Diabetic ketoacidosis in pregnancy

Abstract

The clinical presentation of diabetic ketoacidosis (DKA) includes nausea, vomiting, thirst, polyuria, polydipsia, abdominal pain, tachypnoea, and change in mental status in cases of severe DKA. DKA is similar in pregnant and non-pregnant women, but in pregnant women it can be seen at lower serum glucose levels and symptoms may develop more rapidly. Most, but not all, cases occur in the second or third trimester.

DKA results in reduction in uteroplacental blood flow due to osmotic diuresis, and also in metabolic abnormalities (maternal acidosis, hyperglycaemia, electrolyte imbalance), resulting in fetal hypoxaemia and acidosis. In fetuses with mature cardiac activity, the fetal heart rate may show minimal or absent variability, repetitive deceleration and absence of acceleration. These abnormalities in heart rate usually resolve with resolution of the DKA, which may last for several hours before normalisation.

For the patient reported on here, immediate delivery based on pathological fetal heart rate would have resulted in preterm delivery and jeopardised the maternal clinical condition. However, a holistic clinical approach by the multidisciplinary team to management of the patient led to normal term delivery 5 weeks after presentation with DKA; fetal and maternal outcome were good.

Background

Diabetic ketoacidosis (DKA) results from absolute or relative insulin deficiency combined with counter-regulatory hormone excesses (ie, glucagon, glucocorticoids, catecholamines and growth hormone). ^{1 2} Pregnant women are prone to ketosis because of metabolic changes that have been referred to as ‘accelerated starvation’. Pregnancy also induces a state of insulin resistance in late gestation that may result in greater relative insulin deficiency.³

Triggers for DKA, which include infection, trauma and cardiac ischaemia, are not specific to pregnancy. Additionally, insulin omission, which sometimes occurs when an insulin pump system fails owing to blockage or leakage in the infusion set or connectors, can also contribute to the development of DKA.³

Case presentation

A woman in her mid-20s, primigravida at (31+5) weeks' gestation with type 1 diabetes (on an insulin pump; 60 U basal+10–15 U with meal), presented to the maternity triage after long-haul travel returning from holiday in California.

At the triage, the patient’s main complaint was decreased fetal movement and mild lower back pain. She also reported having vomited once before coming to the triage.

Other than diabetes and a history of multiple surgical removal of pilonidal cyst, she had no significant medical or surgical history. On assessment of the patient the following measurements were obtained: blood pressure 120/85 mm Hg, pulse 85 bpm and O₂ saturation 97% on room air; a dipstick showed a urine ketone result of 3+ and glucose result of 3+, blood glucose was 24.8 mmol/L and blood ketones were 4.3 mmol/l pH 7.3K 5.9 mmol/l and bicarbonate 12.5 mmol/l.

The patient was alert and conscious, the abdomen was soft and lax, the fundal level was at 32 weeks, the uterus was lax and bilateral leg swelling limited to the ankles was noted. A vaginal examination showed that the cervix was closed, no vaginal loss was noted and the Actim-Partus result was negative. The insulin pump was checked by the diabetic team and showed no leakage or block, and no precipitating risk factor was identified.

Fetal cardiotocography (CTG) was started showing a baseline fetal heart rate of 160 bpm, reduced variability, unprovoked deceleration and no uterine activity (figures 1 and 2).

Figure 1.

Fetal cardiotocography (baseline 160 bpm, reduced variability, unprovoked deceleration and no uterine activity).

Figure 2.

Fetal cardiotocography (baseline 160 bpm, reduced variability, unprovoked deceleration, and no uterine activity).

Based on the multidisciplinary team (critical care team and outreach nurse, obstetric anaesthetist, obstetric consultant, obstetric registrar, lead midwives) assessment, pathological CTG was attributed to DKA and immediate management of DKA was needed, with close and continuous monitoring of the mother and fetus.

Treatment

The patient was admitted to the high-dependency unit in the delivery suite with close liaison with the critical care team, and a DKA protocol was initiated, with fluid management as follows:

• 0.9% NaCl 1 L over the first hour (no KCl).

• 0.9% NaCl 1 L with 20 mmol KCl over the next 2 hours (maximum 10 mmol KCl/h).

• 0.9% NaCl 1 L with 20 mmol KCl over the next 2 hours (maximum 10 mmol KCl/h).

• 0.9% NaCl 1 L with 40 mmol/l KCL over the next 4 hours (maximum 10 mmol KCl/h).

