Comparative efficacy of red cell and plasma haloperidol as predictors of clinical response in schizophrenia.

The relative utility of steady-state (SS), plasma (Pl), and red blood cell (RBC) haloperidol levels for predicting clinical response was evaluated in a fixed-dose study in schizophrenic inpatients. There were significant curvilinear relationships between the decrease in BPRS Psychosis Factor Scores by day 24 of haloperidol treatment and both Pl (R2 = 0.34) and RBC (R2 = 0.38) haloperidol levels. Although SS RBC haloperidol levels consistently showed a slightly stronger relationship to clinical response than Pl levels in several comparisons, the differences in R2s between Pl and RBC haloperidol were not statistically significant. Ninety percent confidence intervals for the blood level ranges associated with optimal clinical response in our sample of patients were: 6.5-16.5 ng/ml Pl haloperidol and 2.2-6.8 ng/ml RBC haloperidol.