Diabetes in pregnancy

Diabetes is one of the more difficult conditions to manage, under the conditions for which we write, especially in pregnancy. Insulin supplies are likely to be a major problem, with patients dying when their supplies are interrupted. Even where care is good, it may only be the teachers and health workers who are sufficiently motivated and educated to manage their diabetes well enough to survive for long. Here is what you should do, with some advice as to what you can do, if your facilities are so limited that you can only measure the blood glucose exceptionally, if at all. If so, you will not be able to control a patient's diabetes well enough to prevent its adverse effects on her or her baby, but you should be able to keep her out of coma.

Pregnancy and diabetes adversely affect one another, and the more complicated or long-standing the diabetes, the stronger the effect. Diabetes increases maternal mortality 10 times, as the result of ketoacidosis, hypoglycaemia, and infection. A mother is also at increased risk from polyhydramnios ([mu]50), gestational hypertension ([mu]3), and infection (especially pyelonephritis). Fortunately, pregnancy will not usually harm her health in the long term. If however she has retinopathy, this may get worse during pregnancy, although it may improve afterwards; uropathy on the other hand may worsen irreversibly. Her baby is at increased risk from sudden intrauterine death (usually at 36 to 38 weeks), fetal macrosomia (excessive size), neonatal hypoglycaemia (19.12), prematurity (particularly hyaline membrane disease), and congenital malformations ([mu]3). Careful diabetic management reduces these risks.

At the start of pregnancy she can have: (1) Type One diabetes (IDDM, insulin-dependent diabetes mellitus). (2) Type Two diabetes (NIDDM non-insulin-dependent diabetes mellitus). This is uncommon in women of childbearing age in Africa, and is more common in obese mothers in India and the Middle East. We shall not make any further reference to Types One and Two diabetics; it is the insulin-dependent Type One patients who are important. (3) Impaired glucose tolerance. However, following Chalmers, who concluded that: ''Except for research purposes, all forms of glucose tolerance testing should be stopped (in pregnancy)'', we say nothing more about it here. (4) Renal glycosuria, which is a lowered renal threshold to glucose, and is harmless. If diabetes resolves after pregnancy, it is said to have been ''gestational'.

About 1% of women have glucose in their urine when they are fasting. Most of them have renal glycosuria, and a few have diabetes.

Although testing the urine for glucose in early pregnancy picks up all cases of diabetes, it is an imperfect screening method because it picks up many harmless cases of renal glycosuria. Glucose in the urine may vary greatly from day to day during pregnancy, so that a patient may have glycosuria on one visit and not on the next. Also, testing the urine is also a poor way of monitoring the insulin requirements of a pregnant diabetic, especially after 16 weeks.

Unfortunately, testing the urine for glucose is the only practical screening test, so test the urine of all mothers, and investigate them as best you can.

Diabetic babies tend to be large, so don't let a mother become overdue. Ideally, she should deliver at 36 to 38 weeks. If she is a multip, you can usually induce her fairly easily.

Chalmers I, Enkin M, Keirse MJNC, ''Effective Care in Pregnancy and Childbirth', 1989 OUP. Fig. 17-1a DIAGNOSING DIABETES. Note that the significance of a particular value varies, depending on whether the sample is whole blood or plasma, venous or capillary. ''Diabetes'. 1985. WHO Technical Report Series 227.

DIABETES [s8]DURING PREGNANCY DIAGNOSIS Test the urine of all pregnant mothers.

If a mother has obvious symptoms (polyuria, thirst, pruritus, etc.), and heavy glycosuria, the diagnosis of diabetes is almost certain; but confirm it, if you can, by measuring her blood glucose and consulting Fig. 17-1a.

If she has glycosuria and no symptoms, do a random blood glucose, or a fasting blood glucose. She has renal glycosuria if she has glucose in her urine when her blood glucose is [lt]6.7 mmol/l.

RANDOM BLOOD GLUCOSE. She is diabetic if, when she is not fasting, this is [mt]10.0 mmol/l venous, or [mt]11.1 mmol/l, (200 mg/dl) capillary (both whole blood). If she has no diabetic symptoms (polyuria or polydipsia, etc.), two abnormal levels are preferable for diagnosis.

A FASTING BLOOD GLUCOSE tells you more than a random blood glucose, but it is less convenient, because she does have to fast for 8 hours, and this is likely to mean another visit. If it is [lt]6.7, she has not got diabetes.

THE MANAGEMENT [s7]OF DIABETES A baby's survival and welfare depend absolutely on strict ''normoglycaemia' (his mother's fasting blood glucose should be [lt]4 mmol/l and her postprandial peaks 2 hours after meals [lt]7.5 mmol/l). Ideally, this requires careful blood glucose monitoring, if possible at home also.

If she is a diabetic on oral hypoglycaemic agents she will be better controlled if you can change her to soluble insulin, and manage her as below. These agents are not recommended in pregnancy, although they have not been proved to be teratogenic. They cause neonatal hypoglycaemia, so be sure to stop them by 28 weeks and change to insulin. Stop them at least a week before a planned delivery.

