Intended for healthcare professionals

Rapid response to:

Education And Debate

Discontinuation of thioridazine in patients with learning disabilities: balancing cardiovascular toxicity with adverse consequences of changing drugs

BMJ 2002; 324 doi: (Published 22 June 2002) Cite this as: BMJ 2002;324:1519

Rapid Response:

Committee on Safety of Medicines' gave bad advice on "gradual" withdrawal

Davies et al are right to raise concerns about the discontinuation of
thioridazine in patients with learning disabilities. However, there are
reasons for questioning some of their interpretations of the problems
which have arisen in carrying out advice from the Committee on Safety of
Medicines. Thioridazine’s high anticholinergic activity may be associated
with severe withdrawal syndromes for some people. The Chief Medical
Officer refers to these as “nausea, vomiting, gastric upset, trembling,
dizziness, anxiety, agitation and insomnia, as well as transient
dyskinetic signs or the re-emergence of psychotic symptoms.” ‘Gradual’
withdrawal is advised in pursuance of their avoidance.

Unfortunately, prevalent definitions employed by prescribers refer
to periods of between one week and four weeks as “gradual withdrawal”. For
drugs which may have been taken for decades this is an abrupt and not a
gradual change. In my own observation, discontinuing Thioridazine in
people with learning disabilities after longterm use without a long taper
can have catastrophic effects [1]. Dramatic weight loss, severe
disruption of sleep patterns, pronounced almost incessant agitation and
restlessness with distressed vocalisations are features of these abrupt
withdrawals which may last for months after rapid onset coinciding with
drug discontinuation. Sovner found a similar pattern in his cases of
thioridazine withdrawal, with anxiety and insomnia prominent[2] .

Here is what a veteran pharmacist who spent decades specialising in
psychiatric drugs says about antipsychotic withdrawal, "All the problems,
like benzodiazepines, occur when withdrawal is attempted in a few days or
few weeks"[3] .

There are a number of reasons for suspecting that a proportion of the
patients who had problematic withdrawal from thioridazine in the study by
Davies et al were undergoing withdrawal reactions, with or without
eventual re-emergence of a baseline pathology. One reason is the speed of
onset of disturbed behaviour, one is the appearance of psychotic like
symptoms in people who did not initially have a diagnosis of psychosis,
one is the cases where switching to another antipsychotic did not curtail
the problem behaviours, one is the case of NMS , and finally it would be
surprising if nobody experienced a withdrawal syndrome when after only 37
weeks of antipsychotic treatment even monkeys show disturbed behaviours
after discontinuation for as long as seven weeks before beginning to
return to baseline [4].
This view is also supported by the findings of May et al that “of 23
severely and profoundly mentally retarded adult male patients undergoing
slow "diagnostic" neuroleptic taper, it was determined that at least 60%
could eventually be managed without psychoactive medication. However, many
of these demonstrated a remarkably long, but nonetheless transient, period
of worsening” (my emphasis)[5].

Davies et al remark that, “Although the adverse consequences for the
seven patients who stopped taking thioridazine could be viewed as being
attributable to poor management of change, the clinicians involved felt
they had little option but to follow the Committee on Safety of Medicines'
advice”. A significant defect in that advice was its characterisation of
“gradual withdrawal”. Strangely, when that advice appears as from the
Committee’s Chair, Professor Alan Breckenridge, (online) it is thus:
“When discontinuing thioridazine a gradual reduction in dosage over
several weeksone to two weeks is recommended “[sic][6] . Did someone
alter the original statement and replace “several weeks” with “one to
two”? If so, they may be responsible for widespread distress, often
leading to “prescribing cascades”[7] . An ideally slow taper would be no
faster than 10% every two to three months, especially after chronic use of
thioridazine or other neuroleptics.


1.Murray DKC 1999, “Potions Pills and Human Rights”, Good Autism
Practice April 1999, pp.1-13.

2.Sovner,R. 1995, “Thioridazine withdrawal induced behavioral
deterioration treated with clonidine: Two case reports”. Mental
Retardation, 33, 221-225.

3.Stuart Baker

4. Amore M, Zazzeri N. 1995 Neuroleptic malignant syndrome after
neuroleptic discontinuation. Prog Neuropsychopharmacol Biol Psychiatry

5. May P, London EB, Zimmerman T, Thompson R, Mento T, Spreat S. 1995
“A study of the clinical outcome of patients with profound mental
retardation gradually withdrawn from chronic neuroleptic medication.”
Ann Clin Psychiatry. Dec;7(4):155-60.


7. See Webb, OJ. 1996, Medication use patterns in IHC community
homes. NZ Fam Physician; 23: 4-6. and Murray DKC 1999, “Potions Pills and
Human Rights”, Good Autism Practice April.

Competing interests: No competing interests

30 June 2002
Dinah KC Murray
Distance Education Dept /WebAutism Birmingham University B15 2TT