Editorial III
24 McCarthy JR, Trigg R, John C, Gough MJ, Horrocks M. Patient
satisfaction for carotid endarterectomy performed under local
anaesthesia. Eur J Vasc Endovasc Surg 2004; 27: 654 –9
25 Cassuto J, Sinclair R, Bonderovic M. Anti-inflammatory
properties of local anesthetics and their present and
potential clinical implications. Acta Anaesthesiol Scand 2006; 50:
265 – 82
26 Biki B, Mascha E, Moriarty DC, Fitzpatrick JM, Sessler DI, Buggy DJ.
Anesthetic technique for radical prostatectomy surgery affects
British Journal of Anaesthesia 103 (6): 789– 91 (2009)
doi:10.1093/bja/aep315
Editorial III
Management of diabetes during surgery: 30 yr of the Alberti regimen
In 1979, the British Journal of Anaesthesia published an
article by Alberti and Thomas1 in which they introduced
the i.v. infusion of a premixed bag of glucose – insulin –
potassium for the metabolic management of diabetic
patients in the perioperative period. The Alberti regimen
has been superceded in many centres by the use of separ-
ate infusions of insulin (usually 1 U ml21) and glucose
with or without potassium.2 Although the separate infu-
sions offer greater flexibility, they lack the inherent safety
of the combined regimen.
The article reviewed the management of diabetic surgi-
cal patients in detail, in addition to the description of the
Alberti regimen, and it is apposite to re-examine this
seminal paper when there is great interest currently in
glucose control in surgical patients, both diabetic and non-
diabetic.3 In their introductory remarks, Alberti and
Thomas noted the need for a logical and straightforward
set of guidelines for treating diabetic patients during and
after surgery and that many of the regimens recommended
in 1979 were irrational if not dangerous. It is to be hoped
that 30 yr later, the current guidelines are both logical and
safe, but there is little evidence to support the regimens
described in standard anaesthetic textbooks.4 – 6
The authors discussed first the normal endocrine and
metabolic response to starvation and surgery and then
how this may be modified in the diabetic patient. They
commented that starvation was generally the fate of the
surgical patient, despite the best intentions of the clini-
cians. Although there have been undoubted improvements
in decreasing the duration of preoperative starvation,
postoperative starvation has not received the same atten-
tion in many centres. The detailed description of the
metabolic changes occurring in starvation is striking
even 30 yr later and it is regrettable that this key knowl-
edge is still lacking in some clinicians managing surgical
patients. Alberti and Thomas emphasized that during
789
cancer recurrence: a retrospective analysis. Anesthesiology 2008; 109:
180–7
27 Exadaktylos AK, Buggy DJ, Moriarty DC, Mascha E, Sessler DI.
Can anesthetic technique for primary breast cancer surgery
affect recurrence or metastasis? Anesthesiology 2006; 105: 660– 4
28 Sessler DI, Ben-Eliyahu S, Mascha EJ, Parat MO, Buggy DJ. Can
regional analgesia reduce the risk of recurrence after breast
cancer? Methodology of a multicenter randomized trial. Contemp
Clin Trials 2008; 29: 517 – 26
starvation in type 1 diabetics, the restraining effects of
basal insulin on catabolism are lost and that catabolism
Downloaded from bja.oxfordjournals.org by guest on May 14, 2011
‘runs riot’ when catabolic hormone secretion increased
during surgery. The provision of insulin with sufficient
carbohydrate, 180 g day21, was necessary to mitigate
uncontrolled catabolism.
The description of the metabolic response to surgery
was also comprehensive and it is only in the section on
the effects of anaesthesia, where diethyl ether and chloro-
form are mentioned, that the lack of an anaesthetic coau-
thor is apparent. By 1979, anaesthetic techniques such as
extensive neuraxial block and high-dose opioid anaesthesia
had been shown to reduce the catabolic hormonal response
to some surgical procedures with a consequent decreased
hyperglycaemia.7 8 It had been suggested that these tech-
niques may be appropriate for patients with metabolic
problems such as diabetes.8
Alberti and Thomas summarized the aims of treatment
of diabetes as the rapid recovery from surgery with
minimal metabolic problems, such as hypoglycaemia and
ketoacidosis, and minimal complications such as infection,
delayed wound healing, and cardiovascular events. These
aims are equally appropriate in 2009 and it is only when
the authors state that it is beneficial, if not obligatory, for
the diabetic patient having anything but the most minor
surgery to be admitted to hospital 24– 48 h before the
operation for assessment and stabilization, that the reader
is reminded that the article was published in 1979. The
authors included original experimental data in the review
article to support their recommendations, some of which
were published in more detail in 1984.9 The thoroughness
of their investigation of diabetic patients undergoing
surgery was shown by the measurement of circulating
glucose, NEFA, 3-hydroxybutyrate, pyruvate and lactate
concentrations, and plasma and urinary electrolyte and
urea values for up to 72 h after surgery.
