Pediatric Anesthesia 2004

14: 716–723

Review article

Diabetes mellitus and the pediatric anesthetist V. CHADWICK

MB ChB FRCA

A N D K .A . WIL K IN S O N

FRCA

MRCP DCH

Department of Anaesthesia, Norfolk and Norwich University Hospital NHS Trust, Colney Lane, Norwich, NR4 7UY, UK

Keywords: perioperative; diabetes mellitus; pediatric; anesthetic management

Introduction Diabetes mellitus (DM) is a relatively common problem in the perioperative period in adults. It also occurs in childhood and there is evidence that the prevalence is increasing. Children may present for minor surgery unrelated to their diabetes, often to the nonspecialist center. Even in the most ideal circumstances, children with DM experience marked fluctuations in blood glucose levels. Added to this, the stresses of illness and surgery and their glucose control can be challenging. The rule during the perioperative period is to avoid hypoglycemia or marked hyperglycemia. Therefore, a well thought out scheme of management is essential. An updated review of this important problem is presented in this article.

Classification Recent advances in knowledge of the etiology and pathogenesis of DM has led to revised classifications. The changes have been made in an attempt to describe diabetes on the basis of the pathogenic process that leads to hyperglycemia, as opposed to the criteria such as age of onset or type of therapy (1). Table 1 presents a recent classification of the pediatric population (2).

Correspondence to: K.A. Wilkinson, Department of Anaesthesia, Norfolk and Norwich University Hospital NHS Trust, Colney Lane, Norwich, NR4 7UY, UK (email: kathy.wilkinson@nnuh. nhs.uk).

716

Type 1 diabetes (insulin-dependent diabetes mellitus, IDDM) accounts for 90% of cases in children. It is by far the most common metabolic abnormality in young people and its onset is usually seen in late childhood and early adolescence. Patients present with hyperglycemia, ketoacidosis and are reliant on insulin. Early onset of type 2 diabetes (noninsulin-dependent diabetes mellitus) is very rare in childhood and has previously been reported to account for 2–3% of cases (3). However, recent studies have shown the prevalence of this disorder to be increasing. This may be related to the increasing prevalence of obesity in children (4). Early onset of type 2 diabetes is much more common in particular ethnic groups (e.g. certain Indian populations) and it has a mean age of onset at 12–14 years. Maturity Onset Diabetes of Youth (MODY) is an inherited form of noninsulin-dependent diabetes mellitus. It is transmitted in an autosomal dominant manner and presents in childhood or early adolescence with a relatively mild, slowly progressive form of diabetes. Affected individuals are initially asymptomatic. Hyperglycemia may be treated for long periods with diet or oral hypoglycemics and insulin is rarely required until at least 5 years after diagnosis. More recently it has been recognized that MODY represents a common phenotypic manifestation of several genetic abnormalities, all of which are dominantly inherited (4). To date, five molecular enzyme defects have been described leading to MODY subtypes 1–5. These subtypes vary in severity, requirement for insulin and development of long-term complications. Ó 2004 Blackwell Publishing Ltd

D I A BE T E S M E L LI T U S

Table 1 Types of diabetes in the young and associated conditions 1. Type 1 diabetes (insulin-dependent diabetes mellitus) 2. Early onset type 2 diabetes (noninsulin-dependent diabetes mellitus) 3. Other types Maturity Onset Diabetes of Youth (MODY) MODY types 1–5 Insulin resistance syndromes Idiopathic Known syndromes – Leprechauism, Rabson-Mendenhall Genetic syndromes associated with higher than expected prevalence of diabetes Chromosomal abnormalities: Down’s, Kleinfelter’s, Turner’s Other genetic syndromes: Wolframs, Prader-Willi Pancreatic defects Pancreatic fibrosis – cystic fibrosis, thalassemia Congenital absence or hypoplasia of pancreas Pancreatectomy for hypersplenism Secondary diabetes Drugs – corticosteriod therapy

7 17

environmental influences such as viral infections or dietary trends. There is evidence that the incidence of type 1 diabetes in childhood is increasing in some populations, perhaps as much as two to threefold per decade (6). There is also evidence to indicate an increase in diabetes in the very young (<5 years) (7). These increases may be related to relative affluence, as type 1 diabetes is associated with higher socioeconomic status, although some clusters suggest a localized epidemic of an infectious agent. Infections may also explain the seasonal variation in presentation of type 1 diabetes, which occurs much more commonly in autumn and winter when viral and other pathogens are more prevalent (8). This seasonal variation is consistent in both northern and southern hemispheres.

