A couple of months ago Newsweek magazine published an article questioning the science of mental health services, and in particular, clinical psychology. The article was based on an opinion piece published in the journal of the Association for Psychological Science, in which a team of clinical scientists promoted a new accreditation system for clinical psychology doctoral program, which would ensure that future clinical psychologists are trained as scientists who base their clinical decisions on scientific evidence rather than perceived ‘clinical expertise,’ ideology, or tradition. Some of you may have assumed that therapists practice within specific empirically supported guidelines. Unfortunately this is not the case. Clinical psychologists, social workers, and other therapists have been notoriously resistant, and often hostile, to the adoption of new intervention modalities in response to scientific advances. In fact, many clinicians use therapeutic modalities that simply have no empirical support.
One argument against such criticism of mental health providers is that today’s medical doctors also routinely engage in practices that have little scientific support, such as when using off-label medications. For example, when a psychiatrist prescribes a child an anti-psychotic drug that has only been approved by the FDA for use with adults, she is making a clinical decision that is not supported by scientific evidence (assuming that lack of FDA approval means limited scientific support for the drug’s efficacy and safety). However, there is a major difference between a physician who uses an off-label medication for children and a therapist that refuses to change their therapeutic modality despite evidence that what they do doesn’t work, or that another modality may be more effective. The former usually engages in a non-scientifically supported practice because of the lack of alternatives, while the latter engages in non-supported practices despite scientific evidence that contraindicate such practice. For example, a psychiatrist may use an adult medication with a child if the child does not respond to child-approved medications. In this case, the psychiatrist is not purposely ignoring science. Instead, she is using her clinical judgment to provide the best available alternative. In contrast, a psychologist may use a therapeutic modality that has no empirical support not because there are no alternatives, but because that is the only modality that she knows, or because she is ‘married’ to an ideology that makes her believe that what she does works best. Under this view, psychologists and other clinicians are susceptible to non-scientific pressures (tradition, limited training, ideology) that guide their clinical decidions, while physicians are simply pragmatics who practice outside the scope of scientific evidence only when no alternatives exist. But, is this true?
This past weekend I was reading Pediatrics, the journal of the American Academy of Pediatrics, and encountered an article suggesting that psychiatrist may also be susceptible to non-scientific pressures in their practices. In the article, a group of researchers from Harvard University examined physicians’ use of off-label medications when such medications enter the market. Specifically, the authors examined the use of Ziprasidone (Geodon) with children after Ziprasidone, an atypical anti-psychotic, was approved for use with adults in the early 2000s. During the same years, other atypical antipsychotics, such as risperidon, were already FDA approved for use with children. This allowed the authors of the study to evaluate how often, and under what conditions, physicians would use Ziprasidone with children despite having another similar medication already approved for the use of children.
Before I discuss the results of the study, let me review what are usually considered “acceptable” reasons for the use of off-label medications. One reason is when no empirically supported alternative exist. For example, psychiatrists may use an adult antipsychotic with children if there is no child antipsychotic available. The off-label use of a medication would also be acceptable if the alternative medication is not working. For example, if a physician prescribes an approved medication and this medication does not result in symptom reduction, the physician may then try an off-label medication. Finally, if a physician prescribes an approved medication and this medication results in severe side effects in the child, then the physician may move on and try an off label drug with the hope that this medication provides relief without the side effects.
In the Pediatrics study, the authors examined the prescription practices in Michigan during the early 2000’s by reviewing the Medicaid data for the year following the approval of ziprasidone for adults (2001), which also represents the first time the drug was available for of-label use with children. The authors identified 292 persons under 21 who were prescribed ziprasidone during that year. Close to 60% of these individuals were under the age of 18. Here are some interesting results:
What diagnoses were provided to these individuals?
36% were provided a diagnosis of a psychotic disorder
32% were provided a diagnosis of another mental health disorder
32% were not provided a mental health disorder diagnosis at all
How many kids were prescribed a different antipsychotic before being prescribed ziprasidone?
33% were not prescribed any antipsychotic before ziprasidone.
54% were prescribed only 1 antipsychotic before ziprasidone.
Only 12 % were prescribed 2 or more antipsychotics before ziprasidone.
Who provided the prescription?
36% were general practitioners
36% were psychiatrists
13% were emergency room physicians
A couple of things I found surprising. First, we know that the use of ziprasidone in this sample was not simply due to the lack of alternative because other antipsychotics were already approved for children during this time. That is, physicians used ziprasidone despite the availability of similar medications that had been approved by the FDA for use with children. So the next question is whether ziprasidone was used after the child did not respond to FDA-approved medications, otherwise known as treatment resistance, which is defined as failure with at least 2 previous medications. It seems that only 12% of the sample met treatment resistance criteria. That is, only 12% tried at least 2 other medications before trying ziprasidone. For the other 88%, ziprasidone was prescribed before properly ‘trying’ at least 2 other medications, and for 33% of these patients, ziprasidone was their first medication (being prescribed before trying any child-approved medication!).
Is it possible that ziprasidone was prescribed because it was perceived to be even safer than child-approved medications? The authors indicated that this was not likely the case because at that time there were no published studies on the safety of ziprasidone with children, and there was evidence of dangerous cardiac side effects with adults, which should have been a red flag for these physycians. Therefore, these physicians used ziprasidone with no knowledge as to whether this medication was safe for kids (we now know that ziprasidone is safe as it eventually got FDA approval for children in 2009).
So what explains the rapid and seemingly careless adoption of ziprasidone as a drug of choice for the treatment of psychosis in children before there was any evidence of the drug efficacy or safety with that population?
The authors argued that this early adoption was simply due to marketing. Ziprasidone was the new kid in the block. The manufacturer spent 2.3 million in early marketing before the drug was available and $47 million in advertisement directly to practitioners the year the drug became available. It seems the ad campaign worked.
What this suggests is that physicians do engage in non-evidence based practices but not necessarily because of a lack of alternatives. Instead, just as psychologists, physicians are also susceptible to non-scientific pressures that impact their practices to the detriment of their patients. Psychologists are susceptible to tradition and rigid ideological beliefs. Physicians are susceptible to shinny pens and coffee mugs with flashy ziprasidone logos.
Penfold RB, Kelleher KJ, Wang W, Strange B, & Pajer K (2010). Pediatric uptake of a newly available antipsychotic medication. Pediatrics, 125 (3), 475-82 PMID: 20142282