic acidosis
is a predictable consequence of intraoperative infusion of
0.9% saline. Anesthesiology 1999; 90: 12479.
5 Scheingraber S, Rehm M, Schmisch C, Finsterer U. Rapid
saline infusion produces hyperchloraemic acidosis in
patients undergoing gynaecological surgery. Anesthesiology
1999; 90: 126570.
6 Dunn SR, Farnsworth TA, Karunaratne WU. Hypokalaemic, hyperchloraemic metabolic acidosis requiring ventilation. Anaesthesia 1999; 54: 5668.
7 McFarlane C, Lee A. A comparison of Plasmalyte 148 and
0.9% saline for intra-operative fluid replacement. Anaesthesia 1994; 49: 77981.
8 Baraka A, Taha S, Ghabach M, Sibaii A, Nader A, Matta
M. Hypertonic saline prehydration in patients undergoing
transurethral resection of the prostate under spinal
anaesthesia. British Journal of Anaesthesia 1994; 72: 2278.
9 Yip WC, Joseph VT, Ho TF. Infantile hypertrophic pyloric
stenosis: surgical experience of 24 cases. Annals of the
Academy of Medicine Singapore 1981; 10: 45660.
10 Booth HS, Bixby EM. Fluorine derivatives of chloroform.
Journal of Industrial Engineering and Chemistry 1932; 24: 637
41.
11 Crandell WB, Pappas SG, MacDonald A. Nephrotoxicity
associated with methoxyflurane anaesthesia. Anesthesiology
1966; 27: 591607.
12 Cousins MJ, Mazze RI. Methoxyflurane nephrotoxicity. A
study of dose response in man. Journal of the American
Medical Association 1973; 225: 161116.
13 Nuscheler M, Conzen P, Schwender D, Peter K. Fluorideinduced nephrotoxicity: fact or fiction? Anaesthesist 1996;
45 (Suppl. 1): S3240.
14 Brown BR Jr. Shibboleths and jigsaw puzzles. The fluoride
nephrotoxicity enigma. Anesthesiology 1995; 82: 6078.
15 Chen TL, Wang MJ, Huang CH, Liu CC, Ueng TH.
Difference between in vivo and in vitro effects of propofol on
defluorination and metabolic activities of hamster hepatic
cytochrome P450-dependent mono-oxygenases. British
Journal of Anaesthesia 1995; 75: 4626.
16 Hoffman J, Konopka K, Buckhorn C, Koop DR, Waskell
L. Ethanol-inducible cytochrome P450 in rabbits metabolizes enflurane. British Journal of Anaesthesia 1989; 63:
1038.
17 Cousins MJ, Fulton A, David W, Haynes G, Whitehead R.
Enflurane nephrotoxicity and pre-existing renal dysfunction. Anaesthesia and Intensive Care 1978; 6: 27789.
18 Strube PJ, Hulands GH, Halsey MJ. Serum fluoride levels
in morbidly obese patients: enflurane compared with
isoflurane anaesthesia. Anaesthesia 1987; 42: 6859.
q 2002 Blackwell Science Ltd

Anaesthesia, 2002, 57, pages 140154

D. G. Maloney et al.  Anions and the anaesthetist
................................................................................................................................................................................................................................................

