Психотипные рецепты у пациентов с биполярным расстройством до установленного диагностирования биполярного расстройства.

Introduction

Bipolar I Disorder (BP‐I) is a psychiatric disorder that is characterized by chronic mood episodes of mania that often alternate with major depressive episodes (National Institute of Mental Health, 2024). Differentiating the symptomatic presentation of BP‐I from other common mental health disorders, particularly major depressive disorder (MDD), is challenging and often results in a broad differential diagnosis due to the complexity of presentation and multiple subjective data points involved in establishing a definitive diagnosis. BP‐I is frequently misdiagnosed as MDD because patients are more likely to seek office visits during their depressive episodes compared to their manic episodes [ 1 ]. One study showed that in a primary care clinic, 21% of patients being treated for depression screened positive for bipolar disorder [ 2 ]. Major contributions toward misdiagnosis also include lapses in history‐taking, and presence of psychiatric and medical comborbidities [ 3 ].

Diagnosis of BP‐I is often delayed, with 6.7 years elapsing prior to a conclusive diagnosis being established, oftentimes leading to negative health outcomes due to delayed, ineffective or inappropriate treatments being used [ 4 ]. For example, misdiagnosed patients may be at risk of receiving an antidepressant monotherapy, which can induce a hypomanic or manic episode in patients with BP‐I [ 5 ]. The standard of care for BP‐I patients is an atypical antipsychotic (AA) or mood stabilizer, though an antidepressant may be used in combination with these to address unresolved depressive symptoms [ 6 ]. One study examined the medical costs of misdiagnosed patients to be $6541 greater per patient‐year compared to those initially diagnosed with BP‐I, on account of higher hospitalizations, ER visits, and outpatient visits [ 7 ]. Improving diagnosis of BP‐I can alleviate some of the burden associated with BP‐I, which is estimated to be $219 billion [ 8 ].

While previous studies focused on the burden of misdiagnosing BP‐I, current literature lacks an in‐depth understanding of the treatment patterns during the misdiagnosis stage and the impact that treatment types have on the time until a definitive BP‐I diagnosis. In addition, the impact of MDD diagnosis prior to BP‐I on BP‐I treatment regimen is unknown. Hence, in this study, we aim to analyze psychotropic treatment patterns among BP‐I patients prior to a conclusive BP‐I diagnosis.

Methods

Study Design and Data Source

We conducted a retrospective cohort study, using health insurance claims data, to assess treatment patterns and patient characteristics during the misdiagnosis of BP‐I.

The MarketScan Commercial Claims and Encounters database was used to identify two cohorts: (1) BP‐I patients with the history of being misdiagnosed with MDD prior to BP‐I diagnosis; (2) BP‐I patients without the history of being misdiagnosed with MDD. This database contains de‐identified, patient‐specific health data of reimbursed healthcare claims for employees, retirees, and their dependents of over 250 medium and large employers and health plans, approximately 30 million individuals are covered under private insurance plans. No Medicaid or Medicare data are included.

The study period was from January 1, 2018 through December 31, 2022. The index period, defined to be the date of MDD or BP‐I diagnosis of the study cohorts, was from January 1, 2019, through October 1, 2022, to allow for 12 months of continuous enrollment before the index date and 3 months of follow‐up.

At the time of the analysis, MarketScan data met Health Insurance Portability and Accountability Act of 1996 (HIPAA) requirements for fully de‐identified data sets (IBM MarketScan Research Databases 2022). Institutional Review Board (IRB) approval at the University of Washington was not required, as the study met criteria for non‐human subjects' research as specified by the Human Subjects Division at the University of Washington.

Sample Selection

Misdiagnosed Cohort

Primary group of interest was patients with initial MDD diagnosis who were later diagnosed with BP‐I. We included patients who were at least 18 years old, and who were diagnosed with MDD (ICD‐10‐CM: Appendix 1 ), defined as ≥ 1 inpatient OR ≥ 2 outpatient medical diagnosis followed by ≥ 1 inpatient OR ≥ 2 outpatient medical claims with a diagnosis of BP‐I (ICD‐10‐CM: Appendix 1 ). MDD and BP‐I diagnosis defined the first and second index date in this cohort, respectively. We excluded all patients with prior diagnosis of schizophrenia before the first index date. We included patients with continuous enrollment in medical and prescription benefits 12 months prior to MDD diagnosis, throughout misdiagnosis, and 3 months after BP‐I diagnosis (Appendix 2 ).

Initially BP ‐I Diagnosed Cohort

Our comparator group was BP‐I without history of MDD misdiagnosis. For initially BP‐I diagnosed patients, we included patients who were at least 18 years old and who were diagnosed with BP‐I, defined as ≥ 1 inpatient OR ≥ 2 outpatient medical claims. We defined the index date to be the date of BP‐I diagnosis in this group. We excluded patients with prior diagnosis of MDD and/or schizophrenia before index dates. We required continuous enrollment in medical and prescription benefits 12 months prior and 3 months after BP‐I diagnosis.

