The trusted source for
healthcare information and
Abstracts & Commentary
Synopsis: A prospective study comparing the 3 most commonly used SSRIs for initial treatment of depressed primary care outpatients found no difference in their effectiveness for depressive symptoms, adverse effects, or discontinuation rates.
Sources: Kroenke K, et al. JAMA. 2001;286:2947-2955; Simon G. JAMA. 2001;286:3003-3004.
To evaluate the effectiveness of paroxetine, fluoxetine, and sertraline in actual primary care offices, 2 non-profit volunteer primary care research networks, involving 8600 physicians across the country and an academic clinical organization of 150 physicians, were organized to enroll patients over an 8-month period and follow them for 9 months. In order to duplicate "real-world practice" as much as possible, the decision to initiate antidepressant treatment was based strictly on the physician’s judgment rather than any research scale. Choice of medications was not blinded, and physicians and patients were allowed to change doses, switch medications, or discontinue them as they would in a typical clinical setting.
Six hundred patients were eventually randomized from 63 sites to begin treatment with 1 of the 3 SSRI antidepressants (paroxetine 20 mg, fluoxetine 20 mg, or sertraline 50 mg). Computer-assisted telephone interviews were conducted at baseline and 1, 3, 6, and 9 months later. Study patients received a small stipend in compensation, and interviews were successfully completed in 94% of patients at 1 month, down to 79% at 9 months. Depression was assessed with several measures, including the SF-36 Medical Outcomes Study Mental Component Summary and other psychological outcome measurement scales. Social and work function and other health-related quality-of-life measures were also recorded along with actual medication use. Patients with other medical problems or substance abuse were excluded.
All of the SSRI groups had marked improvement in measures of depression and other health-related symptoms; patients who met criteria for major depression by interview dropped from 74% at baseline to 26% at 9 months, and reports of anxiety attacks declined from 35% to 14%. There was no difference in outcome among the 3 medication groups; recovery was present at 9 months for 81% in the paroxetine group, 77% in the fluoxetine group, and 84% in the sertraline group. These were not substantially different whether patients continued with the same initial antidepressant throughout the study or were switched or dropped out. Statistical analyses did not reveal any differences in patient characteristics or other biases. More than 80% of patients were satisfied with their treatment and their physician’s interest, and this did not differ among the 3 medication groups.
The proportion who stopped or switched to another antidepressant were 13% at 1 month, 32% at 6 months, and 40% at 9 months, and did not differ by drug group. Adverse effects such as GI complaints or insomnia were the most common reason, and also did not differ by choice of medication. Specific questions on 4 areas of sexual function did not reveal any significant concerns or differences among the medications, and in fact showed a slight improvement under treatment rather than worsening.
Comment by Mary Elina Ferris, MD
Despite the claims of drug manufacturers that one SSRI is preferable to another, this groundbreaking study conducted in real primary care offices fails to demonstrate any clear difference in effectiveness among the 3 most commonly prescribed antidepressants. They also failed to find any other patient characteristics that would predict which of the 3 antidepressants would work better in particular situations; no conclusion could be drawn that a specific drug was better for any patient factor.
However, this does not invalidate the clinical necessity to change dosage and switch medications to obtain the best outcome. Nearly 20% of patients in this study switched medications 1 or more times, and in the end two thirds of the patients who began treatment had recovered at the 9-month follow-up, which had been demonstrated in previous academic studies.1
An accompanying editorial in the same issue makes the important point that "equal on average does not mean equal for everyone." Patients who do not respond to the initial SSRI antidepressant will frequently benefit from another in the same class, and side effects may also necessitate a medication change to ensure compliance. New research is suggesting individual genetic variability in liver enzyme metabolism and serotonin transporter proteins, which may enable a more scientifically based approach to drug selection in the future.
In the meantime, there is no evidence-based approach to predict effectiveness in the initial choice of SSRIs, but there is a need to frequently use the other SSRIs when adjusting treatment. The editorial points out that restrictive managed care formularies may be justified indicating a low-cost initial SSRI, but that the other SSRIs clearly need to be available for the predictable frequent medication changes needed for maximal patient benefit.
1. Simon GE. JAMA. 1996;275:1897-1902.
Dr. Ferris, Clinical Associate Professor, University of Southern California, is Associate Editor of Internal Medicine Alert.