Misdiagnosis of Bipolar Disorder

Bipolar disorder is both prevalent and disabling. Surveys suggest that patients with bipolar disorder are often misdiagnosed on initial presentation, most often with major depressive disorder. These patients may receive ineffective treatment, which, in some cases, actually worsens outcome, either by inducing manic or mixed states or by increasing mood cycling. Major contributors to misdiagnosis include incomplete history and lack of patient insight as well as presence of psychiatric comorbidity, such as anxiety or substance use disorders. Careful screening for current and past symptoms of mania or hypomania, as well as close clinical follow-up, can help to reduce misdiagnosis.

(Am J Manag Care. 2005;11:S271-

S274)

Bipolar disorder is a brain disease, or group of diseases, associated with periods of depressed mood as well as periods of particularly elevated or irritable mood. Although generally episodic, it is nearly always recurrent and lifelong. Recent epidemiologic data confirm that bipolar disorder is both chronic and disabling.¹ Bipolar disorder is associated with significant mortality; lifetime risk of death from suicide approaches 20 times that of the general population.² Likewise, bipolar disorder contributes substantially to disability as well as medical costs,³ ranking as the sixth-leading contributor to disability in 1990 World Health Organization figures.

A number of studies also suggest that patients with bipolar disorder are frequently misdiagnosed with other disorders. Two surveys among patients with bipolar disorder, conducted a decade apart, suggested little change in the rate of misdiagnosis.^4,5 The most recent survey⁴ noted that 69% of patients with bipolar disorder reported an initial misdiagnosis, with more than one third experiencing a delay of 10 years or greater before receiving a diagnosis of bipolar disorder. Likewise, a European survey of more than 1000 individuals with bipolar disorder found a mean time to correct diagnosis of 5.7 years.⁶

Of note, such a survey-based approach may overestimate rates of misdiagnosis for several reasons. First, it assumes that the current diagnosis is correct–that is, that the patient actually does have bipolar disorder. A problem in studying bipolar disorder, as with psychiatric illness in general, is the absence of a true gold standard. Second, response rates in these surveys were relatively low, increasing susceptibility to response bias. For example, patients with longer delays in diagnosis might be more likely to join an advocacy group or respond to a survey about treatment. Nonetheless, regardless of the actual rates, these surveys clearly establish that a substantial number of patients with bipolar disorder are misdiagnosed.

Consequences of Misdiagnosis

The consequences of misdiagnosis can be profound. In the absence of effective treatment, patients may experience a greater number of recurrences or more long-term episodes. Not surprisingly, both of these can have profound effects on patient functioning as well as medical costs. Recurrent mood episodes can substantially impair patients' ability to maintain relationships as well as education and employment. Moreover, even after recovery, the episodes may have enduring and cumulative consequences–for example, a patient who loses jobs because depression makes it impossible to get to work on time, or because manic episodes lead to conflict with coworkers or even legal involvement, may find it increasingly difficult to find employment. Family members, friends, or partners may grow tired of unpredictable moods or early morning crises.

In some cases, misdiagnosis may contribute to iatrogenic injury. Debate continues over the magnitude of risk associated with antidepressant treatment in bipolar disorder.^7,8 A subset of patients with bipolar disorder exposed to antidepressants may experience "switch"–that is, induction of a manic or mixed state (simultaneous symptoms of mania and depression). With older antidepressants, such as tricyclics–desipramine or nortriptyline, for example–the risk may be particularly great. Some risk of mania induction exists with newer antidepressants, but the magnitude is difficult to estimate and probably somewhat less.⁸

A second and less widely-appreciated risk with antidepressants is cycle acceleration.^9,10 In this condition, patients experience more frequent episodes than before antidepressant initiation. Without careful monitoring, cycle induction is not always readily apparent–patients may describe a positive antidepressant response initially but loss of response some months later, which may lead to dose increase or use of additional antidepressants. A typical patient might describe episodes every 2 to 3 years before beginning antidepressant treatment, with episodes every year thereafter, despite an "excellent" initial response to treatment.

Finally, although the link between effective treatment and reduction in suicide risk is difficult to establish definitively, by depriving patients of an effective treatment, an opportunity to decrease suicide risk may be missed.¹¹

Contributors to Misdiagnosis

The particular factors that contribute to misdiagnosis in bipolar disorder have not been studied rigorously, but some of them appear to be relatively straightforward. To begin with, patients may provide poor or uneven history, particularly during acute mood episodes. Some patients with depression will state that they have "always" been depressed or cannot recall ever feeling better. Likewise, mildly elevated patients may fail to report important symptoms (such as racing thoughts or a decreased need for sleep) if they fail to perceive them as pathological. In some cases, fear of stigma may lead patients to deliberately underreport symptoms of mania or hypomania: for many, a diagnosis of depression is much more palatable than bipolar disorder. This concern occasionally extends to clinicians as well, who may be reluctant to make this diagnosis in the absence of absolute certainty.

An additional source of diagnostic complexity is psychiatric comorbidity, which is the norm rather than the exception among patients with bipolar disorder.^1,12 More than 50% experience at least 1 comorbid anxiety disorder, including generalized anxiety or panic disorder. Some of the features of anxiety and depression or hypomania/mania may overlap–for example, impaired concentration can be associated with all 3, as can sleep disruption. Anxious patients, as well as manic patients, may report racing thoughts. Thus, a clinician who stops with a patient's chief complaint of anxiety may miss more subtle symptoms of bipolar disorder.

