Management and treatment
practice, but still they have one more message of caution over FGAs, but the side-effect profile was not necessarily
about the nature of the evidence on which we must base advantageous.
our practice. In England, the NICE guidance reflects this
(2009, and under regular review). After a general preamble Delusional disorder is rarely considered separately from
about the context of care, rather than a formulaic approach, schizophrenia in medication trials, mainly because it is so
it recommends for people in a first episode of illness that much less common, perhaps affecting just 0.3% of the pop-
‘the choice of drug should be made by the service user and ulation. Smith and Buckley (2006) have reviewed the sparse
healthcare professional together’. For non-responders, it evidence for effectiveness of medications used specifically
recommends reviewing diagnosis, checking adherence to in this situation.
treatment at an adequate dose for an adequate period,
interference by other drugs – including alcohol or illicit Krakowski et al. (2006) explicitly found no advantage in
drugs, and whether adjunctive psychological treatments psychotic symptom reduction for two second generation
have a role and, where so, have been tried. The advice for antipsychotics – clozapine and olanzapine – over halo-
a trial of at least two antipsychotic medications from two peridol (an FGA) after completing a double-blind RCT
different classes before progressing to clozapine has been with 110 patients, about 80% of whom were men who had
modified only slightly since Kane et al. (1988) defined medi- schizophrenia or schizo-affective disorder. The Clinical
cation refractoriness. NICE guidance suggests at least one Antipsychotic Trials of Intervention Effectiveness (CATIE)
of these medications should be from second generation/ trial similarly failed to find much difference between the
atypical group. two broad groups of antipsychotics (Leibermann et al.,
2005), but this study is controversial on the main grounds
So, what evidence is there from trials? Davis et al. that it has insufficient power (Kraemer et al., 2009). A useful
(2003) completed a systematic review with meta-analysis extended account of the main findings from the trial is to
of clinical trials of antipsychotic mediation, covering be found in a series of articles in Psychiatric Services, May
‘grey’ material as well as peer-reviewed publications dur- 2008. Jones et al. (2006), in England, focused on 227 people
ing 1953–2002. They located 124 RCTs comparing the with schizophrenia or similar psychosis who were already
efficacy of conventional and atypical antipsychotics and on medication, but considered to need a change because
18 comparing antipsychotics within the atypical group. of response failure or side effects. The new prescription
The effect sizes of clozapine, amisulpiride, risperidone, was randomized between conventional or atypical antip-
and olanzepine were significantly greater than those of sychotic groups, but the treating clinician could choose
conventional antipsychotics, but the other six atypicals within group. Blind assessments of outcome were made at
investigated had no such advantage. In a series of reviews, intervals for just over 1 year. There was little difference in
Leucht and colleagues reflect that SGAs do not constitute outcome between the two groups in symptomatic stabil-
a homogeneous class of drugs, and so, perhaps, a simple ity or change, or in quality of life. No clear preference for
answer should not be expected. In 2003, they varied the either drug group emerged from participant reports, and
question to cover relapse prevention and the systematic costs were similar. In 2001–2002, a group of us attempted
review to include RCTs comparing atypicals with placebo to extend this trial into one of the high security hospitals
as well as with conventional antipsychotics, but covered in England. Ethical approval was obtained, but in spite of
a slightly shorter time period (1966–2002). They found 11 some extensive liaison work, no consultant psychiatrist
studies, including over 2,000 patients, comparing one or referred a single patient for the trial, which was formally
more drugs from the atypical and conventional groups. abandoned in 2004.
Rates of relapse and overall treatment failure were signifi-
cantly lower among patients taking atypical antipsychot- Most of the studies which have been included in sys-
ics, but the authors were not impressed with the effect tematic reviews and meta-analyses are of short-term treat-
size and were concerned about aspects of methodology in ment with antipsychotic medication. Tiihonen et al. (2006)
some studies. In 2008, Leucht et al. published a review and reported a substantial Finnish national naturalistic study
meta-analysis of 78 studies of comparisons of the second of longer-term outcome in relation to 10 antipsychotic
generation antipsychotic medications for schizophrenia, medications commonly prescribed, from first and second
concluding that, although there was some evidence for generation groups. For an average of 3.6 years, they followed
an advantage for one or two over others, almost exclu- 2,230 people who had been consecutively hospitalized for
sively in relation to positive symptoms, for the individual schizophrenia or schizo-affective disorder at some time
patient, any small research based advantage for a par- between January 1995 and December 2001. Perphenazine
ticular medication must be weighed against large differ- depot, olanzapine and clozapine were the three medica-
ences in side-effects and cost. By 2009, their review was tions associated with lowest rehospitalization rates, while
of 150 double-blind, but mostly short-term studies with patient mortality was substantially raised among those
no fewer than 21, 533 people. Just four of the SGAs had who were not taking any antipsychotic medication.
small to medium effect size advantage in terms of efficacy
Apart from medication type, questions have been raised
about dose, frequency of medications, their combination
and whether depot or oral forms are to be preferred.
© 2014 by Taylor & Francis Group, LLC 359
Psychosis, violence and crime
1. Dose. In the UK, it is recommended that doses of 4. Depot or oral maintenance treatment? Here too, perhaps
antipsychotic medication should not exceed the unexpectedly, it seems that there are no certainties.
ranges given in the British National Formulary (BNF), Haddad et al. (2009) built on a review of five Cochrane
available online (http://BNF.org). This recommendation reviews by Adams et al. (2001), noted each of the
includes allowing for the possibility of additive effects Cochrane updatings and found one subsequent RCT,
of prescribing more than one antipsychotic and ‘as then went on to consider observational studies too.
necessary’ medications. From time to time, there has No newer data altered the conclusion from meta-
been a vogue for high dose prescribing, particularly analysis of RCT findings that there is no significant
where violence may be an issue. The Royal College of difference in efficacy or in relapse rates between oral
Psychiatrists (2006), which keeps the situation under and depot administration of FGA medications. The
review, notes in its consensus statement that there is four prospective observational studies identified
no evidence of any advantage for high dose medication, offered, between them, mixed results. Two found lower
but some evidence of high risk, for example of cardiac discontinuation rates for the depot groups while two
arrhythmias leading to sudden death. In rare, individual found that outcome was no different or better with
cases, where there is no response to conventional doses oral medication. The 11 mirror-image studies, however,
of a range of medications, then the consensus statement in which each patient is observed for an equivalent
acknowledges that it may be acceptable for a strictly period before and after a change in medication régime,
observed trial of high dose medication on an individual suggested an advantage for long-acting medication
case basis. The principle behind the clinically based in terms of number of hospital admissions and of
observations of lack of effectiveness are that, essentially, inpatient days. In two retrospective observational
relevant neuron receptor sites become saturated, and studies identified, one showed an advantage in terms
so continuing to raise levels of drugs with an effect on of lower readmission rate for long-acting medication
such sites would be pointless. At the other extreme, and one did not, however the study which measured
could low dose medication suffice once symptoms are adverse effects found that patients receiving depot
under control and the issue is mainly one of preventing medications were more likely to need anticholinergic
relapse? The answer, according to a meta-analysis of drugs. This last finding may constitute an important
13 studies with 1,395 participants between them, is that caution when trying to ensure that patients continue
it depends on how low. ‘Very low doses will not suffice, to receive an uninterrupted supply of medication. Ho
but there appears to be no disadvantage to low dose’, et al. (2011) repeated neuroimaging of the brains of 211
although the authors argue that there are insufficient patients over 7–14 years, such that the average number
data for firm conclusions (Uchida et al., 2011). of scans available over time was three (range 2–5). They
compared the four clinical measures of illness severity,
2. Frequency of administration of medication. Could there length, quantity of psychotropic medication and alcohol
be advantage in giving doses of medication more or illicit drug use with brain volume on each occasion.
frequently, or, conversely with ‘medication holidays’? In each case controlling for the other three variables,
Remington and Kapur (2010) discuss the evidence, they found that alcohol and/or illicit drug use were not
finding a suggestion of advantage among stabilized associated with volumetric changes, severity of illness
patients of, say, alternate day prescription of oral had a small association with them, but length of illness
antipsychotics. They recommend more research in this and quantity of medication consumed had a relatively
field. large and significant association with loss of both grey
and white matter. Data are too few to make much
3. Combining antipsychotics. Polypharmacy is common, comment on long-acting SGAs (Fleischhacker, 2009);
with perhaps 30–40% of patients on more than one risperidone has been available in this form for about
antipsychotic medication (Goff and Dixon, 2011). 8 years, and is at least as safe as its oral counterpart.
A systematic review and meta-analysis of 18 RCTs, Use of olanzepine is less well studied, but risk of
with 1,229 participants, indicated an advantage for profound sedation (7/1,000 injections) means that the
polypharmacy, both in terms of effectiveness and depot variant should only be given if the patient can be
maintenance of treatment (Correll et al., 2009). This observed for at least 3 hours afterwards. Results of long-
seems counterintuitive, and against all advice that, in term trials of palperidone palmitate are awaited.
general, treatment with a single drug is to be preferred.
A more recent study (Essock et al., 2011) has, however, Medication, some conclusions
provided limited further endorsement for this. There is little evidence for clinicians’ conventional views
Treatment compliance and consistency was better on choosing medication which will suit most people in
maintained in the polypharmacy group, and the only terms of symptom relief and maintenance of psychological
disadvantage seemed to be a slightly greater weight gain
in the latter, perhaps best accounted for by the specific
drugs involved rather than the combination.