19 Samuelson PN, Merin RG, Taves DR, Freeman RB,

Calimlim JF, Kumazawa T. Toxicity following methoxyflurane
anaesthesia IV. The role of obesity and the effect of low dose
anaesthesia on fluoride metabolism and renal function.
Canadian Anaesthetists' Society Journal 1976; 23: 46579.
20 Komatsu H, Ueki M, Morita J, Chujo K, Ogli K. Clinical
characteristics and biotransformation of sevoflurane in
paediatric patients during antiepileptic drug therapy.
Paediatric Anaesthesia 1996; 6: 37982.
21 Mazze RI. Fluorinated anaesthetic nephrotoxicity: an
update. Canadian Anaesthetists' Society Journal 1984; 31:
S1622.
22 Higuchi H, Arimura S, Sumikura H, Satoh T, Kanno M.
Urine concentrating ability after prolonged sevoflurane
anaesthesia. British Journal of Anaesthesia 1994; 73: 23940.
23 Wiesner G, Wild K, Schwurzer S, Merz M, Hobbhahn J.
Serum fluoride concentrations and exocrine kidney
function with sevoflurane and enflurane. An open,
randomized, comparative phase III study of patients with
healthy kidneys. Anaesthesist 1996; 45: 316.
24 Frink EJ Jr, Malan TP Jr, Isner RJ, Brown EA, Morgan SE,
Brown BR Jr. Renal concentrating function with prolonged sevoflurane or enflurane anesthesia in volunteers.
Anesthesiology 1994; 80: 101925.
25 Kazama T, Ikeda K. The effect of prolonged administration
of sevoflurane on serum concentration of fluoride ion in
patients. Anesthesiology 1991; 75: A347.
26 Oikkonen M. Isoflurane and enflurane in long anaesthesias
for plastic microsurgery. Acta Anaesthesiologica Scandinavica
1984; 28: 41218.
27 Breheny FX. Inorganic fluoride in prolonged isoflurane
sedation. Anaesthesia 1992; 47: 323.
28 Spencer EM, Willatts SM, Prys-Roberts C. Plasma
inorganic fluoride concentrations during and after prolonged (greater than 24 h) isoflurane sedation: effect on
renal function. Anesthesia and Analgesia 1991; 73: 7317.
29 Kong KL, Tyler JE, Willatts SM, Prys-Roberts C.
Isoflurane sedation for patients undergoing mechanical
ventilation: metabolism to inorganic fluoride and renal
effects. British Journal of Anaesthesia 1990; 64: 15962.
30 Truog RD, Rice SA. Inorganic fluoride and prolonged
isoflurane anesthesia in the intensive care unit. Anesthesia
and Analgesia 1989; 69: 8435.
31 Mazze RI, Calverley RK, Smith T. Inorganic fluoride
nephrotoxicity: prolonged enflurane and halothane
anesthesia in volunteers. Anesthesiology 1977; 46: 26571.
32 Kharasch ED, Hankins D, Thummel KE. Human kidney
methoxyflurane and sevoflurane metabolism: intrarenal
fluoride production as a possible mechanism of methoxyflurane nephrotoxicity. Anesthesiology 1995; 82: 68999.
33 Tinker JH, Baker MT. Sevoflurane, fluoride ion and renal
toxicity. Anesthesiology 1995; 83: 2334.
34 Levine MF, Sarner J, Lerman J, et al. Plasma inorganic
fluoride concentrations after sevoflurane anesthesia in
children. Anesthesiology 1996; 84: 34853.
q 2002 Blackwell Science Ltd

