|
A second study examined the charts of all inpatients
prospectively diagnosed with bipolar (n = 44) or schizoaffective
disorder (n = 4) by a psychiatrist with expertise in affective
disorders (12). These patients were diagnosed over 1 year, using
DSM-IV criteria. Patient interviews and chart reviews were used
to obtain referral diagnoses before hospitalization. Patients who
had not previously sought psychiatric treatment, or were currently
experiencing their first manic episode, were excluded. Nineteen
(40%) were identified as having BD previously misdiagnosed as unipolar
depression. Time to bipolar diagnosis after a patient’s first contact
with a mental health professional was 7.5 years (SD 9.8) in the
total sample (vs 0.9 years [SD 2.2] in 25 patients who had already
been diagnosed with BD). Mood stabilizers were underused and antidepressants
overused in this patient population; on admission, only 38% of the
total sample were taking mood stabilizers, and, notably, a similar
number (33%) were taking antidepressants. Thus, systematic application
of DSM-IV criteria identified previously undiagnosed BD in 40% of
a referred population of patients with mood disorders; all these
patients had previously been misdiagnosed with unipolar MDD. Because
the sample consisted only of BD I, the underdiagnosis of BD could
not be attributed to difficulty diagnosing hypomania.
A confirmation study was conducted, with a more detailed assessment
of natural history and the effects of antidepressants on illness
course (13). This outpatient study included patients with BD I as
well as BD II and BD not otherwise specified (NOS) (according to
Akiskal’s criteria of either hypomania or mania occurring only with
antidepressant use or a diagnosis of unipolar disorder and a first-degree
relative with BD I [14]). The study assessed 54 patients with BD
(BD I, n = 27; BD II, n = 11; BD NOS, n = 16)
and found that about 7 years elapsed between the first visit to
a mental health professional and the diagnosis of BD I. For BD II
or BD NOS patients, about 12 years elapsed between first visit and
diagnosis. In the total sample, major depressive episodes (MDEs)
occurred about 5 years earlier than manic episodes and were more
frequent than manic episodes. Patients spent about 50% of their
lives with depression, compared with 11% of their lives experiencing
manic or hypomanic symptoms. Of the sample, 57% had been diagnosed
with unipolar MDD before being diagnosed with BD. When the authors
controlled for patients who had received unipolar diagnoses due
to MDEs occuring before the first manic episodes, 37% of patients
were still misdiagnosed with unipolar MDD after the onset of their
first manic or hypomanic episode. This appears to be the first true
misdiagnosis rate established in a study of BD that took into account
a simultaneous assessment of natural history factors.
Clinician Failure to Recognize BD
As suggested by these previous studies, disparities in clinician
awareness of mania vs depression contribute to misdiagnosis. Sprock
conducted a study of 20 clinicians (mostly psychiatrists) at an
academic institution (15). To assess their diagnostic skill in distinguishing
schizoaffective disorder from other mood disorders, she asked the
clinicians to write all the symptoms of mania and depression that
they could recall in the 3 minutes allotted for each. The clinicians
displayed relatively greater knowledge of symptoms that are DSM
criteria for major depression: 18 clinicians described sleep disturbance,
17 decreased appetite, 15 suicidal ideation, 11 anhedonia, and 10
decreased weight and libido. Conversely, for manic symptoms only
7 clinicians reported euphoria and grandiosity, symptoms that can
be straightforwardly inferred as DSM criteria; 13 described sleep
disturbance and 12, decreased sleep, neither of which reflects the
exact criterion of decreased need for sleep. Twelve described depressed
mood (which is not required for mania), and 8 each described “energy
disturbance,” cycling, and spending sprees. Energy is not always
elevated in mania, cycling is a course criterion, and spending sprees
are a subtype of 1 criterion. Thus, fewer than one-half of the clinicians
described only 2 of the 7 cardinal DSM-IV manic criteria (euphoria
and grandiosity), compared with the fact that most clinicians recalled
most of the major depressive criteria. These results suggest that
clinicians’ ineffectual assessment of manic symptoms results in
misdiagnosis of patients.
Lack of Insight Into Manic Symptoms Among Patients
Apart from the shortcomings of clinicians’ diagnostic skills, patients’
lack of insight into mania also contributes to underdiagnosis of
BD. Empirical studies published specifically on insight in BD were
rare before 1994. Since then, however, 2 groups have noted that
lack of insight is almost as prominent in mania as in schizophrenia,
and it is less impaired in depression (16,17). Using different methods,
the DSM-IV field trials also demonstrated that lack of insight is
a major clinical finding in BD, one that is similar in severity
to that in patients with schizophrenia, and more severe than in
patients with psychotic depression (18). Because insight is more
impaired in mania than in depression, reliance on patient self-report
probably contributes to underdiagnosis of mania (as was alluded
to in the discussion of the Iowa 500 project) and relative overdiagnosis
of unipolar depression. Involving patients’ families and caregivers
in the diagnosis process and extending the collection of data beyond
the patient to third parties is a possible solution to this dilemma.