360 © 2014 by Taylor & Francis Group, LLC
Management and treatment
improvements without major adverse effects. In some ways they tend to be the most pro-socially minded. For these vari-
this is unsurprising, because of the nature of the research ous reasons not only is it unlikely that people with psychosis
studies, RCTs in particular, which are designed with the who also have difficulties with antisocial behaviours would
knowledge that incremental group differences between be recruited to trials of psychological therapies, but a number
treatments are likely to be quite small. In any case, as we of studies explicitly exclude them. A notable exception to the
note elsewhere, group data can only ever provide a basic latter rule is the study by Haddock et al. (2009), which explic-
guide in respect to an individual case. The fairly wide range itly recruited people with psychosis who had violent histories
of medications available, and styles of delivering them, (see also below).
coupled with the lack of clear cut group advantages means
that the clinician really does have to be well informed about Cognitive behavioural therapy
the nuances of both patient and treatment. In the everyday (CBT) for schizophrenia
practice of psychiatry, the availability of adequate time per Jones et al. (2004) conducted a systematic review of
patient, duration of relationship with the patient, and the randomized controlled trials of CBT for schizophrenia
means to actual shared decision-making may be difficult and other functional psychoses, providing they were not
to achieve, but a ‘need to medicate’ must not obscure the primary affective disorders. Their definition of treatment
process of constructing the necessary working alliance for these purposes was: a procedure that would enable
between patient and physician. It is also important to con- the recipient to establish links between thoughts, feelings
sider that some patients will want to minimize medication, and actions, monitor behaviour, and develop alternative
if not avoid it altogether, and that treatment should not ways of coping, the core goal being ‘correction of misper-
hinge on medications alone, or even be the main interven- ceptions, irrational beliefs and reasoning biases’. They
tion. Only if these contingencies are taken into account will identified 30 articles on 19 randomized controlled trials,
the patient be sufficiently well informed and in a position which together included 2,154 people as participants,
to make a real choice of medication in partnership with the more or less equally divided between treatment and
clinician whenever it is possible to do so. Real choice may control groups. The collective evidence from these trials
well prove to be the best predictor of treatment adherence was of only modest impact, with the principal advantages
and a long-term recovery process. in improving global mental state ratings in the medium
term and possibly reducing the length of inpatient stay.
Psychological Treatments None of the studies made any reference to violence dur-
ing the course of the treatment programme, even though
Even after varying medication type and doses, ensuring the CBT was generally lengthy – up to 2 years in duration.
compliance, and, where possible, having confirmation of This is surprising, given the number of people and length
blood levels of active components of the drug concerned, of time involved and a general assumption based on the
for some people the illness and violence will remain refrac- epidemiological studies described earlier that a conserva-
tory. Since the 1990s, there has been increasing optimism tive estimate would be of about 10% of the patients being
that more psychological approaches to treatment may help, violent. As noted, some studies are explicit about exclud-
and cognitive behaviour therapy (CBT) has been particularly ing people who had been violent or who were detained
favoured for people with schizophrenia or similar psychosis in hospital, but it seems unlikely that this alone could
(see also chapter 23). The literature on such psychological account for a wholly violence-free passage for over 2000
approaches is now quite substantial, and includes a number people. The importance of thinking about the possibility
of RCTs as well as observational studies, which, as specific of violence in such circumstances arises from findings
to schizophrenia, we will review briefly below. There are two discussed in more detail above, especially that a key fac-
limitations to the literature which are sufficiently general tor associated with violent action in the context of delu-
and sufficiently important for work with people who have sions was response to hypothetical challenge. People who
both psychosis and a propensity for violence that we will accommodated the challenge and/or strengthened their
introduce them here. The first is that trials of such treatment, belief were more likely to have acted violently in the 28
unlike drug trials, have rarely if ever considered the possibility days under study than those who ignored it (Buchanan
of adverse reactions to the treatment. Any treatment of any et al., 1993). Hypothetical challenge has much in com-
power may be expected to carry risks of unwanted as well as mon with methods of encouraging people to review their
desired effects, so this is a serious omission. The second is an symptoms within the framework of CBT. Among some
extension of the problem of medication trials, that evaluation smaller, uncontrolled studies of treatment in which delu-
tends to be done with people who have problems in the mid- sions appear to have been approached in this way, there
dle range – serious enough for there to be some prospect of are hints at further grounds for concern (e.g. Garety et al.,
change, but not so serious that they would be unable to com- 1994), although there are also anecdotal reports of CBT
ply or likely to break down in some sort of crisis during the being applied successfully and safely among people with
research – and participants in such studies must consent, so
© 2014 by Taylor & Francis Group, LLC 361
Psychosis, violence and crime
schizophrenia who are also serious offenders (e.g. Benn, Given the rate of comorbidities with psychosis when
2002). Notwithstanding its still useful content, the Jones offending is also a problem, in addition to cognitive
et al. (2004) review has now been withdrawn from the behaviourally based treatments more specifically for
Cochrane website, pending update. schizophrenia and similar psychoses, such approaches
could be useful adjuncts in the treatment of comorbidi-
Zimmerman et al. (2005) conducted a systematic ties, particularly problems with substance misuse. A 2010
review of CBT for symptom change in schizophrenia, schiz- Cochrane review found 25 randomized controlled trials
oaffective and delusional psychoses. They identified 15 tri- of various interventions, which had included nearly 2500
als (not all RCTs) in which, collectively, 515 people had had people, some of them offender patients. Drop out rates
CBT and 486 were in comparison groups. Meta-analysis were high. No one psychological intervention, singly or
of 14 trials revealed an advantage for CBT in reducing in integrated combination (e.g. motivational interview-
symptoms, with a higher effect size in acute cases (0.57) ing with CBT within an overall care package) appeared
compared with chronic cases (0.27). Improvement scores to have a particular advantage in reducing substance
were generally, however, given in terms of overall ratings use or improving mental state (Cleary et al., 2008; see
for positive psychotic symptoms, so it was not possible also chapters 8, 23). In addition, there have been calls
to tell which symptoms were most affected. Gaudiano’s for application of cognitive behaviour therapy-based
(2006) systematic review and meta-analysis identified approaches specifically for offending for people with
only 12 studies, and questioned the clinical importance of schizophrenia (e.g. Hodgins and Müller-Isberner, 2004).
statistically significant findings. The proportion of patients While we understand that trials of such interventions
showing significant and reliable scale score change in at may have started, to date, we believe there are none
least one psychotic symptom was significantly higher in reported in print. A comparison has been completed,
the CBT groups than in those receiving routine or alterna- however, between CBT and social activity therapy (SAT),
tive treatments, but clinically important changes did not the former aimed at reducing both psychotic symptoms
distinguish the groups. with their accompanying distress and anger and the
latter to help patients identify things they enjoy doing,
Wykes et al. (2008) tried to resolve any discrepancies and doing them (Haddock et al., 2009). Sixty-eight of 108
between individual and review study findings with a par- eligible patients, all with psychosis and a history of vio-
ticularly rigorous system of six separate meta-analyses, lence, completed at least 10 and up to 25 sessions of the
allowing for treatment and research method differences, therapies, and the follow-up period. Results suggested an
with data extracted from 33 of 34 studies identified. They advantage for the CBT but were quite complex. Delusions
found an overall advantage for CBT on the researcher but not hallucinations improved during CBT but the
defined target problem, on positive symptoms, on negative improvement was not sustained, while anger was unaf-
symptoms, on ‘functioning’ and on mood but not on hope- fected according to any measure. Groups were similar at
lessness, however when only blinded studies were consid- the outset in terms of previously recorded aggressive inci-
ered the effect was small (0.243, 95% CI 0.017–0.428). With dents; CBT had an advantage during but not after treat-
respect to positive symptoms of psychosis specifically ment in terms of numbers of recorded new aggressive
they were able to enter 24 studies, with 1,450 participants incidents, with verbal aggression principally accounting
between them, with a homogeneous treatment approach – for this pattern; physical violence analysed separately did
CBT delivered individually. There was a mean effect size not differentiate the groups during treatment, but CBT
of nearly 0.4 (95% CI 0.243–0.556). As other reviews, how- had an advantage during follow-up both for numbers of
ever, they did not distinguish between positive symptoms people violent and number of incidents.
in describing the impact, nor did they mention violence.
Efforts to modify delusions specifically continue. Ross et al. Other psychosocial treatments
(2011) offer the prospect of a way forward. They compared This is not the place for an exhaustive examination of
34 people with delusions and 24 comparison participants, all possible psychosocial treatments relevant for people
confirming that those with delusions were more inclined with schizophrenia. Some, such as social skills train-
to ‘jump to conclusions’, here measured by the amount of ing, we would regard in some form as simply part of the
information they requested before making a decision. A overall context of (re)habilitation, whether or not the
single session of training about neutral situations and the individual has offending as part of his/her cluster of prob-
idea that it would be preferable not to reach decisions too lems. Cognitive remediation therapy for schizophrenia
quickly was found to have a short-term beneficial effect on is a behavioural training based intervention that aims to
such behaviour. This did not translate into any significant improve cognitive processes (attention, memory, executive
change in flexibility of thinking or less conviction in the function, social cognition or metacognition) with the goals
delusion, but it could be argued that an absence of this of durability and generalization. This reflects observations
further effect is hardly surprising after one session only.
