European Psychiatry 30 (2015) 965974

Contents lists available at ScienceDirect

European Psychiatry
journal homepage: http://www.europsy-journal.com

Original article

Distinguishing bipolar disorder from borderline personality disorder:

A study of current clinical practice
K.E.A. Saunders, A.C. Bilderbeck, J. Price, G.M. Goodwin *
University Department of Psychiatry, Warneford Hospital, Oxford, OX3 7JX, United Kingdom

A R T I C L E I N F O

A B S T R A C T

Article history:
Received 16 July 2015
Received in revised form 1st September 2015
Accepted 5 September 2015
Available online 21 October 2015

Background: Diagnosing mental illness is a central role for psychiatrists. Correct diagnosis informs both
treatment and prognosis, and facilitates accurate communication. We sought to explore how
psychiatrists distinguished two common psychiatric diagnoses: bipolar disorder (BD) and borderline
personality disorder (BPD).
Methods: We conducted a qualitative study of psychiatrists to explore their practical experience. We then
sought to validate these results by conducting a questionnaire study testing the theoretical knowledge
and practical experience of a large number of UK psychiatrists. Finally we studied the assessment process
in NHS psychiatric teams by analysing GP letters, assessments by psychiatrists, and assessment letters.
Results: There was broad agreement in both the qualitative and questionnaire studies that the two
diagnoses can be difcult to distinguish. The majority of psychiatrists demonstrated in survey responses
a comprehensive understanding DSM-IV-TR criteria although many felt that these criteria did not
necessarily assist diagnostic differentiation. This scepticism about diagnostic criteria appeared to
strongly inuence clinical practice in the sample of clinicians we observed. In only a minority of
assessments were symptoms of mania or BPD sufciently assessed to establish the presence or absence
of each diagnosis.
Conclusion: Clinical diagnostic practice was not adequate to differentiate reliably BD and BPD. The
absence of reliable diagnostic practice has widespread implications for patient care, service provision
and the reliability of clinical case registries.
2015 Elsevier Masson SAS. All rights reserved.

Keywords:
Mania and bipolar disorder
Borderline personality disorder
Psychiatry in Europe
Psychiatric assessments

1. Introduction
Both borderline personality disorder and broadly dened
bipolar disorder are common psychiatric diagnoses in the adult
population with similar prevalences of 16% [4,13,17,21,23]. The
two are commonly comorbid [9,11,19] with comorbidity as high as
50.1% of those with bipolar-1 [17], indicating an association well
beyond chance. However, patients with BPD are deemed to require
psychological treatments where medication plays a minor role
[25], whereas those with BD generally require complex medication
and didactic help with self-management [26]. Prognosis is also
very different: 73% of BPD may have remitted in 6 years [45] whilst
BD is usually a life-long relapsing condition [3].
Therefore, psychiatric diagnosis matters as a pragmatic tool for
informing treatment, communicating about patterns in psychiatric
illness, development of appropriate services, and allocation of

* Corresponding author. University Department of Psychiatry, Oxford, OX3 7JX,

United Kingdom. Tel.: +44 1865 226467; fax: +44 1865 204981.
E-mail address: guy.goodwin@psych.ox.ac.uk (G.M. Goodwin).
http://dx.doi.org/10.1016/j.eurpsy.2015.09.007
0924-9338/ 2015 Elsevier Masson SAS. All rights reserved.

resources. Its strength lies in its reliability, which can be estimated

by looking at inter-rater agreement in clinical samples. This can
achieve high values (conventionally described with the Kappa
statistic) when structured interviews are employed; over 0.9 for
bipolar diagnoses [36] and over 0.75 for borderline diagnoses [43].
The similarities and the differences between the disorders and
their co-occurrence are a source of considerable confusion. The
rates of misdiagnosis in BD and BPD in clinical practice are largely
unknown as few studies have sought to explore this systematically.
Patients with BPD have signicantly greater odds of being
diagnosed with BD compared with psychiatric outpatients who
do not have BPD [34,46]. In psychiatric outpatients who had
previously been incorrectly diagnosed with BD, 25% were found to
have BPD when subjected to formal diagnostic assessment using
the SCID-1 and -2 [47] whilst evidence of bipolarity (dened as
bipolar-1 or -2) has been found in 44% of patients who had
previously been diagnosed with BPD [14]. In another study, over
half of supposed BPD participants who had received diagnoses
based on clinical assessment did not meet criteria for BPD when
subjected to a SCID-2 interview [5]. Finally, temporal stability of

966

K.E.A. Saunders et al. / European Psychiatry 30 (2015) 965974

clinical diagnoses is generally poor: in outpatient settings

prospective consistency has been found to be 50.6% for BD but
just 35.6% for personality disorder [7].
We know little about how a diagnosis is made in current
psychiatric practice or why the BD/BPD distinction appears
problematic. There is an overlap of symptoms like impulsivity,
recurrent suicidal behaviour, affective instability, intense anger,
and transient paranoid ideation. While these are diagnostic
features of BPD, they also occur in BD patients during mood
episodes. Discrimination requires detailed exploration at interview
of how such symptoms arise and whether more pervasive
symptoms of BPD like fear of abandonment, unstable personal
relationships, identity disturbance and chronic emptiness are also
present. This requires psychiatrists to collect a substantial history
and enquire systematically about these symptoms. However,
much of the previous research on the diagnostic process has
focussed exclusively on validating clinical diagnoses against
different structured or semi-structured interviews [5,24,35,47].
The recent eldwork trials for DSM-5 are a rare recent example
looking at categorical diagnoses in sequentially recruited rather
than highly selected patient samples and without structured
clinical interviews. BD-I and BPD diagnosis showed very good
reliability (Kappa : 0.75) in some centres, but not in others
[30]. Thus, even when diagnosis is under explicit scrutiny, it cannot
be assumed that diagnostic agreement is high.
We know of no previous qualitative research that explores how
clinicians approach the differential diagnosis of BD or BPD.
Furthermore, we know of no previous research which directly
examines the diagnostic assessment process, including recordings
of assessment interviews, to determine whether clinicians explore
and have available information about diagnostic criteria when
making diagnostic decisions. Here, we report three linked studies,
which aim to understand diagnostic practice for patients
presenting with mood instability. The rst was a qualitative study
of psychiatrists and nurses, aiming to understand their experience
of distinguishing BD and BPD and the factors that inuence their
diagnostic decision-making. A qualitative approach was employed
because it is exible, grounded in individual experiences, and
because we know so little about how diagnoses are actually made.
Using the understanding generated in this study, we developed a
questionnaire used in the second study, an electronic survey of UK
psychiatrists. Finally, the third study comprised a detailed
observational study of the diagnostic assessment process in
ordinary practice. Ethical approval for the study was obtained
from Oxfordshire REC A (11/H0604/8) and practice was informed
by the principles enshrined in the Declaration of Helsinki.

