NIMGU accounts for 70% of glucose uptake under fasting conditions (primarily into the CNS) and for 50% of post-prandial glucose uptake (especially into skeletal muscle)

Impaired glucose-induced insulin secretion

• Obesity

• Physical inactivity

• Reduced dietary carbohydrate

• Diabetogenic drugs (thiazides, glucocorticoids)

Aims of treatment

The management of T2DM and its common comorbidity of cardiovascular disease is complicated because of the added effects of ageing on metabolism and renal function, the use of potentially diabetogenic drugs, and low levels of physical activity. Cardiovascular risk is particularly high because many risk factors of the metabolic syndrome can be present for up to a decade before T2DM is diagnosed. Management strategies for many elderly people with diabetes are the same as those for the younger population, with similar lipid-lowering and antihypertensive treatment schedules, and aspirin or clopidogrel for patients with increased cardiovascular risk. Effective delivery of diabetes care depends on close cooperation between hospital and community, the involvement of diabetes teams and practice nurses, and attention to all causes of disability and ill-health (Table 6.2).

Table 6.2 Special characteristics of older subjects with diabetes

• High level of associated medical comorbidities

• Increased risk of cognitive dysfunction

• Mood disorder

• Varying evidence for impaired activities of daily living and lower limb function

• Increased vulnerability to hypoglycaemia

• Increased risk of inpatient mortality

• Increased need to involve spouses/family and carers in management

Comorbidity

Coexisting cardiovascular and cerebrovascular disease are especially common; polypharmacy is often necessary and carries risks of adverse effects and drug interactions. Degrees of renal impairment, arising for a multiplicity of reasons (e.g. diuretic therapy, prostatic obstruction), are relatively common in the elderly population; care must be exercised with renally excreted drugs. In addition, impaired hepatic function (e.g. hepatic congestion secondary to cardiac failure) may render certain antidiabetic medications inappropriate.

The prevalence of hypertension rises with age together with the risk of cardiovascular events. Attention should be directed towards modifiable cardiovascular risk factors, notably hypertension and dyslipidaemia. Cardiovascular events may present with atypical symptoms in the elderly (e.g. acute confusion).

Oral antidiabetic agents

Care should be exercised with sulphonylureas because severe hypoglycaemia, although uncommon, carries a relatively high mortality rate. Agents with a short duration of action are preferred.

Insulin treatment

Although the majority of elderly patients have type 2 diabetes, insulin therapy becomes necessary in a significant proportion as oral agents lose their capacity to provide adequate glycaemic control. Other patients will require insulin because of contraindications to oral agents arising from comorbidity (see above). Avoidance of insulin-induced hypoglycaemia is often paramount for the very elderly and those living alone. Insulin may be appropriately commenced during an intercurrent illness and continued thereafter when oral agents may suffice.

Institutional care of the elderly diabetic

Elderly people with diabetes in care homes are particularly vulnerable and require greater diabetes specialist input. High levels of disability are common in older people with diabetes and lead to heavy usage of health-care resources and premature death. The burden of hospital care is increased 2–3-fold in those with diabetes compared with the general aged population, with more frequent clinic visits and a 5-fold higher admission rate.

About 16% of elderly people with diabetes in the UK are registered blind or partially sighted (8-fold high than in the non-diabetic population); this justifies regular screening for undetected eye disease.

Risk factors for foot ulceration affect 25% of elderly people with type 2 diabetes.

Older people with diabetes use primary care services two to three times more often than their counterparts without diabetes.

Hospital admissions last twice as long for older patients with diabetes compared with age-matched control groups without diabetes.

Management of diabetes in women of childbearing age

An optimal outcome may be obtained from pregnancy in women with diabetes if excellent glycaemic control is achieved before and during pregnancy. However, type 1 and 2 diabetes are high-risk states for both the woman and her fetus. There is an increased risk of complications of diabetes, including severe hypoglycaemia and progression of microvascular complications.

Ketoacidosis must be avoided.

There are also increased risks of obstetric complications, such as miscarriage, maternal infection, pre-eclampsia, premature labour, polyhydramnios and failure to progress in first or second stage. Fetal and neonatal complications include congenital malformation, later intrauterine death, fetal distress, hypoglycaemia, respiratory distress syndrome and jaundice. Rates of fetal and neonatal loss and congenital malformation are increased by at least 2–3-fold. The prevalence of type 2 diabetes is increasing in women of reproductive age, and outcomes may be equivalent or worse than in those with type 1 diabetes. Management before and during pregnancy should follow the same intensive programme of metabolic, obstetric and neonatal supervision.

An experienced multidisciplinary team, led by a named obstetrician and physician with an interest in diabetes, and including a diabetes specialist nurse, diabetes specialist midwife and dietitian, should provide comprehensive care from pre-pregnancy to postnatal review.

Pre-pregnancy and pregnancy care

Contraception

Contraception should be discussed on an individual basis with all women of childbearing age with diabetes. There is little evidence on choice of contraceptive method specifically for a woman with diabetes. The contraceptive advice should follow that in the general population. The combined oral contraceptive (COC) may be contraindicated in women with diabetes according to the presence and severity of diabetic complications and/or other risk factors for vascular disease. Progestogen-only preparations, oral or intramuscular, are suitable in these women. World Health Organization (WHO) evidence-based guidance for medical eligibility criteria for contraceptive use makes recommendations for women with diabetes.

Long-acting methods such as implants, intrauterine systems (IUS) and copper intrauterine devices (IUD) are safe methods of contraception that may be particularly suitable for use in women with diabetes as they are as effective as sterilization and produce low circulating hormone levels.

Pregnancy should be planned; good contraceptive advice and pre-pregnancy counselling are essential.

Health-care professionals should refer to the WHO medical eligibility criteria for contraceptive use prior to offering contraceptive advice to women with diabetes.

Pre-pregnancy clinic

Attendance at a pre-pregnancy clinic is associated with a reduction in the rate of miscarriage and complications of pregnancy; babies have fewer problems and are kept in special care units for shorter periods than infants of non-attending mothers.

Tip box

Pre-pregnancy care for women with diabetes should be provided by a multidisciplinary team.

The essential components of a pre-pregnancy care programme include review and consideration of the medical (including pharmacological treatment) obstetric and gynaecological history; advice on glycaemic control to optimize HbA1c levels; screening for complications of diabetes; and counselling for maternal and fetal complications.

Tip box

Women contemplating pregnancy should have access to structured education in line with the recommendations for adults for diabetes.

All health-care professionals in contact with women of childbearing age with diabetes should be aware of the importance of pre-pregnancy care and local arrangements for its delivery, and should share this information with the woman.

Pre-pregnancy targets for blood sugar

Observational evidence has found an association between maternal glycaemia (before and during pregnancy) and the increased risk of congenital malformation and miscarriage. The risk of congenital anomaly in the offspring of women with pre-pregnancy diabetes is increased with an increasing level of HbA1c. No HbA1c threshold for such effects has been identified (Table 6.3).

Table 6.3 Derived absolute risk of major or minor congenital anomaly in association with the number of standard deviations (SD) of glycosylated haemoglobin above normal, and the approximate corresponding HbA1c concentration, measured periconceptually

Pre-pregnancy glycaemic control should be maintained as close to the non-diabetic range as possible, taking into account risk of maternal hypoglycaemia.

The target for pre-pregnancy glycaemic control for most women should, as a minimum, be an HbA1c of less than 7% (53 mmol/mol), although lower targets of HbA1c may be appropriate if maternal hypoglycaemia can be minimized.

Oral medication before and during pregnancy

Folic acid

Neural tube defects in high-risk pregnancies are associated with lower levels of folate. All women with diabetes should be prescribed high-dose pre-pregnancy folate supplementation, continuing up to 12 weeks’ gestation.

Folic acid 5-mg tablets are readily available, suitable, and should be provided wherever pre-pregnancy care is delivered.

Metformin and sulphonylureas

Metformin and sulphonylureas are not shown to be associated with an increase in congenital malformation or early pregnancy loss. Sulphonylureas other than glibenclamide are unsafe due to placental passage. Recent evidence suggests that metformin may be used in pregnancy (after discussion with patient)-NICE & IDF guidelines.