 Editorial III
The recommendations for managing type II diabetic
patients were based on a comparison of the glucose –
insulin –potassium regimen with no therapy. They con-
cluded that there was improved metabolic control with the
Alberti regimen in all diabetic patients treated with oral
hypoglycaemic drugs except in those well-controlled
patients undergoing minor surgery. It was recommended
that long-acting sulphonylureas, such as chlorpropamide,
should be stopped at least 3 days before surgery and bigua-
nides also discontinued because of the risk of lactic acido-
sis. A recent evaluation of metformin indicated that the
problems of lactic acidosis may have been exaggerated,10
and the inadvertent failure to stop metformin before
cardiac surgery did not increase mortality and morbidity.11
In type 1 diabetic patients, four regimens were exam-
ined: no insulin – no glucose, insulin s.c. with glucose i.v.,
combined regimen for up to 24 h, and combined regimen
for 72 h. The no insulin –no glucose regimen failed to
control glucose adequately and was associated with a
marked increase in circulating NEFA and ketone body
concentrations, raised urea excretion, and markedly negative
potassium, phosphate, calcium, and magnesium balances.
These results emphasized the importance of examining
markers of lipid and protein metabolism and also blood
glucose when assessing metabolic control in diabetic
patients. Short-term starvation in an unstressed type 1 dia-
betic may be associated with a decrease in blood glucose
but this is at the expense of enhanced catabolism.12 The
regimen of insulin s.c. followed by an infusion of glucose
i.v. was no better than no glucose–no insulin. The continu-
ous infusion regimens provided the best metabolic control
with stable blood glucose values, low circulating ketone
body concentrations, and improved nitrogen and potassium
balances. The recommendations for type 1 diabetic patients
were to use the Alberti regimen in all patients undergoing
surgery and to continue the infusion until the patient was
eating (ambulatory surgery was in its infancy in 1979).
Practical aspects of the use of the combined regimen
were discussed in detail. There were concerns that a vari-
able quantity of insulin would be adsorbed to the plastics in
the infusion bag and giving set but the authors found that
the loss of insulin was only 10–28% which they considered
acceptable. The use of a separate concentrated insulin infu-
sion was being investigated at this time after its success in
the treatment of diabetic ketoacidosis.13 14 However, Alberti
and Thomas advocated the intrinsic safety of the combined
infusion over the possibility of failure of an infusion pump.
Repeated measurement of blood glucose was stressed as the
keystone of successful management of diabetic patients;
before surgery, once or twice during surgery, and 2–3
hourly until the patient was eating. The use of urinary
glucose values to guide treatment was summarily dismissed
as a ‘security blanket by the nervous and ignorant’ because
the values were retrospective, insensitive and urine may be
unavailable. Plasma potassium concentrations were also
measured regularly, about half as frequently as blood
790
glucose. Several factors were noted to increase insulin
requirements: the severity of surgery, particularly cardiac
surgery, the presence of an infection, and the administration
of glucocorticoids and catecholamines. The problem of
obesity with concomitant insulin resistance was recognized
and it was recommended that if the diabetic patient was
more than 50% heavier than their ideal body weight, then
the dose of insulin should be increased two-fold.
The standard i.v. fluid in the Alberti regimen was
glucose 10% infused at a rate of 100 to 125 ml h21. The
risk of hyponatraemia from the prolonged infusion of
glucose solutions is now well recognized.15 The authors
stated that sodium chloride solution 0.9% was an accepta-
ble fluid for i.v. use in diabetic patients but that lactate-
containing solutions such as Hartmann’s solution should
be avoided.16 The evidence to support the contention that
lactate-containing solutions should not be used because
they enhance the glycaemic response to surgery is very
weak and was reviewed recently.17 If the volume of
glucose solution i.v. needed to be restricted, for example,
to give sodium chloride 0.9%, then the use of 20% or
50% glucose infused centrally was recommended.
