DKA and HHS — Clinical Case

DCE Long Case

Patient brief (provided to trainee)

Patient: Mr Liam Chen, 19 years old, university student. [1]

Presenting complaint: Two days of vomiting, abdominal pain, and increasing drowsiness. His mother found him difficult to rouse this morning. [1]

Past history: Type 1 diabetes diagnosed aged 9. No other significant medical history. Has had two previous admissions with DKA in the last 12 months, both related to stopping insulin. [1]

Current medications: Insulin glargine 24 units at night, insulin aspart 8 units with meals (total daily dose 48 units). He takes no other medications. [1]

Social history: Lives with his mother, second-year university student. Reports significant study stress. His mother reports he has been missing insulin doses intermittently over the last month. [1]

Examination findings (trainee elicits):

• GCS 13 (E3 V4 M6), drowsy but rousable, answering slowly
• RR 28, deep sighing Kussmaul respirations, SpO2 97% on room air
• HR 118, BP 96/60, temperature 37.9 degrees Celsius
• Dry mucous membranes, reduced skin turgor
• Soft abdomen with epigastric tenderness, no guarding
• Fruity (ketotic) odour to breath
• Chest: reduced air entry and crackles at the left base [1]

Investigations:

• Capillary glucose 31 mmol/L
• Venous blood gas: pH 6.98, PaCO2 14 mmHg, bicarbonate 6 mmol/L
• Beta-hydroxybutyrate 7.8 mmol/L
• Sodium 130, potassium 5.8, chloride 92, urea 12, creatinine 110
• HbA1c 94 mmol/mol (reference below 42)
• Urine ketones 3+
• Chest X-ray: left lower lobe consolidation
• ECG: sinus tachycardia, normal [1]

Candidate's structured presentation (model)

Opening statement (SASPOP): [1]

"Mr Chen is a 19-year-old university student with an 10-year history of type 1 diabetes, who presents with Symptoms of two days of vomiting, abdominal pain and increasing drowsiness. He stopped his insulin two days ago because he was not eating. On examination he is drowsy with a GCS of 13, has Kussmaul respiration, is tachycardic and hypotensive with dry mucous membranes, and has a fruity breath odour and a left lower lobe pneumonia. His blood gas shows a pH of 6.98 with a bicarbonate of 6, a beta-hydroxybutyrate of 7.8, and an anion gap of 32. His glucose is 31 and his potassium is 5.8." [1]

Problem list: [1]

1. Severe diabetic ketoacidosis (pH below 7.0, bicarbonate below 10) — the immediate life threat. Precipitated by insulin omission, with a lobar pneumonia as an additional trigger.
2. Lobar pneumonia — a precipitant and contributor to the counter-regulatory hormone surge. Needs early antibiotic treatment.
3. Severe dehydration with acute kidney injury (urea 12, creatinine 110).
4. High risk of cerebral oedema — age 19, severe acidosis (low PaCO2 of 14), high urea (12). These are exactly the Glaser risk factors.
5. Poorly controlled type 1 diabetes (HbA1c 94) with recurrent DKA from insulin omission — a pattern suggesting psychosocial distress or an eating disorder that needs multidisciplinary input. [1]

Management plan: [1]

1. Resuscitate and admit to ICU (grade: severe). ABCDE, oxygen, two large-bore cannulae, continuous cardiac monitoring, hourly neurological observations. [1]

2. Fluid resuscitation — 0.9% saline 1 litre in the first hour, then 1 litre over 2 hours, then 1 litre over 2 hours, reassessing at each stage. The corrected sodium is 130 + 0.4 x 25.5 = about 140, so 0.9% saline is appropriate. [1]

3. Fixed-rate intravenous insulin infusion at 0.1 units/kg/hour — for a 70 kg patient, 7 units/hour of Actrapid (50 units in 50 mL saline). Start after the first litre of fluid. No loading bolus. Target a fall in ketones of at least 0.5 mmol/L/hour. [1]

4. Potassium — current level 5.8, so no potassium in the first bag. Check hourly. Once it drops below 5.5, switch to 0.9% saline with 40 mmol/L potassium. Target 4-5.5. [1]

5. Dextrose switch at glucose below 14 — 10% dextrose at 125 mL/hour, continue the FRIII unchanged. Consider reducing the FRIII to 0.05 units/kg/hour if glucose falls faster than 3 mmol/L/hour. [1]

6. Antibiotics for the pneumonia — blood cultures first, then empiric therapy (e.g., benzylpenicillin and azithromycin orally, or ceftriaxone intravenously if he cannot tolerate oral) within one hour. [1]