Table 1
Clinician characteristics of 32 clinicians who participated in qualitative interviews.
Male

Female

Total

16

16

32

Average age (years)

39.1

39.4

39.3

Medical qualication

16

10

26 (81%)

MRCPsych

15

23 (72%)

CMHT

13

20 (63%)

5 (16%)

Self-harm team

2 (6%)

Psychotherapy/therapeutic community

5 (16%)

Specialist training
Bipolar
Borderline
Both

1
1
0

1
3
3

2 (6%)
4 (13%)
3 (9%)

Specialist mood disorders clinic

cases (n = 24), assessments with patients referred for the assessment of mood instability were observed and/or audio-recorded.
2.1.3. Interviews
Clinician interviews were conducted using a topic schedule and
were audio-recorded (for full topic schedule and more information
about qualitative interviews see Supplementary material). Interviews varied in length from 20 to 100 minutes. Some clinicians
completed multiple interviews when patients had multiple
assessments or when clinicians assessed more than one patient
meeting the inclusion criteria, resulting in a total of 38 clinician
interviews of 32 unique clinicians who had assessed 32 patients.
2.1.4. Data analysis
Quantitative data were summarized using standard statistical
approaches; qualitative data coding, management and analysis
were conducted using NVivo software [29]. Semi-structured
interviews were conducted as part of an on-going and iterative
process of data collection and analysis. Audiotaped interviews
were transcribed, reviewed, and uploaded to NVivo. Qualitative
analysis used a framework technique [32]. Data gathering ceased
when understanding of the experience of clinicians in assessing
and diagnosing in patients with mood instability was no longer
being advanced. To reduce researcher bias, we discussed and
maintained an awareness of preconceptions (facilitated by
interviewer note-keeping and memos) and constantly linked the
emerging thematic framework to clinician-derived data.

2. Qualitative study of clinician diagnostic assessments

2.2. Results
2.1. Method
2.1.1. Participants
Participants were 32 psychiatrists and nurses recruited from
secondary mental health services, which included 7 community
mental health teams (CMHTs), a specialist mood disorders clinic, and
a therapeutic community. Purposive sampling was used [8] to ensure
a range of ages, professional backgrounds and geographical locations.
Inclusion criteria included being fully qualied in their
discipline. Written informed consent was obtained from all
participants (both clinicians and patients).
2.1.2. Data gathering
Demographic data were gathered from all participants, including age, gender, qualications and any specialist experience or
training in BD/BPD. The largest proportion of participants were
working in community mental teams (Table 1). In the majority of

2.2.1. Clinician perception of the problem

Most clinicians agreed that distinguishing between the two
diagnoses could be challenging (Box 1). Overlap in diagnostic
criteria between BD and BPD was raised by many clinicians
particularly in regard to mood instability. The need to rely in many
cases on self-reported mood symptoms, the context in which these
symptoms were reported, and inaccurate retrospective recall were
highlighted by most as particularly challenging. Chaotic lifestyles,
including the use of illicit drugs, were reported as additional
challenges, as were the difculties of conducting diagnostic
assessments in crisis situations.
2.2.2. Utility of distinguishing the diagnoses
Many clinicians questioned the validity of the BPD diagnosis
and felt that determining the presence or absence of an axis
1 disorder was more important because this was their primary

K.E.A. Saunders et al. / European Psychiatry 30 (2015) 965974

967

Box 1. Clinician perception of the problem supporting quotations.

Box 2. Utility of distinguishing the diagnosis supporting

quotations.

The main problem is that uh people with emotional unstable

personality disorder do experience significant mood swings
and they describe their mood as unstable, and uh therefore I
mean by definition it becomes a real issue about who is within
the borderline spectrum versus who is in the bipolar spectrum. (CMHT psychiatrist 051).
I challenge anyone to be able to say what their moods like
reliable over that period of time. So there are real problems in
retrospective diagnosis. (Mood disorders specialist 01).
It gets difficult is when people present with quite extreme
behaviours, um people with bipolar can present as very agitated, sometimes they present with sort of psychotic or quasipsychotic symptoms, and they I think they tend to be the ones
where its difficult to differentiate. (CMHT psychiatrist 17).
My experience is that theres a significant subset of people
with bipolar disorder who exhibit a lot of chaotic lifestyle,
some symptoms that would fit within kind of more personality
aspects and this kind of wondering about certain kind of
personality kind of features might predispose you towards
it. (CMHT psychiatrist 25).
So shes on holiday with her friends, theyre going to a
nightclub, theres drink involved, to what extent is it a reflection of the illness, and to what extent is it the outside, what one
might expect from someone going to a, an Ibiza nightclub or I
dont know. (Mood disorders specialist 13).

My feeling is it is not that clear-cut between them is like um

spectrum. (CMHT psychiatrist 07).
I see bipolar affective disorder and borderline PD as being on
a sort of continuum. . . both groups would be using sort of
manic defences at some point. (Psychiatrist in psychotherapy
06).
I dont think that the borderline personality disorder is, it
should be a separate diagnosis I think it exists on a spectrum of
various disorders. (CMHT psychiatrist 27).
To some extent, being able to understand their problem with
the diagnostic category can be helpful, but PD categories are
so useless in many ways. (RMN/group analyst 12).
I think fundamentally I think the diagnoses of personality
disorder mostly lack validity. Theyre not clear-cut discrete
entities. You dont either have it or dont have it. There arent
clear cut-offs of when you have a personality disorder and
when you dont have a personality disorder. (CMHT psychiatrist 18).
My initial attempt is to make the axis 1 diagnosis, as far as Im
concerned if the person meets criteria for bipolar disorder then
that would put a target at my therapeutic endeavours. Whether
or not I think theyve got a personality disorder as well. (Mood
disorders specialist 01).

