I was doing some searching about drug therapy for psychosis in Parkinson's today and came across a few articles which demonstrate the complicated balancing act between evidence and practice.
First, in the article Patterns and trends in antipsychotic prescribing for Parkinson disease psychosis, it is reported that 66% of studied PD patients with psychosis seen between 2002 and 2008 were prescribed quetiapine and 2% received clozapine (the additional 30% were represented mainly by risperidone (17%), aripiprazole (12%), and olanzapine (11%)). Quetiapine is clearly the most prescribed antipsychotic for this population and clozapine the least, even though clozapine has been shown to be clearly effective (albeit the need for monitoring makes it less attractive).
As pointed out in that article, and in greater detail in the review Quetiapine in the treatment of psychosis in Parkinson's disease, the evidence for quetiapine's efficacy is mixed. Unblinded studies found it to be effective, it compared similarly to clozapine (though was less effective), yet 4/5 placebo controlled trials did not show a significant difference between it and placebo. Larger and better designed RCTs are needed to determine its true effectiveness. Despite the the mixed evidence, it may provide the best balance of safety, convenience, and efficacy.
Finally, I just wanted to point out the article Pimavanserin, a Serotonin2A Receptor Inverse Agonist, for the Treatment of Parkinson's Disease Psychosis. Pimavanserin is currently in phase III trials (no locations in Minnesota) and the study shows significant improvements over placebo in some measures, trends for other measures, and no significant differences for adverse effects. It seems like a good drug to keep an eye on.