No behavioural correlates of those changes identified were
reported.
362 © 2014 by Taylor & Francis Group, LLC
Management and treatment
that impairments in cognition may affect longer term out- part of the person with schizophrenia, within or outside
come, including social functions and patients’ concerns the family.
about the impact such deficits may have on their everyday
lives. Given the findings that men who have psychosis and Psychosocial resources provided in settings run by
are also violent may have more structural and functional people with psychosis themselves are increasingly avail-
brain difficulties than those who have psychosis and no able, many with research and evaluation components;
problems with violence, and, in turn, than healthy controls such peer contributions vary considerably, although are
(see chapter 12), one might expect particular benefit from mostly focussed on individual work (e.g. Salzer et al.,
such treatment for people with psychosis who become 2010). Peer provided care emphasizes shared experience,
offenders. It hardly needs saying, but there have been no expertise in coping, and mutuality between the partici-
studies with psychotic offender groups. A number of sys- pants. It is beyond the scope of this chapter to review the
tematic reviews have suggested a general advantage for wide variety of models of service user run interventions,
social function with cognitive remediation, but this finding but they will no doubt continue to spread, and deserve
was not consistent with two major dissenters (Dixon et al., careful attention by researchers. These programmes
2009; NICE, 2009). This may arise partially from the differ- tend to be less likely than professionally directed offer-
ent ways in which the training is delivered and partly from ings to exclude people with histories of violence, yet
research methodology. Wykes et al. (2011) completed a they grapple with similar issues of staff, peer, and setting
systematic review and meta-analysis of all evaluations with safety.
a specified allocation procedure in the context of all partici-
pants receiving ‘standard care’, then allowing for sampling Summary on psychosocial treatments
and other potential methodological biases. They identified There are growing numbers of controlled studies of
40 studies with, collectively, 2,014 participants. They found various psychological treatments for schizophrenia and
that cognitive remediation had a small to medium effect on similar psychoses, generally on the principle that the
both cognition and function, which was durable, with best psychological treatments will be additional to medica-
effects obtained when the interventions were delivered at tion, but applying to people with acute as well as chronic
a time when the patients were clinically reasonably well illnesses. There have been several systematic reviews
and stable, and as part of a rehabilitation package rather of studies evaluating cognitive behaviour therapy for
than in isolation. There was no reference at all to violence schizophrenia, most concluding with an endorsement for
or offending. it, although the effect size is not impressive. Where there
is an effect, it often appears to be more general in terms
Family interventions may also be potentially useful. As of reducing, say, days spent in hospital rather than alter-
described above, for anyone with schizophrenia, social cli- ing specific symptoms of the disorder and, where there is
mate within the family may be affected by and affect the ill- evidence for symptom reduction, it is often in the broad
ness. Much family work has been around these issues, and terms of ‘positive symptoms’. An important issue is that
evaluated in these terms. Pilling et al. (2002) and Pitschel- studies vary both in the detail of the treatment given
Walz et al. (2001) provide evidence from systematic review and of the methods of evaluation. Over time, systematic
and meta-analysis that longer interventions – 6 months reviewers have attempted to take more and more of these
or more – are most likely to be helpful in reducing the risk issues into account. This is increasingly true for other
of patient relapse and/or rehospitalization, while Cuijpers psychosocial treatments too. Insofar as there is remaining
(1999) showed a reduction in family burden as a result of doubt about the strength and consistency of the evidence
such work with families. Subsequent controlled studies (e.g. for the value of such interventions it is more about which
Mueser et al., 2001), have shown that psycho-educational particular approach is more advantageous, delivered
programmes for families improve family relationships to individually or in groups, or for how long, and in com-
the extent that not only do the patients enjoy a better out- parison with which other treatment than whether adding
come, but also the relatives rate their burden as less. Garety some form of psychosocial treatment to medication is
et al. (2008), however, found no effect of family intervention beneficial. A caveat is that so little research, and only one
on symptoms or distress reduction in the patients, or on controlled trial, has been completed specifically with peo-
their duration of hospital stays or relapse rate. A Cochrane ple who have psychosis and are violent. The same sort of
review of RCTs identified a total of 53 of family interven- principle as put forward for medication would seem to be
tions; there were data from 2,981 people included in analy- the best way forward – that choice of psychosocial inter-
sis of frequency of relapse, which was lower in the family vention should, as far as possible, be taken in conjunction
intervention groups, and from 481 people in analysis of with the patient. Particularly for offender-patients, one
hospital (re)admission, which was also lower, and 696 peo- might add that, as far as possible, such decisions should
ple in analysis of medication compliance, which was better include relatives and others in the close social circle of
(Pharoah et al., 2010). We await studies, however, which are the individual.
specific for families in which there has been violence on the
© 2014 by Taylor & Francis Group, LLC 363
Psychosis, violence and crime
Models for a Treatment Framework (2006) found no advantage in terms of reducing misuse
even for enhanced assertive community treatment, when
Complex problems call for complex solutions, and there some of the intensive input was by specialist addictions
is as much need for research that can define and evaluate clinicians. Evaluation of assertive community treatment
the therapeutic framework for delivering specific treat- modified for offender patients similarly showed no advan-
ments as for research into the individual treatments. Few tage in terms of days in jail or hospital or in substance
such models are or will be unique to people with psychosis misuse (Chandler and Spicer, 2006; Lamberti et al., 2004).
who offend; they are commonly designed for people with a
range of mental disorders, and often a multiplicity of them. Economou et al. (2005) described outcomes after an
Nevertheless, as the individuals who have driven service ‘optimal treatment model’, referring to integrated phar-
change and the people who are most commonly treated macological and psychosocial treatments to cover the full
within forensic mental health settings in many countries range of need. In their small study of 50 people, just over a
are people with schizophrenia or similar psychoses, so the third had shown some form of aggressive behaviour at the
frameworks for treatment delivery are often designed with outset, although actual physical violence was rare and no
this group principally in mind. one had used a weapon. No serious violence emerged over
4 years, and aggression generally was reduced. El-Guebaly’s
Tyrer and Simmonds (2003), in a systematic literature (2004) overview of integrated treatment for those with
review, identified three UK trials of different models of a mix of psychosis and substance misuse problems took
care in general psychiatric services – early community account of the fact that outcome studies had come from
intervention versus hospitalization, community-focused a range of different countries, and argued that integration
care versus standard care, and intensive versus standard must reflect, as necessary, issues specific to local culture
case management – for people with severe mental illness and service provision. In the USA, the impact of ‘managed
or psychosis and comorbid personality disorder. Group care’ was evaluated, and found to offer no improvement
allocation did not permit adequate analysis of the lat- over more ad hoc arrangements (Dickey et al., 2003).
ter group, but the other two studies showed that, while Subsequently, a range of new measures has been introduced
outcome for the groups was similar in respect of hospital and evaluated, mainly with the aim of improving treatment
admissions, with an advantage for the community groups, compliance, for example the Mental Health (MH) Courts
both general social and offending outcomes were worse for (Christy et al., 2005; Cosden et al., 2005). Steadman et al.
the comorbid groups in the community treatment condi- (2011) compared 447 people who had been through the MH
tions. Assertive community treatment (ACT) originated Courts with 600 conventionally treated people drawn from
in the USA, particularly for people who had had repeated four counties – two in California, one in Minnesota and one
hospitalizations and/or recent homelessness, and there is in Indiana. They were equivalent on all measures on entry
evidence that it is effective in stabilizing people with such to the study, but the MH Court groups had fewer arrests or
problems to the extent of reducing recurrence of homeless- incarceration days in the following 18 months.
ness or hospitalization (e.g. Coldwell and Bender, 2007;
Nelson et al., 2007). Replication of this approach, with high ‘Outpatient commitment’, also known as ‘mandated
staff:patient ratio, high frequency of patient contact, the community treatment’, similar to a community treatment
multidisciplinary team being available around the clock, order in the UK, has also received considerable attention
seeing people in the community, working with their own in the USA (Erikson, 2005). Hiday (2003) provided a useful
social networks where they have them and actively seeking overview, and there have been a number of important sub-
patients in the community in the event of their failure to sequent studies. A substantial study of 8,752 new orders and
contact staff as agreed, has not shown the same advantages 5,684 renewals were made in New York State under ‘Kendra’s
in the UK (Fiander et al., 2003). The latter group argued that law’ (see also chapters 24 and 25) where the terminology is
the model fidelity was good. Again, as with so many stud- of ‘assisted outpatient treatment’ (AOT) (Swartz et al., 2009).
ies in this field, interpretation of the absence of significant The orders are for a group of high risk people who have
differences from other approaches may better reflect that either been in hospital or jail at least twice in the preceding
the treatment models are as good as each other rather 36 months and/or made acts and/or threats of harm to self
than a failure of the newer approach under trial. Given that or others. Being on an order made no difference to how they
the people needing treatment services do not constitute a viewed mental health services, but advantages with respect
homogeneous group, availability of more than one treat- to treatment engagement, use of medication, maintenance
ment approach with an equivalent prospect of helping in the community and a reduction in arrest rates were not
can only be an advantage. As the approach was applied to only observed during the orders but also sustained if the
patient groups ever more similar to offender patients with orders had been for more than 6 months.
psychosis, however, even in the USA, the results become
more equivocal. With people who have substance misuse Mandated community treatment of any kind, and per-
difficulties as well as psychosis, for example, Essock et al. haps innovative supplementary incentives to treatment
compliance in particular, require ethical debate as well as
364 © 2014 by Taylor & Francis Group, LLC
Management and treatment
evidence for effectiveness. Geller et al. (2006), for example, in terms of numbers needed to treat, it would take 85
discussed the risk that, in the USA, mandated or ‘assisted’ outpatient commitment orders to prevent one readmis-
outpatient treatment is simply ‘deinstitutionalized coer- sion, 27 to prevent one episode of homelessness and 238
cion’, but observe important differences between states, to prevent one arrest.
with some having competency base legislation in this Community treatment orders of various types continue to
respect (e.g. Massachusetts) while others have dangerous- proliferate in US and EU countries, yet there are few data on
ness focused statutes. Winick (2003) analysed the issues which to judge their effectiveness in preventing or reducing
from a jurisprudential position (see also chapters 2, 3 and violence associated with psychosis.