Statins

The British National Formulary recommends that statins (atorvastatin, fluvastatin, pravastatin and simvastatin) should be avoided during pregnancy and lactation. Congenital malformations have been reported and decreased synthesis of cholesterol may further affect fetal development. The malformations include major defects of the central nervous system.

Angiotensin converting enzyme inhibitors

Angiotensin converting enzyme (ACE) inhibitors have been associated with an increased risk of congenital malformation. and both ACE inhibitors and angiotensin receptor blocking medications should be avoided throughout pregnancy.

Tip box

The use of statins, ACE inhibitors and angiotensin receptor blocking medications should be reviewed in women pre-pregnancy, and avoided during pregnancy.

Nutritional management

It is good clinical practice to provide dietary advice to women before, during and after pregnancy.

Advice on diet and exercise should be offered in line with recommendations for adults with diabetes.

Tip box

Dietetic advice should be available in all diabetic antenatal clinics.

Optimization of glycaemic control

Glucose monitoring

Optimal glucose control before pregnancy reduces congenital malformations and miscarriage, while during pregnancy it reduces macrosomia, stillbirth, neonatal hypoglycaemia and respiratory distress syndrome.

All women with pre-gestational diabetes should be encouraged to achieve excellent glycaemic control.

During pregnancy, preprandial testing and where appropriate postprandial testing may be advised. A reduced incidence of pre-eclampsia is associated with postprandial monitoring and targeting.

In women with gestational diabetes, measurement and targeting of postprandial glucose is associated with improved outcomes (including birth weight, reduced perinatal morbidity and macrosomia).

Postprandial glucose monitoring should be carried out in pregnant women with gestational diabetes and may be considered in pregnant women with type 1 or type 2 diabetes.

Tip box

Postprandial glucose monitoring should be carried out in pregnant women with gestational diabetes and should be considered in pregnant women with either type 1 or type 2 diabetes.

In women with type 1 or type 2 diabetes, as long as hypoglycaemia can be minimized, aim to achieve blood glucose:

• between 4 and 6 mmol/L preprandially

• < 8 mmol/L 1 hour postprandially, or

• < 7 mmol/L 2 hours postprandially

• < 6 mmol/L before bed.

There is limited evidence that continuous glucose monitoring may be of benefit to women during pregnancy.

Diabetes specialist nurses and midwives have an important role in educating women on the need for home blood glucose monitoring and intensive insulin regimens. Intensive basal bolus regimens are commonly used and insulin analogues are increasingly used, although research on their role and safety in pregnancy is limited.

Insulin therapy

Intensive insulin therapy is beneficial in terms of maternal and neonatal outcomes. Use of insulin analogues is associated with limited evidence in terms of reduction in hypoglycaemia. The choice of insulin therapy should be discussed as part of pre-pregnancy counselling.

Lispro and aspart are associated with an improved glycaemic profile with a trend towards reduced major hypoglycaemia and nocturnal hypoglycaemia. There is a lack of high-quality evidence regarding outcomes of pregnancy using basal insulin analogue therapy or continuous subcutaneous insulin infusion (CSII) in women who are pregnant. Current evidence also suggests neither benefit nor harm with CSII.

NPH insulin should remain the basal insulin of choice in pregnancy unless the clinical benefit of a basal insulin analogue has been demonstrated on an individual basis. Rapid-acting insulin analogues (lispro and aspart) appear safe in pregnancy and may be considered in individual patients where hypoglycaemia is problematic.

Women should be advised that, although most commonly used regular human insulins are licensed for use during pregnancy, other insulins and oral glucose-lowering agents (e.g. metformin, glibenclamide, other sulphonylureas, detemir) are not.

Complications during pregnancy

Obstetric complications

There is no specific evidence on management of obstetric complications, including pregnancy-induced hypertension and increased risk of thromboembolism, in women with diabetes. These risks should be managed as for other pregnant women.

Metabolic complications

During pregnancy, hypoglycaemic unawareness and severe hypoglycaemia are common. Diabetic ketoacidosis can develop more rapidly, at lower levels of blood glucose and in response to therapeutic glucocorticoids. Women and their partners need education on the management of hypoglycaemia, including the use of glucagon; avoiding hypoglycaemia while driving; and on the recognition and prevention of ketoacidosis. Ketoacidosis may result in fetal death. Local emergency contact arrangements must also be explicit.

Microvascular complications

Diabetic retinal and renal disease can deteriorate during pregnancy. The presence of retinopathy alone is not associated with a poorer pregnancy outcome for the fetus, unless concurrent nephropathy is evident.

Retinopathy

Poor glycaemic control in the first trimester and pregnancy-induced or chronic hypertension are independently associated with the progression of retinopathy.

Nearly 40% of women with baseline retinopathy will progress during pregnancy, although sight-threatening retinopathy is rare (around 2% of pregnancies).

Examination of the retina prior to conception and during each trimester is advised in women with type 1 or type 2 diabetes. More frequent assessment may be required in those with poor glycaemic control, hypertension or pre-existing retinopathy.

Multidisciplinary teams should have locally agreed protocols for the grade of retinopathy required for referral. Owing to the potential for rapid development of neovascularization, early referral of pregnant women with referable retinopathy to an ophthalmologist is recommended.

Parous women with type 1 diabetes have significantly lower levels of retinopathy compared with nulliparous women. Women should be reassured that tight glycaemic control during and immediately after pregnancy can effectively reduce the long-term risk of retinopathy.

Nephropathy

There is an association between pre-existing nephropathy (microalbuminuria or albuminuria) and a poorer pregnancy outcome, although this is not due to any increase in congenital malformations. Proteinuria increases transiently during pregnancy, returning to a pre-pregnancy level within 3 months of delivery. The incidence of worsening chronic hypertension or pregnancy-induced hypertension/pre-eclampsia is high in women with both incipient and overt nephropathy, occurring in over 50% of women where overt nephropathy is present. Nephropathy and superimposed pre-eclampsia are the most common causes of pre-term delivery in women with diabetes.

Women with diabetic nephropathy require careful monitoring and management of blood pressure.

ACE inhibitors and angiotensin receptor blocking medications should be avoided as they may adversely affect the fetus.

Appropriate antihypertensive agents that may be used during pregnancy include methyldopa, labetalol and nifedipine.

Fetal assessment

An early pregnancy scan is considered good practice to confirm viability in women with pre-existing diabetes, particularly when changes in medication are required or diabetic control is suboptimal.

There is evidence of an increased incidence of congenital malformations in women with pre-existing diabetes (type 1 and type 2). In general, the sensitivity of ultrasound scanning for detecting structural abnormalities increase with gestational age. It is not possible to determine when during the second trimester scanning should take place to maximize detection rates. A detailed anomaly scan, including evaluation of the four-chamber heart and outflow tracts, undertaken at around 20 weeks (18–22 weeks) enables detection of many major structural abnormalities.

All women should be offered scanning to include:

• an early viability scan

• a gestational age scan between 11 and 13 weeks (+ 6 days) in association with biochemical screening and nuchal translucency measurement to risk assess for trisomies

• a detailed anomaly scan including four-chamber cardiac view and outflow tracts between 20 and 22 weeks.

In pregnancies complicated by maternal diabetes, the fetus is at risk of both macrosomia and intrauterine growth restriction (IUGR). The risk of macrosomia is greater when there has been poor glycaemic control. The risk of IUGR is greater in women with vascular complications of diabetes (retinopathy, nephropathy) or when pre-eclampsia develops.

Fetal monitoring includes cardiotocography (CTG), Doppler ultrasonography and ultrasound measurement of fetal growth and liquor volume. Although regular fetal monitoring is common practice, no evidence has been identified on the effectiveness of any single or multiple techniques, and therefore the clinical judgement of an obstetrician experienced in diabetic pregnancy is essential.