Numbers following participant description denote participant number.

3. Clinician survey of diagnostic knowledge and experience

3.1. Method

focus for treatment (Box 2). A few clinicians suggested that BD and
BPD were on a continuum.
2.2.3. Use of the diagnostic criteria
While most clinicians stated that the diagnostic features were
helpful in distinguishing the two diagnoses, many of the
presenting features were felt to lack specicity to BD or BPD
(Box 3). Most clinicians felt that limiting assessment to the
diagnostic checklists would lead to incorrect diagnoses being
made, and they emphasized the importance of clinical judgement
and impressionistic approaches.
2.2.4. Other features distinguishing the diagnoses
Many clinicians felt that the nature of the clinician-patient
relationship during the clinical assessment was an important
inuence on their diagnostic decision-making (Box 4). This was
particularly the case with BPD, where hostility and assuming a
childlike relational position were felt to be indicative. Early abusive
experiences and attachment difculties were frequently reported
as more relevant to BPD.
2.2.5. Other non-clinical factors inuencing diagnosis
Several factors unrelated to the clinical presentation or
management were cited as inuencing diagnostic decisions (Box
5). These included systemic factors relating to healthcare targets
and funding. Other clinicians viewed BPD as a diagnosis which was
used when patients difculties did not conform to any specic
mood disorder. The time constraints placed upon clinicians and the
increasing use of diagnosis to manage costs and access to care were
viewed as important external pressures.
2.2.6. Stigma
The perceived stigma associated with BPD was often cited by
clinicians as a reason to avoid giving patients this diagnosis (Box 6).
By contrast, BD was seen by clinicians as a more acceptable
diagnosis for patients.

An online survey of psychiatrists was conducted, which

explored clinicians knowledge and use of diagnostic criteria when
distinguishing BD and BPD; their views of the assessment and
diagnosis of patients who may have BD and/or BPD; and factors
seen to inuence these diagnostic decisions. The majority of survey
items used a 5-point Likert-type scale and there were several
opportunities for respondents to clarify or expand on their
responses using free text (full survey and survey results provided
in the Supplementary material).

Box 3. Use of the diagnostic criteria supporting quotations.

If you go to someone whos got borderline-type problems,
and you just try to find out if theyve got bipolar have had a
hypomanic episode, and you just ask your questions in that
way, youd be very likely to erroneously conclude that they
have had a hypomanic episode. (CMHT psychiatrist 25).
So you could say well look this is, this will if you just take what
she says at face value she ticks a lot of the boxes for a mania
but um I think I wasnt convinced that that was a mania I didnt
think that that was actually what happened from the description that she gave to me. Not that she was lying. (CMHT
psychiatrist 16).
You can refer to diagnostic criteria, particularly with bipolar
disorder, but a large part of it is about a more difficult to pin
down impression which is formed. (CMHT psychiatrist 03).
So I think if youre someone who goes, what do I feel this
patient is presenting to me with? Or youre someone who goes
well, what boxes am I ticking, I think that that then leads you
down two different routes in terms of thinking about diagnosis (CMHT psychiatrist 25).
They are helpful, uh and they give you some sense of direction, um uh but sometimes I um I use I need to use my clinical
judgement rather than uh the diagnostic criteria to diagnose
the borderline personality disorder. (CMHT psychiatrist 26).

968

K.E.A. Saunders et al. / European Psychiatry 30 (2015) 965974

Box 4. Other factors which guide diagnosis supporting

quotations.

Box 5. Other non-clinical factors influencing choice of diagnosis supporting quotations.

. . .Severe difficulties during her teenage years and so theres

abuse as well. . . it will tend to favour borderline diagnosis.
(CMHT psychiatrist 04).
If their problems have been long-term and chronic and you
know and you can clearly go back to difficulties in upbringing
and attachment I think the diagnosis of personality problems
becomes much more likely. (CMHT psychiatrist 18).
He is impulsive and, but I dont, didnt pick up any true
borderline sort of traits. He isnt self-harming, he um seems
to have come from quite a caring, stable-type background.
(Specialty doctor 09).
How the person comes across to me, in the interview its not
all historical. I will look at the way theyre reacting to me, do I
feel that theyre being straight with me, is there an air of
trickiness or hostility about them. (Mood disorders specialist
01).
The patient takes a more childlike position and you, you have
to make the decisions about everything on this persons care.
Thats not the classic interaction I get with people with hypomania. Or mania. They tend to be you know I know everything
and sort of a bit more you get a far more sort of at times the
you get anger, bravado, and sort of uh you know you know
nothing. (CMHT psychiatrist 05).
A bipolar person who has no borderline aspects has basic
trust. Um, unless theyre in the middle of an episode and
theyre psychotic or their thinking is distorted. They have a
basic idea that youre a healthcare professional whos trying to
help them with an illness that they are suffering. And somebody who is borderline does not necessarily have that basic
idea at all. (Psychiatrist in psychotherapy 06).
Again this is not within the sort of diagnostic sort of DSM or
ICD-10 criteria but something that I find useful is the reactions
of the environment towards people and sort of the way people
uh either the GP or the family and the way they kind of respond
to the individual patient that is referred and then how us as
professionals also we kind of interact with this particular
patient. (CMHT psychiatrist 05).