Study Measures and Outcomes

We described the baseline characteristics including patient age, sex, geographic region, insurance plan type, healthcare setting of diagnosis, provider type at diagnosis, Elixhauser comorbidity score, and Diagnostic and Statistical Manual of Mental Disorders (DSM‐5) comorbidity score on the date of BP‐I diagnosis. Elixhauser scores measure overall severity of comorbidities (Appendix 3 ; Agency for Healthcase Research and Quality). DSM‐5 index is a measure of comorbid psychiatric disorders (Appendix 4 ; [ 9 ]). We calculated a modified Elixhauser and DSM‐5 Comorbidity Index, which excluded depressive and bipolar disorders, using ICD‐10 diagnosis codes on medical claims collected during the 12‐month pre‐index period.

The main outcomes of interest were treatment patterns observed during misdiagnosis and assessing the association between treatment type and time until BP‐I conversion among the misdiagnosed cohort. We investigated the five most common treatment regimens that were used during misdiagnosis, which were identified as either monotherapy (1 medication) or combination therapy (2+ medications) from the following seven psychotropic drug classes: (1) antidepressants; (2) anxiolytics; (3) mood stabilizers; (4) psychostimulants; (5) atypical antipsychotics; (6) sedative/hypnotics; and (7) typical antipsychotics. We grouped anticonvulsants with mood stabilizers because they are frequently used as mood stabilizers. Our primary focus was the percentage of patients who received antidepressant monotherapy. Additionally, we compared the first regimen after BP‐1 diagnosis between the two cohorts.

Statistical Analysis

Baseline characteristics and treatment patterns were calculated using means and standard deviations for continuous variables and frequency ( n , %) for all categorical variables.

We assessed the association between the first treatment regimen after MDD diagnosis and patient and provider characteristics in the misdiagnosed cohort by employing multinomial logistic regression. Patient, provider and clinical covariates include age, sex, place of diagnosis, provider type, DSM‐5 comorbidity index score, and Elixhauser comorbidity index score. We defined the first treatment after MDD diagnosis to be all treatments given within 14 days of diagnosis, with at least one subsequent refill taken for ≥ 30 days. We defined the last treatment of the misdiagnosis period as the final medication being taken in the 14 days leading up to BP‐I diagnosis. Additionally, we expanded the multinomial logistic regression analysis to the entire study population (misdiagnosed and initially BP‐I diagnosed cohort) and included misdiagnosis history as an additional covariate to assess the impact of MDD misdiagnosis prior to BP‐I diagnosis on the first treatment regimen after BP‐I diagnosis.

In the misdiagnosed group we employed Cox proportional hazards regression analysis to calculate the differential hazard rate ratio of BP‐I diagnosis based on the first treatment regimen while adjusting for patient and provider characteristics. The outcome variable is the diagnosis of BP‐I (binary). The first treatment regimen is a categorical variable which included the five most common psychotropic drug class combinations and an others category at MDD diagnosis: (1) Antidepressant monotherapy (reference level), (2) Antidepressant + Anxiolytics, (3) Anxiolytics monotherapy, (4) Antidepressant + Mood Stabilizer, (5) Mood stabilizer monotherapy, and(6) Others. Patients and provider covariates include age, sex, place of diagnosis, provider type, DSM‐5 comorbidity index score, and Elixhauser comorbidity index score.

We performed cohort selection and descriptive statistics using SAS version 9.4 (SAS Institute, Cary, NC), and all statistical analyses in R 4.4.0 (R Core Team, 2024).

Results

Baseline Characteristics

We identified 21,771 eligible BP‐I patients who were previously misdiagnosed with MDD (Table 1 ). The average age at MDD diagnosis was 34.8 years old and the average age at BP‐I diagnosis was 36.3 years old. Females comprised 68.1% of the misdiagnosed cohort. In this cohort, 78.0% of the MDD diagnosis prior to BP‐I diagnosis were in an outpatient setting though that number dropped to 64.9% at BP‐I diagnosis. Acute care providers were the most frequent to diagnose MDD (18.4%), followed by family practice (15.4%), and psychiatrists (14.8%). After excluding depressive disorders from the Elixhauser comorbidity index average score was 0.34, with 27% of patients having at least one comorbidity and 6.1% having 2 or more. After excluding depressive and bipolar disorders from the DSM‐5 comorbidity index, the average score was 0.60, with 43.6% having at least one comorbidity and 14.3% having 2 or more. In contrast, our initially BP‐I diagnosed cohort had 50,667 individuals and average age was 37.1 years old with females making up 65% of this population. Psychiatrists were the most frequent to diagnose BP‐I (25.4%) in this cohort. The average Elixhauser comorbidity index was 0.29, while the average DSM‐5 comorbidity index was 0.51.

Association Between Patient and Provider Characteristics and First Treatment Regimen After MDD Diagnosis in the Misdiagnosed Group

Younger patients are more likely to receive an antidepressant monotherapy ( p < 0.001) than all other psychotropic drug classes as a first treatment regimen (Table 2 ) in the misdiagnosed group. Males had a 13% decrease in odds of receiving an antidepressant and mood stabilizer combination therapy instead of an antidepressant monotherapy compared to females (OR = 0.87, p < 0.05). The odds of receiving a mood stabilizer monotherapy instead of antidepressant monotherapy are 61% lower when prescribed by family practice (OR = 0.39, p < 0.001), 37% lower when prescribed by a nurse practitioner (OR = 0.63, p < 0.01), and 45% lower when prescribed by an internal medicine provider (OR = 0.55, p < 0.001) compared to when prescribed by a psychiatrist.