5). Benefits for people with major mental disorder, includ-
ing disability income and housing, are made dependent on Criminal justice system models of management
their attendance for treatment. Swanson and colleagues Reports between 2002 and 2006 on jail diversion pro-
(2006b) have shown, among over 1,000 people, that those grammes and their effectiveness have come almost exclu-
with mental disorder and some evidence of violence per- sively from the USA (Broner et al., 2004; Draine et al.,
petration, but also those with demographic characteristics 2005a; Steadman and Naples, 2005). The latter group stud-
that may just mark them out as unfortunate rather than ied over 1,600 participants in a number of states, focusing
necessarily at risk for violence, are disproportionately particularly on people with psychosis and co-occurring
likely to be subject to such ‘leverage’. Robbins et al. (2006) substance misuse disorders. Costs of the programmes
interviewed 200 outpatients, however, and found that the were similar, but diversion reduced time spent in jail with-
participants who had housing used as leverage were much out compromising public safety. Draine et al. (2005a) com-
more likely than those who had not to believe that housing pared a diversion programme with treatment in jail for a
leverage is effective in helping people to stay well. similar, but smaller and single, state (Pennsylvania) group.
They highlighted likely differences in people referred to
A repeated concern expressed is that, even if there the services. Active psychotic symptoms increased the
are risks of harm to and by people with major mental likelihood of diversion, while those treated within the in-
disorder, the prospect of coercion into treatment may be jail services were more likely to have had a track record of
a deterrent against seeking help, and thus, indirectly, a being managed in some part of the criminal justice sys-
danger in itself. Swartz et al. (2003) evaluated this propo- tem. The take-up of services in a larger (baseline sample
sition among 85 members of staff and 104 patients in 2000), multi-site US study suggested a similar bias (Broner
one state in the USA. Among the staff, only 80% thought et al., 2004). Such issues will need consideration in all such
that coercive measures were more likely to make their studies. Police response is another issue that has been
patients with schizophrenia stay in treatment. Among taken up, at the interface between the criminal justice
the patients, there was considerable lifetime experience system and people with major mental health problems
of some form of coercion (63%), and over 30% reported (Sellars et al., 2005).
that fear of coerced treatment was a barrier to their
seeking help. The barrier principally related, however, Community re-entry models have been considered
to compulsory inpatient treatment; coercion as an out- from the perspective of people leaving jail (Draine et al.,
patient was not such a barrier, whereas threats, or what 2005b). Frustrated by the crude outcome measures in
Swartz and colleagues more tactfully refer to as ‘recent most studies of people with psychosis and other major
reminders or warnings’ of coerced treatment were as bad mental disorders, relying almost wholly on mortality and
as the experience of coerced treatment itself. Link et al. re-offending, we conducted a grounded theory study to
(2008) found somewhat differently among their study of establish a testable theoretical model of the process of
people subject to Kendra’s Law in New York. In particu- discharge from high security hospitals (Jamieson et al.,
lar, they reported: 2006). A substantive theory emerged from open coding and
constant comparative analysis of interviews with a range of
self-reported coercion increases felt stigma (perceived professional people, inclusive of a lawyer and Home Office
devaluation–discrimination), erodes quality of life and officials as well as a range of clinicians. The theory was of a
through stigma leads to lower self-esteem. continuum between pathological dependence and healthy
Phelan et al. (2010) reported a modest improvement for independence, with the role of the various professionals
people who were in outpatient commitment, but con- being facilitation along the pathway towards healthy inde-
cluded that neither the absence of adverse effects nor pendence, and factors such as re-offending entering the
the presence of such modest improvements supports the model as terminators of independence. This could provide
expansion of coerced outpatient treatment. Kisely et al. ’s the basis for improvement in outcome measures of future
(2011) Cochrane Review of outpatient commitment found studies of secure hospital care and treatment, and perhaps
only two trials that met criteria, and concluded that the related treatment packages.
evidence did not establish the efficacy of coerced outpa-
tient treatment. They noted:
© 2014 by Taylor & Francis Group, LLC 365
Psychosis, violence and crime
Summary of impact of models for Specific treatments have occasionally been evaluated
treatment and management within a group of people who have psychosis and prob-
While the specific treatment literature tends to focus on lems with violence, but, more often, inferences still have
disease, and pay little regard to accompanying behaviour to be made from studies in which violence is either not
disorders, including violence, the study of complex treat- mentioned or from which those people with psychosis
ment models, or frameworks within which specific treat- who have been violent have been explicitly excluded.
ments may be delivered, leans in the opposite direction. Complex treatment models, or frameworks for treat-
Work commonly includes people with psychosis but is ment, continue to emerge and be evaluated, here often
not specific to them. The areas of initiative are too wide- including people with psychosis but not exclusive to
ranging for simple conclusions, but treatment models that them. Our tasks are to take our concepts further, make
include an element of coercion or explicit incentive do our measures better, include people who have been vio-
appear to carry advantages in reducing violence or other lent in more research samples, and to learn how to ask
antisocial behaviours not only by the person suffering from more pertinent, potentially useful questions. With both
psychosis but also, as noted earlier, towards him or her. frameworks for treatment and specific treatments, even
in group analysis it is hard to draw definitive conclusions
Conclusions about what works best. A more positive interpretation
is that there is a good deal that works better than doing
Detail about the frequency with which homicide and nothing, and clinical skills lie in working with individual
other violence is associated with schizophrenia has been offender-patients to identify the best combination of
consolidated and developed, together with a good deal treatment framework, specific medication and psycho-
of new insight into how the symptoms of illness and/ social interventions that will be likely to work in that
or comorbid conditions may variously explain, at least case, and then monitor each person’s progress with a
in part, how the violence comes about in this context. good researcher’s rigor.
366 © 2014 by Taylor & Francis Group, LLC
15
Pathologies of passion and related
antisocial behaviours
Edited by Written by
Pamela J Taylor Paul Mullen
Pamela J Taylor
The pathologies of the passions of love, jealousy and sufferer. In his view, the disorder represented ‘an exaggera-
entitlement may all form parts of major mental illness. tion to the extreme limit of the amorous passions’ (p.339).
When so, they are usually without any obvious external Over generations, the concept changed. De Clerambault
provocation, are accompanied by other features of that (1942) probably had the greatest influence on current for-
disorder, and follow a course closely linked to that of the mulations of this disorder. He shifted the emphasis to an
underlying disorder. Sometimes, however, they are the exclusive focus on delusions that the object of the patient’s
sole feature of a person’s disorder, generally emerge after disordered affections loves him/her and initiated the sup-
some real provocation, although not necessarily one that posed relationship. De Clerambault’s viewpoint is parroted
would appear important to a casual observer, and in the in the international (ICD) and US (DSM) disease classifica-
context of some susceptibility in personality trait, a previ- tory systems. This is a pity, as it conforms poorly with clini-
ous sensitizing experience, low mood or some combina- cal reality, which is served better by the wider formulations
tion of these; the course is dependent in part on the actual which preceded his theorizing (Krafft-Ebing and Chaddock
situation and the reactions of others. People with these 1904, originally 1879; Kraepelin 1921, originally 1913).
pathologies of passion have in common associated convic-
tions that one or more other people are abrogating their Erotomanics, like all lovers, endow the loved one with
rights, which must be asserted. The preoccupations and those qualities that make them an object of delight. For
pathologies often overlap; it is a short step for some from the erotomanic, reality plays little part in constraining
love to jealousy and sense of entitlement, although perhaps the process of endowing, so an idealized other can be
jealousy and certainly entitlement may arise in relation to constructed (Mullen, 2008a). The judgments of the eroto-
other desires. These states are of particular importance to manic, again unconstrained by reality, create the certainty
forensic psychiatry, because people who suffer with them of mutuality, or mutuality to come, or even of the lover as
tend at best to frighten others and at worst to harm them, infatuated pursuer. The erotomanic is secure in his/her
and are almost invariably intrusive. Furthermore, these constructed attachment so holding the delights of being
pathologies not uncommonly involve clinicians as the in love uncontaminated by the uncertainties evoked by
object of the passion, whether as ‘love’ object or obstacle real relationships. The desires of the erotomanic are often
to love and/or justice. for an ethereal love, abstracted from the challenges of the
carnal, but not always so. Fantasy plays the major role in
Erotomanias and the world of erotomania. Exploration of the fantasies is
morbid infatuations important, as occasionally they can reveal unexpected and
unpleasant possibilities.