When IUGR is suspected, additional monitoring with serial ultrasonography and umbilical arterial Doppler velocimetry is associated with improved outcomes (fewer inductions of labour and hospital admissions, with a trend to improved perinatal mortality rates).

The evidence for the accuracy of ultrasonography in predicting macrosomia (birth weight above 4000 g) is mixed. The accuracy of fetal weight estimation in women with diabetes is at least comparable to that in women who are not diabetic. The negative predictive value of ultrasonography is high (80–96%), and therefore it is feasible that the true value of ultrasonography in the management of these women is its ability to rule out the diagnosis of macrosomia.

There is evidence to suggest that the incorrect diagnosis of a large-for-gestational-age fetus increases the induction and caesarean section rate without improving clinical outcome. Numerous studies have concluded that the reliability of ultrasound estimation of fetal weight is suboptimal. The ability to predict shoulder dystocia in the non-diabetic population is poor and evidence in the diabetic population limited. However, the determination of polyhydramnios is clinically useful as it may be associated with an adverse clinical outcome.

Trials report either equivalent outcomes or improved outcomes (birth weight, macrosomia, large-for-gestational-age infants) in women with gestational diabetes. Improved outcomes were associated with abdominal circumference (AC) being ascertained early (rather than late) in the third trimester and intensively managed thereafter. Where outcomes were equivalent this was achieved with fewer women requiring insulin or a change of treatment assignment. Although the rates of large-for-gestational-age infants were reduced with insulin therapy, there were no immediate clinical benefits observed from this reduction and there was an increase in the caesarean section rate.

Where IUGR is suspected, regular monitoring including growth scans and umbilical artery Doppler monitoring should be carried out.

Gestational diabetes

Gestational diabetes can be defined as carbohydrate intolerance of variable severity with onset or first recognition during pregnancy. This definition will include women with abnormal glucose tolerance that reverts to normal after delivery, those with undiagnosed type 1 or type 2 diabetes, and rarely women with monogenic diabetes. If type 1 or type 2 diabetes is presumed, for example due to early presentation or grossly raised blood glucose levels, urgent action is required to normalize metabolism.

Randomized controlled trials (RCTs) have shown that intervention in women with gestational diabetes with dietary advice, monitoring and management of blood glucose is effective in reducing birth weight, macrosomia, neonatal hypoglycaemia and the rate of large-for-gestational-age infants, as well as perinatal morbidity. More recent RCTs of the detection and management of GDM have found no change in the rate of caesarean section.

Tip box

A suitable programme to detect and treat gestational diabetes should be offered to all women in pregnancy.

The most appropriate strategies for screening and diagnosing GDM remain controversial. There is a continuous relationship between maternal glucose level (fasting, 1 h and 2 h after a 75-g oral glucose tolerance test; OGTT) at 24–28 weeks and pregnancy outcomes (macrosomia, fetal insulin, clinical neonatal hypoglycaemia and caesarean section). Studies showing a benefit of screening-detected GDM have used a variety of strategies.

Screening for GDM

Early pregnancy

An important aim of screening, particularly in early pregnancy, is to identify women with undiagnosed type 2 diabetes. Clinical suspicion that type 1 or type 2 diabetes is present or developing in pregnancy may be raised by persistent heavy glycosuria in pregnancy (2 + on more than two occasions), random glucose > 5.5 mmol/L 2 hours or more after food or ≥ 7 mmol/L within 2 hours after food.

The International Association of Diabetes and Pregnancy Study Group consensus document suggests that all or high-risk women should be offered screening with HbA1c, fasting or random glucose at the first gestational visit. Although the characteristics of HbA1c in early pregnancy should be close to those outwith pregnancy, it should be noted that normal HbA1c falls in later pregnancy, potentially resulting in false-negative results. Levels of glucose tolerance diagnostic of diabetes (HbA1c ≥ 6.5% (48 mmol/mol), fasting ≥ 7.0 mmol/L or 2-h glucose ≥ 11.1 mmol/L) can be interpreted in early pregnancy, but the clinical interpretation of intermediate levels of glucose tolerance (HbA1c 6.0–6.4% (42–46 mmol/mol), fasting glucose 5.1–6.9 mmol/L, 2-h glucose 7.8–11.0 mmol/L) encompassing current definitions of gestational diabetes, impaired fasting glucose and impaired glucose tolerance remain to be defined.

Later pregnancy

Controversy remains over the best screening strategy for detection of GDM, and the most clinically and cost-effective strategy is likely to vary depending on the population. A number of risk factors can be identified and health economic analysis supports screening of women with risk factors using a 75-g OGTT at 24–28 weeks of gestation. Measurement of fasting glucose provides a good predictor of adverse outcomes.

Strategies to detect previously existing but undetected diabetes in early pregnancy, and strategies to screen for gestational diabetes at 24–28 weeks, will be refined as information on the characteristics of this testing becomes apparent for the local population. Strategies are likely to be simplified for women believed to be at low risk based on risk factors (Table 6.4).

Table 6.4 Risk factors for gestational diabetes

BMI > 30 kg/m²
Previous macrosomic baby weighing 4.5 kg or more
Previous gestational diabetes
Family history of diabetes (first-degree relative with diabetes)
Family origin with a high prevalence of diabetes:

• South Asian (specifically women whose country of family origin is India, Pakistan or Bangladesh)

• Black Caribbean

• Middle Eastern (specifically women whose country of family origin is Saudi Arabia, United Arab Emirates, Iraq, Jordan, Syria, Oman, Qatar, Kuwait, Lebanon or Egypt)

Screening at first antenatal visit

At booking, all women should be assessed for the presence of risk factors for gestational diabetes. All women with risk factors should have HbA1c or fasting glucose levels measured.

Women in early pregnancy with levels of HbA1c ≥ 6.5% (48 mmol/mol), fasting glucose ≥ 7.0 mmol/L or 2-h glucose ≥ 11.1 mmol/L, diagnostic of diabetes, should be treated as having pre-existing diabetes.

Women with intermediate levels of glucose (HbA1c 6.0–6.4% or 42–46 mmol/mol), fasting glucose 5.1–6.9 mmol/L or 2-h glucose 8.6–11.0 mmol/L should be assessed to determine the need for immediate home glucose monitoring and, if the diagnosis remains unclear, assessed for gestational diabetes by a 75-g OGTT at 24–28 weeks.

Screening later in pregnancy

All women with risk factors should have a 75-g OGTT at 24–28 weeks.

Fasting plasma glucose at 24–28 weeks is recommended in low-risk women.

Diagnosis of GDM

There is a continuous relationship between maternal glucose level (fasting, 1 h, 2 h after a 75-g OGTT) and fetal growth.

Appropriate RCTs are not available to guide decision-making regarding the level of glucose at which different health benefits accrue. It is suggested that criteria are set at a level where there is an impact not only on birth weight but also on other outcomes, including shoulder dystocia and caesarean section.

A recent international consensus suggested criteria that result in a diagnosis of gestational diabetes in 16–18% of the pregnant population, where all women are tested with a 75-g OGTT. Women diagnosed using these criteria have a 1.75-fold increased risk of macrosomia. It is suggested that these international consensus criteria be adopted.

Depending on individual clinical circumstance, it is accepted that dietary intervention in women with lower glucose levels (2-h glucose 7.8–8.5 mmol/L) may help to reduce birth weight, and dietary advice and intervention on an individual basis might be considered (for example, in previous macrosomia or previous complicated delivery).

The adoption of internationally agreed criteria for gestational diabetes using the 75-g OGTT is recommended (Table 6.5):

• when fasting venous plasma glucose ≥ 5.1 mmol/L

• when 1-hour glucose is ≥ 10 mmol/L, or

• 2 hours after OGTT ≥ 8.5 mmol/L.

Table 6.5 Screening for and diagnosis of GDM

• Perform a 75-g OGTT, with plasma glucose measurement fasting and at 1 and 2 h, at 24–28 weeks of gestation in women not previously diagnosed with overt diabetes.

• The OGTT should be performed in the morning after an overnight fast of at least 8 h.