Adapted from: Green (26).

Etiology and pathogenesis Down’s, Turner’s and Kleinfelter’s syndromes are associated with the premature development of diabetes. Diseases affecting the exocrine pancreas can also eventually result in glucose intolerance. In children with cystic fibrosis, there is a 2.6% incidence of diabetes under 18 years of age (5). Congenital infections such as rubella greatly increase the risk of DM and endocrinopathies. Other causes of pediatric hyperglycemia include genetic defects in beta cell function and in insulin action, growth hormone secreting tumors and Cushing’s syndrome, all of which are rare but essentially produce a hyperglycemic picture.

Incidence The incidence of type 1 diabetes varies within the UK. In England and Wales, 17 per 100 000 children under 16 years develop diabetes each year and, in Scotland 25 per 100 000 per year (2). A large district general hospital may therefore expect 25–30 new cases of under 16-year olds each year. The incidence of type 1 diabetes also shows marked worldwide variation. Major features are the relatively high incidence in Nordic countries compared with the rest of Europe and the very low incidence in most Asian countries. This variability is unexplained but may relate to genetic susceptibility, Ó 2004 Blackwell Publishing Ltd, Pediatric Anesthesia, 14, 716–723

Type 1 diabetes develops as a result of the synergistic effects of genetic, environmental and immunological factors that ultimately destroy the pancreatic beta cells. The temporal development of type 1 diabetes relates to beta cell mass and this varies among individuals. It is postulated that individual genetic susceptibility coupled with environmental triggers, stimulates an inflammatory response in the pancreas (insulitis). This in turn activates the autoimmune system and results in the attack of beta cells. Beta cells are very much susceptible to the cytotoxic effect of some cytokines such as tumor necrosis factor a, cinterferon and interleukin 1 that are released during the insulitis process. The result is progressive beta cell death. In contrast, early onset of type 2 diabetes arises from the combination of insulin resistance coupled with an inability to maintain adequate compensatory hyperinsulinemia (4). With the exception of specific ethnic groups, inheritance of both disorders has not been linked to any one specific genetic defect and is likely to be polygenic. However, MODY represents a common phenotypic manifestation of several genetic abnormalities, all of which are dominantly inherited. Beta cell dysfunction seems to be the primary pathophysiological abnormality in all the subtypes. This results in varying degrees of impairment of insulin secre-

718

V. CHADWICK AND K.A. WILKINSON

tion and/or a delay in secretion of insulin following a glucose load.

Physiology Insulin is a polypeptide hormone with an overall anabolic effect. It promotes the cellular uptake and utilization of glucose, amino acids, potassium and magnesium. It also induces glycogen and protein synthesis and lipogenesis. Conversely, it inhibits glycogenolysis, protein breakdown and lipolysis. Insulin action is mediated by increasing levels of cyclic adenosine monophosphate. Although insulin deficiency is the overwhelming etiological factor in hyperglycemia and subsequent ketoacidosis, many pathophysiological events observed at presentation or in poorly controlled DM are mediated by increased levels of counter-regulatory hormones. These include glucagons, adrenaline, noradrenaline, growth hormone and cortisol. Insulin deficiency acting in combination with excessive concentrations of these hormones results in: (a) hyperglycemia caused by increased glycogenolysis and gluconeogenesis and decreased peripheral utilization of glucose. Hyperglycemia results in glycosuria, polyuria and loss of water and electrolytes; (b) increased fat lipolysis resulting in increased circulating free fatty acids; (c) the conversion of these free fatty acids to ketone bodies, namely acetoacetic acid and b hydroxybutyric acid, the formation of which exceeds the capacity of the tissue utilization, results in ketonemia, ketonuria and acidosis.

