Risperidone dose for schizophrenia.

BACKGROUND: Risperidone is a widely used antipsychotic drug for people with schizophrenia. It is important to get a balance between gaining the most positive effects for the least negative outcomes. The optimal dose of risperidone is the focus of this review.

OBJECTIVES: To determine risperidone dose response relationships for schizophrenia and schizophrenia-like psychoses.

SEARCH STRATEGY: We searched the Cochrane Schizophrenia Groups Trials Register (July 2008) for all relevant references.

SELECTION CRITERIA: All relevant randomised controlled clinical trials (RCTs).

DATA COLLECTION AND ANALYSIS: Two review authors independently extracted data and resolved disagreement by discussion with a third member of the team. When insufficient data were provided, we contacted the study authors. For homogenous dichotomous data we calculated fixed-effect relative risk (RR) and 95% confidence intervals (CI) on an intention-to-treat basis. For continuous data, we calculated weighted mean differences (MD).

MAIN RESULTS: A consistent finding when risperidone ultra low doses (<2 mg/day) were compared with other doses (short-term data) was that more people left early because of insufficient response (n=456, 1 RCT, RR when compared with standard-low (>==4-<6 mg/day) 12.48 CI 1.43 to 4.30). The insufficient response for this low dose is reflected in measures of mental state. When low doses (>==2-<4 mg/day) are used and compared with standard-higher doses (>==6-<10 mg/day) and the high dose range (>==10 mg/day), more people left early because of insufficient response (>==4-<6 mg/day: n=173, 2 RCTs, RR 4.05 CI 1.09 to 15.07; >==10 mg/day: n=173, 2 RCTs, RR 1.92 CI 1.36 to 2.70). For the outcome of 'no clinically important improvement' results favour standard-higher doses (n=272, 2 RCTs, RR 2.26 CI 0.81 to 6.34). When low doses are compared with other higher doses, we found no differences in terms of cardiovascular, CNS, endocrine or gastrointestinal adverse effects. Unspecified EPS were more frequent with the higher doses (>==10 mg: n=262, 2 RCTs, RR 0.45 CI 0.24 to 0.84). One trial did find that endpoint scores on PANSS significantly favoured a low dose when compared with >==4-6 mg/day (n=124, 1 RCT, MD -12.40 CI -17.01 to -7.79). When >==4-<6 mg/day is compared with high doses, less people left early (n=677, 1 RCT, RR leaving any reason 0.74 CI 0.54 to 1.00; n=677, 1 RCT, RR due to adverse effects 0.56 CI 0.32 to 0.97). >==4-<6 mg/day was no worse than >==6-<10 mg/day for 'no clinically important improvement' (n=39, 1 RCT, RR on CGI-I 0.79 CI 0.29 to 2.17). People allocated >==4-<6 mg/day had more movement disorders than those on a low dose (n=124 1 RCT, RR 2.28 CI 1.67 to 3.11). When >==6-<10 mg/day is compared with standard-lower doses and a high dose range, there is no significant difference in terms of proportions leaving early. >==6-<10 mg/day is better than a low dose for 'no clinical important improvement' (n=172, 2 RCTs, RR 0.76 CI 0.61 to 0.94). Overall >==6-<10 mg/day caused less problems especially in EPS when compared with >==10mg/day (n=261, 2 RCTs, RR unspecified EPS 0.56 CI 0.31 to 0.99). When a high dose was compared with a low dose less people left early (n=70, 1 RCT, RR 0.43 CI 0.26 to 0.71) but not when compared with a standard-lower dose (n=677, 1 RCT, RR leaving due to adverse event 1.78 CI 1.03 to 3.09). >==10 mg/day was better than a low dose in terms of 'no clinical important improvement' (n=257, 2 RCTs, RR 0.64 CI 0.50 to 0.82), but worse than a standard-higher dose (>==6-<10 mg/day: n=255, 2 RCTs, RR 1.22 CI 1.00 to 1.51). >==10 mg/day caused more unspecified EPS adverse effects and any drug for adverse events when compared with a standard-higher dose and with a low dose.

AUTHORS' CONCLUSIONS: There is still lack of strong evidence for an optimal dose for clinical practice. The quality of trials suggests that an over estimate of effect is likely and we think this is most probably for the mid-range doses. One such dose (standard-lower dose range, 4-<6 mg/day) does seem optimal for clinical response and adverse effects. Weak evidence suggests that low doses (>==2-<4 mg/day) may be of value for people in their first episode of illness. High doses (>==10 mg/day) did not confer any advantage over any other dose ranges and caused more adverse effects, especially for movement disorders. Ultra low dose (<2 mg/day) seemed useless. We advise the use of dosages from low dose to standard-lower dose for different kinds of individual patients. Future trials should focus on specific populations, e.g. those in their first episode, with acute exacerbation, in relapse or refractory to treatment, and should also test the optimal dose of risperidone over a longer period of time and in the community.