The emergence of stalking as a category of problem behav-
iour gave erotomania renewed prominence. Previously Erotomania can be characterized as involving:
regarded as a rare syndrome, anti-stalking laws brought to ●● either a conviction of being loved despite the supposed
light many previously unrecognized cases (Mullen et al.,
2008). lover having done nothing to encourage or sustain
that belief but, on the contrary, having either made
Esquirol (1965, originally 1845) coined the term eroto- clear their lack of interest or remained unaware of the
mania for the irrational sentimental attachment to some- claimed relationship;
one who in reality has little or no relationship to the ●● or an intense infatuation without necessarily any
marked accompanying conviction that the affection is
currently reciprocated;
© 2014 by Taylor & Francis Group, LLC 367
Pathologies of passion and related antisocial behaviours
●● a propensity to reinterpret the words and actions of the particularly striking – over half of the 76 men but under 5%
object of their attentions to maintain the belief in their of the 170 women.
supposed romance;
For the forensic clinician, erotomania is chiefly of inter-
●● preoccupation with the supposed love which comes to est for its possible progression into pathological jealousy
form a central part of the subject’s existence. and/or stalking behaviours. Most accounts of stalking
behaviours prior to the twentieth century concern intimacy
These three essential criteria are often accompanied by: seeking by erotomanic or infatuated pursuers rather than
the more aggressive behaviours of the jealous and queru-
●● a conviction that the claimed relationship will lant rights seekers. The historical perspective is considered
eventually be crowned by a permanent and loving union; more fully below.
●● repeated attempts to approach or communicate with
the supposed lover.
Periods of intense infatuation may occur in normal indi- Jealousy
viduals, particularly in adolescence, but fade when it is
clear that no favourable response is to be expected from the Jealousy deserves a privileged place in forensic mental
beloved. The teenage ‘crush’ lacks the conviction of even- health, firstly because of its close relationship to two of the
tual fulfilment, even though fantasies of such fulfilment commonest forms of serious criminal violence: domestic
are common, and acts as a pleasurable embellishment of violence and the killing of sexual partners (Kingham and
life rather than as a preoccupying and disruptive element. Gordon, 2004) and secondly, because of the way laws have
Teenage crushes are often social experiences which are traditionally tended to mitigate the severity of punishment
shared with likeminded peers and pursued through groups accorded jealous men who have killed their partners, by
and clubs. This is in stark contrast to the isolating nature of appeals to justification, provocation and, most recently,
pathological infatuations (Mullen, 2008a). diminished responsibility.
Erotomanic preoccupations are probably more com- Jealousy has failed to find a place among the risk factors
monly secondary to a range of disorders, most frequently which comprise the myriad instruments, computer pro-
in the schizophrenia spectrum; monodelusional states may grams and lowly checklists which claim to be able to assign
account for about a quarter of cases (Rudden et al., 1990) the probability of future violence to an individual (Mullen
but this, like so many of the earlier studies, may be biased and Ogloff, 2009). Experience should, though, shout a
through drawing exclusively on treated patient samples. warning whenever you confront a patient caught up in the
Such cases can often be accommodated under the heading passion of jealousy, particularly if he, or she, has a history
of delusional disorder, or paranoid state. The problem with of assault, or has made threats. Threats constitute another
this is that though the preoccupation may dominate the harbinger of violence usually omitted from the technology
life of the sufferer, it does not always have the phenomeno- of violence prediction (but see chapter 22). Those who have
logical characteristics which establish the experience as killed or maimed in one relationship out of jealousy acquire
delusionally based. In such ‘primary’ erotomanics there is a terrible propensity to repeat the performance in future
usually a pre-existing vulnerability in the form of an over- relationships.
sensitive and self-referential individual, living a life bereft of
intimacy and without close confidantes. Given this potential importance of jealousy and its
pathological extensions, it is perhaps surprising that it
There is a traditional view that erotomania is more currently has a rather low profile in psychiatry in gen-
common among women, although it has always been rec- eral and forensic mental health in particular. This was
ognized as occurring in men too (Taylor et al., 1983). Brüne not always the case, for, in classical psychiatry, jealousy
(2001) confirmed a roughly 7:3 ratio, albeit only among the occupied a prominent position, perhaps never more so
246 cases published in the psychiatric literature between than when Karl Jaspers (1910) used jealousy both in his
1900 and 2000, with an obvious potential for bias in such a initial essay on the phenomenological method and to
highly selected sample. His more interesting observations illustrate the division of mental pathologies into develop-
are about the extent to which, in this series, the different ments, reactions and processes. The decline in the profile
presentations between men and women reflect the sexual of jealousy reflects many of the same qualities which gave
strategies theory of Buss and Schmitt (1993), covering dif- it interest for classical psychiatry. Jealousy fits ill with
ferences in mating strategies between men and women in today’s attempts to fit psychopathology into the commit-
healthy populations. Most of his predictions were sustained tee created boxes of the DSM and ICD systems of defini-
by the data, including the older age of onset among the tion. Attempts to distribute the pathologies of jealousy
women and the greater likelihood that they would involve across a range of today’s sanctioned diagnostic categories
older, higher status men, retain a single love object and as a symptom work only as long as jealousy which exists
remain chaste. The difference between men and women in in, and only in, the pathological aspects of the jealousy
likely progression to ‘forensically relevant behaviour’ was itself, is ignored. Unfortunately for those therefore turned
368 © 2014 by Taylor & Francis Group, LLC
Jealousy
away by psychiatrists as ‘not ill’, it is exactly these types of 2. a state of arousal occasioned by such an intrusion
pathological jealousy which so often manifest in propensi- which may involve fear and sadness at apprehended
ties to violence. These include the intense jealousy which loss, anger at betrayal, pride at possessing a desired
is evoked on the slightest of suspicion, the jealousy which object, but, usually, above all, the distress of uncertainty;
totally preoccupies, the jealousy which generates extraor-
dinary delusions of infidelity (calling that a delusional dis- 3. a conflicting group of desires compounded in varying
order is comforting but obviously circular), the jealousy of proportions of the desire to know, to deny, to expose,
obsessive doubt, the jealousy of paralysing fear, and that of to repossess, the desire for revenge and, ultimately, the
uncontrollable rage, and the absurd but potentially lethal desire to resolve the jealousy either through reassurance
jealousy sometimes found in people who are senile and/or that the suspicions were false, or acceptance of the
have brain damage. infidelity with subsequent reconciliation, or separation;
Normal Jealousy in the Modern World 4. fantasies which, even among the general population
may involve vivid mental images of the partner in sexual
Jealousy is a complex emotion generated by a perceived congress with the actual or supposed rival, or of various
threat to a valued relationship (White and Mullen, 1989). revenge scenarios;
The threat is usually, but not always, experienced as coming
from a rival. Jealousy can and does emerge in close relation- 5. predispositions to behave in ways which serve one
ships which are not sexual (Hill and Davis, 2000). The val- or more of the desire to know, to strike back, or to re-
ued relationship is, however, most frequently sexual and/ establish the lost or threatened intimacy – thus, actions
or intimate and as this is where violence is most frequently follow, which may include spying, cross-questioning,
evoked, such romantic jealousy will be the main focus here. checking, threatening, actual assault (almost always
directed at the partner), making yourself more attractive
In everyday speech, jealousy and envy are occasionally or derogating the rival.
conflated. Jealousy is about that which you believe you pos-
sess but fear you may lose. Envy is about that which you do The problem for mental health professionals lies in decid-
not have but desire. Envy comes in two forms: one positive ing where the limits of normal jealousy end and the realm
and one negative. Aristotle suggested that the positive is of the pathological begins. A simple and appealing solu-
the desire is to become like, or acquire the properties, such tion such as ‘jealousy which leads to violence is patho-
as virtue, success or status of another, perhaps resulting in logical’ could leave clinics overwhelmed, given research
the action of emulation. The negative is destructive envy, in which suggests some 15% of the population report being
which the desire is to dispossess the other of the properties subjected to assault as a result of a partner’s jealousy
you believe make them superior. The critical distinction is (Mullen and Martin, 1994). Jealousy motivated violence
that jealousy is about fighting the apprehended loss, and continues to be normalized as an unfortunate mani-
perhaps destroying the object of desire if it becomes truly festation of love, which attracts far less condemnation
unattainable, but envy is about seeking to gain what others than non-jealousy related aggression (Puente and Cohen,
possess and you desire. 2003). Equally, any attempt to distinguish on the basis of
whether the feared infidelity is based on actual infidelity
Most adults have been in intimate relationships and wilts before the knowledge that jealousy in both its patho-
most have experienced jealousy at some time and to logical and its normal variants may either be in error or
some degree (Mullen and Martin, 1994). It is perhaps the correct in its accusations (Odegard, 1968).