For example, in a study of prodromal symptoms of mania and depression,
families reported behavioural symptoms of mania more than twice
as frequently as patients (47% vs 22%) (19).
|
|
This finding did not hold for depression, where families and patients reported similar symptom rates. Hence, the obfuscating effects of patients’ impaired insight can be counteracted by obtaining family or third-party data (for example, from therapists, nurses, social workers, and residential staff). In our experience, most patients can identify at least 1 close family member or friend to whom they are willing to allow access for vital history taking. Lacking this, even the best psychiatric evaluations can be confounded by a patient’s impaired insight. Concerns about confidentiality may be raised, but it is important to set up an expectation from the very beginning that the patient is entering a medical relationship, in which access to third parties for information is vital to proper treatment. This contrasts with a purely psychotherapy relationship, in which outside contact is commonly avoided.
Is There a Bipolar Spectrum Beyond Type I Illness?
We have just reviewed evidence regarding underdiagnosis
or misdiagnosis, mostly of BD I. We wish to emphasize that the problem
of the misdiagnosis of BD occurs even with classic manic-depressive
illness, what Ketter has termed “Cade’s disease.” In addition, however,
there are possibly many less classic forms of bipolar illness, in
which spontaneous mania or hypomania do not occur.
For over 2 decades, “soft signs” of bipolarity have been studied
and discussed by Akiskal and others (14,20,21). A recent review
of 6 studies done since 1978 suggests that broadening the BD diagnostic
criteria to include other aspects of the bipolar spectrum (hypomania
and cyclothymia) yields a higher prevalence range (3.0% to 8.8%)
than is commonly believed (22). On the other hand, Baldessarini
has pointed out the potential research pitfalls of such a broadening
of the diagnostic spectrum (23). Baldessarini suggests that a broadening
of bipolar diagnosis beyond conventional BD I disease may retard
our understanding of the illness and that biological and genetic
studies may best proceed within more narrow diagnostic parameters.
A consensus has yet to be reached on the approach to (and definition
of) the bipolar spectrum.
Examining the underdiagnosis of BD naturally
leads to a discussion of how broad the spectrum of bipolar diagnosis
should be. Clinical and genetic data suggest that nonclassic parts
of the bipolar spectrum (that is, BD II, NOS, and cyclothymia) may
be more common than classic type I manic-depressive illness (21).
In fact, as Grof has suggested, classic type I manic depressive
illness may differ in many respects from less typical forms of bipolar
illness, especially in being more lithium-responsive. It is this
classic syndrome that Ketter has called “Cade’s disease.” Figure
1 suggests a possible conceptualization of these conditions
on the affective spectrum. Bipolar spectrum conditions exhibit less
severe mania, but they are not less severe in terms of depressive
symptoms. Apart from the major morbidity and substantial suicide
risk that these depressive symptoms present (3), varieties of BD
produce unstable lives, failed careers, high divorce rates, and
stormy biographies. Hence, we believe that the entire bipolar spectrum
needs to be aggressively diagnosed and treated.
The problem of BD underdiagnosis is partly (although not entirely)
related to failure to recognize bipolar spectrum states such as
hypomania, assuming a version of the spectrum beyond full mania
is accepted. Because hypomania is the only major DSM-IV diagnosis
in which the essential criterion of social and occupational dysfunction
is not required (and in fact, one must rule out significant social
and occupation dysfunction), many clinicians find hypomania to be
a difficult condition to diagnose. Thus, hypomania is mainly distinguished
from mania based on function, rather than symptoms. Because the
term “significant” is deliberately vague, psychiatrist identification
of hypomania is not reliable (24). Given this situation, hypomania
may be underdiagnosed as “normality,” and mania may be underdiagnosed
as hypomania.
Also, the complete focus on polarity found
in the diagnostic schema of DSM-III/IV obscures the relation between
bipolar and highly recurrent forms of unipolar depression. BD is
diagnosed when mood elevation is present, and its place in the diagnostic
schema implies a totally separate illness. However, phenomenologic
studies dating back to Kraepelin put primary emphasis on illness
course and considered cycling to be as important as polarity. Cases
of recurrent depression may be more likely to have genetic characteristics
and treatment responses similar to those encountered in BD (3).
Patients presenting with mainly depressive symptoms may exhibit
other clues to possible bipolarity, and these are outlined in Table
1.
Given the debate and confusion surrounding
the bipolar spectrum, we propose here a heuristic definition based
on these clues (Tables 1 and 2). We propose placing all versions
of bipolar illness apart from BD I or II in a single category, labelled
“bipolar spectrum disorder (BSD).” This is in contrast to others
who have suggested types of bipolar illness (III-VI) beyond BD I
and II (21,25). We envision that this BSD diagnosis might replace
the current nonspecific DSM-IV diagnosis of BD NOS. We heuristically
define BSD as a diagnostic category that possesses several of the
potential signs of bipolarity listed in Table 1, with greater weight
given to family history and antidepressant-induced manic symptoms
(26). Even in patients that have not spontaneously experienced a
manic or hypomanic episode, we suggest that BSD can be diagnosed
if they have MDEs with several signs of bipolarity (Table
2).
|
