"Despite there being a lack of evidence of a significant difference in efficacy between older and newer agents, clinicians perceive the newer agents to be more efficacious than the older drugs," said lead author Timothy Petersen, Ph.D., a clinical instructor in psychology at Harvard Medical School and a researcher at the Depression Clinical and Research Program at Massachusetts General Hospital. The hospital’s program is headed by Maurizio Fava, M.D., who cowrote the study.
The survey of prescribing habits was given to attendees at the hospital’s psychopharmacology review course in fall 2000. Similar surveys have been given to attendees for several years.
The 2000 survey, which was published in the January issue of Progress in Neuro-Psychopharmacology and Biological Psychiatry, asked 10 questions of the nearly 800 attendees at the review course. More than 430 returned responses to the survey prior to the start of the course. Questions were geared toward three general areas: whether clinicians believe one class of antidepressants is more efficacious than the others, whether clinicians prefer to prescribe one class of antidepressants over the others, and which particular antidepressants the clinicians associated with particular side effects.
"Overall, the responses indicated that clinicians are using not only more sources of input, but more complex decision trees to make their prescribing decisions than would generally be thought," Petersen told Psychiatric News.
Perhaps the most revealing results were the answers to the first three questions. When asked whether they believed that one antidepressant is more efficacious than another, respondents split almost down the middle, with 51 percent saying no, 49 percent yes.
Of the respondents who indicated yes to the first question, half believed SSRIs are more efficacious, and a quarter believed an SNRI (venlafaxine) was more efficacious (see chart below). Interestingly, after responding in the first question that one antidepressant was more efficacious, 7 percent indicated multiple choices for which one when they answered the second question.
"In the third question, 93 percent of respondents indicated SSRIs when asked whether or not they prefer to prescribe one antidepressant over the others," Peterson said. "But the question asks about preference, not necessarily practice. Just because they’re indicating a preference for SSRIs doesn’t mean they are always prescribing them," he added.
Petersen noted that few head-to-head data exist comparing the efficacy of various antidepressants. He said that neither of the two most widely cited documents, the Agency for Health Care Policy and Research’s (AHCPR) "Treatment of Depression: Newer Pharmacotherapies," published in 1999, or APA’s Practice Guideline for the Treatment of Patients With Major Depressive Disorder, revised last year, support the overwhelming preference seen in the survey.
The AHCPR report concluded that, in its review of more than 300 randomized trials, there was "no evidence to suggest significantly different efficacy between classes of antidepressants or between individual antidepressant agents." APA’s guideline concludes, "The effectiveness of antidepressant medications is generally comparable between classes and within classes of medications." The guideline recommends that medication be selected using other criteria, such as side effects, cost, history of prior response to medications, and presence of comorbid psychiatric and/or general medical conditions.
Most of the strong preference indicated in the survey may be coming from multiple anecdotal accounts within various physician practices, Petersen believes. The majority of the physicians attending the review course surveyed, he noted, were private practice physicians who "often may not have access to or the time necessary to go through the case reports and studies that are so widely known in academic and research circles."
In addition, Petersen noted, most believe that side-effect profiles for SSRIs are much more favorable than for the older TCAs or MAOIs.
The remaining questions in the survey asked respondents about particular side effects commonly associated with antidepressants.
The majority of physicians responding to the survey believed mirtazapine (Remeron) was most likely to cause weight gain. Fluoxetine (Prozac) was considered most likely to cause sexual dysfunction (57 percent), followed by paroxetine (Paxil) (26 percent.)
When asked which antidepressant was most likely to cause a discontinuation syndrome, 48 percent responded paroxetine, and 29 percent said venlafaxine (Effexor).
More than half of those responding said fluoxetine causes agitation, while 26 percent cited bupropion (Wellbutrin).
Petersen noted that these results are a bit of a "mixed bag" with respect to matching the empirical evidence. Two of the respondents’ beliefs are backed up by research—the belief that mirtazapine is the most likely to cause weight gain and that paroxetine is more likely to cause a discontinuation syndrome.
However, he said, research has not distinguished fluoxetine in particular as likely to cause sexual dysfunction, although there is empirical evidence that SSRIs in general are more likely to be associated with this side effect. As for agitation, Petersen noted, a review of empirical literature found that no single antidepressant type is most likely to produce this side effect.
Petersen and his coauthors suggested that "scientific knowledge (empirical evidence) is not the only factor to consider when faced with the decision of what to prescribe a patient." They believe that the prescriber relies on at least four sources of data: accepted scientific knowledge, patient context, own experience, and situational context.
"Prescribing practices are influenced in many ways," Petersen told Psychiatric News, "by economics, by HMO or formulary restrictions, certainly by efficacy, but perhaps most by side effects and situational context. And I think marketing and detailing play a role. You don’t hear these prescribers talking a lot about [older tricyclics] in this survey."
Petersen stressed that future surveys should gather information at the level of the individual agent for all questions and should gather more detailed information about respondents’ characteristics.
An abstract of "A Survey of Prescribing Practices in the Treatment of Depression" is available on the Web through www.sciencedirect.com. ▪