commonplace nature of ordinary jealousy that makes it
so difficult to recognize any shift towards pathology. It The behaviours commonly associated with jealousy in
hardly takes the skills of a psychiatrist to recognize the the general population can provide at least a signpost to
madness in a firmly held belief that mutant rodents are pathology. In a questionnaire survey of over 350 community
circulating in the blood stream because of having eaten residents, representing a higher than 60% return rate on
cheese contaminated by alien rays; the firmly held belief the questionnaires, everyone endorsed at least one jealousy
of a mildly unkempt and obese man about the infidelity of item and 40% acknowledged having experienced jealousy
his attractive wife with one of their sleek and prosperous without good cause (Mullen and Martin, 1994). Nearly 15%
neighbours may hardly seem outlandish, but may none- of men and 19% of women considered that their partner’s
theless be a delusion. The jealousy of the normal popula- jealousy had caused problems. When jealous, 30% of the
tion is primarily about the loved one, not the rival, and participants admitted repeatedly cross-questioning their
usually involves the following: partner in an attempt to catch them out in a lie. Some 10%
1. judgment that an intimate relationship in which you also described phoning to check the partner was where
they said they were, turning up unexpectedly, and even
believe your partner owes you loyalty is threatened by a searching their belongings for evidence. Today, checking
potential, or actual, relationship to another; the partner’s mobile phone SMS data base and recent calls,
plus careful examination of credit card records, would have
to be added to the relatively common behaviours of the
© 2014 by Taylor & Francis Group, LLC 369
Pathologies of passion and related antisocial behaviours
ordinary jealous individual. What was very uncommon, depend primarily on who your father was, and who his father
or never admitted, was opening mail, following, checking was before him. The hierarchy of such societies rests on
underclothes for signs of semen or other incriminating ensuring female fidelity. Infidelity is a challenge to the social
stains, or examining the partner’s body for indicators of order, and jealousy serves the purposes of social stability.
recent sexual activity. These are all behaviours frequently Jealousy may be experienced by the individual (man), usually
described in pathological jealousy. as a challenge to the honour of his family and caste, but is
also understood as a threat to the community. Agrarian soci-
Another area that may help at least suggest when jeal- eties tolerate, or even mandate, the violence of jealousy both
ousy is pathological lies in the nature of the fears evoked as punishment and as a warning to others. Their laws reflect
when infidelity is suspected. In the general population, this reality (Northrop, 1960; Smith and Weisstub, 1983).
jealousy is about a plausible rival and the commonest fear
is loss of the partner, feared especially by men, followed by In mercantile societies, social castes and rigid hier-
loss of attention and alienation of resources. Fear of loss of archies are undermined by the power of wealth and pos-
intimacy was more often cited by women, but loss of sexual sessions. Though birth still constrains the individual’s
exclusivity was not related to gender (Mullen and Martin, possibilities, the acquisition and loss of wealth may radi-
1994). In pathological jealousy, fears of being shamed (made cally alter social relationships. In such societies, marriages
to look a fool), cheated, lied to and generally humiliated for the powerful become less about ensuring blood lines
dominate, although fear of loss of the partner to another is and more about acquiring financial advantage. Infidelity
usually also acknowledged. The egocentricity of pathologi- for the wealthy represents a contractual breach, though for
cal jealousy can be mimicked by the self-absorption of some rich and poor it still encompasses a challenge to honour,
personality types. but now embodied in personal rather than social status.
Jealousy becomes individualized and, for the law, ceases to
In the latter part of the twentieth century, at least in the be relevant to social duty and takes on the guise of a per-
West, it seems that adultery has become a sport indulged sonal provocation which can engender justified violence.
in by the majority of the population (Lawson, 1987). It is
so common that both the statistically sophisticated and In industrial society, the further diffusion and obfusca-
the cynical are likely to assume its existence, even in the tion of the roots of power and influence almost remove
absence of evidence. The ways in which people deal with infidelity from any role in determining social position.
actual sexual infidelity often determine whether relation- Jealousy ceases to be part of the public realm and becomes
ships survive. Jealousy in such a social context is likely to entirely a matter for the citizen’s private life, in which nei-
be viewed as a problem with few positives, particularly as ther society nor law have a legitimate interest. Jealousy is
jealousy is just as likely to fix on the innocent as the guilty, stripped of all social relevance and becomes a private and
and may well precipitate the unjustly suspected into the isolating experience whose utility becomes marginal at
very behaviour of which they stand accused. Jealousy may best. In late industrial society, jealousy has undergone a
still claim the virtue of being a cry of pain at disappointed final stage of its inversion to become a piece of personal
hopes, but even that raises the question of the wisdom of psychopathology, disruptive of the free market in goods,
investing hope in such a fragile quality as sexual constancy. services and people, damaging to happiness, and destruc-
tive of good order. For the law, jealousy comes finally to
The History of Jealousy rest as a potential mental disorder, perhaps mitigating
behaviour, perhaps, in the event of a killing, indicating
The notion that emotions have histories, in the sense diminished criminal responsibility.
of having been experienced differently at different his-
torical periods, is totally foreign to the view of emotions These neat divisions into social societal models, such
taken by orthodox psychology. Emotional tendencies are as agrarian, mercantile, capitalist, or post-industrial, are
regarded as being part of inherited nature, perhaps modi- idealized rather than entirely distinguishable historical
fied by early attachment experiences. They are viewed epochs. Today’s world contains societies in which elements
as representing fixed responses to specific situational of all such systems exist and in which the influences and
triggers, short-circuiting the slower reactions medi- ideas from earlier stages of economic development con-
ated by judgment. There is, however, plenty of evidence tinue to resonate, even after a shift to a new phase. The
to support emotions such as anger, love and jealousy most obvious mediating variable is religion. All the world’s
having been experienced differently with changing eco- great religions developed against the background of agrar-
nomic, social, and cultural realities (Singer, 1966, 1987; ian, or even more basic nomadic, systems. Though some
MacIntyre 1988; Stearns, 1989; Mullen, 1991). reinterpretation of the moral injunctions of such systems
has occurred, in important ways they still reflect the reality
Romantic jealousy is tied to the meaning of infidelity. of the societies in which they were generated. Thus, they
In agrarian economies an individual’s position in society embody powerful injunctions against infidelity, mandate
is largely determined by his/her parentage. Being born to the control, if not active repression, of female sexuality,
power or to servitude and your place in the social system and place great value on sexual continence in general.
370 © 2014 by Taylor & Francis Group, LLC
Jealousy
The greater the influence of these religions in any modern younger, and the general opinion that the husband had
society, the greater are the contradictions which afflict been an angel to tolerate her flagrant infidelities.
the responses to infidelity and the experiences of jealousy
(Vandello and Cohen, 2003). Theology, ideology, and the Pathological jealousy which emerges as the sole or
impact of social and economic realities compete for the primary abnormality may be more difficult to separate
experience of jealousy, making each individual’s jealousy a with confidence from the more flamboyant manifestations
fascinating insight into our culture. of the extreme variates of jealousy which fall just within
normal limits. Indicators that such jealousy has reached a
Pathological Extensions of Jealousy pathological level include
1. The insatiable nature of the passion. A person with
Pathological jealousy, then, may be a psychological state,
in and of itself, in which it forms the only, or the primary ‘normal jealousy’ generally finds resolution, be that
disturbance in the individual’s mental state, and not only through accepting reassurance, accepting infidelity
a symptom emerging out of a pre-existing condition such has occurred and seeking separation or reconciliation,
schizophrenia or chronic amphetamine abuse. The distinc- or even enacting specific elements of revenge or
tion is perhaps best illustrated by a comparison of two reparation. In the pathological state, nothing can satisfy
cases. the jealous desires. The revelations are never sufficient
and some detail, some unacknowledged act or residual
The first case illustrates something of the psychopa- affection is always being sought. For the pathological,
thology of symptomatic jealousy – jealousy which has its no evidence can be sufficiently weighty to establish
origin in mental illness. innocence, separation is impossible, but the extent of
the guilt remains forever in doubt. Pathological jealousy
A man with a long history of intermittent obsessional continues, even beyond the grave of the suspected
symptoms related to fears of contamination began to partner. Those who have killed from jealousy rarely, in our
develop anxieties about his wife’s fidelity. These con- experience, accept they may have been in error, nor cease
cerns would almost always intrude into his thinking searching their memories for further pointers to infidelity.
at work in the form of a vivid mental image of his wife 2. The pattern of behaviour evinced by the jealousy falls
in the arms of another. He would resist these insights outside of that which is acceptable in the individual’s
as absurd and developed a ritual of mental arithmetic cultural and social context.
both as distraction and as ‘proof ’ of the falsity of the 3. The jealousy becomes totally preoccupying, disrupting
intrusive image; falsity that is if the calculation was normal function.
completed correctly. Unfortunately the patient had high 4. The suspicions fall on an implausible rival, or a
level mathematical skills and gradually escalated the multiplicity of implausible rivals, who are often believed
complexity of the calculation to a point where either to consort with the partner under conditions, and
errors occurred, or he could not be sure whether he had within time spans, which are difficult to reconcile with
obtained the right answer via a faulty process. His anxi- the possible let alone the probable.
ety would build to the point when he felt compelled to 5. The jealousy has a course and evolution that is difficult
rush home to reassure himself. His wife was exposed on to relate reasonably to the supposed provoking events
an almost daily basis to her husband bursting in on her and subsequent events, even taking the most generous
at home, in shops, and at friends’ houses, tearful and view of the situation.
apologetic for his absurd behaviour. The jealousy itself
never rose above a suspicion, recognized as almost cer- Jealousy and Domestic Violence
tainly false. There were no associated threats, violence
or even intrusive enquiries. At no point was his partner Domestic violence and spousal homicide emerge from a
exposed to worse than embarrassment. Yet the jealousy complex concatenation of influences in which sexual jeal-
was clearly pathological; here, it formed the content of ousy may be prominent, though of itself rarely sufficient
an obsessional disorder. (Buss, 2000; Jewkes, 2002). Jealousy is common, infidelity is
Selection of jealousy as the content of a delusional or other common, but violence, though far too common, occurs in
disorder of mental state is unlikely to be purely random. only a minority of such cases, and killing is an extreme rar-
The jealousy probably represents aspects of the individual’s ity. Jealousy may well be the prime motivation for an act of
life experience or personality vulnerabilities which predate violence but this still leaves an open question as to why this
the emergence of active symptoms. In another example of individual, on this particular occasion, resorted to force.
