Implementation of antipsychotic selection recommendations at a veterans affairs facility

BACKGROUND

The rising cost of prescription drugs has caused concern in numerous countries, with the United States being no exception. Mental health was the number one class of medication, based on spending, from 2008 to 2011, and only falling to number two in 2012 due to patent expirations in antipsychotics and antidepressants.¹ From 2008 to 2012, $26, $26.1, $28.2, $29.7, and $23.5 billion were spent each year, respectively, for mental health medications. In 2012 alone, 3.4 million were treated with antipsychotics. Of note, the costs of psychotropic medications have risen faster than costs of other medications. Total outpatient prescription expenses for antipsychotics increased over 300% in a ten year span, from $1.7 billion in 1997 to $7.4 billion in 2007. The total number of antipsychotics purchased increased approximately 86% in the same time period, rising from 17.4 million purchases to 32.4 million purchases. The number of people purchasing antipsychotics rose from 2.2 million to 3.9 million as well as the proportion of the population (0.8% to 1.3%). The average total expense per antipsychotic prescription purchase increased from $96 to $228 and the average annual expense increased nearly 150% from $765 to $1,905.²

Similar findings have been reported in England, with 4.7 million prescriptions in 1998 rising to 7.6 million in 2010 resulting in a 250% increase in cost. Olanzapine accounted for 42% of the total expenditure on antipsychotics in 2010 and quetiapine accounted for 34%. Aripiprazole accounted for 13% of the total expenditure despite only making up 5% of the number of prescription items. Risperidone (oral and injectable) accounted for only 6% of the total expenditure in 2010.³

With this in mind, the VA Pharmacy Benefits Management Services, Medical Advisory Panel, and Veterans Integrated Service Network (VISN) Pharmacist Executives created recommendations for antipsychotic selection in schizophrenia and schizoaffective disorders in June of 2012 (Figure 1). The recommendations indicate that a patient be first tried on one of five first line agents as monotherapy (haloperidol, loxapine, quetiapine, risperidone or perphenazine) for at least 6 weeks or intolerable side effects. Should they fail this treatment, they will be offered a second first line agent as monotherapy for 6 weeks. After failing two first line agents, a trial of clozapine should be offered if appropriate. If a trial of clozapine fails, or the patient is not a candidate for use, the patient may be offered a third first line agent, an alternative agent not listed above, or a second line agent (olanzapine, ziprasidone or aripiprazole).

View Figure

• Download as Powerpoint
• Download Figure

These recommendations for antipsychotic selection were based on numerous studies comparing the efficacy of various antipsychotics. The 2003 VA Cooperative Study comparing olanzapine to haloperidol suggested that olanzapine did not demonstrate advantages in compliance, symptoms, extrapyramidal symptoms (EPS), or overall quality of life.⁴ The Clinical Antipsychotic Trials of Intervention Effectiveness (CATIE) study indicated that the efficacy of perphenazine appeared similar to that of quetiapine, risperidone, and ziprasidone and although olanzapine had a longer time to discontinuation than quetiapine and risperidone, it was associated with greater weight gain and increases in measures of glucose and lipid metabolism.⁵ The Cost Utility of the Latest Antipsychotic Drugs in Schizophrenia Study (CUtLASS) concluded that first-generation antipsychotics (FGAs) had lower costs and higher quality-adjusted life-years (QALYs) than second-generation antipsychotics (SGAs) and were more than 50% likely to be cost-effective.⁶ An industry sponsored trial was unable to find a significant difference in the efficacy of aripiprazole versus perphenazine.⁷ A meta-analysis of 150 studies concluded the overall efficacy of aripiprazole, quetiapine and ziprasidone were equal to FGAs. Clozapine, olanzapine and risperidone were found to have greater efficacy than FGAs.⁸

Safety concerns exist for both FGAs and SGAs and are often considered when selecting an appropriate agent. While FGAs are typically thought of as having a higher risk of EPS, perphenazine has a similar rate of EPS as risperidone.⁴ While the risk of tardive dyskinesia (TD) is greater with FGAs, it is not as great as once was held. Differing estimates place the risk of TD with FGAs at 20–32% and 13% for SGAs.^9–12 Based upon these findings, haloperidol, loxapine, quetiapine, risperidone and perphenazine were classified as first line agents by the VA Pharmacy Benefits Management Services, Medical Advisory Panel, and VISN Pharmacist Executives. Clozapine is then to be offered after the failure of two first line agents. Due to their higher cost and lack of superiority over alternative agents, olanzapine, ziprasidone or aripiprazole should be reserved for patients that have failed the progression listed above.

METHODS

This study was approved by the Colorado Multiple Institutional Review Board (COMIRB). A retrospective chart review was performed for patients >18 years of age who were initiated on aripiprazole, ziprasidone, or olanzapine for all indications between July 1, 2011 and July 1, 2012 at the VA ECHCS.