• The diagnosis of GDM is made when any of the following plasma glucose values are exceeded:

• Fasting > 92 mg/dL (5.1 mmol/L)

• 1 h > 180 mg/dL (10.0 mmol/L)

• 2 h > 153 mg/dL (8.5 mmol/L)

OGTT, oral glucose tolerance test; GDM, gestational diabetes mellitus.

Women with frank diabetes by non-pregnant criteria (fasting venous glucose ≥ 7 mmol/L, 2-h glucose ≥ 11.1 mmol/L) should be managed within a multidisciplinary clinic as they may have type 1 or type 2 diabetes and be at risk of pregnancy outcomes similar to those of women with pre-gestational diabetes.

Management of GDM

Management with dietary change to lower blood glucose levels and, if necessary, treatment with insulin improves outcomes in gestational diabetes. Glycaemic management should be tailored to control preprandial and postprandial blood sugar levels. The majority of women with gestational diabetes can be managed with dietary therapy alone. If, after nutritional advice, preprandial and postprandial glucose levels are normal and there is no evidence of excessive fetal growth, the pregnancy can be managed as for a normal pregnancy.

Controlled trials suggested that management strategies using metformin or glibenclamide can achieve similar outcomes to initial management with insulin, although 20–40% of women will still eventually require insulin therapy. Metformin crosses the placenta and glibenclamide appears not to.

If blood glucose levels are in the range for established diabetes, intensive specialist management and initial therapy with insulin is required.

Pregnant women with GDM should be offered dietary advice and blood glucose monitoring, and treated with glucose-lowering therapy depending on fasting and postprandial targets.

Glucose-lowering therapy should be considered in addition to diet where fasting or 2-hour glucose levels are above target, for example where two or more values per fortnight are:

• ≥ 5.5 mmol/L preprandially or ≥ 7 mmol/L 2 h postprandially on monitoring at ≤ 35 weeks

• ≥ 5.5 mmol/L preprandially or ≥ 8 mmol/L 2 h postprandially on monitoring at > 35 weeks, or

• any postprandial value > 9 mmol/L.

After discussion with the patient, metformin should be considered as the initial pharmacological glucose-lowering treatment in women with gestational diabetes.

Delivery

The timing of delivery should be determined on an individual basis. Delivery in women with diabetes is generally expedited within 40 weeks of gestation. However, no clear evidence has been identified to inform the optimal timing for delivery.

Women who are at risk of pre-term delivery should receive antenatal corticosteroids. If corticosteroids are indicated clinically for pre-term labour, supervision by an experienced team is essential to regulate diabetic control.

Women with diabetes have a higher rate of caesarean section even after controlling for confounding factors.

Women with diabetes requiring insulin or oral glucose-lowering medication who have a pregnancy that is otherwise progressing normally should be assessed at 38 weeks’ gestation with delivery shortly afterwards, and certainly by 40 weeks.

Women with diabetes should be delivered in consultant-led maternity units under the combined care of a physician with an interest in diabetes, obstetrician and neonatologist.

Women with diabetes should have a mutually agreed written plan for insulin management at the time of delivery and immediately afterwards.

The progress of labour should be monitored as for other high-risk women, including continuous electronic fetal monitoring.

IV insulin and dextrose should be administered as necessary to maintain blood glucose levels between 4 and 7 mmol/L.

Infants of mothers with diabetes

Labour and delivery should be undertaken only in a maternity unit supported by neonatal intensive care facilities. There is no need for routine admission of the infant to the neonatal unit. There is insufficient evidence on the preferred method of cotside blood glucose measurement in neonates; however, whichever method is used, the glucose values should be confirmed by laboratory measurement. Neonatal hypoglycaemia is defined as a blood glucose level < 2.6 mmol/L, and is associated with adverse short- and long-term neurodevelopmental outcomes.

Neonatal hypoglycaemia has been associated with adverse neurodevelopmental outcomes and impaired cognitive development.

In pre-term infants, significant hypoglycaemia (< 2.6 mmol/L) is associated with reductions in Bayley motor and developmental scores of 13 and 14 points, respectively, at 18 months’ corrected age. An association between recurrent exposure to hypoglycaemia and a 3.5-fold increase in the incidence of cerebral palsy and developmental delay in infants was also found. Recurrent episodes of blood glucose < 2.6 mmol/L in small-for-gestational-age (SGA) pre-term infants were associated with measurable neurodevelopmental deficits affecting fine motor ability and perceptual performance that were still apparent at 5 years of age. Repeated episodes of hypoglycaemia have also been shown to produce a reduction in occipitofrontal circumference, a surrogate marker of brain growth, at 12, 18 and 60 months of age.

Tip box

Early feeding is advised to avoid neonatal hypoglycaemia and to stimulate lactation.

Glycaemic control at 6 weeks in women with type 1 diabetes who exclusively breastfeed their babies has been found to be significantly better that those who bottlefeed. There are further well-documented health benefits for infants who are breastfed. Although breastfeeding is recommended for infants of mothers with diabetes, mothers should be supported in the feeding method of their choice.

Most medicines are not licensed for use in lactation. Specialist reference sources provide information on suitability of medicines in breastfeeding. Insulin, metformin and glibenclamide are considered compatible with breastfeeding, although the infant should be observed for signs of hypoglycaemia. The antihypertensives commonly used in pregnancy (labetalol, nifedipine and methyldopa) are found in breast milk in low concentration and these agents are considered appropriate for use in breastfeeding mothers, although with labetalol the infant should be monitored for bradycardia and hypotension. Of the ACE inhibitors, enalapril and captopril are considered safer. There is no information available on angiotensin-II receptor antagonists. Statins are not recommended when breastfeeding. Information on use of aspirin is conflicting, with some sources advising that low-dose aspirin is safe in breastfeeding, whereas others advise cautious use owing to potential for toxicity (possible Reye syndrome).

Specialist advice should be sought if the baby is premature or unwell.

Postnatal care

Women with type 1 or type 2 diabetes may require adjustment of their treatment regimen postnatally. Women with gestational diabetes should be investigated postnatally to clarify the diagnosis and exclude type 1 or type 2 diabetes. The opportunity should also be taken to provide lifestyle advice to reduce the risk of subsequent type 2 diabetes.

Postnatal follow-up should be seen as an opportunity to initiate pre-pregnancy care for any subsequent pregnancy. Appropriate contraception should be provided and the importance of good glycaemic control emphasized.

Breastfeeding should be encouraged to benefit mother and baby, but it may necessitate insulin dose adjustment and a dietetic review.

Follow-up of women with gestational diabetes

A diagnosis of GDM identifies women at increased risk of developing type 2 diabetes in future. Rates of progression to type 2 diabetes in women with previous GDM vary widely (between 15% and 50% cumulative incidence at 5 years) and will be influenced by other risk factors such as ethnicity, obesity and exercise.

A Cochrane review concluded that diet combined with exercise or diet alone enhances weight loss postpartum. Both intensive lifestyle and pharmacological interventions reduce the onset of type 2 diabetes in people with impaired glucose tolerance, including women with previous gestational diabetes.

Tip box

Women who have developed GDM should be given diet, weight control and exercise advice.

There is no robust evidence to determine when follow-up testing should be carried out. Women who have developed GDM should also be reminded of the need for pre-conception counselling and appropriate testing to detect progression to type 2 diabetes.

Where diabetes is not apparent immediately after delivery, glucose tolerance should be reassessed at least 6 weeks postpartum with a minimum of fasting glucose and with a 75-g OGTT if clinically indicated.

An annual assessment of glycaemia using fasting glucose or HbA1c should be carried out thereafter.

A checklist for provision of information to women with GDM can be found in Appendix 6.2.

Surgery and diabetes

Careful attention needs to be paid to the metabolic status of people with diabetes undergoing surgical procedures. Elective surgery in people with uncontrolled diabetes should preferably be scheduled after acceptable glycaemic control has been achieved. Admission to hospital 1–2 days before scheduled surgery is advisable for such patients. Whenever feasible, emergency surgery should be delayed to allow stabilization of diabetic patients. In addition, to avoid protracted fasting, the operation should be scheduled for early in the day.