35 Kataria B, Epstein R, Bailey A, et al. A comparison of

sevoflurane to halothane in paediatric surgical patients:
results of a multicentre international study. Paediatric
Anaesthesia 1996; 6: 28392.
36 Ni J, Sato N, Fujii K, Yuge O. Urinary excretion of
hexafluoroisopropanol glucuronide and fluoride in patients
after sevoflurane anaesthesia. Journal of Pharmacy and
Pharmacology 1993; 45: 679.
37 Jarnberg PO, Ekstrand J, Irestedt L, Santesson J. Renal
fluoride excretion during and after enflurane anaesthesia:
dependency on spontaneous urinary pH-variations. Acta
Anaesthesiologica Scandinavica 1980; 24: 12934.
38 Oikkonen M, Rosenberg PH, Collan R. Effect of mannitol
and furosemide on urinary fluoride excretion of surgical
patients anaesthetized with enflurane or halothane. Acta
Anaesthesiologica Scandinavica 1982; 26: 827.
39 Klemmer PJ, Hadler NM. Subacute fluorosis. A consequence of abuse of an organofluoride anaesthetic. Annals of
Internal Medicine 1978; 89: 60711.
40 Casthely PA, Villanueva R. Level of consciousness after
cadaveric kidney transplant under halothane anaesthesia.
Canadian Anaesthetists' Society Journal 1983; 30: 58792.
41 Atallah MM, Geddes IC. Metabolism of halothane during
and after anaesthesia in man. British Journal of Anaesthesia
1973; 45: 46470.
42 Johnstone RE, Kennell EM, Behar MG, Brummund W Jr,
Ebersole RC, Shaw LM. Increased serum bromide
concentration after halothane anesthesia in man. Anesthesiology 1975; 42: 598601.
43 Meldgaard OT, Cold GE. Serum bromide after general
anaesthesia with halothane. Acta Anaesthesiologica Scandinavica 1979; 23: 51318.
44 Morrison JE Jr, Friesen RH. Elevated serum bromide
concentrations following repeated halothane anaesthesia in
a child. Canadian Journal of Anaesthesia 1990; 37: 8013.
45 Duvaldestin P, Mazze RI, Nivoche Y, Desmonts JM. Can
the extent of halothane debromination be predicted
preoperatively? Anesthesia and Analgesia 1979; 58: 4704.
46 Plummer JL, Steven IM, Cousins MJ. Metabolism of
halothane in children having repeated halothane anaesthetics. Anaesthesia and Intensive Care 1987; 15: 13640.
47 Duvaldestin P. Serum bromide concentrations in anesthetists. Anesthesiology 1977; 46: 3756.
48 Peduto VA, Napoleone M. Poisoning caused by chronic
exposure to volatile anesthetics. Molecular mechanisms and
risk anesthetics. Minerva Anestesiologica 1989; 55: 487500.
49 Dale O, Nilsen OG. Displacement of some basic drugs
from human serum proteins by enflurane, halothane and
their major metabolites. An in vitro study. British Journal of
Anaesthesia 1984; 56: 53542.
50 Brooks DK, Feldman SA. Metabolic acidosis. A new
approach to `neostigmine resistant curarisation'. Anaesthesia
1962; 17: 1619.
51 Mattar JA, Weil MH, Shubin H, Stein L. Cardiac arrest in
the critically ill. II. Hyperosmolal states following cardiac
arrest. American Journal of Medicine 1974; 56: 1628.
151

D. G. Maloney et al.  Anions and the anaesthetist

Anaesthesia, 2002, 57, pages 140154
................................................................................................................................................................................................................................................

52 Simmons MA, Adcock EW III, Bard H, Battaglia FC.

Hypernatremia and intracranial hemorrhage in neonates.
New England Journal of Medicine 1974; 291: 610.
53 Thomas DB. Hyperosmolality and intraventricular haemorrhage in premature babies. Acta Paediatrica Scandinavica
1976; 65: 42932.
54 Kaplan JA, Bush GL, Lecky JH, Ominsky AJ, Wollman H.
Sodium bicarbonate and systemic hemodynamics in
volunteers anesthetized with halothane. Anesthesiology
1975; 42: 5508.
55 Cooper DJ, Worthley LI. Adverse haemodynamic effects of
sodium bicarbonate in metabolic acidosis. Intensive Care
Medicine 1987; 13: 4257.
56 Willatts SM. Disturbances of acid base balance. In: Willatts
SM, ed. Lecture Notes on Fluid and Electrolyte Balance.
Oxford: Blackwell Science, 1987: 21126.
57 Nahas GG, Sutin KM, Fermon C, et al. Guidelines for the
treatment of acidaemia with THAM. Drugs 1998; 55: 191
224.
58 Bjerneroth G. Tribonat a comprehensive summary of its
properties. Critical Care Medicine 1999; 27: 100913.
59 Voelckel W, Kroesen G. Unexpected return of cardiac
action after termination of cardiopulmonary resuscitation.
Resuscitation 1996; 32: 279.
60 Roberts JE, MacLeod BA, Hollands RH. Improved
peribulbar anaesthesia with alkalinization and hyaluronidase. Canadian Journal of Anaesthesia 1993; 40: 8358.
61 Capogna G, Celleno D, Costantino P, Muratori F,
Sebastiani M, Baldassini M. Alkalinization improves the
quality of lidocaine2fentanyl epidural anaesthesia for
Caesarean section. Canadian Journal of Anaesthesia 1993; 40:
42530.
62 Armstrong P, Watters J, Whitfield A. Alkalinisation of
prilocaine for intravenous regional anaesthesia. Suitability
for clinical use. Anaesthesia 1990; 45: 9357.
63 Stevens RA, Chester WL, Schubert A, Brandon D, Grueter
JA, Zumrick J. pH-adjustment of 2-chloroprocaine
quickens the onset of epidural anaesthesia. Canadian Journal
of Anaesthesia 1989; 36: 51518.
64 McMorland GH, Douglas MJ, Axelson JE et al. The effect
of pH adjustment of bupivacaine on onset and duration of
epidural anaesthesia for Caesarean section. Canadian Journal
of Anaesthesia 1988; 35: 45761.
65 Tetzlaff JE, Yoon HJ, Walsh M. Regional anaesthetic
technique and the incidence of tourniquet pain. Canadian
Journal of Anaesthesia 1993; 40: 5915.
66 Tetzlaff JE, Walsh M, Yoon HJ. pH adjustment of
mepivacaine decreases the incidence of tourniquet pain
during axillary brachial plexus anaesthesia. Canadian Journal
of Anaesthesia 1993; 40: 1336.
67 Berrada R, Chassard D, Bryssine S, Berthier S, Bryssine B,
Bouletreau P. In vitro effects of the alkalinization of 0.25%
bupivacaine and 2% lidocaine. Annales Francaise d'Anesthesie
et de Reanimation 1994; 13: 1658.
152

