Background: This study utilizing pharmacoepidemiologic methods and und
ertaken to determine the prescribing patterns of antidepressants parti
cularly in bipolar depression. Method: From pharmacy records of the Mc
Lean Hospital, the number of patients receiving antidepressants and gi
ven electroconvulsive therapy (ECT) from June 1, 1987, to May 8, 1993,
was determined. We later linked these data bases with patients who we
re diagnosed with DSM-III-R bipolar depression (296.5) during the same
period of time. Results: During the 6-year period, it was determined
that 3829 inpatients had received tricyclic antidepressants (TCAs), 29
81 fluoxetine, 2603 trazodone, 809 bupropion, 743 monoamine oxidase in
hibitors (MAOIs), 592 stimulants, 588 sertraline, 48 paroxetine, and 8
94 ECT. There were significant increases over time in prescriptions of
MAOIs compared with fluoxetine (chi2 = 14.36, p = .006), and bupropio
n compared with TCAs (chi2 = 6.45, p = .04). There was a trend for bup
ropion to be prescribed more over time compared with fluoxetine (chi2
= 5.09, p = .08). There were no significant changes in the prescribing
of other antidepressants or in the use of ECT. Conclusion: At our cen
ter, prescribing of bupropion and MAOIs in bipolar depression has incr
eased significantly. This may be related to the reports in the literat
ure of the low switch rates to mania with the use of these drugs.