Patients with diabetes have an increased requirement for surgical procedures and increased postoperative morbidity and mortality rates.

The stress response to surgery and the resultant hyperglycaemia, osmotic diuresis and hypoinsulinaemia can lead to perioperative ketoacidosis or hyperosmolar non-ketotic (HONK) syndrome (hyperglycaemic hyperosmolar state (HHS)).

The management goal is to optimize metabolic control through close monitoring, adequate fluid and the judicious use of insulin.

Major surgical operations require a period of fasting during which oral antidiabetic medications cannot be used.

The stress of surgery itself results in metabolic perturbations that alter glucose homeostasis; persistent hyperglycaemia is a risk factor for endothelial dysfunction, postoperative sepsis, impaired wound healing and cerebral ischaemia.

The stress response itself may precipitate diabetic crises (DKA, HONK syndrome) during surgery or postoperatively, with negative prognostic consequences. HHS is a well-known postoperative complication following certain procedures, including cardiac bypass surgery, where it is associated with a greater than 40% mortality rate.

The actual treatment recommendations for a given patient should be individualized. However, the recommendations should be based on:

• diabetes classification

• usual diabetes regimen

• state of glycaemic control

• nature and extent of the surgical procedure.

In addition, gastrointestinal instability provoked by anaesthesia, medications and stress-related vagal overlay can lead to nausea, vomiting and dehydration. This compounds the volume contraction that may already be present from the osmotic diuresis induced by hyperglycaemia, thereby increasing the risk of ischaemic events and acute renal failure. Subtle to gross deficits in key electrolytes (principally potassium, but also magnesium) may pose an arrhythmogenic risk, which often is superimposed on a milieu of endemic coronary artery disease in middle-aged or older people with diabetes.

Anaesthesia and surgery cause a stereotypical metabolic stress response, which could overwhelm homeostatic mechanisms in patients with pre-existing abnormalities of glucose metabolism. The features of the metabolic stress response include release of the catabolic hormones:

• adrenaline (epinephrine)

• noradrenaline (norepinephrine)

• cortisol

• glucagon

• growth hormone

• plus the inhibition of insulin secretion and action.

Anti-insulin effects of surgical stress

In addition to insulin resistance induced by circulating stress hormones, surgical stress has a deleterious effect on pancreatic β-cell functions. Plasma insulin levels fall and insulin secretory responses to glucose become impaired during surgery.

These anti-insulin effects of the metabolic stress response essentially reverse the physiological anabolic and anti-catabolic actions of insulin. The important anabolic actions of insulin that may be reversed or attenuated during the stress of surgery include:

• stimulation of glucose uptake and glycogen storage

• stimulation of amino acid uptake and protein synthesis by skeletal muscle

• stimulation of fatty acid synthesis in the liver and storage in adipocytes

• renal sodium reabsorption and intravascular volume preservation.

The anti-catabolic effects of insulin include:

• inhibition of hepatic glycogen breakdown

• inhibition of gluconeogenesis

• inhibition of lipolysis

• inhibition of fatty acid oxidation and ketone body formation

• inhibition of proteolysis and amino acid oxidation.

Direct catabolic effects of stress hormones

The neuroendocrine response to the stress of general anaesthesia and surgery leads to activation of potent counter-regulatory hormones. The catecholamine noradrenaline (norepinephrine) is augmented mostly during surgery, whereas after surgery adrenaline (epinephrine) stimulates gluconeogenesis and glycogenolysis. Catecholamines also inhibit glucose utilization by peripheral tissues and inhibit insulin secretion.

These effects predispose to severe hyperglycaemia, which is further exacerbated by the stimulatory effect of adrenaline and noradrenaline on glucagon secretion. Other catabolic effects of catecholamines include stimulation of lipolysis and ketogenesis. The activated hormone-sensitive lipase promotes lipolysis and release of free fatty acids into the circulation.

Glucagon, the level of which is augmented by catecholamines, exerts catabolic effects similar to those of the catecholamines: stimulation of hepatic glucose production and ketogenesis, and inhibition of insulin action in peripheral tissues. Growth hormone and glucocorticoids potentiate the catabolic effects of catecholamines and glucagon. Glucocorticoids increase hepatic glucose production and induce lipolysis and negative nitrogen balance by stimulating proteolysis. The products of lipolysis and proteolysis provide substrates for increased gluconeogenesis by the liver.

Approaches to management

Operationally, all patients with type 1 diabetes undergoing minor or major surgery and patients with type 2 diabetes undergoing major surgery are considered appropriate candidates for intensive perioperative diabetes management. The management approach in these categories of patients always includes insulin therapy in combination with dextrose and potassium infusion.

The aims of perioperative management are:

• to avoid hypoglycaemia particularly during anaesthesia

• to avoid excess metabolic decompensation

• to achieve a locally agreed care pathway with which health professionals are familiar.

Tip box

Major surgery is defined as one requiring general anaesthesia of ≥ 1 h.

Patients managed by diet alone

People whose diabetes is well controlled by a regimen of dietary modification and physical activity may require no special preoperative intervention for diabetes. The fasting blood glucose level should be measured on the morning of surgery and intraoperative blood glucose monitoring is desirable if the surgical procedure is lengthy (> 1 h). If surgery is minor, no specific therapy is required. If surgery is major or if diabetes is poorly controlled (blood glucose ≥ 11 mmol/L), an IV infusion of insulin and dextrose should be considered (see below) and hourly intraoperative glucose monitoring is recommended.

Patients treated with oral antidiabetic agents

Second-generation sulphonylureas (gliclazide, glipizide and glibenclamide) should be discontinued 1 day before surgery. Other oral agents can be continued until the day of surgery. Although metformin has a short half-life of about 6 h, it is sensible to withhold therapy for 1–2 days before surgery, especially in sick patients and those undergoing procedures that increase the risk of renal hypoperfusion, tissue hypoxia and possible lactate accumulation.

At a minimum, blood glucose should be monitored before and immediately after surgery in all patients. Those undergoing extensive procedures should have hourly glucose monitoring during and immediately after surgery. Bedside capillary blood glucose meters are adequate for these monitoring requirements. However, extremely high or low values should be repeated immediately before instituting remedial action, and a simultaneous blood specimen should be sent for laboratory corroboration.

Patients treated with oral antidiabetic agents

Second-generation sulphonylureas (gliclazide, glipizide, glibenclamide) should be discontinued 1 day before surgery. Other oral agents can be continued until the day of surgery. Although metformin has a short half-life of about 6 h, it is prudent temporarily to withhold therapy 1–2 days before surgery, especially in sick patients and those undergoing procedures that increase the risk of renal hypoperfusion, tissue hypoxia and lactate accumulation.

At a minimum, blood glucose should be monitored before and immediately after surgery. Those undergoing extensive procedures should have hourly glucose monitoring during and immediately after surgery. Bedside capillary blood glucose meters are adequate for these monitoring requirements. However, extremely high or low values should be repeated immediately and a simultaneous blood specimen should be sent for laboratory corroboration before instituting remedial action.

For minor surgery, perioperative hyperglycaemia (> 11 mmol/L) can be managed with small subcutaneous doses (4–10 units) of short-acting insulin. Care must be taken to avoid hypoglycaemia. After minor procedures, most usual antidiabetic medications can be restarted once the patient starts eating. In patients treated with metformin, the drug should be withheld for about 72 h after surgery or iodinated radiocontrast procedures. Metformin therapy can be restarted after documentation of normal renal function and absence of contrast-induced nephropathy. The recommended treatment for patients undergoing major surgery and for those with poorly controlled type 2 diabetes is IV insulin infusion, with glucose, using one of two standard regimens.

Insulin-treated patients

Minor surgery

Patients treated with long-acting insulin (e.g. glargine, detemir) should be switched to intermediate-acting forms before elective surgery. Close perioperative blood glucose monitoring is crucial to avoid extremes of glycaemia. IV insulin–glucose–potassium should be commenced before surgery. Blood glucose levels should be monitored hourly during and immediately after surgery. The infusion should be stopped and usual insulin treatment resumed once oral intake is established. There should be a 1-h overlap between stopping IV insulin and reinstituting subcutaneous insulin.

