Published Online: August 16, 2012
Eric D. A. Hermes, MD; Michael Sernyak, MD; and Robert Rosenheck, MD
Objective: Recent research suggests that secondgeneration antipsychotics (SGAs) may be used more often than clinically warranted. An interven-tion consisting of academic detailing and a prescriber survey was employed to encourage the reduction of newly prescribed on-patent SGAs.
Design: Quasi-experimental quality improvement trial.
Methods: Academic detailing consisted of educational lectures and a pocket guide on the latest effectiveness, safety, and cost data for SGAs and first-generation antipsychotics. Detailing was coupled with a required 20-item survey of provider decision making completed prior to prescriptions for on-patent SGAs at a Veterans Health Administration medical center between October 2007 and May 2009. The survey identified the medication, treated diagnosis, comorbid psychiatric and medical diagnoses, reasons for the medication, prior medications, and provider professional status. The outcome was the number of new SGA prescriptions per month.
Results: The sample included 2176 surveys. The Spearman correlation between the number of prescriptions and the intervention month (range = 1-18) was 0.25 (P = .31), indicating no reduction. The most common medication prescribed was quetiapine (55.8%). The distributions of diagnoses were fairly even among schizophrenia, bipolar disorder, other affective disorders, and posttraumatic stress disorder (17.0, 28.2, 25.8, and 20.4%, respectively). The 3 most common reasons for prescribing an SGA were to improve efficacy (49.8%), reduce side effects (29.1%), and increase sleep or sedation (34.5%).
Conclusions: Academic detailing coupled with a provider survey did not decrease the rate of new prescriptions for on-patent SGAs. Reasons for prescribing SGAs were not consistent with recent research findings regarding efficacy and side effects.
(Am J Manag Care. 2012;18(8):e307-e314)
Second-generation antipsychotics (SGAs) may be used more often than clinically necessary. A nonrestrictive intervention of academic detailing coupled with a provider survey did not decrease the rate of new prescriptions for SGAs. Investigation of provider decision making found:
A high rate of off-label SGA use in posttraumatic stress disorder where there is little evidence of efficacy, as well as for sleep problems where there are many less costly and better tolerated options.
SGAs were commonly prescribed in patients with comorbid metabolic and cardiac disease where their use may exacerbate these conditions.
Second-generation antipsychotics (SGAs) have become the most prescribed psychopharmacologic treatment for schizophrenia.1 However, a growing body of evidence from several large, government- funded trials, such as VA Cooperative Study 451, the Clinical Antipsychotic Trials of Intervention Effectiveness (CATIE), and the Cost Utility of The Latest Antipsychotics in Severe Schizophrenia (CUtLASS), have failed to find an advantage for SGAs, other than clozapine, over first-generation antipsychotics (FGAs) on measures of effectiveness, side effects, or cost-effectiveness.2-5 SGAs had been thought to represent an advantage due to their decreased risk of neurologic side effects6,7 but these advantages have not been confirmed in recent comparative large-scale trials which more closely approximate real-world practice.8,9 In addition, several SGAs are now known to be associated with significant weight gain, impairment of glucose metabolism, and lipid abnormalities.10,11
SGA use has also increased beyond US Food and Drug Administration (FDA)-approved applications, which include, among others, schizophrenia, acute and maintenance treatment of bipolar disorder, and as an adjunct to other depression treatments. Common off-label uses includethe treatment of posttraumatic stress disorder (PTSD), anxiety disorders, and insomnia.12-14 In 2010, SGAs accounted for over $18 billion in sales15 with 75% of US expenditures paid through Medicaid.16 These high costs stem from the fact that most SGAs are still “on patent” and cost up to $10 a day in the United States, 10 to 100 times the cost of FGAs and off-patent SGAs.17 Much of the overwhelming prescriber preference for SGAs is thought to be linked to early impressions of efficacy created through industry-sponsored trials and marketing.18-20
The expense associated with SGAs has led payers to consider diverse strategies for modifying their practice, ranging from more limited measures such as formulary restrictions and prior authorization to less limited measures such as educational outreach.21 Educational strategies, termed academic detailing, link information dissemination activities with 1-on-1 practice reviews and provide alternative information regarding evidencebased prescribing practices to that presented by pharmaceutical manufacturers.22 These interventions, which are sometimes combined with other management strategies,23 provide more balanced evidencebased information to physicians, pharmacists, and patients in order to promote more cost-effective prescribing. Several meta-analyses of academic detailing have
shown moderate effects.24
This study presents data on the impact of an intervention combining both academic detailing and completion of a required prescriber survey on SGA prescribing at a single Veterans Health Administration (VHA) hospital. The administrative assumptions were that academic detailing in combination with a survey prior to completing the prescription would stimulate a more thorough reflection on prescribing decisions and would create a mild disincentive for prescribing SGAs. In addition, survey results would provide insight into the reasoning behind provider decisions regarding SGA prescriptions.
