ENDOCRINE SURGERY

Perioperative needs careful planning, appropriate knowledge and individual-

ized care for each patient.

management of diabetes The traditional concept of type 1 diabetes occurring in young

individuals and type 2 diabetes (T2DM) in older adults is now

and corticosteroid considered obsolete. T2DM is the predominant type of diabetes

and is due to many risk factors including genetic factors, diet,

supplementation lifestyle and behavioural factors.

T2DM is characterized by three important features e insulin
resistance, defective insulin secretion and increase glucose
Saifuddin Kassim production by the liver. In contrast, T1DM is characterized by
Alia Munir defects in beta cell function resulting in absolute insulin defi-
ciency. This is often due to the presence of autoantibodies to
islet cell molecules including insulin, GAD and IA-2. It is
important to be aware that a patient with T1DM is insulin
Abstract
deficient and prone to developing ketosis and acidosis. In
Diabetes is one of the most common endocrinopathies with increasing
addition, ketone prone T2DM or ‘flatbush diabetes’ can occur,
prevalence worldwide. More patients with severe diabetes require elec-
commonly in the black and ethnic minority (BME) groups under
tive and emergency surgery. Diabetes is also associated with increased
physiological stress. There are other specific types of diabetes
postoperative morbidity and mortality. It is essential that good glycae-
such as gestational and monogenic diabetes which are beyond
mic control is maintained to avoid complications related to both hyper-
the scope of discussion of this article. Ultimately, the perioper-
glycaemia and hypoglycaemia. Historically, the means of achieving this
ative management of a patient with diabetes will depend on
has been variable, but there is now a large volume of data underlying
ongoing diabetes therapy, including insulin, oral agent or non-
recommendations for good glycaemic control. The Joint British Dia-
betes Societies for Inpatient Care Group, alongside other societies
insulin injectables.
including Diabetes UK and the British Association of Day Surgery
have provided updated guidance on the management of adults with Diagnosis of diabetes
diabetes undergoing surgery and elective procedures. The manage- The diagnosis of diabetes can be established in the presence of
ment of glycaemia in the intensive care setting is beyond the remit of osmotic symptoms such as polydipsia, polyuria, weight loss or
this article. Ideally, all hospitals should have a diabetes consultant vision blurring AND random blood glucose value of 11.1 mmol/
lead for this service and inpatient diabetes nurse teams to help facilitate litre or higher. However, for asymptomatic individuals, the
optimal management. Corticosteroids are amongst the most common diagnosis of diabetes in a non-pregnant individual can be made
medications prescribed for variety of medical conditions. This can based on: (1) fasting plasma glucose of >7 mmol/litre; (2)
result in suppression of hypothalamicepituitaryeadrenal (HPA) axis plasma glucose in a 2-hour oral glucose tolerance test (OGTT) >
and patients on corticosteroids are unable to mount an effective stress 11.1 mmol/litre; OR (3) glycosylated haemoglobin (HbA1c) > 48
response to surgery. This article aims to give guidance and provide mmol/litre (6.5%). There are certain circumstances where
protocols for the effective perioperative management of diabetes and HbA1c is not appropriate for diagnosing diabetes. These include
glucocorticoid replacement. Data on the need for supra-physiological acutely ill patients, high suspicion of T1DM, acute pancreatic
corticosteroid doses are based on two small randomized controlled damage (e.g. pancreatic surgery), acute high-dose vitamin C or E
trials and other observational studies; highlighting the need for further and haematological factors influencing Hba1c.
research in this area.
Keywords Corticosteroids; diabetes; glucocorticoid replacement; Aims of perioperative diabetes management
perioperative management; surgery
Perioperative hyperglycaemia is associated with increased length
of stay, complications and mortality after surgery.3 The aims of
Prevalence and pathophysiology perioperative management of diabetes include: avoidance of
hypoglycaemia, mainly from metabolic effects of starvation;
Diabetes is one of the most common metabolic disorders with avoidance of ketosis/acidosis especially in patients with T1DM;
increasing prevalence worldwide. Figures from public health avoidance of hyperglycaemia secondary from metabolic stress,
data estimated worldwide prevalence of 415 million; 4.5 million increase in catabolic hormones, inhibition of anabolic hormones;
people with diabetes live in the UK.1 Patients with diabetes and, finally, appropriate use of intravenous fluids. In addition,
occupy around 17% of all beds in hospital for both diabetes- where able, patient autonomy should be maintained.
related and unrelated conditions.2 Good quality diabetes care
Perioperative evaluation