Major surgery

Insulin-treated patients undergoing major elective surgery should preferably be admitted 2–3 days before surgery, if glycaemic control is suboptimal (HbA1c > 8%). If admission is not feasible, a diabetes specialist nurse should work with the patient to optimize self-monitoring of blood glucose (SMBG) values in the days. SMBG should be performed at least before each meal and also at bedtime, with target preprandial values of 80–120 mg/dL and bedtime values of 100–140 mg/dL.

The preoperative evaluation should include a thorough physical examination (with particular focus on autonomic neuropathy and cardiac status), measurement of serum electrolytes and creatinine and urine ketones. The presence of autonomic neuropathy mandates increased surveillance for hypotension, respiratory arrest and haemodynamic instability during surgery. Gross metabolic and electrolyte abnormalities (e.g. hyponatraemia, dyskalaemia, acidosis) should also be corrected before surgery.

Insulin

Several reports have emphasized the advantages of the insulin infusion regimen over subcutaneous delivery.

Two main methods of insulin delivery have been used:

• combining insulin with glucose and potassium in the same bag (the GKI regimen)

• delivering insulin separately with an infusion pump.

The combined GKI infusion is efficient, safe and effective in many patients but does not permit selective adjustment of insulin delivery without changing the bag. The glucose component can be either 5% or 10% dextrose. The latter provides more calories.

Regardless of whether separate or combined infusions are given, close monitoring is required to avoid catastrophe during these infusion regimens. A sample regimen for separate insulin infusion is indicated below.

Blood glucose level (mmol/L)	Infusion rate (units/h)
< 4	0
4.1–7.0	1
7.1–11.0	2
11.1–17.0	3
17.1–22.0	4
> 22	5

Check capillary blood glucose level hourly initially and then 2 hourly.

These recommendations must be interpreted flexibly, given the individual variability in insulin requirements and metabolic profiles. In the absence of strict evidence-based guidelines, the consensus approach is to avoid extremes of glycaemia (aiming for 120–180 mg/dL) and to tailor therapies to individual patients based on feedback from glucose monitoring.

The initial infusion rate can be estimated as between one-half and three-quarters of the patient’s total daily insulin dose, expressed as units per hour. Regular insulin, 0.5–1 units/h, is an appropriate starting dose for most type 1 diabetic patients. Patients treated with oral antidiabetic agents who require perioperative insulin infusion, as well as insulin-treated type 2 diabetic patients, can be given an initial infusion rate of 1–2 units/h.

An infusion rate of 1 unit/h is obtained by mixing 25 units of regular insulin in 250 mL saline (0.1 unit/mL) and infusing at a rate of 10 mL/h. Alternatively, 50 units of regular insulin is made up to 50 mL with saline and given by syringe pump at 1–2 mL/h. Adjustments to the insulin infusion rate are made to maintain blood glucose between 120 and 180 mg/dL.

The duration of insulin (and dextrose) infusions depends on the clinical status of the patient. The infusions should be continued postoperatively until oral intake is established, after which the usual diabetes treatment can be resumed. It is prudent to give the first subcutaneous dose of insulin 30–60 min before disconnecting the IV line.

Glucose

Adequate fluids must be administered to maintain intravascular volume. Fluid deficits from osmotic diuresis in poorly controlled diabetes can be considerable. The preferred fluids are normal saline and dextrose in water. Fluids containing lactate (i.e. Ringer’s lactate, Hartmann’s solution) cause exacerbation of hyperglycaemia.

Adequate glucose should be provided to prevent:

• catabolism

• starvation

• ketosis

• insulin-induced hypoglycaemia.

The physiological amount of glucose required to prevent catabolism in an average non-diabetic adult is approximately 120 g/day (or 5 g/h). With preoperative fasting, surgical stress and ongoing insulin therapy the caloric requirement in most diabetic patients averages 5–10 g/h glucose. This can be given as 5% or 10% dextrose. An infusion rate of 100 mL/h with 5% dextrose delivers 5 g/h glucose.

The usual range of perioperative blood glucose that clinicians are comfortable with is about 6.5–10 mmol/L. The insulin and glucose infusion rates should be adjusted accordingly. Insulin requirements are higher in:

• patients with sepsis

• obese patients

• unstable patients

• patients treated with steroids

• patients undergoing cardiopulmonary bypass surgery.

Potassium

The infusion of insulin and glucose induces an intracellular translocation of potassium, resulting in a risk of hypokalaemia. If renal function is normal and the patient has initially normal serum potassium, potassium chloride (10 mmol/L) should be added routinely to each 500 mL dextrose to maintain normokalaemia. Hyperkalaemia (confirmed with repeat measurement and electrocardiogram) plus renal insufficiency are contraindications to potassium infusion.

Emergency surgery

Unfortunately, many patients who require emergency surgery will have suboptimal glycaemic control. However, this is not necessarily a contraindication to undertaking potentially life-saving surgery. IV access should be secured and immediate blood specimens sent for glucose, electrolyte and acid–base assessment. Gross derangements of volume and electrolytes (e.g. hypokalaemia, hypernatraemia) should be corrected.

Surgery should be delayed, whenever feasible, in patients with DKA, so that the underlying acid–base disorder can be corrected or, at least, ameliorated. Patients with HHS are markedly dehydrated and should be restored quickly to good volume and improved metabolic status before surgery. Blood glucose should be monitored hourly at the bedside, and insulin, glucose and potassium infusion should be administered, as appropriate, to maintain blood glucose in the 6.5–10 mmol/L range. Serum potassium should be checked frequently (every 2–4 h) and potassium supplementation should be adjusted to ensure that the patient remains eukalaemic throughout surgery and postoperatively.

The initial evaluation of a diabetic patient with a suspected surgical emergency must include a thorough medical history and physical examination. Particular care must be taken to exclude DKA and other conditions that are likely to be mistaken for surgical emergencies. For example, patients with DKA and prominent abdominal symptoms have undergone needless surgical exploration for a non-existent acute abdominal emergency. Functional syndromes due to diabetic autonomic neuropathy of the gastrointestinal tract (gastroparesis, gastroenteropathy, intractable or cyclical vomiting) may mimic anatomical surgical emergencies. Similarly, the rare diabetic pseudotabes syndrome, characterized by sharp neuropathic pain along thoracolumbar dermatomes, can be confused with visceral disorders. Patients with pseudotabes typically have pupillary and gait abnormalities from associated cranial and peripheral neuropathy.

The glucose–potassium–insulin (GKI) regimen in adults (≥ 16 years of age)

The GKI regimen should be considered in the following situations:

• patients with insulin-requiring diabetes who need surgery

• patients with type 2 diabetes who need major surgery

• patients with poorly controlled type 2 diabetes who require surgery

• patients with insulin-requiring diabetes who need to fast for procedures (e.g. endoscopy or radiology procedures)

• patients with diabetes who are unable to eat or are nil by mouth (e.g. vomiting due to autonomic neuropathy)

• patients in the later stages of treatment for DKA.

The following regimen will work only if the initial blood glucose level is approximately in the target range of 6–11 mmol/L. If the initial blood glucose is greatly in excess of this range then IV insulin by infusion pump should be given at a rate of 6 units/h, with half-hourly monitoring of glucometer readings, until blood glucose level is less than 15 mmol/L.

Instructions

1. No subcutaneous insulin to be given on morning of operation or procedure.

2. Prior to setting up the GKI infusion, send blood to the laboratory for urgent glucose and urea and electrolytes and glucose estimation (use an emergency form).

3. Set up the GKI infusion at 8 a.m.:

500 mL 10% glucose with 10 mmol KCl (10% glucose and 0.15% potassium chloride IV infusion BP)
plus
15 units Actrapid insulin added to bag and mixed well

To be administered over 5 hours (100 ml/h)

The giving set should be run through with 20–30 mL of the GKI mixture before being connected to the patient.

