Pharmacotherapy for self-injurious behavior: preliminary tests of the D1 hypothesis.

The D1 dopamine hypersensitivity model of self-injurious behavior leads to a testable clinical hypothesis: the mixed D1/D2 dopamine antagonist fluphenazine may improve the symptoms of self-injurious patients. The hypothesis was tested in an open pilot trial in six patients and a partially controlled trial in nine patients. Some degree of clinical improvement was observed in 11 of the 15 patients. The trials represent a partial affirmation of the D1 hypothesis. However, it is also clear that conventional methodology for psychopharmacologic research is inappropriate for the proper clinical evaluation of self-injurious patients. The proper method should include the following elements: 1. An epidemiologically representative sample. 2. A naturalistic study environment. 3. A longitudinal design with long-term follow up. 4. Concurrent behavioral ratings using direct observations and a reliable, treatment-sensitive rating scale. Before subjects enter a clinical trial of an experimental medication, a neuropsychiatric differential diagnosis should be applied to limit the diversity of the sample.