If she is a known diabetic on insulin, she should, ideally, have a planned pregnancy, and be carefully stabilized for 3 months before she becomes pregnant, so that she conceives when she is normoglycaemic, because this reduces the risk of a congenital abnormality.

If she is a known diabetic on insulin, or has been diagnosed as diabetic by the criteria above, try to check her diabetic stability every 2 weeks during pregnancy, and if necessary admit her for the day. Adjust her diet so that it is steady in terms of content and timing, and put her on twice- daily injections of short- and medium-acting insulins (''soluble and isophane', ''soluble and lente', ''Actrapid and Monotard', or regular and NPH). If she is sophisticated, she may be able to adjust her food intake to match this regime; if not, adjust her injections to fit her meals. In early pregnancy, her insulin needs may fall (unusual), so that if she has previously been well controlled, she may become hypoglycaemic. You may have to reduce her dose by about 25% until 16 weeks. After 16 weeks her insulin needs rise again (usual), so that near term she needs 1.5[nd]2 times as much as she did before pregnancy. She is therefore most likely to become ketotic late in pregnancy. Her insulin needs fall immediately the placenta is delivered.

If she is well-motivated and co-operative, aim for a fasting blood glucose of [lt]4.5 mmol/l, and a postprandial level of [lt]8 mmol/l. In achieving this she may be hypoglycaemic sometimes, so make sure she takes a glucose or sugar drink with the first signs of hypoglycaemia, and is prepared for them.

If she is sufficiently frightened by hypoglycaemia to stop taking her insulin, or fears harming herself or her baby, accept less strict control, but aim for a blood glucose of [lt]6 mmol/l, and a postprandial level of [lt]10 mmol/l.

DETERMINING HER GLUCOSE (''sugar') PROFILE. Give her insulin half an hour before breakfast and half an hour before her evening meal. Measure her blood glucose 4 times: at 0600 hrs (reflecting her evening isophane insulin); at 1000 hrs (reflecting her morning soluble insulin; at 1500 hrs (reflecting her morning isophane insulin); and at 2000 hrs, reflecting her evening soluble insulin. Adjust her dose by 2[nd]5 units, increasing it by one unit for each mmol/l (20 mg/dl). If any value is [mt]4 mmol/l higher than you want, or if more than 2 doses need adjusting, keep her for another day and repeat the profile.

Starvation ketosis is bad for a baby, so test her urine for ketones after an overnight fast. If she has ketones, give her a late-night snack.

If only 1 or 2 insulin doses needed changing by [lt]4 units each, repeat the profile in 2 weeks. Otherwise, or if she has had any ketones, check her in one week.

DIABETES IN LATE PREGNANCY. Admit her and adjust her insulin dose on the basis of her blood glucose, or, failing that, her glycosuria. Estimate fetal maturity clinically; if her dates are uncertain, withdraw some amniotic fluid and do a surfactant test (unfortunately, this is less reliable in diabetes). Repeat this weekly and, provided she has no complications, especially no hypertension or gross hydramnios (girth more than 100 cm), induce her as soon as her surfactant test is positive (19.2). Do kick counts (M 28.3), and try to bring her to at least 36 weeks.

THE DELIVERY [s7]OF A DIABETIC MOTHER If her diabetes has been well controlled, and her baby is of average size, and she has no CPD (cephalopelvic disproportion), and she has a positive surfactant test, induce her at 37 weeks.

If she has serious hypertension or polyhydramnios, deliver her earlier, whatever the maturity of her baby.

Section her, if control has been poor, or there is CPD, or she has a malpresentation, or severe gestational hypertension, or a history of a previous difficult delivery, or polyhydramnios or a previous section, or if she is elderly. Your threshold for section should be lower in a diabetic, especially if she wants her tubes tied.

DIABETES DURING DELIVERY. Aim to provide her with glucose, insulin, and fluid, to prevent ketosis, and to keep her normoglycaemic. Don't worry if her blood glucose rises moderately during labour; hypoglycaemia is much more dangerous.

Ideally, give her a continuous constant glucose infusion at a fixed rate using a drip counter, and insulin by a syringe driver, adjusted according to her hourly blood-glucose levels.

Alternatively, give her a continuous glucose, potassium, and insulin (''GKI') infusion: 500 ml of 5% dextrose, with 8 units of soluble insulin (any kind), and 10 mmol of potassium chloride, at 120 ml/hour or 20 drops/min (if you use 10% dextrose give her 15 units of insulin). A 500 ml bottle will then last 4 hours, which will coincide with her 4-hourly urine tests. Test her blood glucose hourly with ''BMstix', or less satisfactorily ''Dextrostix', accepting a range of 4 to 10 mmol/l. If it is [mt]10 mmol/l, take the present bag down and put 12 units in the next one. If it is [lt]4 mmol/l, put 4 units in the next one.

