Current issues of ACP Journal Club are published in Annals of Internal Medicine


Fluoxetine worked as well as imipramine or desipramine for depression in primary care

ACP J Club. 1997 Jan-Feb;126:16. doi:10.7326/ACPJC-1997-126-1-016

Source Citation

Simon GE, VonKorff M, Heiligenstein JH, et al. Initial antidepressant choice in primary care. Effectiveness and cost of fluoxetine vs tricyclic antidepressants. JAMA. 1996 Jun 26;275:1897-902.



To compare the effectiveness and health care costs of initial treatment with fluoxetine with those of imipramine or desipramine in patients with depression.


Randomized, unblinded, controlled trial with a 6-month duration.


Primary care clinics of Group Health Cooperative (a staff model health maintenance organization [HMO]) in Seattle, Washington, USA.


536 patients (median age 41 y, 72% women) who were beginning anti-depressant treatment for depression. Exclusion criteria were use of antidepressant drugs in the previous 90 days, current alcohol abuse, history of mania, current psychotic symptoms, recent use of an antipsychotic medication, or pregnancy.


Patients were allocated to fluoxetine (n = 173), desipramine (n = 181), or imipramine (n = 182). Initial dose, dose changes, and treatment duration for all study drugs were determined by the treating physicians as in usual practice.

Main outcome measures

Scores on the Hamilton Depression Rating Scale (HDRS), the depression subscale of the Hopkins Symptom Checklist (HSCL), and the Medical Outcomes Study SF-36 questionnaire and costs of antidepressant drug use and health care utilization.

Main results

Analysis was by intention to treat. Patients assigned to receive fluoxetine were more likely to continue their original medication (P < 0.001), were more likely to reach adequate doses of study drugs (P = 0.02), and reported fewer adverse effects (P < 0.001) than did patients assigned to receive imipramine or desipramine. The adjusted mean score on the HDRS (combined mean response at 3 and 6 mo) was 8.86 (95% CI 8.31 to 9.41) for patients in the fluoxetine group and did not differ from the adjusted mean scores in the desipramine (8.56, CI 8.01 to 9.11) and imipramine (9.13, CI 8.60 to 9.66) groups (P = 0.49). No differences existed among the treatment groups on the HSCL or the SF-36 questionnaire. Total health care costs over the 6-month follow-up period were similar for all treatment groups.


In primary care, initial treatment with fluoxetine was as effective as initial treatment with desipramine or imipramine in relieving depressive symptoms. No differences in health care costs were seen among the 3 treatments.

Source of funding: Lilly Research Laboratories.

For article reprint: Dr. G.E. Simon, Center for Health Studies, 1730 Minor Avenue, #1600, Seattle, WA 98101-1448, USA. FAX 206-287-2871.


Meta-analyses have shown that selective serotonin-reuptake inhibitors (SSRIs), such as fluoxetine, are about as efficacious as tricyclic antidepressants (TCAs) but produce fewer adverse effects (1, 2). SSRIs also require less dose titration, but they are more expensive than TCAs. Despite the higher cost, decision analyses suggest that SSRIs may be more cost-effective than TCAs (3). The study by Simon and colleagues confirms that in primary care practice, SRIs are associated with better compliance and fewer side effects. These benefits, however, did not translate into better depression outcomes or decreased total health care costs.

The results of this well-designed study are strengthened by the high follow-up rate that was achieved and the careful tracking of health care utilization that was done. The intention-to-treat analysis is complicated by the large and unequal numbers of patients in the 3 groups who crossed over from their initial medication assignments (48% in the desipramine group, 43% in the imipramine group, 20% in the fluoxetine group), potentially minimizing the differences for depression outcomes.

When applying the results of this study, one should consider the patient population and study setting. First, 7% of the study population was aged ≥ 65 years. For older patients who have more coexisting medical illness, the different side effect profiles associated with the different anti-depressant agents may need to be con-sidered. Second, the study was done in an HMO setting that had low copayments for medications; this suggests that these results are best applied to settings in which patients pay low medication fees.

How should we use this information? SSRIs and TCAs have similar effectiveness for important depression outcomes. Health care plans should make both classes of drugs available to physicians for the initial treatment of depression. When choosing an antidepressant drug, physicians should consider the patient's past response to antidepressant drugs, coexisting medical illness, and out-of-pocket drug costs.

John W. Williams Jr., MD
Audie L. Murphy Memorial Veterans Administration HospitalSan Antonio, Texas, USA

John W. Williams Jr., MD
Audie L. Murphy Memorial Veterans Administration Hospital
San Antonio, Texas, USA


1. Anderson IM, Tomenson BM. Treatment discontinuation with selective serotonin reuptake inhibitors compared with tricyclic antidepressants: a meta-analysis. BMJ. 1995; 310:1433-8.

2. Depression Guideline Panel. Depression in Primary Care: Clinical Practice Guidelines. Rockville, MD: U.S. Department of Health and Human Services, Public Health Service, Agency for Health Care Policy and Research; 1993.

3. Jonsson B, Bebbington PE. What price depression? The cost of depression and the cost-effectiveness of pharmacological treatment. Br J Psychiatry. 1994;164:665-73.