The primary outcome was the determination of cost savings that could have been provided had the recommendations been implemented last year. For cost analysis purposes, the mean daily dose used for each second line agent was determined. As the VA ECHCS utilizes tablet splitting, the cost was calculated for both whole tablets and half tablets. The costs of these mean doses were compared to an equivalent mean dose of a first line agent; 4mg of risperidone. VA drug acquisition costs were utilized.

Use of second generation antipsychotics has been associated with weight gain, diabetes, and worsening of lipid profiles which represent major cardiovascular disease risk factors. A secondary aim of this project was to determine the rate of provider adherence to the ADA recommendations. Cardiometabolic monitoring data were collected to include weight, blood pressure, fasting glucose and fasting lipid profiles at baseline and at 12 weeks in accordance with ADA recommendations.¹³ Baseline values were defined as those collected 12 weeks before to 2 weeks after the first fill date. Follow-up 12 week values were defined as those collected 8 weeks to 16 weeks after the first fill date.

RESULTS

A total of 542 patients were identified and assessed for compliance with antipsychotic selection criteria and ADA monitoring guidelines. This retrospective chart review revealed that most patients newly started on aripiprazole, ziprasidone, or olanzapine did not meet criteria for use as described in the recommendations above. Of the 542 patients started on these agents, only 68 met criteria while 77 were started at other facilities and were allowed to continue use of these agents (Figure 2). Of the 77 that were “grandfathered” only 17 would have met criteria. Of the 397 that did not meet criteria, 276 received aripiprazole, 77 received olanzapine, and 44 received ziprasidone (Figure 3).

View Figure

• Download as Powerpoint
• Download Figure

View Figure

• Download as Powerpoint
• Download Figure

Cost analysis data is presented in Table 1. The 267 cases of aripiprazole that did not meet criteria resulted in a majority of the projected cost savings. Had prescribers utilized risperidone as a first line alternative, nearly $750,000 would have been saved. Use of 10mg whole tablets of aripiprazole was nearly $10 more expensive per day than 4mg whole tablets of risperidone. Twice daily dosing of ziprasidone was approximately $6 more expensive per day as compared to 4mg of risperidone. This data also illustrates the cost benefits of utilizing tablet splitting as seen by the approximately $5 per day savings with aripiprazole.

Table 1: Cost analysis

View Fullscreen

This study showed that none of the 542 patients had both baseline and follow-up values for any of the cardiometabolic monitoring data recommended by the ADA (Table 2). While roughly 60% of patients had blood pressure and weight values at baseline and 27% had glucose and cholesterol values, less than 14% received follow-up testing with approximately 2% receiving neither baseline nor follow-up testing.

Table 2: Cardiometabolic monitoring parameters on record in patients taking antipsychotics

View Fullscreen

DISCUSSION

A potential cost savings of over $850,000 may have been realized had these criteria been enacted within the one year time period studied. As most studies do not demonstrate the superiority of one agent and the rates of adverse events between FGAs and SGAs are more similar than previously thought, it is reasonable that patients trial the five listed first line agents before the more expensive second line agents.

Several studies have evaluated the cost-effectiveness of FGAs versus SGAs. A study performed in the United States compared four SGAs to the FGA perphenazine and concluded that the latter was less costly with no significant differences in effectiveness.¹⁴ One study in Germany looked at rehospitalization rates for FGAs and SGAs and discovered that there was no significant difference in the effectiveness of FGAs compared with SGAs on rehospitalization.¹⁵ These studies further support the VA recommendations for antipsychotic selection.

This study also indicates a need for improved cardiometabolic monitoring. None of the 542 patients identified in this study had both a baseline and follow-up value for any of the data. A pharmacist could be utilized to establish a cardiometabolic monitoring clinic to ensure that these patients are appropriately followed in accordance with ADA guidelines and suggest interventions in therapy.

A potential weakness of this analysis is the assumption that all antipsychotics work equally well in all patients. Numerous studies cited above indicate that the efficacy of these agents is similar, yet the interpretation of these studies is muddled by varying durations of study, inclusion and exclusion criteria, and outcome measures. Selection of a particular agent is commonly based upon side effects which may have driven the provider's choice of medication. To address this issue, the VA recommendations provide first line agents with variable side effect profiles.

The results of this study lead to the implementation of these criteria at the VA ECHCS and a consult menu was created for ordering second line agents. Patients receiving second line agents from the VA ECHCS or an outside facility are allowed to continue use. New starts require a prior authorization drug review by a pharmacist to assess for compliance with these criteria. A future project will evaluate the effectiveness of this new consult in curbing drug costs and improving metabolic monitoring.