4. The glucose and insulin infusion must be continued, regardless of whether other IV fluids are required.

5. Subsequent adjustment of the insulin regimen should be as follows:

Blood glucose result	Action
6–11 mmol/L	Continue infusion as above
> 11 mmol/L	Increase insulin concentration by discarding original bag and making up a new bag with 20 units Actrapid added
< 6 mmol/L	Decrease the insulin concentration by discarding original bag and making up a new bag with 10 units Actrapid added

See Section 2, page 87 on adjustment of insulin dose if further dosage change becomes necessary.

Frequency of glucose monitoring:

• In theatre and recovery area: as decided by the anaesthetist

• Thereafter:

• immediately on return to ward

• 2 h and 4 h after return to ward

• then every 2–4 h

• once blood glucose level is stable, every 4 h.

• At least two blood glucose measurements during 24 h must be sent to the laboratory.

Urea and electrolytes should be measured on return from theatre and thereafter once daily. If the patient is on a GKI regimen for more than 24 h, ensure that 0.9% sodium chloride is also given to avoid hyponatraemia. If the potassium replacement requires adjustment on the basis of urea and electrolytes testing, do this by altering the potassium content of other IV fluids, such as sodium chloride 0.9%. Do not adjust the potassium content of the GKI.

Infusion should continue whilst light diet is started. Once the patient is eating reliably, subcutaneous insulin can be restarted. The glucose–insulin infusion should not be stopped until 1 hour after the first subcutaneous injection.

Adjustment of insulin dose in GKI regimen

In clinical trials, 80% of truly insulin-dependent diabetics undergoing surgery were controlled within the target limits of blood glucose using a regimen starting with 15 units insulin per 500-mL bag of 10% glucose with 10 mmol KCl and adjusted if necessary within the following limits:

• Measure blood glucose every 2 h initially – frequency of monitoring is variable and depends on the clinical situation; levels can usually be measured less frequently as time progresses.

• Adjust infusion as follows: continue to adjust in 5 units insulin per 500 mL glucose 10% with 10 mmol KCl (1 unit insulin per h) increments/decrements as necessary. However, if moving outwith the range of 10–20 units insulin/500 mL (2–4 units/h) consider the factors listed below before making change.

Blood glucose result	Action
6–11 mmol/L	Continue usual GKI regimen: 15 units per 500 mL 10% glucose (3 units/h)
> 11 mmol/L	Increase to 20 units per 500 mL 10% glucose (4 units/h)
< 6 mmol/L	Decrease to 10 units per 500 mL 10% glucose (2 units/h)

Where target blood glucose levels are not achieved using the above algorithm, consider:

1. The GKI regimen assumes that metabolism is in a steady state. It will work only if the initial blood glucose level was in the target range. It should not be used for patients who are severely hyperglycaemic at the onset (when a separate insulin infusion is better suited), or when they have started to eat again. If the initial blood glucose level is outwith the target range, appropriate measures must be taken to bring it within the range before using the GKI regimen.

2. Check that the IV cannula (Venflon) is patent and functioning, and consider that the GKI solution may have been constituted incorrectly.

3. The constant infusion of glucose at the rate specified is an essential part of the regimen. The insulin regimen cannot be applied when an infusion fluid other than 10% or 5% glucose administered over 5 h is being used (although additional 0.9% sodium chloride can be infused at the same time as 10% glucose when required).

4. In the following situations, altered insulin requirement is likely to occur:

Condition	Insulin requirement
Patients with subcutaneous insulin requirement < 30 units per day; thin elderly type 2 diabetics	May require reduced insulin – start at 10 units per 500-mL bag 10% glucose and 0.15% KCl IV infusion (2 units/h)
Obesity	May require increased insulin – start at 15 units per 500-mL bag 10% glucose and 0.15% KCl IV infusion (3 units/h), but be aware of possibility of increased requirement
Severe infection	May require increased insulin – may need as much as 25–40 units per 500-mL bag 10% glucose and 0.15% KCl IV infusion (5–8 units/h), or more
Any severe illness, infusion of catecholamines or inotropes	Often marked increase in insulin requirement – consider using infusion pump for insulin rather than adding to bag of glucose. Insulin infusion requirement of ≥ 6–12 units/h possible. More frequent monitoring and adjustment required
Steroid therapy	May require increased insulin

5. Insulin-dependent diabetics never require no insulin (other than very transiently). The half-life of IV insulin is approximately 4 min, and that of its biological effect approximately 20 min. The presence of an apparently zero insulin requirement therefore implies a tissue depot of insulin, such as a recent subcutaneous administration of insulin, or persistence of a depot of long-acting insulin (e.g. Lantus®).

6. Check daily for dilutional hyponatraemia.

‘Sliding scales’ with wide doses ranging from zero to high infusion rates (as published in some clinical handbooks):

• risk ‘roller-coastering’ – wild fluctuations in blood glucose levels with over-frequent adjustments in dosage

• frequently ’work’ satisfactorily by good fortune when they are employed in the middle of their dose range.

It is therefore better to use the GKI dosage algorithm and, where dosages lie outside the algorithm limits, bracket the patient’s requirements as in the notes above.

Sliding-scale intravenous regimen

Add 50 units soluble insulin (Actrapid or Humulin S) to 50 mL 0.9% saline in a 50-mL syringe. Infuse IV using a pump and adjust according to the sliding scale. Sliding-scale regimens need to be re-evaluated frequently as insulin dosages need to be adjusted to achieve target glycaemic control.

If blood glucose IS regularly outwith the range of 6–12 mmol/L, insulin doses should be reassessed and the diabetes/medical team contacted.

Intercurrent illnesses

Type 1 diabetes: sick-day rules

As people with type 1 diabetes are at risk of developing DKA, it is important to take action at the earliest possible sign of any form of illness such as a cold, infection or virus. Regular monitoring of blood glucose levels is advised and insulin doses may need to be increased if necessary. Those with type 1 diabetes are also advised to check for the presence of ketones.

How can I test for ketones?

It is possible to check for ketones in the urine or blood. The ketones build up in the bloodstream and spill over into the urine causing ‘ketonuria’, which can be detected by dipping a test strip in the urine and reading the colour. The test strips are available on prescription.

Many people with type 1 diabetes can benefit from measuring blood ketones using a special meter that can detect levels of ketones in the blood. It is advisable to contact a diabetes nurse or doctor for further advice on this method of testing for ketones. Again, the strips are available on prescription.

People with type 1 diabetes should test their urine or blood for ketones if their blood glucose levels are high (usually over 14 mmol/L) or if they are developing symptoms of ketoacidosis. DKA or ketoacidosis can develop and progress quickly. Treatment for advanced DKA may require admission to hospital. However, it is possible to prevent ketoacidosis by recognizing the warning signs, checking the urine and blood regularly for ketones and glucose, and treating with extra insulin.

What action should be taken if blood glucose levels are high?

Guidelines

Guidelines for extra insulin to treat hyperglycaemia or raised ketones depend upon the total daily dose of insulin taken by the individual. This can be calculated by adding together all the long- and short-acting insulin taken over 24 hours. The table below provides guidance for people with type 1 diabetes who are having four or more injections a day (known as the basal-bolus or basal-prandial regimen).

What action should be taken for a positive ketone test?

• Take extra rapid-acting insulin as soon as possible (see table above).

• Drink plenty of water and unsweetened fluids.

• Test blood glucose every 1–2 h and repeat the extra insulin (see table above) until blood/urine is negative to ketones.

• Try to identify cause of high blood glucose level and seek treatment if necessary.

• Contact diabetes team if high glucose and ketone levels persist.

• Contact GP/accident and emergency department if you are vomiting, as dehydration may occur.

• Continue with the usual amount of background insulin.

REMEMBER – drink at least a cupful of unsweetened fluid every 15 min (about 500 mL per hour) while blood glucose levels are high.
ALWAYS – take insulin during times of illness.

