Research report
The impact of unrecognized bipolar disorders among patients treated for depression with antidepressants in the fee-for-services California Medicaid (Medi-Cal) program: A 6-year retrospective analysis

https://doi.org/10.1016/j.jad.2006.06.018Get rights and content

Abstract

Introduction

The cost of unrecognized bipolar disorders over time is unknown.

Methods

Ten years of data from the California Medicaid program were used to identify depressed patients initiating new episodes of antidepressant therapy and with 6+ years of post-treatment data. Recognized bipolar (RBP) patients received a BP diagnosis or used mood stabilizers in the pre-index period. Unrecognized bipolar (UBP) patients received an initial BP diagnosis or used a mood stabilizer in the post-index period. Depression-only (MDD) patients had no BP diagnosis or mood stabilizer use. Three analyses were conducted: (1) regression models of cost per year, (2) a regression model of aggregate cost over 6 years and (3) a time trend analysis of the costs for UBP patients.

Results

14,809 patients were identified: RBP 14.5%, UBP 28.2% and MDD 57.3%. The growth in costs per month for UBP patients over 6 years (171%) far exceeds the growth for RBP and MDD patients (82% and 95%, respectively). RBP and MDD patients cost $2316 and $1681 less per year in the 6th year relative to UBP patients (p < 0.0001 for both estimates). The cost per month increased by $91 for each month of delayed diagnosis (p = 0.011). Costs for UBP patients increased by $10 per month prior to their initial BP diagnosis (p < 0.001) and by − $1.01 thereafter (p = 0.006 for the change in slope).

Limitations

Classification of patients based on diagnosis or mood stabilizer use using paid claims data is inexact.

Conclusions

Early diagnosis of bipolar disorders may significantly reduce health care cost.

Introduction

Bipolar disorder (BP) is frequently misdiagnosed, taking an average of 8 years before the BP diagnosis is correctly established (Baldessarini et al., 1999). Lish et al. (1994) reported that 70% of BP patients were initially misdiagnosed and half of these patients had consulted three or more physicians or other professionals. Ghaemi et al. (1999) report that 40% of BP patients had previously received an incorrect diagnosis of major depression. Angst et al. (2002) reported that 25–50% of all major depression cases are bipolar, confirming an earlier estimate of 57% reported by Lish et al. (1994). The majority of patients initiate treatment while depressed (Li et al., 2002), likely due to the dominance of depressive symptoms (Judd et al., 2002) and the subjective distress associated with the depressed phase of bipolar disorders (Shi et al., 2002). Akiskal (2003) has argued that the unipolar–bipolar dichotomy found in official nosology fails to account for very common clinical and subclinical conditions that exist in the interface between major depressive disorders and bipolarity. Recent data indicate that the prevalence of bipolar spectrum disorder may be as high as 6.4% in the community (Judd and Akiskal, 2003).

The potential savings to the health care sector from efforts to speed the diagnosis of bipolar disorders may be significant. Misdiagnosis of BP as unipolar depression can cause substantial clinical and economic consequences (Birnbaum et al., 2003). Direct healthcare costs were significantly higher among patients who delayed or did not use mood stabilizers during their first year of bipolar disorder therapy (Li et al., 2002). Mood-stabilizing treatment may be less effective when initiated after unsuccessful treatment for several depressive episodes (Swann et al., 1999). Shi et al. (2004) found that recognized and unrecognized bipolar patients constitute 14.9% and 6.2% of all patients initiating an episode of antidepressant therapy for depression or bipolar disorders. Unrecognized bipolar patients were found to be at higher risk for attempting suicide were estimated to have increased costs over 1 year of $682 (p < 0.05) relative to recognized bipolar patients.

The objective of this study is expand on previous research on the cost of unrecognized bipolar disorder over a period of 6 years following an index prescription for an antidepressant. As in the previous analysis (Shi et al., 2004), treatment outcomes and costs are compared across recognized bipolar (RBP), unrecognized bipolar (UBP) patients and ‘depressed-only’ patients (MDD).