7. Cerebral oedema vigil — hourly neurological observations. If headache, vomiting, drowsiness or bradycardia develop, reduce fluids, give mannitol 0.5-1 g/kg or 3% saline, move to ICU. No bicarbonate (Glaser risk factor). [1]

8. Resolution and transition — when ketones below 0.6, pH above 7.3, bicarbonate above 18, give subcutaneous insulin (restart his usual regimen), wait 30-60 minutes, ensure he has eaten, then stop the FRIII. [1]

9. Long-term — diabetes team and mental health review to address the recurrent DKA and insulin omission. Sick-day rule education. Consider diabetes-related eating disorder screening. [1]

Examiner discussion questions

Q: "How did you calculate the anion gap, and what does it tell you?" [1]

"The anion gap is sodium minus chloride minus bicarbonate: 130 minus 92 minus 6 equals 32. The normal range is 8 to 12, so this is a markedly elevated high-anion-gap metabolic acidosis. In the context of a diabetic patient with ketones, the unmeasured anions are the ketoacids — acetoacetate and beta-hydroxybutyrate. The anion gap confirms that this is a ketoacidosis, not a lactic acidosis or a renal tubular acidosis. I would also note that I do not need to correct the anion gap for albumin here because his albumin is not given, but in a critically ill patient I would correct it: corrected AG equals observed AG plus 2.5 times 4.0 minus albumin in grams per decilitre." [1]

Q: "His potassium is 5.8. Is that hyperkalaemia you need to treat?" [1]

"No — this is the expected initial picture in DKA. The total-body potassium is severely depleted (by 3 to 5 mmol per kilogram) because the osmotic diuresis has been washing it out, but the acidosis has shifted potassium out of cells and the lack of insulin has prevented reuptake, so the serum level reads high. This is a false reassurance. The moment I give insulin, the potassium will shift back into cells and the serum level will fall rapidly. I will not treat this as hyperkalaemia — I will not give calcium gluconate or insulin-dextrose specifically to shift it, because that is the treatment's job. I will start the DKA protocol and watch the potassium fall, replacing it as it does. The danger period is hours 2 to 6, when the level can plummet. I check hourly." [1]

Q: "Why does he have a lobar pneumonia, and does it matter?" [1]

"The pneumonia is likely the additional trigger on top of the insulin omission — an intercurrent infection raises the counter-regulatory hormones (glucagon, catecholamines, cortisol) and increases insulin resistance, which can tip a vulnerable patient into DKA. It may also be that the DKA itself predisposed to infection through impaired immune function. It matters enormously because the DKA will not fully resolve until the infection is treated — the infection is the engine driving the counter-regulatory surge. I will start antibiotics within the first hour, after blood cultures." [1]

Q: "What is the single most important principle of the insulin infusion?" [1]

"That the insulin exists to suppress ketogenesis, not to normalise glucose. The metabolic target is a falling ketone level and a rising bicarbonate, not a falling glucose. The glucose is managed with dextrose once it drops below 14. If I stopped the insulin when the glucose normalised, the ketones would still be circulating and the DKA would recur within hours. The insulin runs until the ketones are below 0.6." [1]

Q: "He is 19. Talk me through cerebral oedema." [1]

"Cerebral oedema complicates about 0.5 to 1 per cent of paediatric and young adult DKAs, with a mortality of 20 to 40 per cent. The Glaser NEJM 2001 study identified three risk factors: a low initial PaCO2 (reflecting severe acidosis and deep compensation — his is 14, very low), a high initial urea (reflecting severe dehydration — his is 12), and bicarbonate administration (the only modifiable one). The signs appear 4 to 12 hours into treatment: headache, vomiting, drowsiness, bradycardia, rising blood pressure, pupillary changes, seizures. The moment I suspect it, I reduce the fluids, give mannitol 0.5 to 1 gram per kilogram or hypertonic 3 per cent saline at 2 to 5 mL per kilogram, and move him to ICU for neuroprotection and possible intubation. I arrange a CT head after stabilisation. I will not give bicarbonate at any point in his treatment, because it is the modifiable risk factor I can control." [1]

Q: "How will you prevent a recurrence?" [1]

"Recurrent DKA from insulin omission carries a high long-term mortality and is often a marker of psychosocial distress, an eating disorder (diabulimia — omitting insulin to lose weight), depression, or chaotic circumstances. I will involve the diabetes team, a diabetes educator, a psychologist or psychiatrist, and a social worker. I will explore the reasons for omission — he reports study stress, and his mother describes intermittent omission over a month. I will teach the sick-day rule clearly and in writing: never stop insulin during illness — you often need more. I will ensure he has a ketone meter and knows to check ketones when unwell and to seek help if ketones are above 1.5 or he cannot keep fluids down. I will arrange close outpatient follow-up with the diabetes team within a week of discharge." [1]