Metabolic effects of surgery and anesthesia Anesthesia, surgery and the contemplation of these events all cause a typical metabolic stress response, which tends to override normal homeostatic mechanisms. These changes are very similar to those found in the early stages of diabetes. The stress response is characterized by increases in the secretion of the catabolic hormones cortisol, catecholamines, glucagon and growth hormone, and inhibition of insulin secretion (9). This leads to increased mobilization of gluconeogenic precursors, in particular amino acids, with increased protein degradation, increased gluconeogenesis, glycogeno-

lysis, and hyperglycemia. Lipolysis is enhanced and there is insulin resistance. Surgery is known to provoke catabolism, the magnitude of which is related to the extent of the surgery and complications such as sepsis. There is an immediate neuroendocrine response with release of ACTH and consequently cortisol, growth hormone and the catecholamines (10,11). Noradrenaline levels show pronounced increases during surgery whereas adrenaline more markedly rises postoperatively (12). Glucagon secretion also increases and insulin levels fall with an impaired responsiveness to glucose (10). In the presence of impaired glucose homeostasis these effects may be more severe, and the diabetic patient is thus more at risk of severe hyperglycemia and ketoacidosis. Surgical patients are usually starved so that measured blood glucose rises may not be massive even in the diabetic patient but this may mask large increases in lipolysis, ketogenesis and proteolysis as well as excessive fluid and electrolyte losses. Anesthesia may contribute to the metabolic response to surgery. Modern inhalation anesthetics produce less marked effects than their predecessors but nevertheless can induce hyperglycemia, fatty acid mobilization and inhibition of insulin secretion. Their effect is dose-dependent and reversible (13). Intravenous induction agents may also affect glucose homeostasis. The effect of propofol on insulin secretion is unknown but diabetic patients have a reduced ability to clear lipids from the circulation. This may be of relevance in diabetic children receiving prolonged total intravenous anesthesia. Benzodiazepines decrease the secretion of ACTH and so production of cortisol when used in high doses (14). Highdose opioid anesthetic techniques not only provide hemodynamic stability, but also hormonal and metabolic stability by effectively blocking the hypothalamic–pituitary axis (15). Spinal anesthesia for surgery and epidural analgesia both intra- and postoperatively have been shown to attenuate the metabolic changes associated with surgical stress (16,17). At present however, there is no evidence that regional anesthesia alone or in combination with general anesthesia, confers specific benefits to the diabetic surgical patient in terms of mortality or major complications. Ó 2004 Blackwell Publishing Ltd, Pediatric Anesthesia, 14, 716–723

D I A BE T E S M E L LI T U S

In conclusion the overall effects of the anesthetic agents are minor in comparison with the metabolic effects of the surgery.

Modern management An intensive insulin program involves insulin regimes that aim to mimic physiological insulin secretion as accurately as possible, i.e. a continuous basal insulin supply with bolus ‘prandial’ dosing. No insulin regime reproduces the precise insulin secretory pattern of the pancreatic islets but modern regimes entail more frequent insulin injections, greater reliance on short acting insulin and frequent capillary plasma glucose measurements. Insulin dosing regimes are plentiful. The three most common include: 1. Twice-daily injections of intermediate acting insulin mixed with short acting insulin before morning and evening meals. Many children with early diabetes manage well with two injections per day. 2. Multiple daily injections: one to two shots of longacting insulin each day to provide basal insulin coverage and three shots of short acting insulin to provide glycemic coverage for meals. 3. Subcutaneous insulin infusions: these allow very fine-tuning of glucose control (18) and have been found to be effective in older children and adolescents. In some centers these are used in up to 20% of patients. Current insulin preparations are manufactured by recombinant DNA technology and consist of the amino acid sequence of human insulin. Animal insulin is rarely used. Biosynthetic techniques have produced human insulin analogues with altered pharmacokinetic properties (19). Two new short acting insulins have been recently produced. Lispro insulin (Humalog; Eli Lilly, Basingstoke, Hampshire, UK) was made by reversing the order of proline and lysine at positions B29 and B30 and aspart insulin (Novolog; Novo Nordisk Pharmaceuticals, Crawley, West Sussex, UK) by substituting aspartic acid for proline at position B29. These amino acid substitutions are reported to have no effect on the metabolic actions of insulin, but merely speed up the absorption of these analogues compared with regular short acting insulin following subcutaneous injection. This leads to early onset, sharper peaks and shorter duration of action. Ó 2004 Blackwell Publishing Ltd, Pediatric Anesthesia, 14, 716–723

