To the Editor: I wish to expand on two points made in my editorial (1), published in the February 2010 issue of the Journal, and correct an error. An accidental injection of depot olanzapine into or near a vein can result from much of the dose being administered as one bolus, producing an overdose, which is manifested as confusion, disorientation, deliria, somnolence, dysarthria, ataxia, and coma or seizure (2). This occurs in approximately 0.07% (the correct value) of individuals per injection, or approximately 1% of patients each year, which cumulates year by year. Hopefully, clinicians will be meticulous about injection techniques, reducing the incidence. Eighty percent of the time, this syndrome starts within 1 hour after injection, 17% of the time within 1—3 hours, and 3% of the time after 3 hours, with the median time to incapacitation being 60 minutes (range: 10—300 minutes). There was no relationship of dose to seriousness of this adverse reaction. In addition, there were no fatalities. Patients completely recovered in a few days, and most agreed to go back on depot medication It is important to prevent the consequences of adverse effects (e.g., auto accidents) by observing the patient for 3 hours after the injection; having the patient leave the clinic with a responsible caregiver; being attentive to the nonspecific prodrome (feeling weak, dizzy, or generally bad); and avoiding sedative medications as well as epinephrine, dopamine, and other beta agonists because they may possibly worsen hypotension as a result of olanzapine's apha-1 properties.

I do not think there is sufficient evidence to recommend tapering or not tapering oral drug doses or using a loading or intramuscular booster dose when switching to depot olanzapine, based on the following evidence. It takes 3—5 injections to reach steady state. Plasma levels decrease after the first injection, to as low as 5%—20% of the levels observed with oral drug formulation, but the half-life of D2 receptor blockade for oral and depot formulations is at least three times as long as that for plasma (3). Kane et al. (4) found the rate of relapse to be approximately 50% greater (not statistically significant) in the first few dosage intervals than the rate observed at steady state. The 405 mg per month dose was almost as effective as the dose of 300 mg every 2 weeks, indicating that monthly injection intervals can be of use.

The dosage of 150 mg every 2 weeks is too low for many patients, and may possibly double the number of relapses when the observed difference is projected over years.