If you cannot measure her blood glucose, test her urine, but make sure that she empties her bladder half an hour beforehand. At 0700 hrs on the morning of delivery (whether for induction or section) give her no food. Instead, start a drip of 5% dextrose in water and run it at 20 drops a minute. If she is having soluble insulin, give her half her normal dose.

Measure and chart her urine glucose 4-hourly, and give her soluble isulin intramuscularly on the scale below. If you find ketones, add 8 units. If you give insulin intramuscularly, it is active for longer, and hypoglycaemia is less common.

Urine glucose nil, add no soluble insulin (blood glucose [lt]6.2 mmol/l, 120 mg/dl).

Urine glucose +, soluble insulin 8 units (blood glucose 6.2[nd]8.4 mmol/l, 120[nd]150 mg/dl).

Urine glucose ++, add soluble insulin 16 units (blood glucose 8.4[nd]11.2 mmol/l, 150[nd]200 mg/dl).

Urine glucose +++, add soluble insulin 24 units (blood glucose 11.2[nd]14 mmol/l 200[nd]250 mg/dl).

Urine glucose ++++, add soluble insulin 32 units (blood glucose [mt]14 mmol/l, 250 mg/dl).

CAUTION ! While she is in the labour ward, test her urine for ketones routinely. If you find them, she needs extra glucose. This is best given with insulin.

INDUCTION. If she is to be induced, rupture her membranes (19.3), and give her an oxytocin drip (M 22.2) in a separate bottle from the dextrose.

If she is not in established labour within 4 to 6 hours of induction, or if delivery is not imminent 10 to 12 hours after induction, section her. In a diabetic there is no place for a ''trial of labour', or a long labour. If necessary, assist vaginal delivery with a vacuum extractor, or outlet forceps.

CAUTION ! In communities where CPD is common shoulder dystocia can be a problem in delivering diabetic babies. You can avoid it by sectioning her, but if you deliver her vaginally, be prepared for it[md]see below.

A DIABETIC BABY should be normal if his mother's diabetes has been well controlled. If it has not been well controlled, see above and Section 19.12.

CAESAREAN SECTION. Give her a ''GKI' infusion or the alternative method, as described above. Put her first on the list and continue the drip through the operation and afterwards until she can take oral fluids.

AFTER DELIVERY of the placenta, her insulin requirements fall sharply (less sharply if she gets an infection). Give her subcutaneous insulin with the next meal. If she has been sectioned, do the same, but she will need the drip up for longer. She should need about a third or half the insulin she needed before delivery, and about the same amount as she needed before she became pregnant. She can start to eat and drink as soon as she wishes, but if this is delayed continue to give her 2 litres of 5% dextrose and one litre of 5% dextrose in 0.9% saline daily (her glucose infusion must be continuous).

If she needed insulin before she became pregnant, she still will still need it and can return to her original dose.

If her diabetes disappears after pregnancy, it was only gestational. If so, her blood glucose will usually have fallen to normal 48 hours after delivery, without any risk of ketosis.

DIFFICULTIES [s7]WITH DIABETES IN PREGNANCY If you CANNOT MEASURE HER BLOOD GLUCOSE, you will have to rely on testing her urine. Make sure she empties her bladder half an hour before producing the next specimen for testing. Aim to make her urine specimens negative without causing hypoglycaemia. If she has hypoglycaemic episodes with glucose in her urine, she probably has renal glycosuria.

If she has SEVERE NAUSEA AND VOMITING OF PREGNANCY, and is receiving insulin, admit her and give her a glucose/insulin infusion.

If her INSULIN REQUIREMENTS FALL ABRUPTLY AT 36 WEEKS, it is probably a sign of impending placental failure, and delivery is indicated.

If her baby's head delivers but his SHOULDERS HAVE STUCK, there is SHOULDER DYSTOCIA, which can be serious (see M 21.4). An episiotomy will not help the dystocia, but it will reduce the risk of perineal or anterior vaginal injury. Ask your assistant to apply firm fundal pressure. Strong traction on his head may disimpact his shoulders, but injure his brachial plexus. Symphysiotomy is safer, so infiltrate local anaesthetic solution round her symphysis at the same time as you do an episiotomy and insert a catheter. If you are very skilled, wait for his head to deliver and then decide if a symphysiotomy is necessary. If so, do it quickly. If you are less confident, do it as the head crowns in any patient in whom you anticipate difficulty. As always, never do a symphysiotomy without also doing an episiotomy.

If he is alive, you can push his head back into her pelvis and do a Caesarean section (the Zavanelli manoeuvre). For example, if his head was born occipito-anterior, and has turned (restituted) to the side, turn his occiput anterior again. Then flex his head. It will probably go back into her uterus easily. Section her without delay.

Fig. 17-2 AN ECLAMPTIC FIT is a [f10]grand mal [f11]epileptic convulsion. This is what you must try to prevent. A tonic phase in which all the muscles of the body contract, so that it becomes rigid, is followed by a clonic phase of rhythmic muscular contractions, as shown here.