7 19

Intermediate and long-acting insulin preparations have historically suffered from two major problems. The first is failure to mix these uniformly in suspensions prior to injection, and the second to provide a ‘peakless’ long-acting effect. Glargine insulin (Aventis Pharmaceuticals, West Malling, Kent, UK) is an analogue of human insulin with C-terminal elongation of the B-chain by two arginines and replacement of asparagine in position A21 by glycine. This molecule is soluble in the acidic solution in which it is packaged but relatively insoluble in the physiologic pH of the extracellular fluid. Consequently the microprecipitates of glargine insulin formed following subcutaneous injection markedly delay its absorption into the systemic circulation. Pharmacokinetic and pharmocodynamic studies have demonstrated that this insulin analog has a very flat and prolonged time-action profile (20,21). Frequent blood glucose monitoring is essential for both the multiple daily injections and subcutaneous insulin infusion regimes. In the perioperative period it is essential we have accurate, quick and easy to use monitoring systems to detect changes in blood glucose as they happen. MiniMed and others have developed new continuous glucose sensing systems. The MiniMed system uses a platinum electrode inside a semipermeable matrix containing glucose oxidase. A needle is used to insert the sensor into the subcutaneous tissue of the anterior abdominal wall. The glucose oxidase in the sensor catalyses the oxidation of glucose in the interstitial fluid, which then generates an electrical current that can be measured and transduced to give a signal. This is on trial (22,23), although not yet specifically in the surgical setting.

Perioperative management The overall aim of perioperative therapy is to avoid hypoglycemia and excessive hyperglycemia (24). Management should be by regimes that are easily understood, widely applicable and that minimize the possibility of error. Preoperative management will vary depending on the type, extent and urgency of surgery and current diabetic control. Minor elective procedures in older diabetic children may be carried out as a day procedure (25). A standard regime of management should be followed and the rationale for this

720

V. CHADWICK AND K.A. WILKINSON

explained to parent and child. Preoperative assessment should pay particular attention to current blood glucose, insulin types and dosing regime should be noted and fluid balance assessed clinically. Serum electrolytes, urea and creatinine should be requested if there is any doubt about hydration and/or renal function and if surgery is major. Patients should ideally be placed first on the operating list and intravenous dextrose/insulin regimes commenced. Plans should be made for the postoperative course. It should be clear to all involved with the care of these patients that cases which might otherwise be suitable surgically, may be impossible to manage on a day stay basis. Early preassessment prior to elective surgery is important in this population. Both the nursing staff and anesthetist have an opportunity to discuss the plan with a parent and child and pass on written information. If scheduled for more major surgery as an inpatient, the child may require high observation facilities at least initially and the pediatric diabetic team should be updated about the child’s current medical status. Locally we have used a specific scheme of management for children for approximately the last 5 years (26). The fluid prescription sheets, which form the basis for this regime are reproduced in Appendix 1. This local guideline has been developed from other similar regimes (27). Our experience is that meticulous attention to detail is essential at all stages. Copies of a protocol should be readily available in all areas managing the child, and surgeons, anesthetists, ward and theatre staff must understand the importance of adhering to the regime even if the theatre stay is short. Intravenous infusions of dextrose and insulin have been shown to provide the most stable glycemic control for diabetic patients undergoing major surgery (28). When compared with subcutaneous insulin infusions in the perioperative period, patients had better glucose control and fewer episodes of hyperglycemia and hypoglycemia. Hence, although there is increasing popularity of subcutaneous insulin infusions for long-term management of diabetes, perioperative control is still better with intravenous regimes. We have found it extremely useful if children arrive in theater with pumps, etc. made up. However, if surgery is short there may be a temptation for infusions to be withheld until the child recovers

or even discharged to the ward. This is risky when the child has received long-acting insulin the evening prior to surgery and a minimal amount of oral dextrose. In children over 5 years who have undergone elective minor surgery early in the day, who have no associated comorbidity, have consumed a normal meal without vomiting and have documented blood sugars in the normal or slightly high normal range, discharge on the same day may be considered. This assumes that parents are given and can follow simple instructions on monitoring blood sugar and are able to return within 1 h and/or obtain telephone advice as necessary.