Section snippets

Data

Data were derived from the 100% paid claims files of the California Medicaid fee-for-services program (Medi-Cal) from the period October 1993 to January 2004. The de-identified patients for this study were selected if they had at least one paid claim with a depression-related diagnosis and at least one prescription for an antidepressant. The depression-related diagnoses used in this study include major depressive disorders (ICD9 codes: 296.2 to 296.3), other affective psychoses (296.9),

Baseline characteristics

The study sample consisted of 14,809 patients with 6 years of data following their index antidepressant episode. Table 1 presents the baseline demographic and prior use characteristics of the study sample across the three study groups. Nearly 43% of long-term Medi-Cal patients treated with antidepressant therapy were bipolar patients, either recognized (14.5%) or unrecognized (28.2%) on the date of the index antidepressant prescription. This is an over-estimate of the proportion of typical

Discussion

This study found significant costs associated with apparent delays in diagnosis for BP patients. These results were not sensitive to alternative methods of classifying patients initiating antidepressant therapy as an unrecognized BP patient. Moreover, while the use of health care services increased significantly in the month in which an UBP patient is diagnosed, the trend in cost per patient per month experienced prior to diagnosis is significantly reduced in the post-diagnosis period.

Acknowledgements

This research was funded by a grant from Eli Lilly and Company, makers of olanzapine and fluoxetine which are used to treat patients with bipolar disorders. Dr. Shi was an employee of Eli Lilly and Company during the initial phases of the research. The University of Southern California maintains publication rights to all findings derived from the research subject to time-limited review and comment by Eli Lilly and Company.

Cited by (49)

  • The real-world health resource use and costs of misdiagnosing bipolar I disorder

    2022, Journal of Affective Disorders
    Citation Excerpt :

    In the United States (US), the lifetime and 12-month prevalence of BP-I is estimated to be 1.0 % and 0.6 %, respectively (Merikangas et al., 2007). The economic burden of bipolar disorder is substantial (Keck et al., 2008), and studies have shown that costs are greater in patients with unrecognized or undiagnosed BP-I (Birnbaum et al., 2003; McCombs et al., 2007). The depressive phase is the most enduring and disabling feature of BP-I (Kupka et al., 2007; Miller et al., 2014), and patients often seek out treatment for depressive symptoms.

  • Classification of unipolar and bipolar depression using machine learning techniques

    2021, Psychiatry Research
    Citation Excerpt :

    Therefore, the presence of mania and hypomania episodes has low sensibility for differentiating between MDD and BD, especially in clinical practice and in non-psychiatric settings. Several issues can arise from the misdiagnosis between these disorders (McCombs et al., 2007). For instance, BD patients may be treated with antidepressant drug therapy, which presents lower treatment response compared to mood stabilizers, increased risk of a manic or hypomanic switch, and increased risk of rapid mood cycling (Bahji et al., 2020; Sidor and Macqueen, 2011).

  • Comparison of the screening ability between the 32-item Hypomania Checklist (HCL-32) and the Mood Disorder Questionnaire (MDQ) for bipolar disorder: A meta-analysis and systematic review

    2019, Psychiatry Research
    Citation Excerpt :

    Bipolar disorder (BD) is a chronic psychiatric illness frequently misdiagnosed in clinical practice (Carvalho et al., 2015). BD is often misdiagnosed as major depressive disorder (MDD) (Phillips and Kupfer, 2013; Culpepper, 2014), which could lead to inappropriate treatment (Calabrese et al., 2003) resulting in adverse consequences, such as increased suicide risk (McCombs et al., 2007), poorer response to antidepressants and antidepressant-induced switch to mania (Smith et al., 2008). On average it takes around 10 years to correctly establish the diagnosis of BD, and around one-third of patients with BD are misdiagnosed at least once (Lish et al., 1994; Drancourt et al., 2013).

  • Exploration of the psychometric properties of the 33-item Hypomania Checklist - external assessment (HCL-33-EA)

    2019, Journal of Affective Disorders
    Citation Excerpt :

    As bipolar disorder (BD) often begins with a depressive episode (Solomon et al., 2006), it is frequently misdiagnosed as major depressive disorder (MDD) (Culpepper, 2014). Misdiagnosis could result in detrimental clinical consequences (Culpepper, 2014), including inappropriate use of antidepressants (Smith et al., 2009) and increased suicidal risk (McCombs et al., 2007). Although standardized and structured or semi-structured diagnostic interviews, such as the Mini-International Neuropsychiatric Interview (MINI) and the Structured Clinical Interview for DSM (SCID), can reliably identify BD, these interviews are time-consuming and require trained personnel to administer (Zimmerman et al., 2004).

View all citing articles on Scopus
View full text