Saifuddin Kassim MBChB MRCP is a Specialist Registrar in Diabetes It is important that during preoperative assessment, a detailed
and Endocrinology at Sheffield Teaching Hospital NHS Foundation history, routine bloods (kidney function test, full blood count
Trust, Sheffield, UK. Conflicts of interest: none declared. and coagulation profile) and updated HbA1c is requested not
Alia Munir MBBCh PhD FRCP is a Consultant Endocrinologist at only to identify potential patients with undiagnosed diabetes but
Sheffield Teaching Hospital NHS Foundation Trust, Sheffield, UK. also to identify the need for glycaemic control in patients with
Conflicts of interest: none declared. established diabetes mellitus. Elevated HbA1c is thought to be a

SURGERY 35:10 596 Ó 2017 Published by Elsevier Ltd.

 ENDOCRINE SURGERY

strong predictor of mortality and morbidity in both cardiac4 and

non-cardiac surgery.5 However, currently there is insufficient Oral agent modification for MINOR perioperative surgical
evidence to suggest an upper limit of HbA1c prior to surgery as management for diabetes patients treated with oral
the urgency of the surgery needs to be taken into account against agent(s)6
the risk associated with poor diabetes control. For elective sur- Oral agents Day of surgery
gery, we recommend consideration of referral to the diabetes
Morning list Afternoon list
team for glucose optimization if HbA1c of more than 69 mmol/
litre (8.5%). This is in line with recommendation from Joint
Metformin No change No change
Diabetes Societies for Inpatient Care Group (JBDS-IP). Ideally,
Sulphonylureas (e.g. gliclazide) Omit morning dose Omit morning
when possible, patients with diabetes should be prioritized on
and evening
the operating list in order to minimize starvation time that can
Pioglitazone No change No change
predispose patients to unnecessary hypoglycaemia and also Dipeptidyl peptidate Omit dose Omit dose
promote early resumption of patients’ normal diabetes medica- inhibitors (DDP-4) or ‘gliptins’
tions. We also encourage referral to individual trust guidelines on Sodium glucose co-transporter Omit dose Omit dose
perioperative management of diabetes and involvement of the 2 (SGLT-2) inhibitors or ‘gliflozin’
diabetes team in ongoing care of complex diabetes patients. (e.g. empagliflozin)
Non-insulin injectables Omit dose Omit dose
Principles of perioperative management of diabetes (e.g. glucagon-like peptide
mellitus 1 (GLP-1) agonist)
The choice of perioperative management has to be individualized
and is dependent on various factors including timing of surgery Table 1
(morning versus evening), complexity of surgery, ongoing
medications and extent of control prior to admission. The target
Type 1 diabetes or insulin treated diabetes
for glucose level is capillary glucose level between 6 and 10
mmol/litre (allow between 4 and 12 mmol/litre). Minor surgery (anticipated one missed meal)
As mentioned previously, these patients should ideally be oper-
Diet controlled diabetes ated on the morning list to minimize starvation. Patients should
This group of patients do not usually require any therapy continue taking their usual insulin dosage the day before the
preoperatively and are not at risk of developing hypoglycaemia. surgery. The dose of insulin on the day of surgery will depend on
In the event of hyperglycaemia (glucose >12 mmol/litre), pa- timing (morning list versus afternoon list) and insulin regime of
tients could be given 0.1 units/kg of subcutaneous rapid acting the individual patient. This is summarized in Table 2.
insulin and have the glucose rechecked in 1 hour’s time. If
hyperglycaemia persists, variable rate intravenous insulin infu- Complex surgery (anticipated > 1 missed meal)
sion (VRII) should be commenced. The diabetes team should All patients in this group should have VRII started as part of
then be consulted post-surgery as the patient may require further perioperative diabetes care. Long-acting insulin can be continued
modification in their diabetes management. whilst on VRII. There are no comparable studies showing benefit
or harm in the practice of continuing long-acting insulin whilst
T2DM treated with oral agent or/and non-insulin on VRII (at either full or reduced dose). Many centres now
injectables advocate the practice of continuing long-acting insulin in order to
In this group of patients, perioperative management will depend make the transition to subcutaneous insulin smoother. It also has
on the anticipated starvation period before and after surgery and the added advantage of reducing the incidence of ketosis in the
the complexity of surgery (minor versus complex). event of accidental discontinuation of VRII in patients suffering
from T1DM or ketone prone diabetes.
Minor surgery (anticipated one missed meal)
Patients having minor surgery should ideally be scheduled for Variable rate intravenous insulin infusion protocol
surgery on the morning list to minimize the period of starvation.
The detailed management and modification of the oral medica- There are two independent separate components to VRII e in-
tion and non-insulin injectables are described in Table 1. sulin infusion and fluid infusion. Normal daily fluid intake and
electrolyte requirements for adults is 25e30 ml/kg/day of water,
Complex surgery (anticipated > 1 missed meal) 1 mmol/kg/day of sodium, potassium and chloride, and 50e100
All patients in this sub-group should have VRII started as part of g/day of glucose. Fluid requirements will vary significantly in
the perioperative diabetes care. Generally, the VRII can be surgical patients depending on the need for resuscitation or
commenced early morning (e.g. 6 am) for a patient on the replacement of ongoing abnormal losses.7 In the context of VRII,
morning list of surgery or late morning (e.g. 11 am) for the af- we recommend the use of 0.45% saline with 5% glucose with
ternoon list. All oral hypoglycaemic agents with non-injectable 0.3% KCL (40 mmol/litre) at 125 ml/hour or 5% dextrose with
insulin should be omitted on the day of surgery as patients will 40 mmol KCL at 125 ml/hour if serum potassium is between 3.5
be commenced on VRII. These can be stopped once the patient and 5.5 mmol/litre. Additional requirement of fluid and/or
starts eating and drinking normally. adjustment (including fluid rate) can be made individually based