jealousy as a symptom of mental disorder, an elderly man
with dementia developed delusions about the infidelity of A community study of jealousy found that 15% of both
his wife, bizarre because she was even more elderly and men and women had, at some time, been subjected to phys-
disabled than he was. A colleague who knew the family ical violence at the hands of a jealous partner (Mullen and
was able to describe the wife’s colourful reputation when Martin, 1994). In a study carried out in Scotland, nearly half
of 109 battered women interviewed identified the excessive
© 2014 by Taylor & Francis Group, LLC 371
Pathologies of passion and related antisocial behaviours
possessiveness and sexual jealousy of their partner as the Careful and repeated questioning of the jealous indi-
precipitant of violence (Dobash and Dobash, 1980). Two- vidual and his/her partner is advisable. Specific enquiry
thirds of the women at a refuge for battered women in the should be made about:
London area reported that their partner’s excessive jealousy ●● threats;
was the primary cause of the violence and that, in many ●● damaging the partner’s, or rival’s, personal property;
cases, the partner’s suspicions were entirely without foun- ●● throwing objects;
dation (Gayford, 1975, 1979). Studies from North America ●● pushing, shoving or shaking;
produced similar results. Hilberman and Manson (1977), for ●● blows with hands, fists or feet;
example, reported that extreme jealousy contributed to the ●● threatening with a (potential) weapon;
violence in most of their group of 60 battered women, and ●● throttling;
Rounsaville (1978) noted similar findings with just over half ●● the possession of firearms or other weapons;
of the battered women listing jealousy as the main problem ●● attacks with weapons;
and no less than 94% naming it as a frequent cause. In one of ●● any other action which could have inflicted harm (e.g.
the few studies in which men were asked why they battered
their partners, the men most frequently nominated anger driving at him/her with a vehicle or trying to produce
at supposed infidelity (Brisson, 1983). Whitehurst (1971), an accident whilst partner was a passenger; poisoning).
reporting on 100 cases of spousal violence, noted that in The last point is important as, in this area, the improbable
nearly every case the husband appeared to be responding does occur. One elderly lady caught up in jealous suspicions
out of frustration at his inability to control his partner but about her husband of over 30 years was so dismissive of
that the overt justification, and accusation, was that the earlier enquires about violence it seemed silly to ask if she
partner was sexually unfaithful. From such studies, jealousy had ever tried to harm or even kill her husband by other
would appear to be both a motive for domestic batterers means. In the event she happily described the unsuccessful
and also an excuse, or rationalization, for the violence. attempts to kill him with rat poison in his meals.
When violence has occurred, the intent, the context and
The statistics on violence and jealousy in the general the nature of the damage inflicted must be noted. The level
community are grim, but evidence from samples of people of the victim’s fear is only a guide when it appears high in
who are pathologically jealous is worse. A UK study of all relation to acknowledged actions; it should not be taken as
138 psychiatric patients with pathological jealousy admitted reassuring if it is apparently low, despite escalating aggres-
to the Bethlem Royal and Maudsley hospitals over a 14-year sion, indeed, in such cases it is arguable that the victim is
period (1967–80), none referred from the criminal justice sys- exceptionally vulnerable, as s/he is unable to perceive the
tem and only six referred because of a violent incident, found risks for her/himself. In most cases, violence is preceded by
that more than half had a history of having assaulted their clear indicators of mounting danger. These may be ignored
partners and only 15% of the men and 27% of the women or downplayed by the partner who cannot believe they are
had neither threatened or enacted violence against partners at risk from their loved one. A prudent clinician should not
(Mullen and Maack, 1985). In a later, US sample of 19 men make the same mistake.
and one woman, Silva et al. (1998) similarly found that most The features in a jealous individual which increase con-
had harmed their spouse. cern about violence include:
Jealousy has emerged as one of the more frequently ●● escalating conflict between the couple;
identified motives for homicide in a number of studies ●● threats;
(Gibbens, 1958; West, 1968; Wolfgang, 1958). Daly and col- ●● a history of violence in the domestic context in
leagues (1982) concluded that male sexual jealousy is the
commonest motive for killing in domestic disputes. Jealous particular;
homicides are usually perpetrated by men, with the usual ●● fantasiesofviolentretributionorstrongimpulsestoattack
victim being their female partner, although same sex homi-
cidal jealousy is increasingly recognized as an important (however reassuring s/he is about never ‘really doing it’);
issue, and may be a particular concern in secure settings, ●● depression, especially in the presence of suicidal
where the sexes are segregated. Where domestic disputes
generated by jealousy lead to women killing, it is claimed preoccupations or behaviour;
this is typically an act of self-defence to ward off the male ●● substance abuse;
partner’s jealous rage (Daly and Wilson, 1988). ●● where the cultural and social background of the jealous
Assessment of the possibility that violence will occur in individual is one which tends to condone resort to
the context of jealousy is important, always. In the context violence in the face of infidelity.
of pathological jealousy rates of violence toward the partner In general, jealous preoccupations tend to be more intense in
are so high that there have to be very good – and clearly doc- younger people, who also resort more readily to violence. In
umented – reasons not to make the assumption that it will pathological forms of jealousy, however, advancing years
happen, and to devise a management strategy accordingly. do little to ameliorate the risks of violence. Gender differ-
ences are largely in terms of the seriousness of the damage
inflicted rather than the frequency of assaultive behaviours.
372 © 2014 by Taylor & Francis Group, LLC
Stalking
Apprehending violence is one thing, preventing its name, unsolicited gifts, spreading unfounded rumours,
realization another. The option of admission to hospital, vexatious complaints and legal actions, ‘cyber-terrorism’,
whether voluntarily or compulsorily, offers one protec- and, ultimately, assault.
tive route but, when pathological jealousy is not clearly
secondary to a severe mental illness, this may be difficult. Stalking has only emerged as a significant social prob-
Both mental health review tribunals/boards, as well as fel- lem since the 1990s. Initially it claimed public attention
low clinicians, on occasion, take a broad view of ‘sanity’, with respect to celebrities, especially after the killing
particularly in face of delusional disorders. This may result in 1989 of television star Rebecca Shaefer by her long-
in a detained patient returning early to the community, time stalker Richard Bardo. Later, stalking became more
untreated, but now with additional ‘cause’ for grievance, associated with domestic violence and the pursuit of
and thus more convinced than ever of the evil machina- ex-partners, but a more balanced view now prevails. The
tions of the partner. Negotiating separation with a couple behaviour is recognized as occurring in a range of con-
embroiled in conflict over jealousy is usually difficult and texts and against a wide range of victims (Lowney and
frustrating, even at the limits of normality, but certainly Best, 1995; Mullen et al., 2008).
when jealousy becomes pathological, it often reflects a
relationship which, although conflicted, is also intensely History
involved. Partners in such a relationship are not easily sepa-
rated; even with less over-involved couples there may be, Stalking is not a new behaviour. There are clear descrip-
despite clear warnings, little appreciation of the risks. It is tions in novels, in psychiatric texts, in law reports, and
helpful to consider also the needs and, perhaps the pathol- even in autobiographies dating back centuries (Alcott,
ogy of the victim-partner. Perhaps attraction to jealous men 1997; Esquirol, 1965 orig 1845; Kierkegaard, 1987 orig 1848).
reflects a fear of being unlovable; the fact that these men do What made it emerge in the 1990s as a matter of public
take a real interest in their partners and most particularly in and forensic concern? One possibility is a decreasing tol-
what they do and how they spend their days may at first be erance of threatening or otherwise disruptive behaviours,
very reassuring. By contrast, people who are less prone to another the increasing value placed on personal privacy,
jealousy show the more usual spectrum of concern for their but the possibility that its newfound prominence reflects
partners – ranging from mild interest to polite indifference. an increasing prevalence of the behaviour in today’s society
is worth considering; community studies suggest far higher
In those not labouring under frank delusions of infidelity lifetime rates are reported by younger cohorts (e.g. Budd
it is sometimes possible to alter behaviour and de-escalate and Martinson, 2000).
tensions simply by enumerating the risks and sharing one’s
anxieties about the future conduct. Where delusions are The commonest form of stalking has become that by
present, then medication for the jealous partner may help a rejected ex-partner. Such stalking usually emerges in
pave the way for psychosocial interventions to have some the context of the breakdown of a sexual relationship in
prospect of assisting the couple, and easing the torment of which one partner, almost always the man, either refuses
both. Separation, at least until the ideas/delusions are under to accept that the relationship is at an end or embarks on
some control, must always be considered, although it may be a course of harassment to express his rage at rejection. The
hard to enforce as the object of the jealous ideas may some- probability of such stalking is likely to increase as the fre-
times be as hard to persuade as the jealous. (For more on quency increases of people entering and, more importantly,
management, see Crow and Ridley, 1990; Dolan and Bishay, exiting relationships. Stalking by a rejected male may also
1996; Kingham and Gordon, 2004; Pines, 1992; Mullen, 1995; reflect the changed balance of power between men and
de Silva, 2004; White and Mullen, 1989; chapter 22.) women in the Western world where, in sexual relationships,
female choice is beginning to replace female acceptance
Stalking and subservience. In societies with high separation and
divorce rates, the risk of one partner refusing to accept the
Stalking is a problem behaviour characterized by one per- end of the relationship gracefully is higher too, and this
son repeatedly imposing unwanted contacts and/or com- brings with it the possible resort to stalking.
