Surveillance of hyponatremia in psychogenic polydipsia

To the Editor

Psychogenic polydipsia is an important cause of hyponatremia in psychiatric inpatients. However, although there are protocols for the surveillance of symptomatic hyponatremia, evaluating asymptomatic patients at risk of hyponatremia is less well-described. We present this case to discuss the complex assessment of psychogenic polydipsia.

John is a 33-year-old man with schizophrenia and borderline intellectual impairment. He is normally managed with 550mg clozapine, has co-morbid asthma, impaired glucose tolerance, chronic renal failure, and has an ileal conduit for the management of obstructive uropathy.

John was initially admitted for management of a subdural haemorrhage, caused by a seizure with head-strike secondary to hyponatremia. After several similar admissions, John was transferred to the Psychiatry Inpatient Unit for management of his ongoing psychogenic polydipsia.

John continued to drink up to twelve litres of water per day, and was resistant to suggestions that he reduce his fluid intake. John noted that he drank to reduce his anxiety and boredom, and his symptoms were assessed to be associated with low-level anxiety in the context of his other psychiatric conditions. As clozapine and enalapril have some efficacy in the management of psychogenic polydipsia (Dundas et al., 2007), John remained on clozapine and was commenced on enalapril 20mg, later titrated to 10mg due to hyperkalemia. John was also encouraged to attend structured ward activities in order to distract him from his need to drink.

The frequency of electrolyte monitoring was an ongoing management dilemma. Accepted practice in the surveillance of asymptomatic patients at risk of hyponatremia is to monitor them for symptoms of hyponatremia, such as confusion and lethargy. Routine electrolyte surveillance is infrequent, such as the recommended weekly surveillance of patients at risk of carbamazepine-induced hyponatremia (Palmer et al., 2003). However, there are no recommendations available for the surveillance of psychogenic polydipsia, which poses unique clinical challenges compared to other causes of hyponatremia.

Firstly, hyponatremia may be difficult to clinically identify in the presence of co-morbid psychiatric disease. John’s unwillingness to disclose his fluid intake, alongside a baseline low level of confusion, rendered it difficult to identify early symptoms of hyponatremia.

Secondly, using symptoms of hyponatremia as a warning of impending seizure activity is of limited value in patients on clozapine, as clozapine may lower the seizure threshold (Kumlien and Lundberg, 2010). This may have been demonstrated by John, who developed a seizure while having a sodium level of 118 mmol/L, which is within the safety bracket for hyponatremia-induced seizures in some patients (Dundas et al., 2007).

Thirdly, hyponatremia in patients with psychogenic polydipsia exhibits diurnal variation. Patients drink water upon waking, leading to hyponatremia, and subsequent diuresis normalises sodium levels by the following morning. This could indicate that patients with psychogenic polydipsia warrant bi-daily electrolyte monitoring, although the practical implications of bi-daily venepuncture may limit the utility of this observation.

Considering these factors, John’s final management plan involved twice-weekly electrolyte monitoring with additional tests when clinically indicated, which was reduced to weekly monitoring in preparation for discharge. John did not have any further seizures, and was discharged to community care.