68 Ron D, Taitelman U, Michaelson M, Bar-Joseph G,

Bursztein S, Better OS. Prevention of acute renal failure in
traumatic rhabdomyolysis. Archives of Internal Medicine 1984;
144: 27780.
69 Bugg NC, Jones JA. Hypophosphataemia. Pathophysiology,
effects and management on the intensive care unit.
Anaesthesia 1998; 53: 895902.
70 Smith CL, Snowdon SL. Anaesthesia and glucose-6phosphate dehydrogenase deficiency. A case report and
review of the literature. Anaesthesia 1987; 42: 2818.
71 Macdonald R, Guillou PJ, Kester RC. Plasma phosphate
and red cell 2,3-diphosphoglycerate following abdominal
surgery. Anaesthesia 1980; 35: 33944.
72 Berkelhammer C, Bear RA. A clinical approach to
common electrolyte problems: 3. Hypophosphatemia.
Canadian Medical Association Journal 1984; 130: 1723.
73 Andersen PT, Toft E, Hansen AK, Sinkjaer T. Postoperative
hypophosphataemia and muscle function. British Journal of
Surgery 1991; 78: 11416.
74 Gravelyn TR, Brophy N, Siegert C, Peters-Golden M.
Hypophosphatemia-associated respiratory muscle weakness
in a general inpatient population. American Journal of
Medicine 1988; 84: 8706.
75 Lewis JF, Hodsman AB, Driedger AA, Thompson RT,
McFadden RG. Hypophosphatemia and respiratory failure:
prolonged abnormal energy metabolism demonstrated by
nuclear magnetic resonance spectroscopy. American Journal
of Medicine 1987; 83: 113943.
76 Hasselstrom L, Wimberley PD, Nielsen VG. Hypophosphatemia and acute respiratory failure in a diabetic patient.
Intensive Care Medicine 1986; 12: 42931.
77 Newman JH, Neff TA, Ziporin P. Acute respiratory failure
associated with hypophosphatemia. New England Journal of
Medicine 1977; 296: 11013.
78 Brown EL, Jenkins BA. A case of respiratory failure
complicated by acute hypophosphataemia. Anaesthesia
1980; 35: 425.
79 Knochel JP. Disorders of phosphorus metabolism. In: Fauci
AS, Braunwald E, Isselbacher KJ et al., eds. Harrison's
Principles of Internal Medicine. New York: McGraw-Hill,
1998: 225963.
80 Fiaccadori E, Coffrini E, Fracchia C, Rampulla C,
Montagna T, Borghetti A. Hypophosphatemia and phosphorus depletion in respiratory and peripheral muscles of
patients with respiratory failure due to COPD. Chest 1994;
105: 13928.
81 Batchelor GN, Asbury AJ, Greenfield AA. Laboratory data
on ICU admission. Its relationship to mortality. Anaesthesia
1982; 37: 56570.
82 Frans A, Clerbaux T, Detry B, et al. Effect of inorganic ions
on the oxyhemoglobin dissociation curve of severely ill
patients. Pathologie Biologie (Paris) 1998; 46: 814.
83 Vannatta JB, Andress DL, Whang R, Papper S. High-dose
intravenous phosphorus therapy for severe complicated
hypophosphatemia. Southern Medical Journal 1983; 76:
14246.
q 2002 Blackwell Science Ltd

Anaesthesia, 2002, 57, pages 140154

D. G. Maloney et al.  Anions and the anaesthetist
................................................................................................................................................................................................................................................