SURGERY 35:10 597 Ó 2017 Published by Elsevier Ltd.

 ENDOCRINE SURGERY

Measurement of capillary blood glucose should be taken

Insulin modification for MINOR perioperative surgical hourly initially to ensure accurate insulin infusion is achieved.
management for diabetes patients treated with insulin VRII can be discontinued once patient is able to eat and drink.
(for both T1DM and T2DM (insulin treated) The VRII can normally be discontinued 30e60 minutes after
Insulin Day of surgery administration of subcutaneous insulin.

Morning list Afternoon list

Perioperative corticosteroid management
Once daily insulin No change No change There are three main types of hormone produced by adrenal
(both morning or cortex including glucocorticoids, mineralocorticoids and
evening) androgen (sex steroids). Glucocorticoids are secreted in relatively
Twice daily mixed Halve the usual morning Halve the usual high amounts of around 5.7e7.4 mg/m2 per day8 under the
insulin (e.g. dose morning dose control of adrenocorticotropic hormone (ACTH), produced and
Humulin M3, secreted by anterior pituitary gland. ACTH is secreted in a pul-
Novomix 30) satile manner with a circadian rhythm with levels highest on
Multiple daily Basal long acting e No Basal long acting wakening in the morning and reaching nadir values in the eve-
injection (MDI) change e No change ning. The hypothalamicepituitaryeadrenal (HPA) axis is an
regime/‘basal- Omit morning and lunchtime Take usual important component of the response to stress in the human
bolus’ regime short acting insulin morning short body.
acting but omit Glucocorticoids have various principal sites of action in the
afternoon short human body. It is antagonistic to insulin and can contribute to a
acting insulin diabetogenic effect by increasing hepatic glycogen deposition,
Continuous Contact diabetes team for Contact diabetes increasing gluconeogenesis and increasing peripheral insulin
subcutaneous further advise team for further resistance. In relation to surgery and stress, glucocorticoids play
insulin infusion advise an essential role in blood pressure homeostasis. They increase
(CSII) or also blood pressure by increasing sensitivity to pressor agents such as
known as ‘insulin catecholamines, reducing nitric oxide mediated endothelial
pump’ relaxation and causing sodium retention and potassium loss in
the kidneys.
Table 2 The anti-inflammatory effects of glucocorticoids were first
discovered in the late 1940s, when it was found to relieve
symptoms in patients with rheumatoid arthritis.9 Since then, it
on regular assessment of patients both clinically and has been used for a variety of inflammatory and immunological
biochemically. conditions such as inflammatory bowel disease, haematological
The infusion of insulin will consist of 50 units of Actrapid malignancy, chronic obstructive pulmonary disease and vascu-
insulin (Insulin Aspart, Humulin S or Insulin Lispro can also be litis. The use of glucocorticoids at supra-physiological doses and/
used) that is added to 49.5 ml of 0.9% sodium chloride in a 50 ml or long-term therapy will suppress the HPA axis. Other factors
Luer lock syringe. This infusion should run along the fluid that could influence HPA axis suppression includes potency/type
regime described above and titrated to achieve glucose level of glucocorticoid (see Table 4), systemic versus compartmental
between 6 and 10 mmol/litre (see Table 3). therapy,10 daily versus alternate day therapy, day versus night
administration and duration of therapy. The full recovery of the
HPA axis after cessation of steroids can take as long as one year
or more.
Glucocorticoid deficiency can also occur in primary adrenal
Suggested variable rate infusion of insulin (VRII) disease (primary hypoadrenalism), secondary hypoadrenalism
(ACTH deficiency) and tertiary hypoadrenalism. The cause of
Capillary glucose Initial rate of insulin infusion (unit/hour) primary hypoadrenalism includes Addison’s disease, intra-
(mmol/litre) adrenal haemorrhage, bilateral adrenalectomy, metastatic
Regime one Regime two (for insulin
resistant patients) tumour/infiltration and infection. Secondary hypoadrenalism
can be caused by any disease process that interferes with ACTH
<4 0 0 production such as pituitary lesions/tumours, trauma, genetic
4.1e7.0 1 2 disorders and drugs. Tertiary hypoadrenalism is caused by any
7.1e9.0 2 3 process that involves the hypothalamus and disturbs the CRH
9.1e11.0 3 4 secretion such as hypothalamic tumours (e.g. craniophar-
11.1e14.0 4 5 yngioma) or an infiltrative process (e.g. sarcoidosis).
14.1e17.0 5 6
17.1e20 6 7 Evaluation of HPA axis
>20 7 8
We suggest that any patient who has potential HPA axis sup-
Table 3 pression undergo as evaluation using the ACTH stimulation test,