Other common comorbidities in patients with bipolar disorder are substance abuse and dependence, which are often present at initial presentation.¹³ Alcohol and stimulants can produce symptoms that mimic mood episodes; for patients with ongoing substance use, it can therefore be difficult to discern the presence of an underlying mood disorder.

Finally, some misdiagnosis likely arises from the lag between episodes. Patients may experience more subtle mood elevation (ie, hypomania) between or overlapping with depressive episodes. In some cases, patients may experience depressive episodes without a manic episode for 5 years or more thereafter–for the first few years of their illness, they are therefore "misdiagnosed" with recurrent major depressive disorder.^14,15

Improving Diagnosis of Bipolar Disorder

In many cases, careful history-taking allows recognition of bipolar disorder. Two aspects are particularly important: the use of collateral informants, such as friends or family members wherever possible, and systematic inquiry about symptoms of depression, hypomania, and mania. For hypomania, some clinicians begin by looking for any period of abnormally elevated or irritable mood or change in behaviors, such as driving, spending, or sexual activity. These are all quite nonspecific, but if any such period can be defined, additional questions can focus on typical associated symptoms.

Diagnostic and Statistical Manual

of Mental Disorders, 4th Edition (DSMIV)

Although several diagnostic instruments exist, none can replace careful diagnostic evaluation.¹⁶ A helpful tool in the diagnostic process can be a patient-rated instrument, such as the Mood Disorder Questionnaire.^17-19 This set of questions includes typical symptoms of mania and hypomania based on the criteria; a typical question asks whether there has ever been a period of time when "you were much more social or outgoing than usual, for example, you telephoned friends in the middle of the night?" Obviously, although not sufficient to diagnose bipolar disorder in and of itself, this instrument can lay the groundwork for more detailed follow-up questions to clarify diagnosis.

DSM-IV

Although the diagnostic criteria for mania have been demonstrated to have high interrater reliability–ie, 2 clinicians faced with a manic patient applying criteria would be likely to arrive at the same diagnosis–those for hypomania show somewhat less reliability.²⁰ Apart from the use of structured interviews or questionnaires, perhaps the best opportunity to improve diagnosis is through careful follow-up. Over time, hypomania or mood cycling may become more apparent, and the possible deleterious effects of antidepressants may be detected. Some form of mood charting¹⁶ by patients can also help where initial diagnosis is unclear.

The phase of illness that presents the most difficulty in diagnosis is typically depression: patients present in a depressive episode, and the clinician must assess the likelihood of bipolar disorder versus major depressive disorder. Unfortunately, although many studies have tried to identify clinical features of bipolar depression that distinguish it from major depressive disorder, none are particularly specific. For example, a number of studies describe reverse neurovegetative symptoms (increased sleep and appetite, rather than insomnia and loss of appetite) as being more common in bipolar depression.^1-9 Irritability^10,11 and anger^12,13 may also be suggestive of bipolar disorder. However, it is important to emphasize that all of these symptoms are quite prevalent among patients with major depressive disorder as well.

Similarly, certain features of illness course have been suggested to indicate bipolarity. Bipolar disorder typically has earlier age at onset than major depression.¹⁴ Indeed, when it presents in childhood or adolescence it may be particularly difficult to identify, because mood episodes may be less distinct (with persistent mood lability or irritability more common), or symptoms may be attributed to attention-deficit/hyperactivity disorder. Bipolar disorder is also highly familial: having a first-degree relative with bipolar disorder certainly increases a patient's likelihood of having bipolar disorder by approximately 7-fold. On the other hand, bipolar patients frequently have unipolar family members and vice versa.^15,16 Finally, highly recurrent depression, a history of abrupt antidepressant response followed by impersistence of response, or a history of numerous failed antidepressant trials, has been suggested to indicate bipolarity, although the evidence supporting this assumption is limited.

In sum, none of these individual clinical features are diagnostic for bipolarity. Rather, their presence should increase the clinician's suspicion of bipolar disorder and lead to even greater scrutiny (both in terms of past history and future course) for manic and hypomanic symptoms.

Consequences of Overdiagnosis

Of course, any diagnostic tool for bipolar disorder represents a tradeoff between sensitivity and specificity. Decreasing the rate of underdiagnosis generally entails some increase in the rate of false-positives–in this case, patients with major depressive disorder treated for bipolar disorder. Risks include exposure to a greater medication burden (in some cases requiring additional monitoring) as well as lesser likelihood of clinical improvement. Unfortunately, the tolerability of most bipolar pharmacotherapies is relatively poor compared with antidepressants, with possible medical consequences including diabetes and weight gain.²¹ Moreover, the evidence in controlled trials that bipolar pharmacotherapies, such as atypical antipsychotics or valproate, are effective antidepressants for individuals with major depressive disorder is generally lacking, with the possible exception of lithium.^22-24 Having committed a patient to mood stabilizer treatment, clinicians are often loathe to reverse themselves, so a misdiagnosis is likely to persist. However, as with any other diagnosis in psychiatry, periodic reconsideration of both diagnosis and treatment regimen is warranted.

Conclusion

The accurate discrimination of bipolar disorder from major depressive disorder, with which it is most often confused, is receiving increasing attention as its clinical and economic consequences are recognized. Ultimately, the use of biological markers–brain imaging or genetic tests, for example–may facilitate early and accurate diagnosis. While such markers are developed, careful clinical evaluation remains the most useful tool for recognizing bipolar disorder.

Download PDF: Misdiagnosis of Bipolar Disorder