84 Rosen GH, Boullata JI, O'Rangers EA, Enow NB, Shin B.

Intravenous phosphate repletion regimen for critically ill
patients with moderate hypophosphataemia. Critical Care
Medicine 1995; 23: 120410.
85 Vincent JC, Sheikh A. Phosphate poisoning by ingestion of
clothes washing liquid and fabric conditioner. Anaesthesia
1998; 53: 10046.
86 Stokes DN. The tumour lysis syndrome. Intensive care
aspects of paediatric oncology. Anaesthesia 1989; 44: 1336.
87 Andersen PT, Nielsen LK, Hansen AK, Toft E, Jensen BV.
Serum and urinary phosphate during and after prolonged
muscular ischaemia in non-exercising men and women.
Journal of Internal Medicine 1989; 226: 14955.
88 Levy G. Sulphate conjugation in drug metabolism: role of
inorganic sulphate. Federation Proceedings 1986; 45: 2235
40.
89 Medical disease influencing anaesthesia. In: Atkinson RS,
Rushman GB, Davies NJH, eds. Lee's Synopsis of
Anaesthesia. Oxford: Butterworth-Heinemann, 1994: 373
418.
90 Lasser EC. Sulphated surfaces in sensitivity: an `on-off '
switch for anaphylaxis? Medical Hypotheses 1991; 34: 1319.
91 Kirschbaum B. Effect of hemodialysis on the hypersulphatemia of chronic renal failure. American Society for
Artificial Internal Organs Journal 1998; 44: 31418.
92 Michalk D, Klare B, Manz F, Scharer K. Plasma inorganic
sulphate in children with chronic renal failure. Clinical
Nephrology 1981; 16: 812.
93 Walmsley RN, Guerin MD. Chloride-anion gap. In:
Walmsley RN, Guerin MD, eds. Disorders of Fluid and
Electrolyte Balance. Bristol: Wright, 1984: 7683.
94 Gabow PA, Kaehny WD, Fennessey PV, Goodman SI,
Gross PA, Schrier RW. Diagnostic importance of an
increased serum anion gap. New England Journal of Medicine
1980; 303: 8548.
95 Figge J, Jabor A, Kazda A, Fencl V. Anion gap and
hypoalbuminemia. Critical Care Medicine 1998; 26: 1807
10.
96 Reilly RF, Anderson RJ. Interpreting the anion gap. Critical
Care Medicine 1998; 26: 17712.
97 Brenner BE. Clinical significance of the elevated anion gap.
American Journal of Medicine 1985; 79: 28996.
98 Walmsley RN, White GH. Normal `anion gap' (hyperchloraemic) acidosis. Clinical Chemistry 1985; 31: 30913.
99 Edwards JD. Measuring the anion gap implications for
the bedside. International Journal of Intensive Care 1997; 4:
41.
100 Jacobsen D. Diagnostic work-up of severe metabolic
acidosis. In: Vincent J-L, ed. Yearbook of Intensive Care and
Emergency Medicine. Berlin: Springer-Verlag, 1994: 7004.
101 Zonszein J, Baylor P. Diabetic ketoacidosis with alkalemia:
a review. Western Journal of Medicine 1984; 77: 1924.
102 Kruse JA. Clinical utility and limitations of the anion gap.
International Journal of Intensive Care 1997; 4: 5166.
q 2002 Blackwell Science Ltd

