The researchers compared haloperidol, one of an older class of schizophrenia drugs called typical antipsychotics, to olanzapine, the most expensive among the newer atypical antipsychotics. Used alone, the older medications are more likely to cause troubling side effects such as tremors and twitches. But the study had doctors prescribe haloperidol as they would ideally in actual practice-accompanied from the outset by another drug, benztropine, to minimize side effects.
"We gave the benztropine prophylactically along with the haloperidol, as is needed for a fair and clinically informative comparison," said lead author Robert Rosenheck, MD, director of VA's Northeast Program Evaluation Center in West Haven, Connecticut, and a professor of psychiatry and public health at Yale University Medical School.
"That's why we think this study is more relevant to everyday practice. We wanted to compare the two drugs in the way they are used in the real world."
The randomized, double-blinded study, which followed patients for one year, found no differences between the drugs in reducing schizophrenia symptoms or improving quality of life. As for side effects, olanzapine tended to cause weight gain. It resulted in slightly less akithesia, or restlessness, and somewhat better cognitive status, but not enough to improve patients' quality of life or overall functioning.
Though the drugs produced similar results overall, they come with a whopping difference in price: Olanzapine costs more than $8 per day per patient, based on VA figures, compared to just about 6 cents per day for haloperidol. And the higher-priced drug didn't lead to any significant reduction in hospital or outpatient costs.
Rosenheck said he does not see the study as prompting a return to the older class of schizophrenia drugs. The newer atypical antipsychotics have become widely accepted over the past 10 years as the first-line choice for treating schizophrenia. But he did say the findings sharply challenge the perception that olanzapine, while costlier at the pharmacy, more than pays for itself by lowering overall health-care and social service costs for its users. In the study, olanzapine was associated with $3,000 to $9,000 in greater annual VA costs per patient-mostly due to the higher cost of the drug.
"This study suggests that the advantages of olanzapine may be limited, while costs are considerably greater," said Rosenheck. "As a nation we are spending $2 billion annually on a treatment whose advantage over less expensive treatments is questionable and which may incur adverse health effects related to weight gain."
He said the findings support VA's current prescribing guidelines-"that physicians should try cheaper drugs before more expensive ones, as long as they are similarly effective."
VA guidelines recommend risperidone or quetiapine, two other atypical antipsychotics, as first choices for treating schizophrenia. Olanzapine, along with the atypicals ziprasidone and clozapine, is a second-line drug. Typical antipsychotic drugs such as haloperidol are only recommended when patients fail to respond to the above treatment. Physicians are free in all cases to use their discretion, based on the individual patient.
Risperidone costs about half as much as olanzapine, and has shown equal benefits in most clinical trials. According to 2002 figures from VA's National Psychosis Registry, more than 80 percent of VA patients with schizophrenia are on atypical antipsychotics, with about 37 percent on risperidone and 38 percent on olanzapine. In fiscal 2003, VA spent $208.5 million overall on atypical antipsychotics- $106.6 million for olanzapine alone.
Besides its cost, another concern over olanzapine is that many users tend to gain weight. The drug has been linked in some studies with an increased risk for diabetes.
Rosenheck said his team is eager to pursue studies analyzing the benefits of a new long-acting, injectable form of risperidone that is expected to become available and may result in better compliance.
In additional research, four VA hospitals are among 53 sites currently participating in a $42-milllion nationwide study, funded by the National Institute of Mental Health, comparing five atypical antipsychotic drugs to each other and to an older, traditional antipsychotic drug. Rosenheck was named director of service use and economic assessment for the study. Results are expected in 2006.
Schizophrenia, a biological disease of the brain, is the most common form of psychotic illness, affecting about 2.2 million Americans. It usually develops between ages 16 and 30. Contrary to popular notion, the disease is marked not by a "split personality," but by delusions, hallucinations, and confused thinking. People with schizophrenia may have trouble carrying on a conversation or focusing on a task, and usually show flat emotions and a lack of interest in life. They are more likely to die early, because of a higher suicide risk and other issues that arise from their mental state, such as automobile accidents, medical problems or homelessness.
VA provides health care for about 200,000 veterans with psychosis. Of these, about half have schizophrenia and more than a third have bipolar disorder. VA spends 15 percent of its total health care budget on medical and psychiatric care for this population. In fiscal year 2002, VA spent more than $35 million on research focused on the diagnosis, treatment and prevention of mental illnesses, including schizophrenia.
Rosenheck noted that while Eli Lilly and Company, the maker of olanzapine, did supply drugs and placebos for his study, the analysis was conducted with complete independence on the part of VA.
The Eli Lilly Company and VA's Cooperative Studies Program, a unit of the department's Office of Research and Development, supported this study.
NOTE FOR REPORTERS: Robert Rosenheck, M.D., director of VA's Northeast Program Evaluation Center, is the principal investigator for this study, and is available for press interviews. For telephone or on-site interviews, please call his office at 203-937-3850 or contact Pamela Redmond at 203-937-3824 or firstname.lastname@example.org. For additional assistance, please contact Jim Blue at 212-807-3429 or email@example.com.