Theyre actually saying I feel this person has difficulties that

probably dont fit with any kind of affective illness framework,
but they dont feel very confident in dealing with that. (CMHT
psychiatrist 25).
There is a sort of lack of incentive for people to make a
diagnosis of bipolar disorder if they dont feel theyre comfortable with what to do about it. Ah, whereas a diagnosis of
borderline personality disorder kinda solves the problem really because theyre not going to do anything for it. . . people
dont like to be manipulated. . . An exclusive diagnosis of
personality disorder is quite a good way of saying I dont want
anything to do with this patient. (Mood disorders specialist
01).
In a busy outpatient clinic you dont have time. So my
consultant will tell please discharge her back to the GP. I know
she will say that because she knows how busy and I am and
she knows we cannot have the luxury to see patients without
active problems, problems happening here and now for more
than a couple of times. (CMHT psychiatrist 29).
I think people clinically start to work in different ways. . . I think
around that, the system has investments in research, in financial interests, in simplicity for treatment, in the labelling of
people. . . its categories and its boxes and its money. . . pushes
clinicians into not seeing the individual, seeing the category,
and I think it is hugely detrimental. . . its partly how much
money is around all of this, means that the patient sitting in the
room is not getting an unbiased diagnosis. . . borderline personality disorder. . . it often means in some CMHTs, not all, that
it is seen as or it becomes a reason not to see the person.
(Psychiatrist in psychotherapy 06).
Unlike in the States or other countries where you get paid per
treatment given, or diagnosis, or clinical activity, in this country you just get paid regardless of how much youre doing,
theres no incentive therefore to over diagnose or over treat.
Which is a good thing but on the other hand there is an
incentive to underdiagnose and undertreat. (Mood disorders
specialist 1).
I think that the diagnosis of personality disorder often seems
like an excuse for sort of therapeutic nihilism, so you know
nothing, its felt that well we can do nothing to help this person
and so lets you know lets just leave them. (CMHT psychiatrist 24).

The survey was hosted online using Limesurvey (www.

Limesurvey.org) between September and November 2012. Email
invitations were sent out to roughly 8000 members of the Royal
College of Psychiatrists on behalf of the research team. Responses
were received from 648 members of the Royal College of
Psychiatrists, a response rate of 8.1%.
3.1.1. Statistical analysis
Statistical analysis was carried out using SPSS v. 20 for
Windows [37]. Data were analysed using univariate analyses with
between-subjects factors of gender and clinician grade. Nonparametric Chi2 tests were used in analysis of categorical data. Free
text responses were coded in a simple descriptive way.
3.2. Results
Six hundred and forty-eight responses were received of which
546 (84%) were completed in full. The majority of respondents
were consultant grade psychiatrists (70%); the remaining were
staff grade (7%), associate specialist (3%) and trainee (20%) doctors.
Most respondents (73%) reported that differentiating between BD
and BPD formed a signicant part of their clinical practice (dened as
5% of their case load or more). The majority of doctors reported
feeling condent in making this discrimination (76%); only 1.8%
reported feeling not at all condent. A greater proportion of doctors
in consultant grades were condent about distinguishing the

diagnosis than those in non-consultant grades (79% vs. 64%,

x2 = 16.90, P < 0.001). Female doctors reported feeling condent
less often than their male counterparts (69% vs. 79%, x2 = 8.26,
P = 0.004). Fifty-two percent of respondents indicated that distinguishing BD and BPD was frequently a source of disagreement in
clinical practice, with just 0.7% reporting that this was never the case.
Respondents broadly matched the DSM-IV-TR diagnostic
features of the two disorders with the factors they actually used
to distinguish them. Factors which overlap in the diagnostic
criteria (such as impulsivity and affective instability) were
correctly viewed as less indicative. Historical factors were also
reported as important in guiding diagnostic decisions, with 84%
endorsing early life trauma as strongly or very strongly indicative
of BPD compared with <1% for BD, and 72% endorsing a positive
family history as being strongly or very strongly indicative of BD
compared with 13% for BPD. In free text comments, 50% of
respondents highlighted the need to focus on obtaining a clear
history while 23% highlighted the role of interpersonal dysfunction
in identifying BPD. For BD, 38% highlighted the need to focus on
mood changes particularly the presence of elated mood, and a
further 10% mentioned the need to identify inter-episode
euthymia.

K.E.A. Saunders et al. / European Psychiatry 30 (2015) 965974

Box 6. Stigma supporting quotations.

Nobody welcomes. . . a diagnosis of personality disorder. . .
unlike bipolar disorder it carries more stigma. . . so it does feel
like a much more judgemental diagnosis. . . the person sort of
has a different response to be given an illness diagnosis versus
something that feels like a moral judgement. (CMHT psychiatrist 16).
Bipolar is a more acceptable diagnosis to have than borderline personality disorder and I think um partly thats our fault as
mental health professionals that we can be quite pejorative
about personality disorder. (GP trainee 23).
Im not saying bipolar doesnt exist. Clearly, you know, we
think it does. Um, I think that a diagnosis of bipolar disorder or
being a manic-depressive is somewhat more romantic and
acceptable, you know because along with it goes creativity,
eccentricity of a particular kind, thats kind of acceptable,
whereas being a PD, or a borderline, doesnt carry the same
kind of strokes. (CMHT psychiatrist 08).

3.2.1. Use and utility of the diagnostic criteria

Eighty-four percent of participants agreed or strongly agreed
that the diagnostic criteria correlated with their clinical observations of BD compared with 75% for BPD. Around half of respondents
reported a preference for using diagnostic criteria compared with
an impressionistic approach for both BD and BPD (54% and 42%,
respectively). Almost a third of clinicians expressed no preference
for diagnostic approach for either diagnosis (31.0% and 31.7% for
BD and BPD, respectively) and the remainder endorsed an
impressionistic approach (15% for BD and 27% for BPD).
Respondents were equally likely to consult a third party, with
84% vs. 79% (for BD and BPD respectively) consulting a third party
in half or more of their assessments.
The majority (70%) of psychiatrists did not view BD and BPD as
being part of the same illness spectrum. Comorbid diagnoses of
BPD in those with BD were reported as being made infrequently,
with just 6% endorsing this as a fairly frequent occurrence in their
practice.
4. Observed clinical practice
4.1. Method
4.1.1. Data sources
Participants were 20 of the 26 psychiatrists whose assessments
were observed in the qualitative study. Most assessing clinicians
(17; 85%) had membership of the Royal College of Psychiatrists
indicating that they had at least 3 years of specialist experience and
had received postgraduate training in psychiatry [11].
A subsample of eighteen assessments conducted by fteen
clinicians were audio-recorded. The remaining 25% were not
recorded because of technological issues (n = 2) or because the
patient did not consent (n = 3). Written and informed consent was
obtained from all participants as well as from their patients, and
both clinicians and patients gave consent for us to access electronic
patient records.
4.1.2. Analysis of diagnoses
GP referral letters and assessment recordings were analysed to
see which diagnostic criteria were reported, using a checklist of the
symptoms of BD and BPD as dened by DSM-IV-TR. Clinical
diagnoses were obtained from the assessment letters sent back to
GPs.
Research diagnoses were generated using OPCRIT+ [33]. OPCRIT
(Operational CRITeria) is a diagnostic system that automates the

