A decision analysis approach to neuroleptic dosing: insights from a mathematical model

Although several published studies suggest that little benefit accrues from raising doses of conventional antipsychotic drugs above 500-800 chlorpromazine equivalents per day (CPZeq/day), institutionalized patients with schizophrenia often receive larger doses. Decision analysis could alter this practice by helping clinicians select doses through use of quantitative models that incorporate the consequences of each dose, the likelihood of those consequences, and explicit risk/benefit weightings.

This study uses representative published data to develop equations and graphs that describe dose-associated likelihoods of treatment response, side effects, and balances between benefits and incidence of side effects.

Response rates fit a sigmoid curve that flattens at 500 CPZeq/day; a hyperbolic curve describing side effects reaches a plateau at much higher doses. Combining these curves shows that higher drug doses yield ever diminishing returns, because as the dose increases, the number of side effects per benefited patient also increases. A table and graphs show clinicians how to use these results to critique their current practices and make explicit risk/benefit judgments about dosages.

Mathematical expressions for dose-related side effect and response rates are potentially useful tools for evaluating low-, intermediate-, or high-dosage neuroleptic treatment regimens.