1118 — Rate and Predictors of Tricyclic Antidepressant Monotherapy to Patients with Bipolar I Disorder
Schutte KK, Hu KU, and Balasubramanian V, Palo Alto COE; Schmitt SK, HERC; Dally SK, Palo Alto COE; Phibbs CS, HERC;
The prescription of TriCyclic Antidepressant medication (TCA) to patients with BiPolar I disorder (BPI), especially when prescribed as an Antidepressant Monotherapy (ADM) without an accompanying mood stabilizer medication, has been associated with deleterious patient outcomes. We determined the rate and predictors of this risky healthcare practice.
Utilizing MedSAS, Fee Basis, and DSS Pharmacy inpatient and outpatient data, we identified six unique cohorts of pharmacy-using newly-diagnosed BPI patients. Cohort members attained sufficient pieces of evidence (POE) of BPI (at least one inpatient or two outpatient, primary, or secondary 296.0x-1x, 296.4x-296.7 ICD-9-CM diagnoses) within each fiscal year FY03-FY08. For each cohort, we evaluated the rate of TCA prescription in the year subsequent to attaining BPI POE. Next, we identified the FY08 start date for all “new” TCA ADM prescriptions (“new” = no TCA coverage for at least 90 preceding days). We identified factors associated with new TCA ADM prescriptions.
Each new cohort of BPI patients comprised, on average, 15,300 unique individuals. The rate of TCA prescription (1 or more/30 or more days’ coverage) among patients with newly diagnosed BPI decreased over time, from 8.12%/4.14% for the FY03 cohort to 6.25%/3.22% for the FY08 cohort. Among those prescribed TCAs, most received TCA ADM part of the time. We identified n = 1,299 BPI patients who received a new FY08 TCA ADM prescription. These prescriptions occurred more often among patients who, in the prior 7 days, were treated within non-psychiatric rather than in specialty psychiatric settings (66% v. 34%, respectively). Examination of diagnoses proximal (14 days plus/minus) to the TCA ADM prescription date revealed that only 50% of patients had any depressive diagnosis. Among the 33% who had no proximal psychiatric diagnosis, the most frequently recorded diagnoses were for pain-related conditions (e.g., lumbago). Utilizing more conservative criteria for defining BPI (minimum two POE; no subsequent schizophrenia diagnosis) did not change study conclusions.
Efforts to reduce TCA ADM to BPI patients should target prescription of TCAs for non-depressive conditions, including off-label TCA prescription, especially outside specialty psychiatric settings.
Results provide guidance for reducing use of a healthcare practice that has the potential to do harm.