969

generation of putative diagnoses using a checklist constructed

from the major psychiatric diagnostic classications. Clinical data
derived from electronic medical records and recorded patient
assessments are entered into the OPCRIT computer programme,
which generates diagnoses. It has been used in a wide number of
clinical epidemiological and biological research settings
[6,33,39].
4.2. Results
All 30 patient participants were referred from primary care for
assessment, the majority for the assessment of possible BD (n = 21;
70%). The remaining 30% reported mood instability in the
presentation. The majority of these referrals were made to CMHTs
(n = 28; 93%), with further two participants assessed at the
specialist mood disorders clinic.
Spider-diagrams showing the proportion coverage of symptoms (in GP referral letters, assessments and psychiatrist letters)
for depression, mania and borderline PD are shown in Figs. 13
respectively.
4.2.1. GP referral letters
All but one referral letter mentioned low mood (n = 29; 97%).
Other depressive symptoms, features of mania/hypomania, and
features of BPD were was much less consistently described
(Figs. 13).
4.2.2. Assessment audio-recordings
Most assessed patients were female (21/30; 70%) with mean
age of 34 (2) and 33 (3) for females and males respectively.
The numbers of symptoms explicitly covered at interview are
shown in Tables 2 and 3. Overall, depressive symptoms were the
most thoroughly assessed with the exception of feelings of
worthlessness and psychomotor agitation/retardation, which were
rarely explored. In 61% cases an adequate number of criteria were
assessed to afrm the presence of a diagnosis according to DSM-IVTR.
Manic symptoms were less likely to be explored than
depressive symptoms, and in 20% of cases elevated mood was
not addressed at all. In only 38% of cases were an adequate number
of symptoms explored to afrm the presence of the diagnosis.
Borderline symptoms were rarely explored. In only 4 cases (22%)
were an adequate number of symptoms explored to afrm the
diagnosis.
In just two cases (11%) were an adequate number of symptoms
explored for depression, mania and BPD in the same assessment.
4.2.3. Diagnoses
Letters to the general practitioner from the psychiatric team
contained diagnoses in 24/30 cases (80%). Table 4 summarizes the
symptom count towards different diagnoses recorded in the
psychiatrist assessment and in the return letter to the GP.
The reliability of the diagnoses against OPCRIT criteria (Table 5)
was poor. Of the 7 patients given a diagnosis of BD or cyclothymia
by their psychiatrist only 3 had BD when ICD-10 criteria were
applied via OPCRIT. OPCRIT diagnoses accorded exactly with the
positive clinical diagnosis for only 5/18 patients. In those given no
diagnosis by a clinician 5/6 had an axis 1 disorder when the OPCRIT
checklist was applied.
OPCRIT does not generate DSM-IV-TR or ICD-10 criteria for
BPD so we were unable to assess the extent to which BPDspecic information was objectively available for diagnosis.
However, of the 5 patients given a diagnosis of BPD by the
psychiatrist, the OPCRIT checklist suggested 1 had BD, 2 had
major depressive disorder, and 2 had a non-organic psychotic
syndrome.

970

K.E.A. Saunders et al. / European Psychiatry 30 (2015) 965974

Fig. 1. Proportion of diagnostic criteria for depression addressed in GP referral letters, assessment, and psychiatrist letters.

Fig. 2. Proportion of diagnostic criteria for mania addressed in GP referral letters, assessment, and psychiatrist letters.

K.E.A. Saunders et al. / European Psychiatry 30 (2015) 965974

971

Fig. 3. Proportion of diagnostic criteria for borderline personality disorder addressed in GP referral letters, assessment, and psychiatrist letters.

5. Discussion
To our knowledge, this is the rst study to explore the views,
experience and clinical practice of clinicians attempting to
distinguish BD and BPD. There was broad agreement in both the
qualitative and questionnaire studies that the two diagnoses can
be difcult to distinguish from one another, and that this
differential diagnosis can be a source of disagreement amongst
clinical staff. The majority of psychiatrists demonstrated a
comprehensive understanding of the criteria recommended in
Table 2
Proportion of depressive symptoms recorded when all sources (n = 30 referrals)
examined.
Depression

Depressed mood

93

Anhedonia

67

Signicant weight loss or increase in appetite

Poor appetite
Weight loss
Increased appetite
Weight gain

70
27
33
27

Insomnia or hypersomnia
Initial insomnia
Middle insomnia
Early morning wakening
Hypersomnia

47
33
20
70

Psychomotor agitation or retardation

Agitation
Retardation

7
3

Fatigue or loss of energy

77

Feelings of worthlessness or excessive or inappropriate guilt

73

DSM-IV-TR in our survey study; however in qualitative interviews

and free text responses to our online survey many expressed the
view that the diagnostic criteria did not necessarily assist
diagnostic differentiation. Indeed, a quarter of respondents in
the questionnaire study held a neutral attitude or thought
diagnostic criteria failed to correlate with the clinical phenomena
in BPD. Additionally, over a quarter of responders (27%) expressed
a preference for using an impressionistic approach rather than
diagnostic criteria in diagnosing BPD (15% in BD).
This scepticism about diagnostic criteria appears to strongly
inuence actual practice in the sample of clinicians whose
behaviour was directly observed. Notably, we found that in only
a minority of assessments were symptoms of mania or BPD
sufciently assessed to establish the presence or absence of each
diagnosis. Signicant numbers of symptoms were unexplored even
when drawing on all available sources of clinical information
(including patient records, letters and assessment recordings). Yet,

Table 3
Proportion of manic symptoms recorded when all sources examined.
Mania/hypomania

Elevated, or irritable mood, lasting at least 1 week

(or hospitalization) for mania, or 4 days for hypomania
Elevated mood
Irritable mood