Association Between Misdiagnosis History and First Treatment Regimen After BP ‐I Diagnosis

Prescribing patterns after BP‐I diagnosis were substantially different depending on whether the patients initially were diagnosed with MDD (Table 3 ). Compared to those who were misdiagnosed with MDD, patients initially diagnosed with BP‐I were 84% more likely (OR = 1.84, p < 0.001) to be prescribed an antidepressant and a mood stabilizer combination therapy. They also had 2.63 times the likelihood (OR = 2.63, p < 0.001) of being prescribed mood stabilizer monotherapy over antidepressant monotherapy.

Treatment Patterns and Time From MDD Misdiagnosis to BP ‐I Conversion

In our misdiagnosed cohort, we identified the 13 most common treatment regimens at MDD diagnosis (Appendix 5 ). Among these, antidepressant was the most common, with 25.1% of patients starting on an antidepressant monotherapy, and 8.7% continuing to take it after BP‐I diagnosis (Figure 1 ). Surprisingly, in our initially BP‐I diagnosed cohort, 7.6% received an antidepressant monotherapy at index. At MDD diagnosis, only 8.7% of patients were on 3+ unique drug classes, and after BP‐I diagnosis that number rose to 16.6% (Figure 2 ). At BP‐I diagnosis, AA monotherapy was utilized in approximately 4.3% of the misdiagnosed cohort and 9.4% of the initially BP‐I diagnosed cohort. Similarly, mood stabilizers were used in about 6.4% of the misdiagnosed cohort and 11.1% of the initially BP‐I diagnosed cohort, despite both being primary guideline‐recommended therapies. The five most common treatment regimens in both of our cohorts were Antidepressant monotherapy, Antidepressant + Anxiolytic, Antidepressant + Mood Stabilizer, Mood Stabilizer monotherapy, and Atypical Antipsychotics.

In the misdiagnosed group, time until BP‐I conversion varied based on their initial treatment regimen (Figure 3 ). Mood stabilizer monotherapy was associated with the highest hazard compared to antidepressant monotherapy (1.26, 95% CI: 1.19–1.34, p < 0.001), implying that mood stabilizer monotherapy as the first post‐MDD diagnosis treatment is associated with an earlier BP‐I conversion after misdiagnosis (Table 4 ). Compared to females, males had a significantly higher rate of BP‐I conversion (1.09, 95% CI: 1.06–1.12, p < 0.001). Additionally, higher comorbidity scores were associated with a reduced rate of BP‐I conversion. Specifically, the DSM‐V comorbidity score was associated with a rate ratio of 0.95 (95% CI: 0.93–0.97, p < 0.001), while Elixhauser comorbidity score had a rate ratio of 0.98 (95% CI: 0.95–1.00, p < 0.05), suggesting that higher comorbidity may delay the time to BP‐I diagnosis.

Discussion

We conducted a retrospective cohort study to describe patient characteristics, treatment patterns and assess time until BP‐I conversion in patients who were initially misdiagnosed with MDD. We found that most patients were initially diagnosed with MDD in an outpatient setting (78.0%) though that number was lower (64.9%) at the conversion to BP‐I. Consistent with that finding, we had seen over a 2‐fold increase in converted BP‐I diagnoses in an inpatient setting and through telehealth, which may be explained by a more severe event taking place that led to BP‐I diagnosis (i.e., manic episode leading to a hospitalization; [ 10 ]). The increase in telehealth diagnosing may also be due to our study timeline overlapping with the Covid‐19 pandemic.

Bipolar‐I disorder presents a diagnostic conundrum in clinical practice, often masquerading as MDD during initial presentation. Previous literature has highlighted liabilities with misdiagnosis, particularly the risk of initiating ineffective treatments, such as antidepressant monotherapy in individuals with BP‐I [ 11 ]. Literature has shown that antidepressant monotherapy, which 25% of our misdiagnosed cohort was taking at their MDD diagnosis, lacks efficacy in treating bipolar depression [ 12 ]. Our study underlines these findings, emphasizing the need for accurate identification of BP‐I to mitigate risks associated with inappropriate treatments. We noticed that even after BP‐I diagnosis following an initial MDD diagnosis, patients were still receiving treatments that do not align with evidence‐based treatment guidelines, including antidepressant monotherapy [ 13 ]. Our study also revealed that BP‐I patients with a history of MDD diagnosis are more likely to continue antidepressant monotherapy than those initially diagnosed with BP‐I.

Given the risk of mania associated with antidepressant monotherapy, it was initially expected that patients would convert to BP‐I more quickly [ 14 ]. Contrarily, recent literature has shown that antidepressant‐induced mania is negligible [ 15 ]. We found that patients on antidepressant monotherapy had faster time to BP‐I conversion compared to those taking antidepressants in combination with either anxiolytics or mood stabilizer, and those taking mood stabilizer monotherapy. This outcome might be explained by patients with an MDD diagnosis displaying clinical manifestations or histories indicative of a potential BP‐I diagnosis. Additionally, this study did not differentiate the MDD cohort by specifiers, and according to the DSM‐5, mixed features associated with a major depressive episode have been found to be a significant risk factor for the development of bipolar I or bipolar II disorder [ 16 ]. Consequently, patients within this cohort who were already at increased risk for BP‐I may have more frequently received mood stabilizer monotherapy instead of antidepressant monotherapy.

