Diabetic Emergencies Case 2

Answer the questions at the bottom of the page when all data has been reviewed

History & Examination

A 22 year old woman with a history of insulin dependent diabetes mellitus (IDDM) is admitted to West Hampton Hospital generally unwell. Her IDDM had been diagnosed at the age of 15 years and despite excellent compliance and heroic efforts by both the patient and her mother, the diabetes had proven both brittle (prone to erratic variation) and difficult to control (manifesting as a persistently high HbA1c).
Over the seven years since diagnosis, she had developed multiple complications of diabetes including significant diabetic retinopathy and autonomic neuropathy. She was troubled with gastrointestinal symptoms secondary to intractable gastroparesis. Indeed, by the time of the events described here, these problems had reached a level sufficient to render her housebound.

In the year prior to this event, she had undergone repeated admissions to the Hospital with complications of her diabetes. She is well known to the medical and nursing staff. However, she remained remarkably stoical and cheerful throughout her difficulties.

On admission, she was drowsy and little history was available. She had been well a few days earlier but in the days prior to admission she had increasing episodes of nausea and vomiting. Fed up with repeated admissions, she had been reluctant to come into Hospital again.

On examination she was dehydrated. Her chest was clear. Abdominal examination was entirely normal. CNS examination showed no focal neurology.


BK bloods CSi

ST master.jpgnowcsi

CXR report: Portable AP, poor quality. No gross abnormality.


BK Obs anon



Q1. What is the likely metabolic diagnosis?

Q2. What other tests will establish the metabolic diagnosis in Q1?

Q3. What is the underlying trigger, if any, of the metabolic diagnosis in Q1?

Q4. Her paO2 is high, how do you explain this?

Q5. Can you explain her hyponatremia?

Q6. Is the total body potassium increased or decreased?

Q7. How do you explain the FBC abnormalities?

Q8. What are the key points in her management?


Q1. What is the likely metabolic diagnosis?

Metabolic diagnosis: Diabetic ketoacidosis (DKA)

Bernadette Kelly ABG AG accurate Csi

ABG analysis

She is a known type I diabetic who has been vomiting, is confused and unwell. She is markedly hyperglycemic (urinalysis shows glucose ++++). So clinically, even without seeing the blood gas, we suspect she may have a metabolic acidosis (diabetic ketoacidosis). When we analyse the ABG we note a low pH (an acidosis). The level of ‘metabolic base’ (bicarbonate) is low. This is consistent with a ‘metabolic acidosis’. Is this metabolic acidosis due to the addition of fixed acid to the extracellular fluid (ECF) or loss of bicarbonate from the ECF? The anion gap is grossly elevated, [Na+] – ([HCO3-] + [Cl-]) = 31 mEq/l (mmol/l) (normal < 18). This combined with the clinical context implies that addition of fixed acid to the ECF is the likely mechanism. In fact, there is evidence that ketones are the offending fixed acids. The high degree of ketonuria (++++) recorded on the fluid balance chart is consistent with high levels of these acids in the blood stream.

So, we have a raised anion gap metabolic acidosis. The pCO2 is low consistent with respiratory compensation. Compensation is partial (pH has not returned to normal). However, applying Winter’s formula, we see that, although partial, the compensation is adequate.

Winter’s formula tells us that the range of expected pCO2 in a metabolic acidosis is related to the observed bicarbonate concentration as follows:

Range of expected pCO2 = 1.5[bicarbonate] + 8 +/-2 = 20 – 24 mmHg

Observed pCO2 = 23 mmHg  (kPa x 7.5)

This is a simple acid-base disturbance.

Q2. What other tests will establish the metabolic diagnosis in Q1?

This condition is due to lack of insulin action and a cellular switch from carbohydrate to ketone metabolism. We need to establish the presence of high levels of ketones in the serum. We can do this by analysing the serum or urine.

Plasma ketones will be markedly elevated

Urinary ketones will be markedly elevated (see fluid balance chart)

We should also confirm the presence of hyperglycemia

Plasma Glucose will usually be markedly elevated, in this case it was 703 mg/dl (39 mmol/l)

Urinary Glucose will be strongly positive (see fluid balance chart)

Q3. What is the underlying trigger, if any, of the metabolic diagnosis in Q1?


She has a low grade pyrexia, an elevated white cell count  and her urine is positive for leukocytes and nitite. She may well have a UTI. She must have blood cultures taken and a CSU sent for culture.


Q4. Her paO2 is high, how do you explain this?