57
53

Inated self-esteem or grandiosity

87

Decreased need for sleep

43

More talkative than usual or pressure to keep talking

20

Flight of ideas or subjective experience that thoughts are racing

23

Distractibility
Diminished ability to think or concentrate, or indecisiveness

50

Suicidal thoughts or behaviour

90

Increase in goal-directed activity or psychomotor agitation

27

Excessive involvement in pleasurable activities that have a high

potential for painful consequences

30

K.E.A. Saunders et al. / European Psychiatry 30 (2015) 965974

972

Table 4
Diagnoses given by psychiatrists together with the number of symptoms explored in assessments and described in GP letters.
Diagnosis given

Bipolar disorder
Borderline personality disorder
Bipolar disorder and borderline
personality disorder
Major depressive disorder
Cyclothymia
PTSD
No diagnosis given
No assessment letter written
a
b

Average number of symptoms assessed (n = 18)

recordings (clustera)

Number of symptoms reported in letter to GP

(n = 30)

Depression

Mania

BPD

Depression

Mania

BPD

5
4
1

5.5 (2)
3.0 (1)
N/Ab

6.0 (1)
3.5 (0)
N/Ab

3.0 (1)
5.5 (1)
N/Ab

4.4
3.4
4

2.80
2.6
7

0.0
2.8
5

10
1
1
6
2

4.5 (2)
3.0 (0)
N/Ab
5.5 (3)
7.0 (2)

3.0 (2)
5.0 (1)
N/Ab
4 (1)
7.0 (2)

2.5 (1)
4.0 (0)
N/Ab
1.5 (0)
4.0 (1)

5.2
3.0
6.0
3.67
N/A

4.7
4.0
6.0
3.0
N/A

1.8
1.0
2.0
0.83
N/A

This is the number of assessments where adequate symptoms were explored to make a diagnosis.
This assessment was not recorded.

diagnoses were offered in most cases. Perhaps unsurprisingly,

when such diagnoses were checked against OPCRIT criteria, only a
minority were supported.
Our survey data suggests that clinicians view a number of
factors aside from diagnostic criteria as relevant to the diagnostic
differentiation, and this appeared to correspond well with
observed clinical practice. Early abusive experience was believed
to favour a BPD diagnosis, while positive family history was
believed to favour BD. Whilst these associations are supported by
research [15,20,42,44], abuse during childhood is common in BD
[16] and the prevalence of BPD in rst degree relatives is
considerable, at 14.1% [18]. As such, that these factors show poor
discriminatory specicity. Impressions formed of patients, as
emphasized by some interviewed clinicians and preferred by some
survey responders, may have an important role in clinical
assessment. That the nature of the interaction with the patient
is important is supported by attachment theories of personality
disorder. However, the use of impressionistic approaches raises
questions regarding the basis upon which their impressions are
formed, the utility of diagnostic labels to inform treatment and
communication, and their medico legal defensibility. Clinicians
ambivalence about the use of diagnostic criteria, and the poor
coverage of diagnostic criteria in observed assessments, may relate
to the perception that both BD and BPD diagnostic categories lack
validity [2]. An alternative hybrid model for BPD was proposed for
inclusion in DSM-5 but was rejected by the committee on the basis
that the transition from a categorical diagnostic system of
individual disorders to one based on the relative distribution of
personality traits has not been widely accepted [38].

Findings from our qualitative study suggest that diagnosis is

also subject to external and arbitrary factors relating to service
provision, cost and stigma. While stigma is associated with most
diagnoses, it is a particular issue for personality disorder which is
often thought to be a non-biological illness [28]. In addition,
celebrity endorsement of BD has come to see its status rise by an
implicit association with creativity [12]. However, avoiding
diagnosis because of concerns about stigma only serves to deny
patients access to appropriate treatment and better understanding
of their illness. We have previously observed that patients with
mood instability can value a diagnosis of either BD or BPD, when it
enhances their understanding and lends meaning to their
experience [8].
These ndings contribute to a body of work which points to
potential inconsistencies in the diagnosis of BD and BPD
[27,31,34,46,47] and suggest that there is considerable scope for
improving diagnostic practice. Our survey results demonstrate
that most psychiatrists know which diagnostic criteria contribute
to diagnosis. Based on the assumption that greater symptom
coverage can improve the assessment process, a change in
application of this knowledge is required. Standardized or semistructured interviews may provide more systematic coverage of
different symptom dimensions [10] but have limitations [41] and,
our results suggest, would be met with resistance amongst
clinicians concerned about false-positives. The majority of
practising psychiatrists maintain that a diagnostic system based
upon clinical descriptions is more useful than a list of operationalized criteria [1] despite a wealth of evidence suggesting that,
used in isolation, it is associated with inaccurate diagnoses

Table 5
ICD-10 diagnoses generated by OPCRIT.
ICD-10 diagnosis

Bipolar disordera
Major depressive disorder
Non-organic psychotic syndrome
PTSD
Agoraphobia
Social phobia
Harmful use of alcohol
Alcohol dependence
Harmful use of cannabis
Cannabis dependence
Substance misuse disorder
Other personality disorders
No diagnosis
a

Including manic episodes.

Clinical diagnosis
Bipolar
disorder

Borderline
personality
disorder

Major
depressive
disorder

Cyclothymia

PTSD

Number of
diagnosis
given

Number of
assessment
letter

3
1
1
0
0
0
0
0
0
0
0
1
1

1
2
2
0
0
0
0
2
0
0
1
2
0

0
1
3
2
1
1
0
3
1
1
0
1
2

0
0
1
0
0
0
0
0
0
0
0
0
0

1
1
1
0
0
0
1
0
1
0
0
0
1

0
0
2
0
0
0
0
0
0
0
0
0
0

1
0
0
1
0
0
0
0
0
0

K.E.A. Saunders et al. / European Psychiatry 30 (2015) 965974

[22,24,31,40]. The widespread availability of new technologies and

the move towards patient-accessible and patient-annotated health
care records may bring signicant opportunities for systematically
enhancing data collection and diagnosis.

973

approval of the manuscript; or the decision to submit the

manuscript for publication.