There are several limitations of our study. First, the diagnosis of BP‐I is not possible until the first presence of manic or mixed symptoms, thus many patients first present with an acute depressive episode, without experiencing a manic episode [ 17 ]. This can be indicative that some patients were not misdiagnosed through poor assessment, but that their illness trajectory had not converted to a definitive BP‐I diagnosis yet. Additionally, the study's criteria may not fully reflect real‐world clinical settings, and the timeline may not have been long enough to capture the full duration of misdiagnosis, which existing literature estimates to average around 6.7 years. Importantly we grabbed snapshots in time via claims, and therefore cannot make conclusion about what was happening in between these time points. Finally, this analysis was on full‐time employees with employer sponsored private insurance which may not be generalizable to the full U.S. population. This is especially important in our population because low levels of household income are associated with several lifetime mental disorders including bipolar disorder [ 18 ].

BP‐I can be extremely challenging due to the complexity of presentation and subjective diagnostic criteria, though there are more common presentations of symptoms in BP‐I patients [ 19 , 20 ]. For example, earlier onset of first depressive episode, shorter depressive episodes, and family history of BP‐I are all factors that can guide clinicians toward a BP‐I diagnosis. Previous literature also reveals disparities in screening tools between MDD and BP‐I. Most healthcare providers rely on tools to screen for depression but do not use any to assess BP‐I [ 21 ]. There is a need for more frequent use of screening tools like the Mood Disorder Questionnaire (MDQ) or the Rapid Mood Screener (RMS) in patients suspected to have MDD to increase the likelihood of capturing BP‐I. Our study highlights the need to correctly diagnose BP‐I early and efficiently. By prioritizing initial correct diagnosis of BP‐I, we can mitigate the risks associated with misdiagnosis, optimize treatment selection, and improve overall patient care.

Conclusion

Among the 21,771 BP‐I patients who were misdiagnosed with MDD, a significant proportion (28.5%) were started on antidepressant monotherapy at MDD diagnosis, in contrast to the 11.2% in the initially BP‐I diagnosed cohort of 50,667 patients. Additionally, 13.3% continued this treatment even after an established BP‐I diagnosis. About 1% of misdiagnosed patients received an Atypical Antipsychotic When these patients received mood stabilizer monotherapy they had a 26% higher hazard for earlier onset of a BP‐I diagnosis (HR: 1.26, 95% CI: 1.19–1.34, p < 0.001).

Odds of receiving the first treatment regimen also significantly differed based on initial diagnosis. Patients initially correctly diagnosed with BP‐I had 61% greater odds of being prescribed anxiolytic monotherapy (OR = 1.61, p < 0.001), 84% greater odds of being prescribed an antidepressant combined with a mood stabilizer (OR = 1.84, p < 0.001), and were 2.63 times more likely to be prescribed mood stabilizer monotherapy (OR = 2.63, p < 0.001) compared to those initially misdiagnosed with MDD.

Acknowledgments

We would like to thank Mousam Parikh for her valuable contributions to this work. We also acknowledge the support of AbbVie for facilitating this research by providing access to MarketScan data, which was instrumental in conducting our analyses. Special thanks to Lakshmi Kandukuri and Shihan Jin for their insight and feedback throughout this project.

Data Availability Statement

The data that support the findings will be available in IBM MarketScan Commercial Claims and Encounters Database at https://www.ibm.com/products/marketscan%E2%80%90research%E2%80%90databases following an embargo from the date of publication to allow for commercialization of research findings.

References

1 C. de Almeida , J. Renner , and M. L. Phillips , “ Distinguishing Between Unipolar Depression and Bipolar Depression: Current and Future Clinical and Neuroimaging Perspectives ,” Biological Psychiatry 73 , no. 2 ( 2013 ): 111 – 118 , 10.1016/j.biopsych.2012.06.010 . 22784485 PMC3494754

2 R. M. Hirschfeld , A. R. Cass , D. C. Holt , and C. A. Carlson , “ Screening for Bipolar Disorder in Patients Treated for Depression in a Family Medicine Clinic ,” Journal of the American Board of Family Practice 18 , no. 4 ( 2005 ): 233 – 239 . 15994469 10.3122/jabfm.18.4.233

3 T. Singh and M. Rajput , “ Misdiagnosis of Bipolar Disorder ,” Psychiatry (Edgmont) 3 , no. 10 ( 2006 ): 57 – 63 . PMC2945875 20877548

4 J. Scott , A. Graham , A. Yung , C. Morgan , F. Bellivier , and B. Etain , “ A Systematic Review and Meta‐Analysis of Delayed Help Seeking, Delayed Diagnosis, and Duration of Untreated Illness in Bipolar Disorders ,” Acta Psychiatrica Scandinavica 146 , no. 5 ( 2022 ): 389 – 405 . 36018259 10.1111/acps.13490