103 Iberti TJ, Liebowitz AB, Papadakos PJ, Fischer EP. Low
sensitivity of the anion gap as a screen to detect
hyperlactataemia in critically ill patients. Critical Care
Medicine 1990; 18: 2757.
104 Elisaf MS, Tsatsoulis AA, Katopodis KP, Siamopoulos KC.
Acid2base disturbances in patients with diabetic ketoacidosis. Diabetes Research and Clinical Practice 1996; 34: 237.
105 Kruse JA. Use of the anion gap in intensive care and
emergency medicine. In: Vincent J-L, ed., Yearbook of
Intensive Care and Emergency Medicine. Berlin: SpringerVerlag, 1994: 68596.
106 DiNubile MJ. The increment in the anion gap: overextension of a concept? Lancet 1988; 2: 9513.
107 Fencl V, Leith DE. Stewart's quantitative acid2base
chemistry: applications in biology and medicine. Respiration
Physiology 1993; 91: 116.
108 Kellum JA. Metabolic acidosis in the critically ill: lessons
from physical chemistry. Kidney International 1998; 66
(Suppl.): S816.
109 Kellum JA, Kramer DJ, Pinsky MR. Strong ion gap: a
methodology for exploring unexplained anions. Journal of
Critical Care 1995; 10: 515.
110 Gilfix BM, Bique M, Magder S. A physical chemical
approach to the analysis of acid2base balance in the clinical
setting. Journal of Critical Care 1993; 8: 18797.
111 Balasubramanyan N, Havens PL, Hoffman GM. Unmeasured anions identified by the Fencl-Stewart method
predict mortality better than base excess, anion gap, and
lactate in patients in the pediatric intensive care unit.
Critical Care Medicine 1999; 27: 157781.
112 Severinghaus JW. Siggaard-Andersen and the `great transAtlantic acid2base debate'. Scandinavian Journal of Clinical
Laboratory Investigation 1993; 214 (Suppl.): 99104.
113 Siggaard-Andersen O, Fogh-Andersen N. Base excess or
buffer base (strong ion difference) as measure of a nonrespiratory acid2base disturbance. Acta Anaesthesiologica
Scandinavica 1995; 107 (Suppl.): 1238.
114 Cochrane Injuries Group Albumin Reviewers. Human
albumin administration in critically ill patients: systematic
review of randomised controlled trials. British Medical
Journal 1998; 317: 23540.
115 Nel MR. Human albumin administration in critically ill
patients. Critical analysis of original studies has to take
place. British Medical Journal 1998; 317: 882.
116 Beale RJ, Wyncoll DL, McLuckie A. Human albumin
administration in critically ill patients. Analysis is superficial
and conclusions exaggerated. British Medical Journal 1998;
317: 884.
117 Soni N. Human albumin administration in critically ill
patients. Validity of review methods must be assessed.
British Medical Journal 1998; 317: 8834.
118 Boldt J. The good, the bad, and the ugly: should we
completely banish albumin from our intensive care units?
British Journal of Anaesthesia 2000; 91: 88795.
119 Nicholson JP, Wolmarans MR, Park GR. The role of
albumin in critical illness. British Journal of Anaesthesia 2000;
85: 599610.
153

D. G. Maloney et al.  Anions and the anaesthetist

Anaesthesia, 2002, 57, pages 140154
................................................................................................................................................................................................................................................

120 Baldo-Enzi G, Baiocchi MR, Vigna G, Andrian C,

Mosconi C, Fellin R. Analbuminaemia: a natural model of
metabolic compensatory systems. Journal of Inherited Metabolic Disease 1987; 10: 31729.
121 Stacpoole PW, Wright EC, Baumgartner TG, et al. Natural
history and course of acquired lactic acidosis in adults.
DCA 2 Lactic Acidosis Study Group. American Journal of
Medicine 1994; 97: 4754.
122 Cowan BN, Burns HJ, Boyle P, Ledingham IM. The
relative prognostic value of lactate and haemodynamic
measurements in early shock. Anaesthesia 1984; 39: 7505.
123 Stacpoole PW, Wright EC, Baumgartner TG, et al. A
controlled clinical trial of dichloroacetate for treatment of
lactic acidosis in adults. The Dichloroacetate2Lactic
Acidosis Study Group. New England Journal of Medicine
1992; 327: 15649.
124 Chamberlain JH, Lis MT. Measurement of blood lactate
levels and excess lactate as indices of hypoxia during
anaesthesia. Anaesthesia 1968; 23: 52133.
125 Kellum JA. Lactate and pHi: our continued search for
markers of tissue distress. Critical Care Medicine 1998; 26:
17834.
126 Gutierrez G, Wulf ME. Lactic acidosis in sepsis: a
commentary. Intensive Care Medicine 1996; 22: 616.
127 Fink MP. Does tissue acidosis in sepsis indicate tissue
hypoperfusion? Intensive Care Medicine 1996; 22: 11446.
128 Hayes MA, Timmins AC, Yau EH, Palazzo M, Watson D,
Hinds CJ. Oxygen transport patterns in patients with sepsis
syndrome or septic shock: influence of treatment and
relationship to outcome. Critical Care Medicine 1997; 25:
92636.
129 Kellum JA, Kramer DJ, Lee K, Mankad S, Bellomo R,
Pinsky MR. Release of lactate by the lung in acute lung
injury. Chest 1997; 111: 13015.
130 White SA, Goldhill DR. Is Hartmann's the solution?
Anaesthesia 1997; 52: 4227.
131 Bacher A, Mayer N, Mittlboeck M, Zadrobilek E.
Anaesthesia and systemic oxygenation. Acta Anaesthesiologica
Scandinavica 1996; 40: 86975.
132 Shende D, Sharma S, Kaushik S. Effects of spinal anaesthesia
on metabolic processes. Indian Journal of Medical Sciences
1998; 52: 196200.
133 O'Hara JF Jr, Tetzlaff JE, Smith MP. Severe perioperative
lactic acidosis: how clinically significant is it? Cleveland
Clinical Journal of Medicine 1994; 61: 31416.
134 Barton IK, Hilton PJ, Treacher DF, Bradley RD. Treatment
of combined renal failure and lactic acidosis by haemofiltration. Clinical Intensive Care 1992; 3: 1968.

