Diversity versus concentration in antipsychotic drug selection

It is unknown to what extent providers utilize the full range of available antipsychotic agents, or tend to prescribe just a few with which they are familiar.

This study uses the Herfindahl index to measure the degree to which antipsychotic prescribing is diverse or concentrated within medical centers in the Department of Veterans Affairs.

Patients with schizophrenia who received prescriptions for antipsychotic medications from the Department of Veterans Affairs in fiscal year 2005 were identified. For each facility, the proportion of prescriptions that were written for each antipsychotic medication was computed. The Herfindahl index, which is defined as the sum of the square of the proportionate shares of each medication (i.e., 50% squared = 0.25), was computed for each facility, and its distribution across facilities was examined.

We identified 785,485 prescriptions for antipsychotic medications across 76,787 patients within 128 facilities. With 7 drugs to choose from (first-generation antipsychotics were not differentiated as separate drugs), the Herfindahl could range from 0.14 (if the drugs were prescribed equally) to 1 (if only 1 drug was prescribed to all patients). The Herfindahl averaged 0.20 to 0.21 and ranged from 0.16 to 0.31 across facilities.

The Herfindahl index is a useful metric for describing the degree to which providers use a range of antipsychotic medications in treating patients with schizophrenia and suggests that VA facilities use a diversity of agents, thus taking advantage of the potentially unique drug-patient matches.

The extent to which providers tailor their choice of antipsychotic medication to the particular characteristics of the patient may lead to better quality of care.

The fact that providers do use a range of antipsychotic medications suggests that formulary policies should not reduce access to these drugs.

Further research is needed to construct physician-level measures of prescribing diversity and to examine whether clinical outcomes are improved when providers use a wider range of medications in their practice.