Evidence Based Review Article

The Journal of Informed Pharmacotherapy 2002;9:203.

Risperidone or Haloperidol for Schizophrenic Patients: What Does a Focus on Relapse Rate Reveal?

Reviewer: Colette Raymond
Reviewer's email address:
Reviewer's profession/specialty:
Clinical Pharmacotherapeutic Specialist - Psychiatry

Original Citation

Csernansky JG, Mahmoud R, Brenner R; The Risperidone-USA-79 Study Group. A comparison of risperidone and haloperidol for the prevention of relapse in patients with schizophrenia. N Engl J Med 2002;346:16-22.  PubMed Citation

Overall Study Question

The objective of this trial was to compare risperidone to haloperidol for reduction of relapse in outpatients with schizophrenia or schizoaffective disorder.

Patients included in the study were adults with clinically stable schizophrenia (82%) or schizoaffective disorder (18%).  Patients had to have experienced outpatient crisis management, and treatment in a psychiatric hospital or emergency room within 24 months of study entry.  Patients were taking a stable dose of antipsychotic medication for ≥30 days, and were judged as clinically stable by investigators.  Patients were excluded from the study if they had another primary psychiatric disorder, a personality disorder, a history of refractoriness to antipsychotic drugs, or treatment with depot neuroleptics.

Patients were randomized to receive haloperidol or risperidone.  Over seven days the dose of previous antipsychotic medication was decreased as the dose of study medication increased.  Doses were increased to 4 mg of risperidone or 10 mg of haloperidol over three days.  Starting on day eight, doses were adjusted by investigators to target daily doses of 5-20 mg of haloperidol or 2-8 mg of risperidone.

Primary outcomes were relapse rate and time to first relapse. Secondary outcome measures included: Positive and Negative Syndrome Scale (PANSS) score and subscale scores and adverse effects as assessed by physical exam, laboratory screening and the Extrapyramidal Symptom Rating Scale.

Patients were assessed every week for four weeks, and then every four weeks until all patients had completed one year of treatment.

Are the Results of the Study Valid?

1. Was assignment of patients randomized?

Yes. Patients were randomly assigned to receive haloperidol or risperidone. Details of the randomization procedure were not provided.

2. Were all patients who entered the trial properly accounted for and attributed at its conclusion?

Yes. A total of 397 patients were enrolled. Patients who were assessed at least once (N=365) were included in Kaplan-Meier analysis of relapse rate. 

3. Were patients, their clinicians, and study personnel 'blind' to treatment?

Yes. The study was conducted in a double-blind manner. Tablets were identical appearing in order to maintain blinding.  Investigators did not report the effectiveness of blinding. It is assumed that this was not evaluated.

4. Were the groups similar at the start of the trial?

Yes. Both groups had similar baseline demographics including DSM-IV diagnosis, and total and subscale scores on PANSS.  Since group size was reasonably large and allocation was random, it is unlikely that the two groups would have been different in this regard.

5. Aside from the experimental intervention, were the groups treated equally?

Yes. There was no difference in compliance as assessed by pill counts (97% in the risperidone group and 96% in the haloperidol group).

What were the Results?

1. How large was the treatment effect? 

Patients were followed for a median of 364 days (range 3 to 799) in the risperidone treatment group and 238 days in the haloperidol group (range 4 to 794). The mean doses were 4.9 ± 1.9 mg of risperidone and 11.7 ± 5.0 mg of haloperidol. At the end of the study, 25.4% of patients in the risperidone group had relapsed and 39.9% of patients in the haloperidol group had relapsed, ARR = 14.5%, NNT = 9. The Kaplan-Meier estimate of the risk of relapse at the end of the study was 34% for the risperidone group and 60% for the haloperidol (p<0.001), ARR = 26% (95% CI 13.8% to 38.3%), NNT = 4 (95% CI 3 to 8). In other words, nine patients would have to be tried on risperidone and four patients would have to comply with risperidone for one year instead of haloperidol in order to prevent one relapse. Although the Kaplan-Meier estimate is less conservative, it reflects the favorable tolerability profile of risperidone.

A greater proportion of patients receiving haloperidol discontinued therapy for reasons other than relapse as compared to patients receiving risperidone, (44.1% and 52.7% respectively) relative risk 1.52 (1.18-1.96).

Patients randomized to risperidone experienced small improvements in PANSS scores as compared to haloperidol, however the differences were small and of minimal clinical significance.