Glucose/insulin regimes Absorption of insulin onto the surface of syringes and intravenous giving sets is an unavoidable problem. In particular, problems are seen with high surface area sets, high volume infusions and low insulin concentration regimes. More consistent delivery can be achieved with more concentrated solutions of lower volume administered from a syringe (29). With an insulin concentration of >400 ngÆml)1 (10 UÆl)1) the effect is minimal. Patients require between 0.01 and 0.2 UÆkg)1 h)1 of insulin. This combined with a glucose infusion, potassium replacement and hourly glucose monitoring appears to achieve our goals. The aim is to maintain blood glucose between 7 and 10 mmolÆl)1 (126–180 mgÆdl)1). Separate infusions may have potentially disastrous consequences should one of the infusions stop inadvertently. However, improved perioperative glycemic control seems to be achieved (30).

Emergency surgery In the emergency setting swings in blood glucose will be much more pronounced. Very close observation is essential and early instigation of insulin and dextrose infusion regime titrated by hourly blood glucose measurements is essential prior to induction of anesthesia. Diabetic ketoacidosis (DKA) is a medical emergency that still carries a considerable mortality rate of up to 15% (31). Scrupulous attention to fluid, electrolyte and insulin management is essential. Cerebral Ó 2004 Blackwell Publishing Ltd, Pediatric Anesthesia, 14, 716–723

D I A BE T E S M E L LI T U S

edema is the principal cause of death and disability. Other complications include aspiration and cardiac arrhythmias. Risk factors for cerebral edema include young age (<5 years), hypocapnia, elevated urea at presentation, large amounts of fluid resuscitation (>40 mlÆkg)1), use of bicarbonate and overly rapid fall in blood glucose. Children requiring large volumes of fluid resuscitation (>40 mlÆkg)1) require transfer to a pediatric intensive care unit. Full guidelines on management are beyond the scope of this article and can be accessed elsewhere (32). DKA may present with symptoms of an acute abdomen. Anesthetists should ensure that these patients have been well fluid resuscitated (initially with normal saline) prior to surgery with blood glucose lowered to 17 mmolÆl)1 (306 mgÆdl)1) or less and any abnormal electrolytes vigorously corrected. However it is imperative (because of the risk of cerebral edema) that rehydration, acid–base and blood sugar correction does not occur too rapidly, and is optimally achieved over 36–48 h. In particular sodium bicarbonate should not be administered to treat acidosis unless the pH is <7.0 (33). Patients should always receive close neurological observation. Delayed gastric emptying should be assumed and requires continuous nasogastric drainage and rapid sequence induction should anesthesia/intubation be required. Children should not consume until abdominal signs and symptoms resolve. Blood glucose estimation should occur at least hourly intraoperatively with potassium and acid– base estimation every 3–4 h. It should be remembered that hypoglycemia may also occur in the sick diabetic patient and rigorous observation and urgent treatment of this is essential. Our protocol for management of hypoglycemia is instigated when there is any blood glucose measurement of 2 mmolÆl)1 (36 mgÆdl)1) or less. If insulin is being infused, it should be immediately stopped. If the child is conscious and tolerating oral fluids they are given 50 ml of a sugary drink or dextrose tablets. This is repeated every 3–4 min until symptoms disappear. If the child is conscious and not tolerating oral fluids, ‘hypostop’ is administered to the gums. If the child is unconscious or ‘hypostop’ is not available then we give 5 mlÆkg)1 10% dextrose intravenously over 2–10 min or glucagon by intramuscular injection (1 mg if >12 years, 0.5 mg if 6–12 years old and 0.25 mg if 5 years old and less). Ó 2004 Blackwell Publishing Ltd, Pediatric Anesthesia, 14, 716–723

7 21

It is also important to remember that postoperative glucose control can be as challenging as pre- and intraoperative control and requires close observation. The postoperative phase involves the continuing stresses of recovering from tissue injury, pain, gastroparesis and any ongoing fluid and electrolyte derangement. Patients start oral ingestion of food, and this transition is not always successful. Glucose levels may remain unstable for some time after even a minor operation and this must be recognized by all involved in the patients care.

Future prospects for IDDM management Much effort has been devoted to arrest, reverse or ideally prevent autoimmune B cell damage (34,35). The search for more physiological ways to replace insulin continues. A number of systems for the pulmonary delivery of insulin are in development and preliminary results of phase 3 studies have been promising (36). Progress is being made on the development of a practically applicable artificial pancreas and considerable advances have been made in the technology of transplanting either the pancreas or preparations of islet tissue (37). Major problems remain however in obtaining donor tissue and in preventing immune rejection of the graft.