SURGERY 35:10 598 Ó 2017 Published by Elsevier Ltd.

 ENDOCRINE SURGERY

Relative potency of synthetic steroids11

Steroids Equivalent Anti-inflammatory Mineral-corticoid Hypothalamic e
doses (mg) action action pituitary e
adrenal
suppression

Hydrocortisone 20 1 1 1
Prednisolone 5 4 0.75 4
Methylprednisolone 4 5 0.5 4
Dexamethasone 0.75 25 0 17

Table 4

Indications for steroids cover

Exogenous steroid use Expected HPA Axis activity Recommended perioperative steroid use

Steroid dose 5 mg or less (taken in the Normal or maybe suppressed activity Yes. We recommend SST to assess HPA axis
morning) of uncertain duration prior surgery. (If in doubt, give additional
steroid cover)
Steroid use at any dose for less than 3 weeks Normal No additional steroid cover advised
Steroid dose more than 5 mg for more than 3 Suppressed activity Yes
weeks
Any steroid dose with clinical features of Suppressed activity Yes
Cushing’s syndrome
Adrenal insufficiency secondary to primary Suppressed activity Yes
adrenal disease, pituitary disease or
hypothalamus disease

Table 5

or ‘Synacthen Test’. A normal cortisol response confirms corticosteroids can lead to an Addisonian crisis and death.13
adequate cortisol reserve and rules out the need for additional However, the precise doses of corticosteroids that would be
perioperative replacement. The only exception where steroids adequate in the perioperative period is still unclear. Very high
may be required in a normal response to ACTH is in patients with doses of corticosteroid replacement has been shown to cause
recent pituitary insult/ACTH deficiency. In this case, it will be detrimental side effects such as cardiac arrhythmia,14 potential
safer to assume that the patient has adrenal insufficiency. for delayed healing and infections15 and myopathy in prolonged
stay in intensive care unit.
Corticosteroid response to surgery
Recommendation for perioperative steroid
Surgery is a potent activator of ACTH, which can be measured
supplementation
within minutes of the start of surgery. The increase of plasma
cortisol peaks within 4e6 hours after surgery and can be sus- Given the lack of clarity on adequate dosage of perioperative
tained up to 72 hours postoperatively with a value of more than corticosteroids, a sensible approach in providing adequate but
1500 nmol/l.12 It is clear that inadequate perioperative not excessive corticosteroid replacement should be based on type
and duration of surgery, duration and type of ongoing gluco-
corticoid medications and aetiology of cortisol insufficiency
Recommended glucocorticoid coverage based on type of
(adrenocortical insufficiency versus suppression secondary to
surgery (in case of uncertainty or concerns regarding
chronic corticosteroid therapy). The recommendations are sum-
high BMI, high dose may be used)
marized in Tables 5 and 6. Patients with uncertain history highly
Induction Post-surgery suspicious of HPA axis suppression should receive empirical
glucocorticoid treatment.
Minor 25 mg hydrocortisone Nil needed
Moderate 50 mg hydrocortisone 25 mg hydrocortisone Conclusion
QDS for 24 hours
Major 100 mg hydrocortisone 100 mg hydrocortisone In summary, patients with diabetes and those needing cortico-
QDS for 24e72 hours steroid cover need careful perioperative planning to reduce the
morbidity of surgery. Health professionals must be aware of the
Table 6 basic principles of perioperative management as outlined in this

