The more psychotic you are, the more benefit there is in taking antipsychotics


Antipsychotic medications are currently the treatment of choice for someone who is experiencing psychotic symptoms. There efficacy is very well established and has been widely blogged about on the Mental Elf site, for example, the recent post by Raphael Underwood (2014) highlighting the efficacy of high vs. low-potency first-generation antipsychotics for schizophrenia.

However, antipsychotics also produce a range of unpleasant and potentially disabling side effects. Prescribing continues to be based predominantly on custom and practice, rather than following NICE recommendations or symptomology.

This study aimed to explore how prescribing antipsychotics or placebo relative to symptoms at baseline influences improvements in mental state.


The authors (Furukawa et al, 2015) undertook a systematic review and identified 97 trials. Of these they were able to get the patient level data for only six studies. For this data a participant level meta-analysis was conducted.

The authors describe this as a ‘3-level mixed-effects model repeated-measures analysis (MMRM) with maximum likelihood estimation’.

The authors specifically looked to examine the relationship between positive and negative symptoms and outcome using the Positive and Negative Syndrome Scale (or converted scores from the Brief Psychiatric Rating Scale), or the Scale for Assessment of Negative Symptoms.


  • 6 trials with patient level data were identified:
    • 3 for positive symptoms (611 patients)
    • 3 for negative symptoms (475 patients)
  • These studies used 3 antipsychotics Olanzapine (x2), Risperidone (x1), and Amisulpride (x3) for negative symptoms
  • All the included studies were placebo-controlled studies but only one olanzapine trial and one risperidone trial included haloperidol arms
  • Data was collected approximately 20 years ago
  • Outcomes were assessed several times up to 6 weeks, i.e. the maximum follow-up time common to the included trials. This is why the authors could use repeated-measures analyses
  • Baseline symptoms were related to outcome (p<0.01) so the worse your symptoms, the more the improvement
  • Less benefit for less symptoms, and greater risk of side effects

The authors found that patients with the most severe baseline illness were likely to benefit the most from antipsychotic treatment.


The clinical implications of our findings may be as follows: we can expect the benefits of antipsychotics for patients with full spectrum of severity who are likely to treat for acute schizophrenia and for highly symptomatic patients with predominantly negative symptoms, and the severer the illness is at baseline, the bigger the benefits will be.


Towards the mildest end of the spectrum, judicious clinical consideration of trade-offs between benefits and risks of the antipsychotic treatment is required.

The authors

The authors conclude that caution is required when prescribing antipsychotics to patients at the milder end of the spectrum of illness.


What surprised me most about this paper was the use of data from old trials to develop and test this model. The included data ranged from papers published between 1993 and 1999 and the data would have been collected before this. This is not clearly identified in the paper. For the authors to only be allowed access or get access to 6 of the 97 trials is terrible. It is clear that this paper was supported via drug companies and their role in this is unclear, particularly whether the authors could only access certain data.

From a clinical perspective it is useful to know that those with the worst symptoms (either negative or positive) are likely to improve more as a result of prescribing an antipsychotic vs placebo. But the review only included three of the commonly prescribed antipsychotic drugs, and there was no discussion about optimum dosages. However, the outcome of these trials were measured at 6 weeks, so this does not help clinicians understand the longer term implications of antipsychotic prescriptions.


Does the inclusion of data from 20 years ago affect the applicability of this study?


Primary paper

Furukawa TA, Levine SZ, Tanaka S, Goldberg Y, Samara M, Davis JM, Cipriani A, & Leucht S. (2015) Initial Severity of Schizophrenia and Efficacy of Antipsychotics: Participant-Level Meta-analysis of 6 Placebo-Controlled Studies. JAMA Psychiatry ;72(1):14-21. doi:10.1001/jamapsychiatry.2014.2127. [PubMed abstract]

Other references

Underwood R. (2014) Efficacy of high vs. low-potency first-generation antipsychotics for schizophrenia. The Mental Elf, 10 Oct 2014.

Huda S. (2015) Psychosis and schizophrenia in adults: new quality standard from NICE. The Mental Elf, 19 Mar 2015.


Readers who want to learn more about the research in this area may like to watch this Google Hangout organised recently by the Evidence-Based Mental Health journal.

Share on Facebook Tweet this on Twitter Share on LinkedIn Share on Google+
Mark as read
Create a personal elf note about this blog
Profile photo of John Baker

John Baker

John Baker was appointed to Chair of Mental Health Nursing in 2015. John's research focuses on developing complex clinical and psychological interventions in mental health settings. He is particularly interested in i) acute/inpatient mental health services and clinical interventions; ii) medicines management in mental health care; iii) the attitudes and clinical skills of mental health workers, iv) the mental health workforce. The good practice manuals which he developed have been evaluated, cited as examples of good practice, and influenced clinical practice in the UK and abroad. The training package for patients, service users and carers to promote research awareness and understanding has been cited by the MHRN and NICE as an exemplar of good practice.

John is a member of the NIHR post-doctoral panel, sits on the Editorial boards for Journal of Psychiatric and Mental Health Nursing & International Journal of Mental Health Nursing. He is a Registered Nurse Teacher with the Nursing, Midwifery Council (NMC) and is active within Mental Health Nursing Academics (UK).

More posts

Follow me here –