Summary and conclusions With the increasing prevalence of DM in our pediatric population it is inevitable that these patients will be presenting more frequently for surgery. Over recent years evidence has accumulated indicating that improving glycemic control in the short and long-term improves their outcome. Attention to detail in the daily management of the disease is essential. Likewise, an aggressive approach to glycemic control in the perioperative period results in better wound healing and lower incidences of infection, overall morbidity and shorter hospital stays. While there can be little argument about the management of diabetic children undergoing major procedures, the management of minor surgery may be equally challenging and preparation is essential. Tight metabolic control in the perioperative period is imperative and is a goal that is attainable in most patients.

722

V. CHADWICK AND K.A. WILKINSON

Acknowledgements We would like to thank Dr. Nandu Thalange, Consultant Paediatric Endocrinologist, Norfolk and Norwich University Hospital NHS Trust, for his help and advice in preparing this text.

References 1 Powers A. Diabetes Mellitus. In: Braunwald E, Fauci AS, Kasper D, Hauser SL, eds. Harrison’s Principals of Internal Medicine, 15th edn, Vol. 2. McGraw-Hill, 2001. 2 Green S. Diabetes in Childhood and Adolescence. In: Heller SR, Dunneen SF, eds. Medicine. The Medicine Publishing Company Ltd, 2002: 60–65. 3 Pinhas-Hamiel O, Dolan LM, Daniels SR et al. Increased incidence of non-insulin dependent diabetes mellitus among adolescents. J Pediatr 1996; 128: 608–615. 4 Glaser NS. Non-insulin-dependent diabetes mellitus in childhood and adolescence. Ped Clin North Am 1997; 44: 307–325. 5 Koch C, Cuppens H, Rainisio M et al. European Epidemiological Registry of Cystic Fibrosis: comparison of major disease manifestations between patients with different classes of mutations. Pediatr Pulmonol 2001; 31: 1–12. 6 Stewart-Brown S, Haslam M, Butler N. Evidence for increasing prevalence of diabetes mellitus in childhood. Br Med J 1983; 286: 1855–1857. 7 Gardener SG, Bingley PJ, Sawtell PA et al. Rising incidence of insulin dependent diabetes in children under 5 years in the Oxford region: time trend analysis. Br Med J 1997; 315: 713–717. 8 Atkinson MA, Maclaren NK. The pathogenesis of insulindependent diabetes mellitus. N Engl J Med 1994; 331: 1428– 1436. 9 Alberti KG, Gill GV, Elliott MJ. Insulin delivery during surgery in the diabetic patient. Diabetes Care 1982; 5(Suppl. 1): 65–77. 10 Allison SP, Tomlin PJ, Chamberlain MJ. Some effects of anaesthesia and surgery on carbohydrate and fat metabolism. Br J Anaesth 1969; 41: 588–593. 11 Clarke RS. The hyperglycaemic response to different types of surgery and anaesthesia. Br J Anaesth 1970; 42: 45–53. 12 Halter JB, Pflug AE. Relationship of impaired insulin secretion during surgical stress to anaesthesia and catecholamine release. J Clin Endocrinol Metab 1980; 51: 1093–1098. 13 Desborough JP, Jones PM, Persaud SJ et al. Isoflurane inhibits insulin secretion in isolated rat pancreatic islets of Langerhans. Br J Anaesth 1993; 71: 873–876. 14 Desborough JP, Hall GM, Hart GR et al. Midazolam modifies pancreatic and anterior pituitary secretion during upper abdominal surgery. Br J Anaesth 1991; 67: 390–396. 15 Hall GM, Lacoumenta S, Hart GR et al. Site of action of fentanyl in inhibiting the pituitary adrenal response to surgery in man. Br J Anaesth 1990; 65: 251–253. 16 Engquist A, Brandt MR, Fernandes A et al. The blocking effect of epidural analgesia on the adrenocortical and hyperglycaemic responses to surgery. Acta Anaesthesiol Scand 1977; 21: 330–335. 17 Wolf AR, Eyres RL, Laussen PC et al. Effect of extradural analgesia on stress response to abdominal surgery in infants. Br J Anaesth 1993; 70: 654–660. 18 Pickup JC, Keen H, Parsons JA et al. Continuous subcutaneous insulin infusion: an approach to achieving normoglycaemia. Br Med J 1978; 1: 204–207.