5 T. Singh and M. Rajput , “ Misdiagnosis of Bipolar Disorder ,” Psychiatry (Edgmont (Pa.: Township)) 3 , no. 10 ( 2006 ): 57 – 63 . PMC2945875 20877548

6 K. R. Connolly and M. E. Thase , “ The Clinical Management of Bipolar Disorder: A Review of Evidence‐Based Guidelines ,” Primary Care Companion for CNS Disorders 13 , no. 4 ( 2011 ): PCC.10r01097 , 10.4088/PCC.10r01097 . PMC3219517 22132354

7 R. S. McIntyre , F. Laliberté , G. Germain , S. D. MacKnight , P. Gillard , and A. Harrington , “ The Real‐World Health Resource Use and Costs of Misdiagnosing Bipolar I Disorder ,” Journal of Affective Disorders 316 ( 2022 ): 26 – 33 , 10.1016/j.jad.2022.07.069 . 35952932

8 M. Cloutier , M. Greene , A. Guerin , M. Touya , and E. Wu , “ The Economic Burden of Bipolar I Disorder in the United States in 2015 ,” Journal of Affective Disorders 226 ( 2018 ): 45 – 51 , 10.1016/j.jad.2017.09.011 . 28961441

9 R. C. Kessler , W. T. Chiu , O. Demler , K. R. Merikangas , and E. E. Walters , “ Prevalence, Severity, and Comorbidity of 12‐Month DSM‐IV Disorders in the National Comorbidity Survey Replication ,” Archives of General Psychiatry 62 , no. 6 ( 2005 ): 617 – 627 . 15939839 10.1001/archpsyc.62.6.617 PMC2847357

10 L. Culpepper , “ The Diagnosis and Treatment of Bipolar Disorder: Decision‐Making in Primary Care ,” Primary Care Companion for CNS Disorders 16 , no. 3 ( 2014 ): PCC.13r01609 , 10.4088/PCC.13r01609 . PMC4195640 25317368

11 L. Culpepper , “ The Role of Primary Care Clinicians in Diagnosing and Treating Bipolar Disorder ,” Primary Care Companion to the Journal of Clinical Psychiatry 12 , no. 1 ( 2010 ): 4 – 9 , 10.4088/PCC.9064su1c.01 . PMC2902189 20628500

12 M. M. Sidor and G. M. Macqueen , “ Antidepressants for the Acute Treatment of Bipolar Depression: A Systematic Review and Meta‐Analysis ,” Journal of Clinical Psychiatry 72 , no. 2 ( 2011 ): 156 – 167 , 10.4088/JCP.09r05385gre . 21034686

13 CAMH , “Mood Stabilizing Medications,” ( 2012 ), https://www.camh.ca/en/health‐info/mental‐illness‐and‐addiction‐index/mood‐stabilizing‐medication .

14 R. Patel , P. Reiss , H. Shetty , et al., “ Do Antidepressants Increase the Risk of Mania and Bipolar Disorder in People With Depression? A Retrospective Electronic Case Register Cohort Study ,” BMJ Open 5 , no. 12 ( 2015 ): e008341 , 10.1136/bmjopen-2015-008341 . PMC4679886 26667012

15 C. Rohde , S. D. Østergaard , and O. H. Jefsen , “ A Nationwide Target Trial Emulation Assessing the Risk of Antidepressant‐Induced Mania Among Patients With Bipolar Depression ,” American Journal of Psychiatry 181 , no. 7 ( 2024 ): 630 – 638 , 10.1176/appi.ajp.20230477 . 38946271

16 American Psychiatric Association , Diagnostic and Statistical Manual of Mental Disorders , 5th ed. ( American Psychiatric Association , 2013 ), 10.1176/appi.books.9780890425596 .

17 N. Shah , S. Grover , and G. P. Rao , “ Clinical Practice Guidelines for Management of Bipolar Disorder ,” Indian Journal of Psychiatry 59 , no. 1 ( 2017 ): S51 – S66 , 10.4103/0019-5545.196974 . 28216785 PMC5310104

18 J. Sareen , T. O. Afifi , M. M. KA , and A. GJG , “ Relationship Between Household Income and Mental Disorders: Findings From a Population‐Based Longitudinal Study ,” Archives of General Psychiatry 68 , no. 4 ( 2011 ): 419 – 427 , 10.1001/archgenpsychiatry.2011.15 . 21464366

19 L. N. Yatham , T. Chakrabarty , D. J. Bond , et al., “ Canadian Network for Mood and Anxiety Treatments (CANMAT) and International Society for Bipolar Disorders (ISBD) Recommendations for the Management of Patients With Bipolar Disorder With Mixed Presentations ,” Bipolar Disorders 23 ( 2021 ): 767 – 788 . 34599629 10.1111/bdi.13135

20 P. B. Mitchell , G. M. Goodwin , G. F. Johnson , and R. M. Hirschfeld , “ Diagnostic guidelines for bipolar depression: a probabilistic approach ,” Bipolar Disorders 10 ( 2008 ): 144 – 152 , 10.1111/j.1399-5618.2007.00559.x . 18199233

21 M. E. Thase , S. M. Stahl , M. I. RS , et al., “ Screening for Bipolar I Disorder and the Rapid Mood Screener: Results of a Nationwide Health Care Provider Survey ,” Primary Care Companion for CNS Disorders 25 , no. 2 ( 2023 ): 22m03322 . 10.4088/PCC.22m03322 37115145

ICD‐10 Codes for Bipolar‐I Disorder and Major Depressive Disorder.