SURGERY 35:10 599 Ó 2017 Published by Elsevier Ltd.

 ENDOCRINE SURGERY

article. Helpfully, there is clear guidance on the nature and extent 6 JBDS-IP. Management of adults with diabetes undergoing sur-
of interventions for perioperative diabetes and steroid manage- gery and elective procedures: improving standards. 2015. Avail-
ment for these patients. There is a paucity of good quality data on able from: http://www.diabetologists-abcd.org.uk/JBDS/JBDS.
the absolute levels of corticosteroid replacement. Corticosteroid htm (accessed 17 Mar 2017).
supplementation is essential but there is no evidence to support 7 Powell-Tuck J, Gosling P, Lobo DN, et al. British consensus
overzealous treatment in patients undergoing surgery. Early guidelines on intravenous fluid therapy for adult surgical patients
liaison with the local diabetes and endocrine teams and ensuring GIFTASUP. 2009, http://www.bapen.org.uk/pdfs/bapen_pubs/
compliance with local protocols will help ensure optimum care of giftasup.pdf.
these patients to improve outcomes and minimize hospital 8 Debono M, Ross RJ, Newell-Price J. Inadequacies of glucocorti-
stay. A coid replacement and improvements by physiological circadian
therapy. Eur J Endocrinol 2009; 160: 719e29.
9 Munck A, Guyre PM, Holbrook NJ. Physiological functions of
REFERENCES
glucocorticoids in stress and their relation to pharmacological
1 Diabetes UK. Facts and stats. 2015. Available from: https://www.
actions. Endocr Rev 1984; 5: 25e44.
diabetes.org.uk/About_us/What-we-say/Statistics (accessed 15
10 Mortimer K, Tata L, Smith C, West J, Harrison T. Oral and inhaled
Jul 2016).
corticosteroids and adrenal insufficiency: a case-control study.
2 NHS digital. National diabetes audit. 2015. Available from: http://
Thorax 2006; 61: 405e8.
content.digital.nhs.uk/searchcatalogue (accessed 17 Mar 2017).
11 Brunton L, Lazo J, Parker K. The pharmacological basis of ther-
3 Gale S, Sicoutris C, Reilly P. Poor glycemic control is associated
apeutics. Goodman and Gilmans, 2006.
with increased mortality in critically ill trauma patients. Am Surg
12 Nicholson G, Burrin JM, Hall GM. Peri-operative steroid supple-
2007; 73: 454e60.
mentation. Anaesthesia 1998; 53: 1091e104.
4 Subramaniam B, Lerner A, Novack V. Increased glycemic vari-
13 Udelsman R, Ramp J, Gallucci W. Adaptation during surgical
ability in patients with elevated preoperative HbA1C predicts
stress. A reevaluation of the role of glucocorticoids. J Clin 1986;
adverse outcomes following coronary artery bypass grafting sur-
77: 1377.
gery. Anesth Analg 2014; 118: 277e87.
14 Erstad B. Severe cardiovascular adverse effects in association
5 Gustafsson UO, Thorell A, Soop M, Ljungqvist O, Nygren J.
with acute, high-dose corticosteroid administration. DICP Ann
Haemoglobin A1c as a predictor of postoperative hyperglycaemia
Pharmacother 1989; 23: 1019e23.
and complications after major colorectal surgery. Br J Surg 2009;
15 Miller RD. Miller’s anesthesia. Elsevier, 2010.
96: 1358e64.

SURGERY 35:10 600 Ó 2017 Published by Elsevier Ltd.