Acid-Base Physiology

Case 8 : An ill diabetic patient with vomiting and polyuria

History

A 23 year old 53kg female was admitted with persistent vomiting, polyuria and thirst. She had been ill for about 16 hours. She had been an insulin dependent diabetic for 11 years but her health was usually excellent. There was no dysuria and no evidence of chest, pelvic or skin infection. She had omitted several doses of insulin in the previous 3 days.

She was drowsy but easily roused and able to give a clear history. BP 140/80. Pulse 108/min. Resp rate 48/min, temp 37C. Lungs were clear, heart sounds were normal, and abdomen was soft. Chest wall tenderness was ascribed to frequent vomiting and dry retching. Urine testing: 3+ glucose, 'large' ketones. Serial pathology results are listed below.

She was admitted to Intensive Care. Management included oxygen by mask, normal saline, insulin infusion, antibiotics and potassium replacement. No sodium bicarbonate was given. Recovery was uneventful.

Biochemistry & Arterial Gases for Case 8

Hours since admission:

0

1

2.5

4

6.5

8

16 hrs

Na+

134

142

142

141

138

134

134

K+

7.7

6.2

4.3

4.6

4.7

4.1

4.5

Cl-

104

115

123

123

120

117

110

HCO3-

4.2

5.9

4.5

6.3

8.6

11

15

Glucose

30

22.2

10.7

6.4

8.7

8.5

5.8

Urea

7.9

6.9

5.4

4.8

3.7

2.9

1.8

Creatinine

251

239

245

218

156

100

78

Osmolarity

296

302

289

282

270

270

266

Anion gap

25.8

21.1

14.5

11.7

9.4

6

9

pH

6.99

7.03

7.12

7.22

7.3

7.34

pCO2

12.6

12.3

19.1

24

25

26

HCO3-

3.1

3.3

6.3

9.8

12.2

14

Assessment

A critical comment here is that the [K+] is very high and life-threatening. Management of this has immediate priority and treatment should not be delayed.

Firstly: Initial clinical assessment

The clinical diagnosis is obviously severe diabetic ketoacidosis. Patients with DKA may also have a lactic acidosis consequent to hypovolaemia and poor perfusion. They may also have a coexisting normal anion gap acidosis but this more commonly develops following treatment. Vomiting of acidic gastric content can also add a component of metabolic alkalosis and this can be difficult to distinguish in this setting.

Secondly: The acid-base diagnosis

Proceeding systematically:

  1. pH: Severe acidaemia so severe acidosis must be present
  2. Pattern: Low pCO2 & bicarbonate along with the acidaemia confirm a severe metabolic acidosis
  3. Clues: The high anion gap & hyperglycaemia are expected in diabetic ketoacidosis. The urinary ketones confirm the diagnosis. The creatinine is elevated but the urea is not: this can occur in ketoacidosis because the ketones interfere with the common laboratory method (Jaffe reaction) used for creatinine causing it to be measured artificially high. The decrease in the bicarbonate is larger than the increase in the anion gap so there is no evidence of any pre-existing metabolic alkalosis despite the history of vomiting.
  4. Compensation: The predicted pCO2 is 12.6 (= 1.5 x 3.1 + 8) plus or minus 2. The actual pCO2 exactly matches this! The conclusion is that respiratory compensation (Kussmaul respiration) is at its maximum level
  5. Formulation: Severe high anion gap metabolic acidosis with maximal respiratory compensation. No evidence of a respiratory acid-base disorder or of a pre-existing metabolic alkalosis.
  6. Confirmation:

Finally: The Clinical Diagnosis

This patient has diabetic ketoacidosis.

Comments

With treatment, the acidosis has changed to become a hyperchloraemic normal anion gap acidosis. This is not uncommon during therapy and patients may present with some hyperchloraemic component to their acidosis. The mechanism here is the renal loss of ketoacids and their replacement by chloride and this increases following fluid resuscitation with normal saline. This causes a delay in the correction of the acidosis because fewer ketoacids are now available to be metabolised to generate bicarbonate. Glucose levels have returned to normal much more quickly than the correction of the acidosis. Sodium bicarbonate was not administered. It will improve the appearance of the blood gases but this has no proven benefit and it can cause other problems (eg hyperkalaemia, rebound alkalosis) during the recovery period.

The initial Delta ratio of about 0.7 has fallen to about 0.1 by 2.5 hours after admission. This is the expected result as low Delta ratios are found with hyperchloraemic acidosis (see Section 3.3).