Downloaded from bja.oxfordjournals.org by guest on May 14, 2011
The review ended with an Appendix that was a recipe
of guidelines for the care of diabetic patients during and
after surgery. The paper ended with ‘Special Notes’ which
included the following advice that is still highly relevant
today:
Continuous therapy. If the infusion stops, the
patient’s ECF will contain no effective insulin within
30 min. Therefore, ensure that therapy is continuous.
Normoglycaemia. Aim to maintain blood glucose
between 5 and 10 mmol l21.
Common sense. These are guidelines. The
starting regimen (10% glucose 500 mlþinsulin
10 unitsþKCl l g) is based on experience and a com-
parative study, but it is not meant to be unthinkingly
adhered to. Patients will always vary, so therapy must
be flexible and common sense applied.
After 30 yr, this article still has the sparkle of a precious
gem; it contains the science, practical details, and wisdom
necessary to manage surgical diabetic patients successfully.
Close reading of the text should be obligatory for all anaes-
thetists and intensivists. Without doubt it is a citation classic.
G. M. Hall*
Department of Anaesthesia and Intensive Care Medicine
Division of Cardiac and Vascular Sciences
St George’s, University of London
London SW17 0RE
UK
*E-mail: ghall@sgul.ac.uk
 Editorial III

References 9 Thomas DJ, Platt HS, Alberti KG. Insulin-dependent diabetes

during the perioperative period. An assessment of continuous
1 Alberti KGMM, Thomas DJB. The management of diabetes during
glucose – insulin – potassium infusion, and traditional treatment.
surgery. Br J Anaesth 1979; 51: 693 – 710
Anaesthesia 1984; 39: 629 – 37
2 McAnulty GR, Robertshaw H, Hall GM. Anaesthetic management
10 Holstein A, Stumvoll M. Contraindications can damage your
of patients with diabetes mellitus. Br J Anaesth 2000; 85: 80 – 90
health—is metformin a case in point? Diabetologia 2005; 48:
3 Webster NR, Galley HF. Does strict glucose control improve
2454 – 9
outcome? Br J Anaesth 2009; 103: 331 – 4
11 Duncan AI, Koch CG, Xu M, et al. Recent metformin ingestion
4 Blanshard H. Endocrine and metabolic disease. In: Allman G,
does not increase in-hospital morbidity or mortality after cardiac
Wilson IH, eds. Oxford Handbook of Anaesthesia. Oxford: Oxford
surgery. Anesth Analg 2007; 104: 42 – 50
University Press, 2006; 150 – 5
12 Alberti KGMM. Role of glucagon and other hormones in devel-
5 Pinnock CA, Haden RM. Special patient circumstances. In: Pinnock
opment of diabetic ketoacidosis. Lancet 1975; 1: 307– 11
C, Lin T, Smith T, eds. Fundamentals of Anaesthesia. Cambridge:
13 Page MMcB, Alberti KGMM, Greenwood R, et al. Treatment of
Greenwich Medical Media, 1999; 122 – 4
diabetic coma with continuous low-dose infusion of insulin. Br
6 Grant IS, Nimmo GR, Nimmo S. Intercurrent disease and anaes-
Med J 1974; 2: 687 – 90
thesia. In: Aitkenhead AR, Smith G, Rowbotham DJ, eds. Textbook
14 Leslie RDG, Mackay JD. Intravenous insulin infusion in diabetic
of Anaesthesia. London: Elsevier, 2007; 466 –8
emergencies. Br Med J 1978; 2: 1343 –4
7 Bromage PR, Shibata HR, Willoughby HW. Influence of prolonged
15 Chung HM, Kluge R, Schrier RW, Anderson RJ. Postoperative hypo-
epidural blockade on blood sugar and cortisol responses to oper-
natraemia: a prospective study. Arch Intern Med 1986; 146: 333–6
ations on the upper part of the abdomen and thorax. Surg
16 Thomas DJB, Alberti KGMM. Hyperglycaemic effects of
Gynaecol Obstet 1971; 132: 1051 – 6
Hartmann’s solution during surgery in patients with
8 Hall GM, Young C, Holdcroft A, Alagband-Zadeh J. Substrate
maturity-onset diabetes. Br J Anaesth; 1978; 50: 185 – 8
mobilisation during surgery. A comparison between halothane and
17 Simpson AK, Levy N, Hall GM. Perioperative iv fluids in diabetic
fentanyl anaesthesia. Anaesthesia 1978, 33: 924 – 30
patients—don’t forget the salt. Anaesthesia 2008; 63: 1043– 5

Downloaded from bja.oxfordjournals.org by guest on May 14, 2011

791