Appendix A. Supplementary data

5.1. Limitations
The number of clinicians included in the qualitative study was
small. However, they were purposively sampled to ensure that we
were able to reect a number of different contexts and
geographical spread, and data saturation was reached. The male
gender bias in the sample reects local gender ratios in more senior
clinical staff. All clinical staff initially approached about the study
agreed to participate, such that a self-selection bias was unlikely to
inuence data collection. We cannot determine whether the poor
quality of assessments is specic to BD and BPD or true of
psychiatric assessment more generally. While diagnosis is not
always the reason for GPs to refer to psychiatric services it was in
the cases studied here. The majority of referrals were for the
assessment of BD as opposed to BPD. Given that the CMHTs are the
gatekeepers for access to other psychiatric services (including
psychological services) the nature of these referrals is likely to be
broadly representative. The reasons why GPs tend to query BD as
opposed to BPD in those presenting with mood instability is not
known.
The response rate for the questionnaire study was low which
raises the possibility of signicant respondent bias. As we did not
conduct an independent diagnostic assessment of participating
patients, no rm conclusions can be drawn as to the true
diagnosis. The OPCRIT + diagnostic system has not been developed
to generate BPD diagnoses (or indeed any axis II diagnoses),
limiting the extent to which we could evaluate clinical diagnoses.
However, that clinicians did not have available sufcient evidence
to support their decision may be a cause of concern in itself.
Furthermore, our OPCRIT analysis indicated clinicians may have
failed to make correct diagnoses, and/or recognise the presence of
comorbidities in a substantial proportion of cases, raising
questions about the value of the diagnostic process.
Disclosure of interest
Dr Saunders, Dr Bilderbeck and Dr Price declare that they have
no competing of interest.
Prof Goodwin holds grants from Bailly Thomas, Medical
Research Council, NIHR, Servier, holds shares in P1Vital and has
served as consultant, advisor or CME speaker for AstraZeneca, BMS,
Boehringer Ingelheim, Cephalon/Teva Janssen-Cilag, Eli Lilly,
Lundbeck, Otsuka, P1Vital, Roche, Servier, Shering Plough, Shire,
Sunovion, Takeda.
Authors contribution
Drs Saunders, Price and Prof Goodwin devised and designed the
study. Drs Saunders and Bilderbeck collected the data. Primary
data analysis was conducted by Dr Saunders with support from Drs
Bilderbeck, Price and Prof Goodwin. Drs Saunders and Bilderbeck
drafted the manuscript which was reviewed by all the authors.
Funding
This study was funded by an NIHRResearch for Patient Benet
Grant (PG-PB-0909-19070). The funder did not have any role in
the design and conduct of the study; collection, management,
analysis and interpretation of the data, preparation, review, or

Supplementary data associated with this article can be found, in

the online version, at http://dx.doi.org/10.1016/j.eurpsy.2015.09.
007.
References
[1] Aboraya A. Use of structured interviews by psychiatrists in real clinical
settings: results of an open-question survey. Psychiatry 2009;6(6):248.
[2] Akiskal HS. Demystifying borderline personality: critique of the concept and
unorthodox reections on its natural kinship with the bipolar spectrum. Acta
Psychiatr Scand 2004;110(6):4017.
[3] Angst J. Course of unipolar depressive, bipolar manic-depressive, and schizoaffective disorders. Results of a prospective longitudinal study. Fortschr
Neurol Psychiatr Grenzgeb 1980;48(1):330 [authors translation].
[4] Angst J. The emerging epidemiology of hypomania and bipolar II disorder. J
Affect Disord 1998;50(23):14351.
[5] Atre-Vaidya N, Hussain SM. Borderline personality disorder and bipolar mood
disorder: two distinct disorders or a continuum? J Nerv Ment Dis
1999;187(5):3135.
[6] Azevedo MH, Soares MJ, Coelho I, Dourado A, Valente J, Macedo A, et al. Using
consensus OPCRIT diagnoses. An efcient procedure for best-estimate lifetime
diagnoses. Br J Psychiatry 1999;175:1547.
[7] Baca-Garcia E, Perez-Rodriguez MM, Basurte-Villamor I, Fernandez del Moral
AL, Jimenez-Arriero MA, Gonzalez de Rivera JL, et al. Diagnostic stability of
psychiatric disorders in clinical practice. Br J Psychiatry 2007;190:2106.
[8] Bilderbeck AC, Saunders KEA, Price J, Goodwin GM. The experience of assessment by patients referred to secondary care with mood instability: a qualitative study. Br J Psychiatry 2014;204:2349.
[9] Brieger P, Ehrt U, Marneros A. Frequency of comorbid personality disorders in
bipolar and unipolar affective disorders. Compr Psychiatry 2003;44(1):2834.
[10] Brugha TS, Bebbington PE, Jenkins R. A difference that matters: comparisons of
structured and semi-structured psychiatric diagnostic interviews in the general population. Psychol Med 1999;29(5):101320.
[11] Carpenter D, Clarkin JF, Glick ID, Wilner PJ. Personality pathology among
married adults with bipolar disorder. J Affect Disord 1995;34(4):26974.
[12] Chan D, Sireling L. I want to be bipolar. . . a new phenomenon. Psychiatrist
2010;34:1035.
[13] Coid J, Yang M, Tyrer P, Roberts A, Ullrich S. Prevalence and correlates of
personality disorder in Great Britain. Br J Psychiatry 2006;188:42331.
[14] Deltito J, Martin L, Riefkohl J, Austria B, Kissilenko A, Corless CMP. Do patients
with borderline personality disorder belong to the bipolar spectrum? J Affect
Disord 2001;67(13):2218.
[15] Distel MA, Trull TJ, Derom CA, Thiery EW, Grimmer MA, Martin NG, et al.
Heritability of borderline personality disorder features is similar across three
countries. Psychol Med 2008;38(9):121929.
[16] Garno JL, Goldberg JF, Ramirez PM, Ritzler BA. Impact of childhood abuse on
the clinical course of bipolar disorder. Br J Psychiatry 2005;186:1215.
[17] Grant BF, Chou SP, Goldstein RB, Huang B, Stinson FS, Saha TD, et al. Prevalence,
correlates, disability, and comorbidity of DSM-IV borderline personality disorder: results from the Wave 2 National Epidemiologic Survey on Alcohol and
Related Conditions. J Clin Psychiatry 2008;69(4):53345.
[18] Gunderson JG, Zanarini MC, Choi-Kain LW, Mitchell KS, Jang KL, Hudson JI.
Family study of borderline personality disorder and its sectors of psychopathology. Arch Gen Psychiatry 2011;68(7):75362.
[19] Kay JH, Altshuler LL, Ventura J, Mintz J. Prevalence of axis II comorbidity in
bipolar patients with and without alcohol use disorders. Ann Clin Psychiatry
1999;11(4):18795.
[20] Kendler KS, Aggen SH, Czajkowski N, Roysamb E, Tambs K, Torgersen S, et al.
The structure of genetic and environmental risk factors for DSM-IV personality disorders: a multivariate twin study. Arch Gen Psychiatry 2008;65(12):
143846.
[21] Lenzenweger MF, Lane MC, Loranger AW, Kessler RC. DSM-IV personality
disorders in the National Comorbidity Survey Replication. Biol Psychiatry
2007;62(6):55364.
[22] McGorry PD, Mihalopoulos C, Henry L, Dakis J, Jackson HJ, Flaum M, et al.
Spurious precision: procedural validity of diagnostic assessment in psychotic
disorders. Am J Psychiatry 1995;152(2):2203.
[23] Merikangas KR, Akiskal HS, Angst J, Greenberg PE, Hirschfeld RM, Petukhova
M, et al. Lifetime and 12-month prevalence of bipolar spectrum disorder in the
National Comorbidity Survey replication. Arch Gen Psychiatry 2007;64(5):
54352.
[24] Miller PR, Dasher R, Collins R, Grifths P, Brown F. Inpatient diagnostic
assessments: 1. Accuracy of structured vs. unstructured interviews. Psychiatry
Res 2001;105(3):25564.
[25] National Collaborating Centre for Mental Health. Borderline personality disorder: treatment and management. Full guideline. NICE; 2009.