Q: "What if he had been on an SGLT2 inhibitor instead of having type 1 diabetes?" [1]

"Then I would consider euglycaemic DKA — the SGLT2 inhibitor causes glucosuria that lowers glucose and suppresses insulin while raising glucagon, driving ketogenesis with a near-normal glucose. The clue would be a glucose of, say, 9 or 10 with a pH of 7.1 and ketones of 5. The management is the same DKA protocol, but with dextrose started from the outset because the glucose is already below 14. The key teaching point is the sick-day rule: stop the SGLT2 inhibitor during any acute illness with vomiting, diarrhoea, or reduced intake." [1]

DCE Short Case — Coma Assessment and Hyperglycaemic Emergency

Instruction

"This 72-year-old woman with type 2 diabetes is drowsy. Assess her and discuss your findings and management. You have 8 minutes." [1]

Systematic assessment routine

1. Airway and breathing — is she protecting her airway? What is the respiratory rate and pattern? (HHS: normal breathing, no Kussmaul. DKA: Kussmaul.)
2. Circulation — pulse, blood pressure, perfusion, signs of dehydration (dry mucous membranes, reduced skin turgor).
3. Glucose — check capillary glucose immediately.
4. Neurological — GCS, pupils, focal signs. (HHS: depressed GCS proportional to osmolality, possible focal signs mimicking stroke. DKA: often alert or mildly confused.)
5. Breath — ketotic (fruity) odour? (DKA: yes. HHS: no.)
6. Precipitant clues — fever (infection), chest pain (infarction), signs of a source. [1]

Key findings

• GCS 10 (E3 V3 M4), drowsy
• RR 16, normal pattern, SpO2 95%
• HR 96, BP 110/70, dry mucous membranes, reduced skin turgor
• No ketotic breath odour
• Capillary glucose 48 mmol/L
• No focal neurological signs [1]

Presentation to the examiner

"This woman has a depressed level of consciousness with profound dehydration and a capillary glucose of 48. Her breathing is normal in rate and depth — there is no Kussmaul respiration — and there is no ketotic breath odour. This pattern is most consistent with a hyperosmolar hyperglycaemic state rather than DKA. I would confirm with a venous gas, beta-hydroxybutyrate, electrolytes and a calculated osmolality, and I expect to find a pH above 7.3, a bicarbonate above 15, ketones below 3, and an osmolality above 320. My immediate priority is fluid resuscitation with 0.9% saline, low-dose insulin, prophylactic anticoagulation, and an urgent search for the precipitant — most commonly infection or a silent myocardial infarction in an elderly type 2 diabetic." [1]

Discussion

Examiner: "What bedside feature distinguishes HHS from DKA?" [1]

"The breathing. A patient in DKA has Kussmaul respirations — deep, sighing, rapid — as the body compensates for the metabolic acidosis by blowing off CO2, and often a fruity breath odour from exhaled acetone. A patient in HHS has a normal breathing pattern because the pH is preserved and there is no metabolic acidosis to compensate for. The sensorium also differs: the DKA patient is often alert, while the HHS patient is depressed in proportion to the osmolality." [1]

Examiner: "Why is anticoagulation routine in HHS but not always emphasised in DKA?" [1]

"Because the severe dehydration and hyperviscosity in HHS — the osmolality is very high, the patient is profoundly volume-depleted — create a much higher risk of venous and arterial thrombosis. Deep vein thrombosis, pulmonary embolism, myocardial infarction, and stroke are among the leading causes of death in HHS. DKA causes dehydration too, but the patients are usually younger and the risk is lower. Unless there is a contraindication, I give prophylactic low-molecular-weight heparin to every HHS patient." [1]

Examiner: "How rapidly should you correct her glucose and osmolality?" [1]

"The target is a glucose fall of about 3 mmol per litre per hour and an osmolality fall of 3 to 8 mOsm per kilogram per hour. Too rapid a correction risks cerebral oedema from osmotic shifts — water moves into the brain faster than the brain can equilibrate. I monitor the glucose hourly and the osmolality every 2 to 4 hours, and I adjust the fluid rate to stay within the target. In the elderly, I also monitor for fluid overload — a rising oxygen requirement or new crackles mean I need to slow down." [1]