• 0.9% NaCl 1 L with 40mmol/l KCL over the next 4 hours (maximum 10 mmol KCl/h).

• 50 Units human of soluble insulin (Actrapid) made up to 50 mL with 0.9% NaCl solution started at a continuous IV fixed rate (0.1 U/kg/h) via an infusion pump.

• Anti-sickness medication prescribed.

• Arterial line inserted, hourly monitoring of blood glucose, blood gases and ketones.

Outcome and follow-up

After 2 hours of starting the DKA protocol, the patient's clinical and biochemical condition (ketones decreased to 1.6 mmol/l and blood glucose to 12 mmol/L) improved. The patient's perception of fetal movement increased. CTG variability improved and there was only sporadic unprovoked deceleration, which was not consistent during fetal monitoring (figure 3).

Figure 3.

Fetal cardiotocography (variability improved and only sporadic unprovoked deceleration, not consistent during fetal monitoring).

Three hours later, the patient's ketones went down to 0.5 mmol/l, blood glucose was 11 mmol/L and the insulin pump was stopped. In accordance with medical team advice, basal insulin (Lantus) was started in addition to the fixed-rate insulin. Fetal movement returned to normal and fetal CTG showed further improvement (baseline fetal heart rate settled from 160 bpm to 150 bpm, there was no deceleration and normal variability). However, Dawes-Redman criteria were not met owing to the absence of acceleration (figure 4).

Figure 4.

Fetal cardiotocography (baseline from 150 to 160 bpm, no deceleration, normal variability).

Four hours later, the patient's biochemical values were as follow (ketones 0.1mmol/l, blood glucose 10 mmol/L, bicarbonate 16.6mmol/l). fetal movement was normal and the CTG showed baseline 145 bpm, no deceleration, acceleration, normal variability and Dawes-Redman criteria met (figures 5 and 6). Accordingly, observation of the patient was de-escalated to two hourly observation, fixed rate was discontinued and a sliding scale started, intermittent fetal monitoring was eight hourly.

Figure 5.

Fetal cardiotocography (baseline 145 bpm, no deceleration, acceleration, normal variability and Dawes Redman criteria met).

Figure 6.

Fetal cardiotocography (baseline 145 bpm, no deceleration, acceleration, normal variability and Dawes-Redman criteria met).

Twenty-four hours after the initial presentation the patient was transferred to the maternity ward and a follow-up plan was set up by the diabetic and obstetric team. The insulin pump was restarted and no maternal or fetal compromise was recorded during her antenatal admission. The patient was discharged after 4 days with a follow-up appointment in the diabetic antenatal clinic.

After 5 weeks the patient presented to the triage with labour pain when she was at 36+5 weeks of gestation. She progressed well during her labour and was delivered by ventouse for a prolonged second stage of labour. She delivered a 3220 g baby girl with an Apgar score of 8/9/9. Her postnatal period was uneventful.

Discussion

Occurrence of DKA during pregnancy is exceedingly rare, with a reported incidence of between 0.5% and 3% of all diabetic gestations.^3–17 A study of pregnant women in the UK by Diguisto et al reported a local prevalence of DKA of between 0.1% and 1.6%.²

Our patient presented at 30 weeks' gestation. DKA developed during the third trimester of gestation in three other case reports of DKA in pregnancy.^5–15

Development of DKA is attributed to increased insulin resistance during pregnancy from several hormones, including human placental lactogen, placental insulinase and progesterone. These hormones peak in the second and third trimesters and can inhibit the effects of maternal insulin, resulting in insulin deficiency.¹⁵

Precipitating factors for DKA in pregnancy are infection, including COVID-19⁷; failure of, or non-adherence to, insulin therapy; steroid use for fetal lung maturity; dehydration and unrecognised new-onset diabetes mellitus, which accounts for up to 30% of cases of DKA.^{1 2} In our patient, no risk factor for DKA was identified, but the long-haul travel and relative dehydration might have contributed to DKA development.