154

135 Levraut J, Ciebiera JP, Jambou P, Ichai C, Labib Y, Grimaud

D. Effect of continuous venovenous hemofiltration with
dialysis on lactate clearance in critically ill patients. Critical
Care Medicine 1997; 25: 5862.
136 Mehta AK, Ticku MK. An update on GABA-A receptors.
Brain Research Reviews 1999; 29: 196217.
137 Barnard EA, Skolnick R, Olsen RW, et al. International
Union of Pharmacology. XV. Subtypes of gamma-aminobutyric acid A receptors: classification on the basis of
subunit structure and receptor function. Pharmacology
Reviews 1998; 50: 291313.
138 Sieghart W, Fuchs K, Tretter V, et al. Structure and subunit
composition of GABA(A) receptors. Neurochemistry International 1999; 34: 37985.
139 Paul SM, Purdy RH. Neuroactive steroids. FASEB Journal
1992; 6: 231122.
140 Short TG, Chui PT. Propofol and midazolam act
synergistically in combination. British Journal of Anaesthesia
1991; 67: 53945.
141 Fan SZ, Liu CC, Yu HY, Chao CC, Lin SM. Lack of effect
of flumazenil on the reversal of propofol anaesthesia. Acta
Anaesthesiologica Scandinavica 1995; 39: 299301.
142 Krasowski MD, Koltchine VV, Rick CE, Ye Q, Finn SE,
Harrison NL. Propofol and other intravenous anesthetics
have sites of action on the gamma-aminobutyric acid type
A receptor distinct from that for isoflurane. Molecular
Pharmacology 1998; 53: 5308.
143 Mihic SJ, Ye Q, Wick MJ, et al. Sites of alcohol and volatile
anaesthetic action on GABA(A) and glycine receptors.
Nature 1997; 389: 3859.
144 Franks NP, Lieb WR. Molecular and cellular mechanisms
of general anaesthesia. Nature 1994; 367: 60714.
145 Murray TF. Basic pharmacology of ketamine. In: Bowdle
TA, Horita A, Kharasch ED, eds, The Pharmacological Basis
of Anesthesiology: Basic Science and Practical Applications. New
York: Churchill Livingstone, 1994: 33755.
146 Bliss TV, Collingridge GL. A synaptic model of memory:
long-term potentiation in the hippocampus. Nature 1993;
361: 319.
147 Dickenson AH. A cure for wind up: NMDA receptor
antagonists as potential analgesics. Trends in Pharmacological
Sciences 1990; 11: 3079.
148 Stannard CF, Porter GE. Ketamine hydrochloride in the
treatment of phantom limb pain. Pain 1993; 54: 22730.
149 Werling LL, Fiskum G. The roles of calcium in ischaemic
brain injury. In: Stamford JA, Strunin L, eds, Neuroprotection. London: Bailliere Tindall, 1996: 44559.

q 2002 Blackwell Science Ltd