BACKGROUND: Patients with bipolar disorder typically present to physicians in the depressed rather than the manic or hypomanic phase of illness. Because the depressive episodes in bipolar disorder may be indistinguishable from those in major depressive disorder (MDD), misdiagnosis may occur.OBJECTIVES: To estimate from administrative claims data and a telephone survey the prevalence of potential misdiagnosis of bipolar disorder among patients with MDD and the humanistic (health-related quality of life [HRQOL] and disability) effects associated with misdiagnosis in a managed care setting.METHODS: Administrative claims data were used to identify patients with medical claims for MDD from a database of 9 million members of commercial health plans from 3 U.S. regions. The inclusion criteria were as follows: Patients with ICD-9-CM codes for bipolar disorder at any time throughout the study period (January 1, 2000, through March 31, 2005) were excluded from this cohort. This cohort was targeted for a telephone survey that was conducted from August 1 through October 30, 2006. From the telephone survey sampling frame of 5,777, a total of 1,360 interviews were completed for a response rate of 23.5%. Respondents were screened for potential bipolar disorder using the Mood Disorder Questionnaire (MDQ). The Medical Outcomes 12-Item Short Form Survey (SF-12), Version 2, a widely used and validated instrument that assesses health-related functioning, and the Sheehan Disability Scale (SDS), which measures depression-related disability, were administered to a convenience subsample of 112 survey respondents to collect HRQOL and disability information, respectively. RESULTS: Of 1,360 adult patients aged 18 years or older with a diagnosis of MDD but without a medical claim for diagnosis of bipolar disorder, 94 (6.9%) screened positive for bipolar disorder on the MDQ. More patients with a positive screen for bipolar disorder reported lifetime histories of obsessive compulsive disorder (24.5% vs. 8.2%, P < 0.001), psychotic disorders or hallucinations (9.6% vs. 2.4%, P < 0.001), suicidal ideation (61.7% vs. 29.4%, P < 0.001), and drug abuse (34.0% vs. 11.1%, P < 0.001) than did patients with a negative screen for bipolar disorder. In the subgroup of patients who completed the SF-12 and SDS, patients with a positive screen for bipolar disorder (n = 33) had lower scores (i.e., greater impairment) on the social functioning, role emotional, and overall mental component summary scales of the SF-12 than did patients with a negative screen for bipolar disorder (n = 79, P < 0.001), but did not significantly differ on the physical component summary scale. Patients with a positive screen for bipolar disorder on the MDQ were more likely than patients who screened MDQ-negative to report severe depression-related impairment (scores of 7 and higher on the SDS scale) with work life (54.5% vs. 24.1%, respectively, P = 0.002), social life (66.7% vs. 39.2%, P = 0.008), and family life (66.7% vs. 34.2%, P = 0.002) on the SDS. CONCLUSIONS: In t...