High-Cost Use of Second-Generation Antipsychotics Under California's Medicaid Program

Medi-Cal pharmacy claims from May 1, 1999, through August 31, 2000, were analyzed to identify outpatients who had received at least one of the three second-generation antipsychotics—risperidone, olanzapine, or quetiapine—for at least 60 days out of any 75-day period. (At that time neither ziprasidone nor aripiprazole was on the formulary.) The primary study population was then broken down into three groups: patients who had received only one second-generation antipsychotic during the 16-month period (group 1), patients who had received more than one second-generation antipsychotic for no more than 59 out of 75 days concomitantly (group 2), and patients who had received more than one of the three drugs simultaneously for at least 60 days out of any 75-day period (group 3).

The population that received any antipsychotic (first- or second-generation) during the 16-month period was also evaluated for polypharmacy. The criteria for concomitant use were the same as for group 3—that is, at least two agents had to have been used for at least 60 days out of any 75-day period.

The cost analysis was based on the amount paid per prescription within Medi-Cal.

The breakdown of the total population into groups 1, 2, and 3 is shown in Table 1, along with the gender and age compositions of the groups. A total of 116,114 patients received at least one of the three second-generation antipsychotics for at least 60 out of 75 days between May 1, 1999, and August 31, 2000. A total of 91,969 of those patients (79.2 percent) received only one agent during the study period; 19,350 (16.7 percent) received at least two agents with no period of concomitant use greater than 59 out of 75 days, and 4,795 (4.1 percent) received at least two agents simultaneously for at least 60 out of 75 days.

The total amount paid (pharmacy costs) through Medi-Cal for patients who received a second-generation antipsychotic was $309,644,640. Group 1 accounted for $219,123,216 (70.8 percent) of the total pharmacy costs, group 2 accounted for $54,380,659 (17.6 percent), and group 3 accounted for $36,134,508 (11.7 percent). The mean amount paid per patient for group 1 was $2,382, for group 2 was $2,810, and for group 3 was $7,536.

Secondary analysis showed that 218,731 patients received either a first-generation or a second-generation antipsychotic for at least 60 days out of any 75-day period during the 16 months. Of these, 21,332 (9.8 percent) received at least two agents concomitantly for at least 60 out of 75 days.

Second-generation antipsychotics are the most expensive drug class in Medi-Cal, and the costs of polypharmacy can be twice as much as the costs of monotherapy at therapeutic dosages. In this analysis of prescription data from Medi-Cal, the mean cost per patient for the polypharmacy group was more than three times that for the monotherapy group. If antipsychotic polypharmacy were more effective than monotherapy, this practice may actually reduce total costs despite the higher pharmacy costs. However, there are currently no data available to support the effectiveness of antipsychotic poly-pharmacy (2,3). Thus there is currently no reason to believe that antipsychotic polypharmacy would reduce total costs.

Despite the high economic costs, there are some patients for whom antipsychotic polypharmacy may be appropriate, and thus this option should remain open to clinicians. However, it may be prudent to reserve antipsychotic polypharmacy for patients who have not responded to adequate trials of monotherapy with first-line second-generation antipsychotics and who either do not respond to or are unable to tolerate clozapine (4,8). The drug use review board of the Medi-Cal division of the California Department of Health Services recommends that at least two—if not all five—first-line second-generation antipsychotic monotherapies be attempted before other treatment options (9). Monotherapy with first-generation antipsychotics, clozapine monotherapy, or augmentation with divalproex are also recommended before antipsychotic polypharmacy. This proposal is consistent with the revised recommendations of the Texas Medication Algorithm Project (TMAP)—that is, at least two trials of first-line second-generation antipsychotic monotherapy as well as a trial of clozapine before combination treatment is used (10).

The increasingly high costs of second-generation antipsychotics have raised the threat that use of these valuable resources will be restricted, most likely by removing certain agents from the formulary solely on the basis of economic factors. Offsetting the high costs of second-generation antipsychotics by reducing particularly costly yet poorly researched uses, such as polypharmacy, may be one means of eliminating the threat of imposed formulary restrictions.

This study was supported by AstraZeneca, Eli Lilly and Company, Janssen Pharmaceutica, and Pfizer Incorporated.

The authors are affiliated with the Neuroscience Education Institute, 5857 Owens Avenue, Suite 102, Carlsbad, California 92008 (e-mail, [email protected]).