It is imperative to have a high suspicion of ketoacidosis in an acidotic pregnant patient with diabetes mellitus or gestational diabetes. Placental sequestration of blood glucose can make DKA a diagnostic and therapeutic challenge. Serum ketones must be checked in any diabetic patient during periods of illness.¹⁵ In this patient the high blood sugar level and ketosis helped in detecting DKA, but a small percentage of patients with DKA may be euglycaemic despite developing sever acidosis, adding to the difficulty of diagnosing DKA. A case report describes a type 1 diabetic patient in her early 30s who was pregnant at 30 weeks, developed euglycaemic DKA and her pregnancy ended with intrauterine fetal death.¹⁵

DKA in pregnancy has immediate fetal complications, including high fetal mortality rates of 27–35%. Several reasons for fetal lethality have been discussed in previous studies: DKA-associated hypovolaemia causes decreased uteroplacental blood flow and increased concentrations of catecholamines, leading to fetal hypoxaemia; transfusion acidosis with electrolyte imbalance and hyperlactacidaemia worsens fetal hypoxaemia; lethal cardiac arrhythmias, might occur due to fetal hyperinsulinaemia-induced hypokalaemia.^{11 12}

The metabolic derangements that take place during DKA result in fetal hypoxaemia and acidosis, which may influence all modes of fetal testing. Monitoring of fetal heart rate during an acute episode of DKA often reveals minimal or absent variability, absent accelerations, repetitive variable and late decelerations.^4–18 In this case, fetal monitoring showed reduced variability, unprovoked deceleration and raised baseline fetal heart rate.

After correction of the metabolic derangements in DKA and maternal stabilisation, these fetal abnormalities will usually improve; however, it may take 4–8 hours for the fetal heart rate tracing to become normal.¹⁸ In our case, the trace showed gradual improvement with correction of the patient's acidotic condition until complete normalisation and reactivity after 9 hours.

The fetal biophysical profile can also be abnormal, and Doppler studies may show signs of blood ¹⁹ flow redistribution (ie, increased umbilical artery pulsatility index and reduced middle cerebral artery pulsatility index). The frequency and severity of these abnormalities will depend on the severity and duration of DKA.^4–18

In our case Doppler ultrasound was done after maternal stabilisation and showed normal growth and a normal Doppler result. Follow-up antenatal ultrasound showed no abnormality. In one case report study, ultrasound was used at the time of acute DKA presentation and showed a normal Doppler result and normal growth of the baby. The rationale for using ultrasound was to differentiate the symptoms from other possible causes of patient abdominal pain and abnormal trace; follow-up antenatal ultrasound showed no abnormality.²⁰

DKA during pregnancy has long-term complications.⁸ Although few publications have dealt with long-term outcomes, impairment of brain development (lower IQ) and the development of autistic children have been described.^{8 9}

It should be noted that acute severe DKA should not result in immediate emergency delivery, as this would harm the mother and most probably not save the child. However, when delivery is inevitable, as in cases where the maternal condition worsens despite aggressive treatment, it is associated with high maternal morbidity and mortality.^7–10

Our case shows how fetal heart trace is affected by DKA. The trace showed criteria of hypoxaemia and acidosis, the patient's symptoms were not severe, but her laboratory findings confirmed DKA. Rapid management of DKA led to stabilisation of both the maternal and fetal condition, which was reflected by the patient’s clinical, laboratory findings and by the fetal heart trace, and prevented the decision to carry out a rapid delivery that might have resulted in increased risk to both mother and fetus.

A case series, which included eight cases of pregnant diabetic women whose pregnancy was complicated by either hyperglycaemic or euglycaemic DKA, stated that treatment of DKA in pregnant women must be started immediately and that intravenous fluids, insulin and electrolyte replacement must be increased. Prevention, early recognition, immediate, hospitalisation and aggressive management remain the cornerstones of DKA management in pregnancy.¹⁴

A case report of one patient with two pregnancies complicated by DKA, demonstrates that increased awareness of DKA and a successful multidisciplinary approach changed the pregnancy outcome from intrauterine fetal death in her first pregnancy to a positive pregnancy outcome in her second pregnancy.⁷

Learning points.

• Pregnant women are more prone to diabetic ketoacidosis (DKA), particularly in their second and third trimester; vigilant monitoring of diabetes in pregnancy is of utmost importance in avoiding DKA development.

• It is important to note that absence of the typical clinical picture of DKA (nausea, vomiting, decreased level of consciousness) does not exclude the development of DKA during pregnancy.

• Serum ketones and blood sugars should be checked in any diabetic patient during periods of illness to avoid missing DKA diagnosis.

• Rapid appropriate multidisciplinary treatment of DKA improves fetal and maternal outcome and can avoid unnecessary emergency delivery.

Ethics statements

Patient consent for publication

Consent obtained directly from patient(s).