The study began as a quality improvement project which evolved into a quasi-experimental study to test the effect of a novel intervention on the rate of SGA prescriptions over a 20-month period, from October 2007 until May 2009. The intervention consisted of 2 components: an academic detailing effort and an electronic survey completed at the time of every new prescription for an on-patent SGA.
Academic detailing consisted of 4 formal detailing sessions with the entire mental health service delivery section of the medical center, psychiatry residents, nurse practitioners, and the division of the outpatient clinic responsible for treating patients with schizophrenia, respectively. Each session included a 45-minute didactic presentation followed by a discussion. The sessions highlighted background on the high level of expenditures on SGAs nationally and in the VHA, as well as a number of specific trials including the VHA Cooperative study, CUtLASS, CATIE, and others.2-5 The public policy issues regarding the cost-effectiveness of SGAs were also discussed.25 Finally, various strategies for cost containment were reviewed in addition to the objectives, hypotheses, and procedures of the current cost-containment study. Sessions were given by professors of psychiatry from the Yale University School of Medicine who had joint appointments at the VHA.
The required survey was completed at the time any on-patent SGA on formulary (olanzapine, long-acting injectable [LAI] risperidone, aripiprazole, ziprasidone, or quetiapine) was ordered as a new, nonrefill prescription for an outpatient who had not previously been receiving the SGA prescribed. This study was approved by the institutional review boards of the Yale University School of Medicine and the VA Connecticut Healthcare System.
Setting and Participants
The intervention was implemented throughout the VA Connecticut Healthcare System, which provides out-patient and inpatient treatment to approximately 10,000 veterans with mental illness annually,26 and targeted at prescribers of SGAs who are limited at this institution as a utilization management strategy to psychiatry residents, fellows, attendings, advance practice registered nurses, and physician assistants. Primary care providers and other specialists did not have privileges for SGA prescribing but could place prescriptions for SGAs after pharmacy review.
The survey was electronically delivered as part of the medication ordering system at the time of any new on-patent SGA prescription (risperidone was the only generic SGA at the time of the intervention). Survey completion was required before an SGA prescription could be electronically sent to the pharmacy. Thus, the survey completion rate was presumed to be 100%.
The survey consisted of 20 questions which documented the proposed medication, the primary psychiatric diagnosis under treatment, important comorbid psychiatric and medical diagnoses, the clinician’s reasons for using the selected medication, prior medications attempted, and the professional background of the provider. A list of forced choice selections and write-in answers was given for each question.
Exclusion Criteria. Surveys in which information on the SGA prescribed was missing were excluded. Some surveys documented oral risperidone use even though it was off patent at the time and these were excluded.
Outcome. The outcome for this study was the number of new SGA prescriptions measured monthly as the number of surveys completed between November 1, 2007, and April 30, 2009. Results from the first and last month of the survey period were truncated, as the project started
and finished mid month.
Statistical Analysis. Time trends in the prescription of atypical antipsychotics were evaluated by calculating the Spearman correlation between the study month (range 1-18) and total number of prescriptions per month. Analysis of differences in the number of SGA prescriptions over time for selected subgroups derived from survey questions was calculated using χ2. Data management and statistics were performed using SAS, version 9.2 (SAS Institute, Cary, North Carolina).
The original data set contained 2648 observations of which 30 were excluded for missing data on the medication prescribed, while 442 were excluded for documenting oral risperidone prescriptions. The final sample consisted of 2176 surveys.
The average age of individuals receiving an SGA prescription was 54.5 years (standard deviation [SD] = 14.9) while the mean body mass index (BMI) was 29.2. (SD = 6.0) (Table 1). The most common medication prescribed was quetiapine (55.8% of prescriptions) while olanzapine and aripiprazole were less common (18.8% and 17.9%, respectively) and ziprasidone and LAI risperidone were least common (5.9% and 1.6%, respectively). Of the new prescriptions, 54.8% represented a new prescription of an SGA in the absence of prior antipsychotic therapy, 25.0% represented a change in antipsychotic to an SGA, and for 20.2% the prescription represented an addition to an FGA.
The distribution of diagnoses for which a prescription was made was fairly even among schizophrenia, bipolar disorder, other affective disorders, and PTSD (17.0%, 28.2%, 25.8%, and 20.4%, respectively). The most common reasons given for prescribing a particular SGA were improved efficacy (48.9%), improved sleep or sedation (34.5%), minimizing side effects as a group (29.1%), and patient preference (27.9%). Prescriptions were made relatively evenly among psychiatric provider types with 28.0% coming from advance practice nurses or physician assistants, 28.3% from psychiatric residents or fellows, and 30.2% from attending psychiatrists.
PDF is available on the last page.