19 Tamborlane W. Recent advances in treatment of youth with type 1 diabetes: better care through technology. Diabet Med 2001; 18: 864–870. 20 Lepore M, Pappanelli S, Fanelli C et al. Pharmacokinetics and pharmacodynamics of subcutaneous of long-acting human insulin analog glargine, NPH insulin, ultralente human insulin and continuous subcutaneous infusion of insulin lispro. Diabetes 2000; 49: 2142–2148. 21 Heinemann L, Linkeschova R, Rave K et al. Time-action profile of the long-acting analog glargine in comparison with those of NPH insulin and placebo. Diabetes Care 2000; 23: 644–648. 22 Boland EA, Delucia M. Limitations of conventional methods of self blood glucose monitoring: lessons learned from 3 days of continuous glucose monitoring in pediatric patients with type 1 diabetes. Diabetes 2000; 49(Suppl. 1): A98. 23 Gibson LC, Halvorson J. Short-term use of the Minimed continuous monitoring system to determine patterns of glycaemia in pediatric patients with type 1 diabetes mellitus. Diabetes 2000; 49: A108. 24 Gavin L. Perioperative management of the diabetic patient. Endocrinol Metabol Clin N Am 1992; 21: 457–475 25 Kirschner R. Diabetes in Pediatric ambulatory surgical patients. J Post Anaesth Nurs 1993; 8: 322–326. 26 Available at: http://www.norfolk-norwich-hospitals.net/depart/ anaes/. 27 Kirk J. Paediatric Vade Maecum. In: Barrett T, Diwakar V, Lander T, eds. Guidelines on the Management of Diabetes During Surgery, 14th edn. London: Edward Arnold, 2002. 28 Kaufman F, Devgan S, Roe TF et al. Perioperative management with prolonged intravenous insulin infusion versus subcutaneous insulin in children with type 1 diabetes mellitus. J Diab Comp 1996; 10: 6–11. 29 Robertshaw HJ, McAnulty GR, Berge D et al. Pre-prepared glucose–insulin–potassium (‘GIK’) syringes: a practical alternative to ‘GIK’ bags. Br J Anaesth 1992; 82(Suppl. 1): 129. 30 Simmons D, Morton K. A comparison of 2 intravenous insulin regimes among surgical patients with insulin dependent diabetes mellitus. Diab Educator 1994; 20: 422–427. 31 Basu A, Close CE, Jenkins D et al. Persistent mortality in diabetic ketoacidosis. Diabet Med 1993; 10: 282–284. 32 Consensus Guidelines 2000–ISPAD. In: Swift PGF, ed. Consensus Guidelines for the Management of Type 1 Diabetes Mellitus in Children and Adolescents. Netherlands: Medical Forum International. www.d4pro.com/diabetesguidelines. 33 Advanced Paediatric Life Support (APLS), 3rd edn. London: BMJ Publishing, 2001. 34 Pickup J, Williams G. Textbook of diabetes. Insulin-dependent Diabetes Mellitus: An Overview, 2nd edn, Vol. 1. Oxford: Blackwell Publishing, 1997. 35 Scott A, Donnelly R. Improving outcomes for young people with diabetes: use of new technology and a skill-based training approach is urgently needed. Diabet Med 2001; 18: 861–863. 36 Scyler JS, Cefalin WT, Kourides IA et al. Efficacy of inhaled human insulin in type 1 diabetes mellitus: a randomised proof of concept study. Lancet 2001; 357: 331–335. 37 Soon–Shiong P, Heintz RE, Merideth N. Insulin independence in a type 1 diabetic patient after encapsulated islet transplantation. Lancet 1994; 343: 950–951.

Accepted 21 November 2003

Ó 2004 Blackwell Publishing Ltd, Pediatric Anesthesia, 14, 716–723

D I A BE T E S M E L LI T U S

Appendix 1

Ó 2004 Blackwell Publishing Ltd, Pediatric Anesthesia, 14, 716–723

7 23