Acid-Base Physiology

Case 3 : A weak old lady

Clinical details

An elderly woman from a nursing home was transferred to hospital because of profound weakness and areflexia. Her oral intake had been poor for a few days. Current medication was a sleeping tablet which was administered by nursing staff as needed.

Admission biochemistry (in mmol/l): Na+ 145, K+ 1.9, Cl- 86, bicarbonate 45, anion gap 14 and a spot urine chloride 74 mmols/l.

Arterial Blood Gases

pH 7.58

pCO2 49 mmHg

pO2 - not given

HCO3 44.4 mmol/l


First: Initial clinical assessment

The history does not indicate any particular acid-base disorder. There is no respiratory distress. There is suggestion of a drug-related disorder. The areflexia could indicate a hypokalaemia. There is no evidence of diarrhoea, vomiting or polyuria. The poor oral intake suggests the possibility of dehydration and maybe lactic acidosis due to poor perfusion but there is no indication of increased respiratory effort (as expected ie Kussmaul respirations). Simply put, the history is not indicative of a particular acid-base disorder.

The biochemistry results tell a more interesting acid-base story. The high bicarbonate, hypochloraemia & severe hypokalaemia suggests a significant metabolic alkalosis.

Second: The acid-base diagnosis

  1. pH: Alkalaemia so an alkalosis is present
  2. Pattern: A high pCO2 & a high bicarbonate occurs with either a respiratory acidosis or a metabolic alkalosis. In this case then, a metabolic alkalosis is present
  3. Clues: The hypokalaemia & hypochloraemia are typical findings with a metabolic alkalosis. The anion gap is 14 so there is no evidence of a coexisting high anion gap acidosis. The urinary chloride level is high and points towards possible aetiologies for the alkalosis (see below)
  4. Compensation: For a metabolic alkalosis, rule 6 is used to calculate an "expected pCO2". Quick mental arithmetic gives a value of about 51mmHg which is very close to the "actual" (or measured) value of 49mmHg. Compensation is appropriate and there is no indication of a respiratory acid-base disorder.
  5. Formulation: Severe metabolic alkalosis with appropriate respiratory compensation. The [K+] is low enough to be life-threatening and emergency management is necessary
  6. Confirmation: There are no further tests needed to confirm the acid-base diagnosis

Finally: The Clinical Diagnosis

Severe metabolic alkalosis with life-threatening hypokalaemia. The cause is not yet determined. The high urinary chloride suggests a cause in the volume-resistant group (ie the 'chloride' resistant group).


The severe hypokalaemia is the cause for the weakness and requires urgent therapy. Intravenous K+ replacement is urgently indicated. Hypokalaemia can cause serious arrhythmias. It can also cause rhabdomyolysis which can result in hyperkalaemia (& malignant arrhythmias) and renal failure.

About 90% of cases of metabolic alkalosis are due to diuretic therapy or loss of gastric secretions (vomiting or nasogastric suction)

These two causes are not present in this case so other causes must be considered.

Consider the 2 major groups of causes of metabolic alkalosis: these groups are differentiated by measurement of the urinary chloride level.

The 2 major divisions of Metabolic Alkalosis

Chloride responsive’ group (urine chloride < 10 mmol/l)

Key Feature: Chloride Deficiency

Typical causes in the low urine chloride group are:

  • Loss of gastric juice (eg vomiting esp if pyloric obstruction, or nasogastric suction)
  • Diuretic therapy

‘Chloride resistant’ group (urine chloride > 20 mmol/l)

Key Feature: Excess Steroids or Current Diuretic Use

Typical causes:

  • Excess adrenocortical activity (eg primary aldosteronism, Bartter’s syndrome, Cushing’s syndrome, other causes of excess adrenocortical activity)
  • Current diuretic therapy
  • ‘Idiopathic’ group

The urine chloride at 74 mmol/l is very high in this patient and this suggests a diagnosis in the second group.

Biphasic action of diuretics: Diuretics cause a high urine chloride while they are causing a diuresis, but a low urine chloride when measured after their pharmacological action has passed. As diuretic use is common, this relationship to the timing of a dose should be known to assist in interpretation of the urine chloride result. The basic problem is that recent diuretic use by increasing urinary chloride is interfering with the usefulness of spot urine chloride measurement when attempting to sort out the cause of a metabolic alkalosis.


This patient recovered with correction of the potassium deficit. Further investigation failed to find a cause for the metabolic alkalosis. Hyperfunctioning of the adrenal cortex was not present. So excluding diuretic therapy, loss of gastric juice and excess adrenocortical activity leaves a diagnosis in the ‘idiopathic’ group.

The major aetiologic factor in this group seems to be the presence of severe potassium deficiency with plasma [K+] < 2 mmol/l (as in this case). Saline solutions (ie chloride replacement) alone do not correct the alkalosis but adding potassium replacement invariably does in these patients. Potassium depletion usually does not cause a severe alkalosis unless chloride depletion is also present.

Studies have reported high mortality rates associated with severe alkalosis but it has not been established that the relationship is directly causal.