Abstract
Case study
AB is a 34-year-old lady, who was admitted to hospital by her GP with an episode of acute vertigo complicated by severe vomiting. Blood investigations revealed severe hypokalaemia with a potassium of 1.7 mmol/l (3.4–5.6 mmol/l) and metabolic alkalosis with a venous bicarbonate of greater than 40 mmol/l (22–32 mmol/l). Blood pressure was normal at 130/85. A working diagnosis was made of acute labyrinthitis with vomiting-related hypokalaemia and AB was treated with an intravenous potassium infusion. At the time of discharge, her potassium levels were at the lower end of normal at 3.7 mmol/l.
Subsequently, AB had persistent hypokalaemia with values between 2.2 and 2.7 mmol/l despite potassium supplementation and required several further admissions on account of this. A detailed history and investigations via the local endocrine clinic did not reveal any obvious cause, although they did reveal high urinary potassium of 91.1 mmol/l and a persistent metabolic alkalosis.
Clinical features and causes of hypokalaemia
Symptoms of hypokalaemia generally manifest at values less than 3 mmol/l. Serum potassium levels of less than 2.5 mmol/l can have serious cardiac and neurological consequences. A variety of atrial and ventricular arrhythmias can occur at such levels. Respiratory muscle weakness, ileus associated with nausea and vomiting, muscle cramps and tetany are features of severe hypokalaemia. Unexplained hypokalaemia in a normotensive patient has a few differential diagnoses (Table 1).
Differential diagnosis of unexplained hypokalaemia in normotensive patients
The most important parameters that aid in diagnosis in this scenario are acid-base balance (measured by checking serum bicarbonate) and urinary potassium excretion. Other causes of hypokalaemia that may be relevant to general practice include cell shift causes, e.g. insulin, beta-agonists administration and occasionally hypomagnesaemia. AB had a persistent hypokalaemic metabolic alkalosis of seemingly recent onset with a high urinary potassium excretion, raising the suspicion of diuretic abuse. Vomiting may have compounded the hypokalaemia during her initial admission.
Patients with bulimia have a persistent over-concern with body shape and weight. This is characterized by recurrent episodes of binge eating followed by recurrent inappropriate compensatory behaviour such as self-induced vomiting, diuretic and laxative abuse, fasting or excessive exercising. These disorders are difficult to identify as patients, most frequently women, avoid seeking medical attention and try to hide their condition. The physical consequences of such a delay, as in this case, can be potentially life threatening (Whittal et al., 1999). In cases of severe refractory unexplained hypokalaemic alkalosis, it is important to have a high index of suspicion for bulimia and concomitant diuretic abuse. Requesting a urine diuretic assay will usually clinch the diagnosis leading to appropriate referral for specialist psychological treatment. This is rewarding, as appropriate referral for cognitive behavioural and antidepressant therapy in bulimia has been shown to be of benefit (NICE, 2004; Treasure et al., 2010)
The ethics of testing
A diagnosis of diuretic abuse was suspected but AB strongly denied this. Could a urine diuretic assay be undertaken without the patient's consent? The General Medical Council has published guidance documents on patient's consent (GMC, 2008). It is for the patient, not the doctor, to determine what is in the patient's best interests. The doctor can recommend a course of action but must not put pressure on the patient to accept the advice. Importantly, a test done without consent would be unlikely to improve patient outcome (if AB was indeed taking diuretics), as she might not accept findings of diuretic abuse or simply disengage with the investigating team. Furthermore, she might be able to bring a case against the investigating clinician in negligence claiming distress caused by a non-consensual test and possibly in confidentiality for sharing of personal information (with laboratory staff) without her consent.
Therefore, AB was advised about the reasons for advising a urinary diuretic assay and of the consequences of not taking the test. She reluctantly agreed to testing. A urine diuretic screen was requested that was reported positive for furosemide.
AB was confronted with this result during a subsequent review. She admitted to buying furosemide tablets through the Internet to manage problems with her ‘body image’. She was referred to a clinical psychologist who made a formal diagnosis of bulimia nervosa and offered appropriate treatment with cognitive behavioural therapy.
Footnotes
Acknowledgement
We would like to thank Dr Wing May Kong, Vertical Theme Head Ethics, Leadership, Professionalism and Management, Imperial College, London, for her help with the ethical perspectives of this case.