ICD‐10‐CM Code Code description BP‐I Disorder F30 Manic episode F30.1 Manic episode without psychotic symptoms F30.10 Manic episode without psychotic symptoms, unspecified F30.11 Manic episode without psychotic symptoms, mild F30.12 Manic episode without psychotic symptoms, moderate F30.13 Manic episode, severe, without psychotic symptoms F30.2 Manic episode, severe with psychotic symptoms F30.3 Manic episode in partial remission F30.4 Manic episode in full remission F31 Bipolar disorder F31.0 Bipolar disorder, current episode hypomanic F31.1 Bipolar disorder, current episode manic without psychotic features F31.10 Bipolar disorder, current episode manic without psychotic features, unspecified F31.11 Bipolar disorder, current episode manic without psychotic features, mild F31.12 Bipolar disorder, current episode manic without psychotic features, moderate F31.13 Bipolar disorder, current episode manic without psychotic features, severe F31.2 Bipolar disorder, current episode manic severe with psychotic features F31.3 Bipolar disorder, current episode depressed, mild or moderate severity F31.30 Bipolar disorder, current episode depressed, mild or moderate severity, unspecified F31.31 Bipolar disorder, current episode depressed, mild F31.32 Bipolar disorder, current episode depressed, moderate F31.4 Bipolar disorder, current episode depressed, severe, without psychotic features F31.5 Bipolar disorder, current episode depressed, severe, with psychotic features F31.6 Bipolar disorder, current episode mixed F31.60 Bipolar disorder, current episode mixed, unspecified F31.61 Bipolar disorder, current episode mixed, mild F31.62 Bipolar disorder, current episode mixed, moderate F31.63 Bipolar disorder, current episode mixed, severe, without psychotic features F31.64 Bipolar disorder, current episode mixed, severe, with psychotic features F31.7 Bipolar disorder, currently in remission F31.71 Bipolar disorder, in partial remission, most recent episode hypomanic F31.72 Bipolar disorder, in full remission, most recent episode hypomanic F31.73 Bipolar disorder, in partial remission, most recent episode manic F31.74 Bipolar disorder, in full remission, most recent episode manic F31.75 Bipolar disorder, in partial remission, most recent episode depressed F31.76 Bipolar disorder, in full remission, most recent episode depressed F31.77 Bipolar disorder, in partial remission, most recent episode mixed F31.78 Bipolar disorder, in full remission, most recent episode mixed Major Depressive Disorder F32 Major depressive disorder, single episode F32.0 Major depressive disorder, single episode, mild F32.1 Major depressive disorder, single episode, moderate F32.2 Major depressive disorder, single episode, severe without psychotic features F32.3 Major depressive disorder, single episode, severe with psychotic features F32.4 Major depressive disorder, single episode, in partial remission F32.5 Major depressive disorder, single episode, in full remission F32.9 Major depressive disorder, single episode, unspecified F33 Major depressive disorder, recurrent F33.0 Major depressive disorder, recurrent, mild F33.1 Major depressive disorder, recurrent, moderate F33.2 Major depressive disorder, recurrent severe without psychotic features F33.3 Major depressive disorder, recurrent, severe with psychotic symptoms F33.4 Major depressive disorder, recurrent, in remission F33.40 Major depressive disorder, recurrent, in remission, unspecified F33.41 Major depressive disorder, recurrent, in partial remission F33.42 Major depressive disorder, recurrent, in full remission F33.9 Major depressive disorder, recurrent, unspecified Schizophrenia F20 Schizophrenia F20.0 Paranoid schizophrenia F20.1 Disorganized schizophrenia F20.2 Catatonic schizophrenia F20.3 Undifferentiated schizophrenia F20.5 Residual schizophrenia F20.8 Other schizophrenia F20.81 Schizophreniform disorder F20.89 Other schizophrenia F20.9 Schizophrenia, unspecified F25 Schizoaffective disorders F25.0 Schizoaffective disorder, bipolar type F25.1 Schizoaffective disorder, depressive type F25.8 Other schizoaffective disorders F25.9 Schizoaffective disorder, unspecified Note: A detailed list of ICD‐10 diagnosis codes used to identify BP‐I and MDD in the study population.

Study Population Attrition and Cohort Selection.

Flowchart and accompanying table detailing the criteria for inclusion/exclusion at each stage of cohort selection, including the number of patients removed at each step.

Elixhauser Comorbidity Index: ICD‐10 Codes.