974

K.E.A. Saunders et al. / European Psychiatry 30 (2015) 965974

[26] National Institute for Health and Care Excellence. Bipolar disorder: the assessment and management of bipolar disorder in adults, children and young
people in primary and secondary care. London: National Institute for Health
and Care Excellence; 2014.
[27] Oiesvold T, Nivison M, Hansen V, Sorgaard KW, Ostensen L, Skre I. Classication of bipolar disorder in psychiatric hospital. A prospective cohort study.
BMC Psychiatry 2012;12:13.
[28] Paris J. Why psychiatrists are reluctant to diagnose: borderline personality
disorder. Psychiatry (Edgmont) 2007;4(1):359.
[29] QSR International. QSR Nvivo. 7th ed. Doncaster, Australia. 2006
[30] Regier DA, Narrow WE, Clarke DE, Kraemer HC, Kuramoto SJ, Kuhl EA, et al.
DSM-5 eld trials in the United States and Canada, part II: test-retest
reliability of selected categorical diagnoses. Am J Psychiatry 2013;170(1):
5970.
[31] Rettew DC, Lynch AD, Achenbach TM, Dumenci L, Ivanova MY. Meta-analyses
of agreement between diagnoses made from clinical evaluations and standardized diagnostic interviews. Int J Methods Psychiatr Res 2009;18(3):
16984.
[32] Ritchie J, Spencer L. Qualitative data analysis for applied policy research. In:
Bryman A, Burgess R, editors. Analyzing qualitative data. Routledge; 1994. p.
17394.
[33] Rucker J, Newman S, Gray J, Gunasinghe C, Broadbent M, Brittain P, et al.
OPCRIT+: an electronic system for psychiatric diagnosis and data collection in
clinical and research settings. Br J Psychiatry 2011;199(2):1515.
[34] Ruggero CJ, Zimmerman M, Chelminski I, Young D. Borderline personality
disorder and the misdiagnosis of bipolar disorder. J Psychiatr Res 2010;44(6):
4058.
[35] Shear MK, Greeno C, Kang J, Ludewig D, Frank E, Swartz HA, et al. Diagnosis of
nonpsychotic patients in community clinics. Am J Psychiatry 2000;157(4):
5817.

[36] Simpson SG, McMahon FJ, McInnis MG, MacKinnon DF, Edwin D, Folstein SE,
et al. Diagnostic reliability of bipolar II disorder. Arch Gen Psychiatry
2002;59(8):73640.
[37] SPSS Inc. SPSS for Windows Release 15.0.1.12007.
[38] Trestman RL. DSM-5 and personality disorders: where did axis II go? J Am Acad
Psychiatry Law 2014;42(2):1415.
[39] Van Os J, Gilvarry C, Bale R, Van Horn E, Tattan T, White I, et al. A comparison of
the utility of dimensional and categorical representations of psychosis. UK700
Group. Psychol Med 1999;29(3):595606.
[40] van Praag HM. Over the mainstream: diagnostic requirements for biological
psychiatric research. Psychiatry Res 1997;72(3):20112.
[41] Wittchen HU, Ustun TB, Kessler RC. Diagnosing mental disorders in the
community. A difference that matters? Psychol Med 1999;29(5):10217.
[42] Zanarini MC, Gunderson JG, Marino MF, Schwartz EO, Frankenburg FR. Childhood experiences of borderline patients. Compr Psychiatry 1989;30(1):1825.
[43] Zanarini MC, Frankenburg FR, Vujanovic AA. Inter-rater and test-retest reliability of the Revised Diagnostic Interview for Borderlines. J Pers Disord
2002;16(3):2706.
[44] Zanarini MC, Frankenburg FR, Yong L, Raviola G, Bradford Reich D, Hennen J,
et al. Borderline psychopathology in the rst-degree relatives of borderline
and axis II comparison probands. J Pers Disord 2004;18(5):43947.
[45] Zanarini MC, Frankenburg FR, Reich DB, Fitzmaurice G. Time to attainment of
recovery from borderline personality disorder and stability of recovery: a 10year prospective follow-up study. Am J Psychiatry 2010;167(6):6637.
[46] Zimmerman M, Galione JN, Ruggero CJ, Chelminski I, Young D, Dalrymple K,
et al. Screening for bipolar disorder and nding borderline personality disorder. J Clin Psychiatry 2010;71(9):12127.
[47] Zimmerman M, Ruggero CJ, Chelminski I, Young D. Psychiatric diagnoses in
patients previously overdiagnosed with bipolar disorder. J Clin Psychiatry
2010;71(1):2631.