# 29 - First episode of psychosis

# First episode of psychosis

28
The Maudsley® Prescribing Guidelines in Psychiatry
CHAPTER 1
Antipsychotic prophylaxis
First episode of psychosis
Antipsychotics provide effective protection against relapse, at least in the short to 
medium term,1 and the introduction of antipsychotics in the 1950s seems to have 
improved outcomes overall.2 A meta-­analysis of placebo-­controlled trials found that 
26% of patients with first-­episode schizophrenia on maintenance antipsychotic relapsed 
after 6–12 months compared with 61% on placebo.3 Although the current consensus is 
that antipsychotics should be prescribed for 1–2 years after a first episode of schizophrenia,4,5 one study6 found that withdrawing antipsychotic treatment in line with this 
view led to a relapse rate of almost 80% after 1 year medication-­free and 98% after 2 
years. A 2019 Swedish population study revealed that the longer the treatment with 
antipsychotics, the lower the risk of hospitalisation (e.g. those with 5 years’ treatment 
had half the hospitalisation rate of those treated for less than 6 months).7
Other studies in first-episode schizophrenia confirmed that only a small minority of 
patients who discontinue remain well 1–2 years later8–11 (e.g. a small study found 94% 
of patients with first-­episode schizophrenia relapsed within 2 years of stopping risperidone long-­acting injection; 97% at three years).12 A 2018 meta-­analysis of eight RCTs 
was rather more optimistic and found relapse rates averaged 35% (treated) and 61% 
(discontinued) at 18–24 months.13
A 5-­year follow-­up of a 2-­year RCT, during which patients either received maintenance antipsychotic treatment or had their antipsychotic dose reduced or discontinued 
completely, found that while there was a clear advantage for maintenance treatment 
with respect to reducing short-­term relapse this advantage was lost in the medium term. 
Further, the dose-­reduction/discontinuation group were receiving lower doses of anti­
psychotic drugs at follow-­up and had better functional outcomes.14 There are numerous 
interpretations of these outcomes but the most that can be concluded is that dose reduction is a possible option in first-episode psychosis. The study has been heavily criticised15 and there are certainly other studies showing disastrous outcomes from 
antipsychotic discontinuation,16 albeit over shorter periods with fewer patients. 
Nonetheless, some patients with first-­episode psychosis will not need long-­term anti­
psychotics to stay well – figures as high as 18–30% have been put forward.17
There are no reliable patient factors linked to outcome following discontinuation of 
antipsychotics in patients with first-­episode psychosis (other than cannabis use)18 and 
there remains more evidence in favour of continuing antipsychotics than for stopping 
them.19 There are indications that very prolonged discontinuation regimens using 
hyperbolic tapering (see section on stopping antipsychotics in this chapter) may offer 
the best chance of successfully withdrawing from antipsychotic treatment.20,21
Definitions of relapse usually focus on the severity of positive symptoms and largely 
ignore cognitive and negative symptoms: positive symptoms are more likely to lead to 
hospitalisation while cognitive and negative symptoms (which respond less well, and in 
some circumstances may even be exacerbated by antipsychotic treatment) have a greater 
overall impact on quality of life.
With respect to antipsychotic choice, in the context of an RCT, clozapine did not offer 
any advantage over chlorpromazine in the medium term in patients with first-­episode 
non-­refractory schizophrenia.22 However, in a large naturalistic study of patients with