ICD‐10‐CM Code Code description Congestive heart failure I09.9, I11.0, I13.0, I13.2, I25.5, I42.0, I42.5–I42.9, I43, I50, P29.0 Cardiac arrhythmias A52.0, I05–I08, I09.1, I09.8, I34–I39, Q23.0–Q23.3, Z95.2–Z95.4 Valvular disease A52.0, I05–I08, I09.1, I09.8, I34–I39, Q23.0–Q23.3, Z95.2–Z95.4 Pulmonary circulation disorders I26, I27, I28.0, I28.8, I28.9 Peripheral vascular disease I70, I81, I73.1, I73.8, I73.9, I77.1, I79.0, I79.2, K55.1, K55.8, K55.9, Z95.8, Z95.9 Hypertension (combine uncomplicated and complicated) Hypertension, uncomplicated: I10 Hypertension, complicated: I11–I13, I15 Paralysis G04.1, G11.4, G80.1, G80.2, G81, G82, G83.0–G83.4, G83.9 Other neurological disorders G10–G13, G20–G22, G25.4, G25.5, G31.2, G31.8, G31.9, G32, G35–G37, G40, G41, G93.1, G93.4, R47.0, R56 Chronic pulmonary disease I27.8, I27.9, J40–J47, J60–J67, J68.4, J70.1, J70.3 Diabetes without chronic complications E10.0, E10.1, E10.9, E11.0, E11.1, E11.9, E12.0, E12.1, E12.9, E13.0, E13.1, E13.9, E14.0, E14.1, E14.9 Diabetes with chronic complications E10.2–E10.8, E11.2–E11.8, E12.2–E12.8, E13.2–E13.8, E14.2–E14.8 Hypothyroidism E00–E03, E89.0 Renal failure I12.0, I13.1, N18, N19, N25.0, Z49.0–Z49.2, Z94.0, Z99.2 Liver disease B18, I85, I86.4, I98.2, K70, K71.1, K71.3–K71.5, K71.7, K72–K74, K76.0, K76.2–K76.9, Z94.4 Chronic peptic ulcer disease K25.7, K25.9, K26.7, K26.9, K27.7, K27.9, K28.7, K28.9 HIV and AIDS B20–B22, B24 Lymphoma C81–C85, C88, C96, C90.0, C90.2 Metastatic cancer C77–C80 Solid tumor without metastasis C00–C26, C30–C34, C37–C41, C43, C45–C58, C60–C76, C97 Rheumatoid arthritis/collagen vascular diseases L94.0, L94.1, L94.3, M05, M06, M08, M12, M12.3, M30, M31–M31.3, M32–M35, M45, M46, M46.8, M46.9 Coagulation deficiency D65–D68, D69.1, D69.3–D69.6 Obesity E66 Weight loss E40–E46, R63.4, R64 Fluid and electrolyte disorders E22.2, E86, E87 Blood loss anemia D50.0 Deficiency anemias D50.8, D50.9, D51–D53 Alcohol abuse F10, E52, G62.1, I42.6, K29.2, K70.0, K70.3, K70.9, T51, Z50.2, Z71.4, Z72.1 drug abuse F11–F16, F18, F19, Z71.5, Z72.2 psychoses F20, F22–F25, F28, F29, F30.2, F31.2, F31.5 Depression F20.4, F31.3–F31.5, F32, F33, F34.1, F41.2, F43.2 Note: A comprehensive list of ICD‐10 codes used to define the comorbidities included in the Elixhauser Comorbidity Index, which measures the overall burden of comorbid conditions in the study population.

DSM‐5 Comorbidity Index: Psychiatric Disorders.