Bariatric surgery

Efficacy of weight loss health outcomes

Bariatric surgery is an effective weight loss intervention. Patients with a body mass index (BMI) ≥ 35 kg/m² receiving bariatric surgery (laparoscopic banding, biliopancreatic diversion ± Roux-en-Y gastric bypass) will lose between 50% and 80% excess weight at 10 years postsurgery. Laparoscopic adjustable banding in diabetic patients with a BMI > 35 kg/m² results in greater excess weight loss at 2 years compared with intensive diet, lifestyle and medical therapy (87.2% versus 21.8%, respectively; P < 0.001).

The overall mortality rate is 29–40% lower 7–10 years after bariatric surgery (adjustable or non-adjustable gastric banding, vertical banded gastroplasty or gastric bypass) compared with that in BMI-matched subjects not having surgery. More than 70% of recently diagnosed type 2 diabetic patients (less than 2 years since diagnosis) undergoing bariatric surgery with adjustable gastric banding are likely to go into remission with reduction of other cardiovascular risk factors, compared with 10% of the control group (where the focus is on weight loss by lifestyle change). Surgical and control/lifestyle groups lose approximately 20% and 2% of their weight, respectively, at 2 years, with no serious complications in either group. Remission of diabetes appears to be related to the degree of weight loss and lower baseline HbA1c levels. There is also a significantly higher level of withdrawal of diabetic medications following surgery.

Rates of many adverse maternal (e.g. gestational diabetes and pre-eclampsia) and neonatal (e.g. macrosomia and low birth weight) outcomes are lower in women who become pregnant after having had bariatric surgery, compared with rates in pregnant women who are obese.

There are greater improvements at 10 years in current health perceptions, social interactions, obesity-related problems and depression in patients who had bariatric surgery compared with those having the best available medical weight management. However, individuals who chose to have bariatric surgery had worse health-related quality-of-life scores at baseline.

There is a significantly higher mortality rate from non-disease causes (accidents, poisoning, and suicide) at 10 years postsurgery in patients receiving bariatric surgery compared with severely obese individuals from the general population. The reasons are not clear, although the individuals who seek bariatric surgery have differing baseline psychological status (e.g. increased anxiety levels) than those with similar obesity levels who do not seek surgery.

Factors influencing the efficacy of surgery

Predictors of efficacy of bariatric surgery, particularly laparoscopic adjustable gastric banding, include lower age of patient, lower BMI and male sex. Surgical experience is also a predictor of successful outcome. Predictors of mortality from gastric bypass include BMI > 50 kg/m², male sex, hypertension, high risk of pulmonary thromboembolism and age above 45 years.

Presence of depression does not influence efficacy of surgery. Increased psychological dysfunction, dysfunctional eating behaviour, binge-eating disorder and a past history of intervention for substance misuse are not associated with poorer weight loss outcomes. Data are available for up to 6 years postoperatively.

Harms and the balance of risk

Postoperative mortality in the bariatric surgery group is approximately 0.25% at 90 days. This included bleeding (0.5%), thromboembolic events (0.8%), wound complications (1.8%), deep infection, abscess or leak (2.1%), and pulmonary complications (6.2%); reoperation was necessary in approximately 2% of patients.

Men who undergo bariatric surgery (adjustable or non-adjustable gastric banding, vertical banding gastroplasty or gastric bypass) have a 4.2-fold increased incidence of cholelithiasis, 4.5-fold increased incidence of cholecystitis and a 5.4-fold increased incidence of cholecystectomy. There is no difference in the frequency of biliary disease in women following bariatric surgery.

Nutrition/supplementation

Plasma vitamin D concentrations are low in obese patients before bariatric surgery and are associated with increased parathyroid hormone levels (secondary hyperparathyroidism). Postsurgery vitamin D concentrations are either unchanged or occasionally increased.

Obese adults with type 2 diabetes should be offered individualized interventions to encourage weight loss (including lifestyle, pharmacological or surgical interventions), in order to improve metabolic control.

Male and female sexual dysfunction

Erectile dysfunction (Tables 6.6–6.9)

The prevalence of erectile dysfunction (ED) in men with diabetes increases with age and is about 35–50% overall. Tumescence is a vascular process under the control of the autonomic nervous system. The erectile tissue of the corpus cavernosum behaves as a sponge and erection occurs when it becomes engorged with blood. This leads to compression of the outflow venules against the rigid tunica albuginea. Smooth muscle relaxation is the key phenomenon in this process. The process is under the control of parasympathetic fibres; the neurotransmitter involved is now known to be nitric oxide (NO). NO is produced in the parasympathetic nerve terminals and is generated by NO synthase in the vascular endothelium. Within the smooth muscle cell of the corpus cavernosum, NO stimulates guanylate cyclase, leading to increased production of the second messenger, cyclic guanosine monophosphate (cGMP), which induces smooth muscle relaxation, probably by opening up calcium channels. There is also evidence that neuronally derived NO is important in initiation, whereas NO from the endothelium is responsible for maintenance of the erection.

Table 6.6 Key features in the clinical history of erectile dysfunction in diabetes

• Onset usually gradual and progressive

• Earliest feature often inability to sustain erection long enough for satisfactory intercourse

• Erectile failure may be intermittent initially

• Sudden onset often thought to indicate a psychogenic cause

• Preservation of spontaneous and early morning erections does not necessarily indicate a psychogenic cause

• Loss of libido is consistent with hypogonadism

• Impotent men often understate their sex drive for a variety of reasons

Table 6.7 Medications associated with erectile dysfunction

Antihypertensives

• Thiazide diuretics

• Beta-blockers

• Calcium-channel blockers

• Angiotensin converting enzyme (ACE) inhibitors

• Central sympatholytics (methyldopa, clonidine)

Antidepressants

• Tricyclics

• Monoamine oxidase inhibitors

(NB: selective serotonin reuptake inhibitors can cause ejaculatory problems) Major tranquillizers

• Phenothiazines

• Haloperidol

Hormones

• Luteinizing hormone-releasing hormone (goserelin, buserelin)

• Oestrogens (diethylstilbestrol/stilbestrol)

• Anti-androgens (cyproterone)

Miscellaneous

• 5α-reductase inhibitors (finasteride)

• Statins (simvastatin, atorvastatin, pravastatin)

• Non-steroidal anti-inflammatory agents

• Fibrates

Drugs of ’abuse’/’social’ drugs

• Alcohol, tobacco, marijuana

Table 6.8 Conditions associated with erectile dysfunction

Psychological disorders

• Anxiety about sexual performance

• Psychological trauma or abuse

• Misconceptions

• Sexual problems in the partner

• Depression

• Psychoses

Vascular disorders

• Peripheral vascular disease

• Hypertension

• Venous leak

• Pelvic trauma

Neurological disorders

• Stroke

• Multiple sclerosis

• Spinal and pelvic trauma

• Peripheral neuropathies

Endocrine and metabolic disorders

• Diabetes

• Hypogonadism/hyperprolactinaemia/hypopituitarism

• Thyroid dysfunction

• Hyperlipidaemia

• Renal disease/liver disease

Miscellaneous

• Surgery and trauma

• Smoking

• Drug and alcohol abuse

• Structural abnormalities of the penis

Table 6.9 Investigation of erectile dysfunction in diabetes

Serum testosterone if libido reduced or hypogonadism suspected (ideally taken at 9 a.m.)
Serum prolactin and luteinizing hormone if serum testosterone subnormal
Assessment of cardiovascular status if clinically indicated:

• Electrocardiography (ECG)

• Serum lipids

Glycosylated haemoglobin, serum electrolytes if clinically indicated

In men with diabetes there is evidence that ED is caused by failure of NO-induced smooth muscle relaxation caused by both autonomic neuropathy and endothelial dysfunction.

More recently, other potential abnormalities have been described that may contribute to the development of ED in diabetes:

• Endothelium-derived hyperpolarizing factor (EDHF) has a role in endothelium-dependent relaxation of human penile arteries, and this is significantly impaired in penile resistance arteries in men with diabetes.

• The formation of products of non-enzymatic glycosylation to produce advanced glycosylation end-products generates reactive oxygen species that impair NO bioactivity.

• Non-enzymatic glycation of proteins has been reported to impair endothelium-dependent relaxation of aorta in rats.