Dialysis with bipolar affective disorder: experience of competing treatments

A 64-year-old male – James Jenkins – was referred to the liaison psychiatry team due to challenges in managing his bipolar affective disorder, controlled successfully with sodium valproate, and renal failure concurrently. The patient had been receiving haemodialysis three times per week due to renal failure, and had a longstanding diagnosis of bipolar affective disorder controlled successfully with sodium valproate. Valproate is used widely for treatment of mania, either as valproic acid or sodium valproate, although only valproate semisodium is licensed for this.¹ The accepted wisdom regarding the effect of haemodialysis on blood sodium valproate levels is that there should be no effect on levels, and therefore dosing should be able to continue as before.

There is also limited literature regarding the effect of haemodialysis on other psychotropic medication. Again, the accepted wisdom, dating back to the 1980s, is that antipsychotic medications are not removed by dialysis,⁵ but this was challenged in a review published in 2021, which found published cases where pre-existing psychiatric disorders were exacerbated in patients undergoing haemodialysis.⁶ In addition, chronic kidney disease is known to reduce cytochrome P450-mediated drug metabolism, which is the most important metabolic pathway in many psychotropic drugs. This has been found to be reversed completely by renal transplant and partially by haemodialysis,⁷ suggesting strongly that haemodialysis has the potential to change the plasma levels of many psychotropic medications.

Presentation and past medical/psychiatric history

James was a retired electrician, and much-loved husband and father. He first became unwell with bipolar affective disorder in the mid 1970s, in his late teens, resulting in an admission to hospital. He was started on lithium, which normalised his mood, and he continued to take this as maintenance therapy after discharge.

James was first told that he had ‘kidney trouble’ in 1992. It was explained to him that this was likely a consequence of the lithium and he switched to sodium valproate as a mood stabiliser, which also proved effective at maintaining his bipolar affective disorder. About eight years later he was made aware that his kidneys were continuing to deteriorate. This continued over the following years, and in July 2020 he started dialysis three times per week. He was able to walk and was fully independent at this point (summer 2020), before this next extreme episode.

James had very little recollection of the next few weeks; his mental state deteriorated following starting dialysis, and he became very manic, requiring admission to hospital under Section. He was started on the mood-stabilising antipsychotic quetiapine in addition to the sodium valproate, the quetiapine was titrated up slowly due to concerns about his renal failure. However, his mental state continued to deteriorate and it became increasingly difficult to effect transfer to the dialysis unit due to James’ elevated mood. He became more confused and was transferred to the acute hospital, where he was cared for by the renal physicians, still under Section 3 of the Mental Health Act (MHA), for which the liaison team consultant became the responsible clinician.

As his symptoms of mania worsened, James refused all interventions from the acute trust staff, including dialysis, medication, blood testing and observations. Over the course of a week he became bed bound, and had been without dialysis for over a week. Due to thought disorder and delusional beliefs, he lacked capacity to make decisions about treatment for his mental and physical disorders. The liaison psychiatrist recommended a course of Clopixol Accuphase (a medium-term acting formulation of zuclopenthixol, given as an injection) and intramuscular lorazepam, which allowed dialysis to be given and James to be transferred back to the mental health unit.

The legalities of the treatment for this disorder are complex; the MHA allows treatment of a mental disorder but not physical disorder unless caused by or causing the mental disorder. While there may be a range of opinion as to whether this case is covered by that particular piece of the code of practice, it was the opinion of the liaison psychiatrist at the time (September 2020) that either the MHA or MCA would cover treatment given in order to sedate James so that he could have dialysis. In fact it was the level of sedation and intervention required to achieve this that caused such extreme anxiety in the acute hospital staff, requiring reassurance that as clinicians our prime directive is to protect the best interests of an incapacitated patient.

Treatment of dialysis in the context of mental disorder does have precedent; in a 2020 case highly similar to this one,⁹ Mrs Justice Lieven stated that ‘there is a very real prospect that if X was not mentally ill he would self-care in a way that would have not led to the need for dialysis. Further, that X is refusing dialysis is very obviously a manifestation of his mental disorder’. So dialysis is therefore treatment within the scope of Section 63 MHA 1983. While well, James had consented to dialysis and to taking his antipsychotic and mood stabiliser medication, so sedating him in order to carry out dialysis was deemed to be what he would choose to do if not mentally unwell.

Clearly something had changed when the dialysis started; as psychiatrists we reflect on the impact of life events in precipitating acute episodes of mental disorder, but the sudden change coinciding with the start of dialysis treatment was highly suggestive of an effect on the levels of mood stabiliser in the patient's body. Advice from the renal team was that sodium valproate is not excreted by the kidney and therefore, as the dialysis is mimicking the kidney, the levels should not be affected by it. However, the mental health team continued to notice that James’ mood deteriorated regularly following dialysis. Was this simply the response to the continuing stress of attending the acute trust at a time when anxiety levels were high due to COVID-19, or was there a more physical reason?

In the light of our suspicions, we tested his valproate levels pre- and post-dialysis, and found a significant drop in valproate levels (the therapeutic window for treatment for bipolar affective disorder is 50–100mg/L). In consultation with the ward pharmacist, we increased the regular twice daily dosage of valproate but also added a large ‘reloading dose’ immediately after each episode of dialysis in order to counteract the effect of the dialysis on the valproate levels. Aripiprazole was started concurrently to treat and prevent recurrence of the ongoing symptoms of mania.

James’ mood responded well to the increases in sodium valproate dosage, but there followed a period of several months of physiotherapy and rehabilitation to enable him to be able to walk enough to return home to his family. It was suggested by the renal physicians that he increase the dialysis to four days per week, but he described that as ‘taking up my whole life’ as well as being afraid of the effect on his bipolar affective treatment of an extra day of dialysis.

Reflection and discussion

This case demonstrates a clear correlation between starting haemodialysis and the deterioration in a bipolar affective disorder, which had previously been well controlled by a mood-stabiliser drug. This directly challenges the perceived and accepted views about the effects of haemodialysis on these drugs. More literature is becoming available to support this challenge.

Titration of the treatment was frustrating for both the treating team and for James and his family. It is impossible to know other than in very broad terms what the absorption rate is of sodium valproate from a patient's body, and both this and the rate at which it is metabolised will change; this is dependent on many variables, including what they have eaten, other medications they are taking, their current hydration level and current body fat level.¹⁰ The only thing known for definite was the dose per kg being administered and the valproate levels we were able to obtain from James’ blood testing. On reflection, this level of uncertainty is something we encounter in many aspects of psychiatric treatment, and this added to the fear experienced by James when trying to plan his future.

At the time of writing James had been out of hospital for over a year and had been able to slowly reduce his medication, with close oversight from his community mental health team and input from physiotherapy and occupational therapy for his mobility. His experiences have made him keen to share his story and to challenge the accepted wisdom regarding the use of psychotropic medications with haemodialysis.

Declaration of interests

Neither Mr Jenkins nor Dr Abbey have any conflict of interest.