Sleep–wake disorders G47.00, G47.09, G47.10, G47.19, G47.8, G47.9, G47.419, G47.411, G47.429, G47.33, G47.31, R06.3, G47.37, G47.34, G47.35, G47.36, G47.21, G47.22, G47.23, G47.24, G47.26, G47.20, F51.3, F51.4, F51.5, G47.52, G25.81 Anxiety disorders F93.0, F94.0, F40.218, F40.228, F40.230, F40.231, F40.232, F40.233, F40.248, F40.298, F40.10, F41.0, F40.00, F41.1, F06.4, F41.8, F41.9 Trauma‐ and stressor‐related disorders F94.1, F94.2, F43.10, F43.0, F43.21, F43.22, F43.23, F43.24, F43.25, F43.20, F43.9, F43.8 Neurodevelopmental disorders F70, F71, F72, F73, F79, F88, F80.9, F80.0, F80.89, F80.81, F80.9, F84.0, F90.2, F90.1, F90.8, F90.9, F90.0, F81.0, F81.81, F81.2, F82, F98.4, F95.2, F95.1, F95.0, F95.8, F95.9, F89 Medication‐induced movement disorders and other adverse effects of medication G21.11, G21.19, G21.0, G24.02, G24.09, G25.71, G25.79, G24.01, G25.1, T43.205A, T43.205D, T43.205S, T50.905A, T50.905D, T50.905S Bipolar and related disorders F31.11, F31.12, F31.13, F31.2, F31.73, F31.74, F31.9, F31.0, F31.31, F31.32, F31.4, F31.5, 31.75, F31.76, F31.9, F31.81, F31.89, F34.0, F06.33, F06.34, F06.31, F06.32, F31.9 Neurocognitive disorders F05, R41.0, G31.84, G31.9, F02.81, F02.80, F01.50, F01.51, R41.9 Obsessive‐compulsive and related disorders F42, F63.2, L98.1, F06.8 Sexual dysfunctions F52.32, F52.21, F52.22, F52.31, F52.6, F52.0, F52.4, F52.8, F52.9 Schizophrenia spectrum and other psychotic disorders F21, F22, F23, F20.81, F20.9, F25.0, F25.1, F06.2, F06.0, F06.1, F28, F29 Depressive disorders F34.8, F32.0, F32.1, F32.2, F32.3, F32.4, F32.5, F32.9, F33.0, F33.1 F33.2, F33.3, F33.41, F33.42, F33.9, F34.1, N94.3, F32.8, F32.9 Elimination disorders F98.0, F98.1, N39.498, R32, R15.9 Personality disorders F60.0, F60.1, F60.3, F60.4, F60.81, F60.6, F60.7, F60.5, F07.0, F60.89, F60.9 Somatic symptom and related disorders F45.1, F45.9, F45.21, F45.22, F44.4, F44.5, F44.6, F44.7, F54, F68.10, F45.8 Dissociative disorders F44.81, F44.0, F44.1, F48.1, F44.89, F44.9 Feeding and eating disorders F98.3, F50.8, F98.21, F50.01, F50.02, F50.2, F50.9 Disruptive, impulse‐control, and conduct disorders F91.3, F63.81, F91.1, F91.2, F63.3, F91.9, F91.8, F60.2, F63.1, F91.9 Gender dysphoria F64.2, F64.8, F64.9, F64.1 Paraphilic disorders F65.3, F65.2, F65.81, F65.89, F65.51, F65.52, F65.4, F65.0, F65.1, F65.9 Gambling disorder F63.0 Other mental disorders F09, F99 Other conditions that may be a focus of clinical attention Z62.820, Z62.891, Z62.29, Z63.8, Z62.898, Z63.0, Z63.5, Z63.4, T74.12XA, T74.12XD, T76.12XA, T76.12XD, T74.22XA, T74.22XD, T76.22XA, T76.22XD, T74.02XA, T74.02XD, T76.02XA, T76.02XD, T74.32XA, T74.32XD, T76.32XA, T76.32XD, Z69.010, Z69.020, Z69.011, Z62.812, Z62.811, Z91.412, Z91.411, T74.11XA, T74.11XD, T76.11XA, T76.11XD, T74.31XA, T74.31XD, T76.31XA, T76.31XD, T74.21XA, T74.21XD, T76.21XA, T76.21XD, T74.01XA, T74.01XD, T76.01XA, T76.01XD, Z69.11, Z69.81, Z69.12, Z70.9, Z69.81, Z69.82, Z69.021, Z62.810, Z91.410, Z55.9, Z56.82, Z56.9, Z59.0, Z59.1, Z59.2, Z59.3, Z59.4, Z59.5, Z59.6, Z59.7, Z59.9, Z60.0, Z65.4, Z65.8, Z64.4, R41.83 V, Z60.2, Z60.3, Z60.4, Z60.5, Z65.0, Z65.1, Z65.2, Z65.3, Z64.0, Z64.1, Z65.5, Z91.82, Z60.9, Z65.9, Z71.9, Z91.49, Z91.5, Z91.89, Z72.9, Z72.811, Z72.810, Z75.3, Z75.4, Z91.19, E66.9, Z76.5, Z91.83 Mental retardation F70, F71, F72, F73, F74, F75, F76, F78, F79 Note: A list of DSM‐5 psychiatric comorbidities (e.g., anxiety disorders, neurodevelopmental disorders) included in the study's comorbidity analysis, with corresponding ICD‐10 codes.

Extended Treatment Patterns by Regimen and Diagnosis Stage.

Misdiagnosed ( n = 21,771) Control ( n = 50,667) 1st Regimen after MDD(%) Final regimen of MDD(%) 1st Regimen after BP‐I(%) 1st Regimen after BP‐I(%) Treatment Regimen Antidepressant 25.1 14.0 8.7 7.6 Antidepressant + Anxiolytic 7.7 4.3 3.1 2.3 Antidepressant + Mood Stabilizer 5.5 6.7 6.8 6.5 Mood stabilizer 5.1 8.3 6.4 11.1 Anxiolytic 5.5 6.6 4.3 5.6 Antidepressant + Atypical Antipsychotic 4.6 6.1 6.0 4.9 Atypical Antipsychotic 4.3 7.4 6.0 9.4 Antidepressant + Anxiolytic + Mood Stabilizer 3.0 4.1 5.4 3.8 Psychostimulant 2.7 2.7 2.7 3.2 Antidepressant +Psychostimulant 2.3 1.7 2.0 1.3 Anxiolytic + Mood Stabilizer 2.1 3.3 3.3 4.0 Atypical Antipsychotic + Mood Stabilizer 1.6 3.2 4.5 6.0 Sedative/Hypnotic 0.7 0.6 0.5 0.7 Other 29.8 31.0 40.3 33.6 Note: *Other consisted of 5000+ potential combinations of psychotropics for the seven drug classes: Antidepressant, Anxiolytics, Mood Stabilizers, Psychostimulants, Atypical Antipsychotics, Typical Antipsychotics, and Sedative/Hypnotics. This appendix provides a detailed breakdown of psychotropic treatment patterns in patients misdiagnosed with MDD and those correctly diagnosed with BP‐I, comparing treatment changes before and after BP‐I diagnosis.