Bernadette Kelly ABG AG accurate Csi

The paO2 is elevated above the normal range. This is expected. Annoyingly, neither the time nor the level of oxygen therapy has been written on the ABG result. We assume it was taken shortly after her presentation in view of her history. Looking at her observation chart, she is on 36% (4L) oxygen at that time. She is also hyperventilating (low paCO2) in order to compensate for the metabolic acidosis (‘blowing off’ CO2). Remember, we have to interpret the paO2 in terms of the level of inspired oxygen (the FiO2) and the degree of alveolar ventilation (indicated by the paCO2). In fact, in this case when we apply the alveolar gas equation,

(Alveolar) pAO2 = FiO2(Atmospheric pressure – airway vapour pressure) – paCo2 (mmHg)/RQ

(FiO2 = 0.36, RQ = respiratory quotient = 0.8)

This yields a pAO2 of 228 mmHg

With an A-a gradient normal for her age (age/4 + 4 = 10, and adding 7.5 mmHg for every 10% rise in FiO2 = 21 mmHg), we would expect a paO2 of approximately 200 mmHg. She is close enough to this, paO2 = 202 mmHg. The paO2 on her ABG is appropriate to her level of oxygen therapy and her degree of alveolar ventilation.

Q5. Can you explain her hyponatremia? 

Her hyponatremia may be ‘appropriate’. In the absence of insulin action, glucose cannot enter cells and becomes an ‘effective osmolyte’ raising the osmolality of the extracellular fluid (ECF). This can result in a shift of water from the intracellular fluid (ICF) to the ECF with dilution of ECF sodium. If the patient is able to drink, this effect will be compounded by increased fluid intake in response to the glucose-related rise in plasma osmolality.

In contrast, in elderly patients, the thirst mechanism is often impaired. Loss of large amounts of free water in the urine due to the osmotic diuretic effect of glucose filtered at the glomerulus may not be associated with increased fluid intake. Sodium levels in the ECF may become concentrated with marked hypernatremia (see Case 1 in this course). Equally, if a patient with DKA can no longer maintain fluid intake, due to a reduced level of consciousness or the emetogenic effects of ketones, hypernatremia may develop due to continuing free water loss in the urine.

Q6. Is the total body potassium increased or decreased? 

If you said ‘increased’, don’t take up poker. The serum K+ level at presentation in a diabetic ketoacidosis (DKA) is misleading. Remember, when we measure the ‘serum’ level of an electrolyte we are, in effect, measuring it’s concentration in the entire extracellular fluid (ECF). Under normal circumstances, K+ enters large intracellular stores in skeletal muscle in conjunction with glucose. The vast majority of potassium in the body is present in intracellular stores. The co-transporter mediating the uptake of potassium into skeletal muscle is critically dependent on the presence of insulin for its action. In the absence of insulin, therefore, potassium accumulates outside cells in the ECF.

Ketones induce vomiting via stimulation of the chemoreceptor trigger zone in the brainstem. Our patient has been vomiting repeatedly for some days. Vomiting is associated with loss of large amounts of potassium in the urine. She is, therefore, likely depleted of potassium. This is disguised by the shift of potassium from intracellular stores into the ECF secondary to the loss of insulin action. As soon as we instigate treatment, insulin will drive potassium into the cells and she may develop a dangerous hypokalemia. Therefore, close monitoring of potassium levels with eventual replacement of this cation is essential during the treatment of DKA.

Q7. How do you explain the FBC abnormalities?

It is well worth knowing that hyperglycemia per se may be associated with an elevated white cell count. However, in this case, the patient has a low grade pyrexia and evidence of urosepsis on dipstick analysis of the urine (see fluid balance chart). This infection probably accounts for the raised white cell count.

Q8. What are the key points in her management? 

As in any medical emergency we first pay attention to securing the patient’s airway and circulation. After this initial assessment, we immediately obtain IV access and start high flow oxygen.

The key to managing DKA is replacing fluids (she is profoundly dehydrated) and insulin. Most hospitals will have a standard protocol of insulin infusion and fluid replacement in this situation (at least they should have). Significant hypokalemia must be avoided by regular monitoring of serum potassium levels and appropriate potassium replacement. The metabolic response to treatment may be monitored by VBG analysis.

If an underlying trigger is identified it should be treated. In this case, urosepsis was treated with IV antibiotics.

Subsequent Course


She made a satisfactory recovery in the subsequent days. Her pyrexia settled and her inflammatory markers, including her white cell count, drifted down towards normal. On the day after admission, her MSU confirmed the presence of urosepsis. The isolate was fully sensitive to the prescribed antibiotics. The organism was also isolated from her blood cultures and, therefore, IV antibiotic therapy was continued necessitating a longer admission. She was taken off her sliding scale of insulin and continued on her regular subcutaneous insulin. However, her blood glucose proved difficult to control. While an inpatient, despite extreme fatigue and ongoing severe intractable difficulties with diabetic complications, including abdominal pain, postural hypotension and nausea, she volunteered to participate in research projects. She subsequently developed an in-patient complication which can teach us a great deal. We will go through this in the Course on ABG analysis (link).