There were similar overall incidences of adverse effects in both patient groups (90% in the risperidone group and 91% in the haloperidol group). Patients in the risperidone group gained an average of 2.3 kg, while patients in the haloperidol lost 0.73 kg (p < 0.001). There was a greater incidence of extrapyramidal side effects in the haloperidol group as compared to the risperidone group. Patients in the risperidone group improved from baseline, while patients in the haloperidol showed slight worsening on the Extrapyramidal Symptom Rating Scale (p=0.02). Fewer patients in the risperidone group required antiparkinsonian medications for more than 30 consecutive days than in the haloperidol group (9% vs. 17.2%, respectively p=0.02).

2. How precise was the estimate of the treatment effect?

The risk of relapse at the end of the study for the risperidone group was estimated to be 34% (95% CI 28% to 43%) and for the haloperidol group was 60% (95% CI 50 to 70%). The 95% confidence intervals for the ARR of 26% are 13.8% to 38.3%.

Will the Results Help Me in Caring for My Patients?

1. Can the results be applied to my patient care?

This trial indicates that stable, medication-compliant outpatients with chronic schizophrenia or shizoaffective disorder have a better chance of remaining relapse-free over a period of one year if maintained on risperidone as compared to haloperidol.

2. Were all clinically important outcomes considered?

The primary outcome was clearly defined and is of clinical importance in the long-term management of psychiatric patients. Possible clinically important outcomes that were not considered include quality of life and costs.

3. Are the likely treatment benefits worth the potential harms and costs?

This study shows a reduction in relapse rate for the treatment of schizophrenia with risperidone as compared to haloperidol, with fewer extrapyramidal adverse effects, but greater incidence of weight gain. The full impact of the risk of antipsychotic-induced weight gain on cardiovascular morbidity and mortality has yet to be fully described.

Although the acquisition cost of risperidone is greater than that of haloperidol (approximately 65 times), the decrease in relapse rate would likely trivialize the acquisition cost difference. A pharmacoeconomic evaluation has shown that risperidone does not increase, and may reduce costs of the treatment of schizophrenia through a reduction in hospitalizations as compared to conventional neuroleptics.(1)


Risperidone has been compared to haloperidol in many short term and few long term (>1 yr) studies in patients with schizophrenia. (2) As compared to conventional antipsychotics (most commonly haloperidol), risperidone has shown to be beneficial in terms of clinical improvement, with a lower incidence of movement disorders, somnolence, greater patient acceptability, (2) but greater incidence of weight gain. (2,3)  Usual study endpoints include symptom and adverse effect scales, as well as clinical global improvement scores. (2)  Few studies have assessed relapse rates.

This study offers a unique perspective owing to its long-term follow-up and focus on relapse rate, rather than clinical symptoms as a primary endpoint. The breakdown of the relapse endpoint would have been useful information to report. For example, it would have been interesting to determine how many patients in each group experienced a need for hospitalization, as compared to a worsening of symptoms.

A limitation to the interpretation of this study is the high rate of withdrawal. Despite the fact that doses were adjusted according to response and adverse effects, it has been suggested that the dose of haloperidol is higher than what is generally recommended, and that this may be a factor contributing to the tolerability, effectiveness, and relapse rate observed. (4,5)

Overall, this study indicates that risperidone is associated with a lower risk of relapse than is haloperidol for outpatients with stable schizophrenia or schizoaffective disorder over one year.


  1. Foster RH, Goa KL. Risperidone. A pharmacoeconomic review of its use in schizophrenia. Pharmacoeconomics 1998;14:97-133.
  2. Kennedy E, Song F, Hunter R, et al. Risperidone versus typical antipsychotic medication for schizophrenia (Cochrane Review). In: The Cochrane Library, Issue 4, 2001. Oxford: Update Software.
  3. Allison DB, Mentore JL, Heo M, et al. Antipsychotic-induced weight gain: a comprehensive research synthesis. Am J Psychiatry 1999;156:1686-96.
  4. Geddes J. Prevention of relapse in schizophrenia. N Engl J Med 2002;346:56-8
  5. Geddes J, Freemantle N, Harrison P, et al. Atypical antipsychotics in the treatment of schizophrenia: a systematic review and meta-regression analysis. BMJ 2000;321:371-6

Copyright © 2002 by the Journal of Informed Pharmacotherapy. All rights reserved.