munications on another person in a manner which creates
fear, or at least significant distress. The contacts can be in Another currently common form of stalking involves
the form of approaches, following, loitering nearby, and men, usually lonely and unattached, pestering and follow-
keeping under observation. The communications are com- ing young women to whom they are strangers, in the hope
monly by phone calls, SMS, letters, email, notes attached of starting a relationship. Such behaviour may reflect a
to property and graffiti. The imposed contacts and com- society in which less socially adept and prepossessing men
munications form the core of stalking behaviour, but there are finding it more difficult to establish relationships, while
are a range of associated behaviours. These include threats, being immersed in a culture which places a high value
ordering or cancelling goods and services in the victim’s on visible indications of sexual relationships, or at least
activity. It may also reflect increasing suspicion between
neighbours in urban environments, where we live among
© 2014 by Taylor & Francis Group, LLC 373
Pathologies of passion and related antisocial behaviours
strangers. Finally, those who stalk out of resentment for became virtual prisoners in their own homes, afraid
some perceived injustice or mistreatment may actually to go out or answer the telephone or doorbell;
be becoming more frequent in societies which emphasize 7. resilience – some victims report heightened appreciation
individual rights and promote unattainable expectations. of family networks or development of self-confidence
through learning new skills, such as self-defence;
The Impact of Stalking 8. resource impact; in the USA, efforts have been made
to estimate the costs from medical, psychiatric
The central reason to take stalking behaviours seriously and/or social care needed, through time lost from
is the damage they can inflict, or presage, for the victim. work to the judicial system costs of pursuing the
Data are now available from studies both of selected sam- stalker. In the early part of the new millennium, the
ples and random community samples. Initially, studies societal cost in these terms was estimated to be
focused on self-identified victims of stalking who sought $342 million (Centers for Disease Control, 2003).
help, joined stalking support organizations, or responded The impact of stalking on the stalker is rarely considered, but
to adverts placed by researchers (e.g. Bjerregaard, the stalker usually also pays a high price for what is generally
2000; Blaauw et al., 2002; Hall, 1998; Kamphuis and a futile, time wasting, and resource consuming enterprise.
Emmelkamp, 2001; Pathé and Mullen, 1997). The studies Stalkers can disrupt their own lives to an almost similar
report broadly similar results. A number of random com- degree as they disrupt the lives of their victims. Stalking
munity samples have not only determined the prevalence can become an all encompassing preoccupation which puts
of stalking but investigated aspects of the impact on vic- the stalker at risk of alienating those social supports that
tims (e.g. Budd and Matttison, 2000; Dressing et al., 2005; s/he has, and eventually brings significant legal sanctions. In
Kuehner et al., 2007; Purcell et al., 2002, 2005). Though managing the stalker remembering that stalkers also suffer
rates of overt anxiety, post-traumatic and depressive from their behaviour is important in reconciling one’s role as
disorders are lower than in studies of clinic samples, the agent of the community and agent of the patient.
psychological and social impact of stalking is neverthe-
less considerable. Among a random sample of German The Epidemiology of Stalking
women, for example over 40% acknowledged anxiety
and sleep disturbance, over a third psychosomatic symp- The answer to the question ‘how common is stalking?’
toms, and only slightly lower proportions acknowledged depends on how the behaviour is defined, the methods of
depression (28%) and/or panic attacks (12%) (Dressing ascertainment, and the population investigated. In practice
et al., 2005). The impact of threats and violence more what is counted is the victim recall of stalking, usually in
specifically is discussed below. terms both of lifetime experiences and of 1-year prevalence
(period prevalence).
Stalking, then, if it persists for longer than a few weeks,
will create psychological and social problems for the vic- The earliest study confined itself to women stalked by
tim, which, in some cases, may be both long-lasting and men (Australian Bureau of Statistics (ABS), 1996). A lifetime
disabling. In their systematic review and meta-analysis of rate of 15% for women was reported, with 2.4% having been
175 studies of stalking, Spitzberg and Cupach (2007) found stalked in the previous year. The ABS repeated this study
eight main categories of effect: a decade later using very similar methodology except, on
1. general disturbance, in which they included PTSD; this occasion, men were included. In this later study, 19% of
2. affective health, referring to such problems as anxiety, women and 9.1% of men reported having ever been stalked
(Australian Bureau of Statistics, 2005). The increase is
anger and depression; partly explained by explicitly including same sex stalking in
3. social health, covering relationship difficulties; the second survey, which was not done in the first study. A
4. cognitive health, incorporating such states as confusion, first US study was on a larger scale and included a random
community sample of men and women (US Department
suspiciousness, self-esteem or suicidal ideation; as of Justice, 1997; Tjaden and Thoennes, 1998). Using a more
with members of so many victim groups, their distress restrictive definition, they reported lifetime rates of 8% for
is often compounded by self-blame – for choosing women and 2% for men. The British Crime Survey (BCS)
such a partner in the first place, for not managing first incorporated enquiries about having been the victim
the separation more adroitly, for not recognizing the of stalking in 1998, when lifetime rates of 11.5% were found
problem, for failing to deal appropriately with the (women 16.1%, men 6.8%) with 2.8% stalked in the year
unwelcome advances, and much, much more; prior to interview (Budd and Mattinson, 2000). Dressing
5. physical health, including appetite or sleep disturbance, and colleagues (2005) reported a similar lifetime rate of
but also addictions; 11.6% with an annual rate of 1.6% in a random community
6. behavioural disturbance – victims often react to sample in Mannheim, Germany.
an ongoing sense of threat by decreasing social
outings and work attendance, or changing residence
and/or place of employment; the worst affected
374 © 2014 by Taylor & Francis Group, LLC
Stalking
The various studies generally agree about the sex distri- the stalking is sustained by the communications
bution of victims (70–80% females) and perpetrators (80– and enforced contacts becoming a substitute for
85% males). The apparent numerical discrepancy here is the lost relationship, and a parody of past intimacy.
accounted for by same sex stalking, which occurs in some
20% of cases, and more frequently involves male on male Rejected stalkers are usually men (80–90%). The lost
than female on female. There is more variation between relationship was usually sexual and intimate, though
studies in the frequency with which victims are pursued by occasionally another family relationship, or close friend-
ex-partners, strangers, or acquaintances. In the BCS, 29% of ship. If these are broken unilaterally, they may engender
victims were pursued by former partners, 32% by acquaint- such stalking. Rejected stalkers rarely have psychotic
ances and 34% by strangers (Budd and Mattison, 2000). disorders, though personality traits of overdependence,
Studies are consistent in reporting higher lifetime rates narcissism, or obsessiveness are frequently encountered.
of stalking among younger respondents. This could be an The influence of jealousy and domestic violence in the
issue of recall or even willingness to construct experiences relationship prior to separation and subsequent stalking
of harassment in terms of stalking. The discrepancy per- remains uncertain, in part because several studies have
sists however even for those exposed to very lengthy (more conflated stalking-like behaviours during the time the
than a year) and often distressing experiences of stalking, partners are living together with such behaviours after
which suggests the probability that rates of stalking have separation. Though this approach is defensible it seems
increased since the 1990s. to ignore an essential element: that stalking is about
imposing one’s presence in situations where one has no
The duration of stalking varies widely from a matter of legitimate right to be (Mullen et al., 2008).
days to years. A careful analysis of duration suggested there 2. The Intimacy Seeking Stalker. Here, the behaviour
may be two separable types of stalking (Purcell et al., 2002, emerges in a context of loneliness, in which the stalker
2004). One group stalk for less than 2 weeks, with a modal experiences him/herself) as bereft of love or com-
duration of a couple of days; people in the other group who panionship. The initial motivation is to establish a
persist beyond 2 weeks have a modal duration of 12 months. close relationship, usually romantic, though occasion-
The briefer episodes of often intense harassment are usually ally one of friendship, or maternal or child-like inter-
perpetrated by strangers (75%) who predominantly employ dependence is the goal. The victim is usually a stranger
approaches and following. The longer episodes of stalking or acquaintance selected from celebrities, public fig-
are usually by prior intimates and acquaintances (80%) and ures, or casual and professional contacts. Health pro-
involve a range of unwanted communications and contacts. fessionals are favoured targets of such misplaced
As might be expected, significant social and psychological affections. The stalking is sustained despite the lack of
damage is virtually confined to those victims pursued for response, or outright rejection, because the fantasy or
longer than 2 weeks (Purcell et al., 2004). delusion of a close relationship is so much better than
no intimacy at all.
Stalking Classifications and Typologies
This is the only type of stalking in which women
A variety of stalking classifications have been advanced predominate. These stalkers have high rates of psycho-
on the basis of the supposed underlying psychopathology, pathology, including pathologically erotomanic infatu-
nature of the prior relationship or variety of stalking behav- ations. This may be the most persistent of all types of
iours employed as well as mixing elements of the behaviour stalking. Violence is uncommon but not unknown.
– for example motivation – together with psychopathol- 3. Incompetent Suitors. This type of stalking also comes out
ogy (Boon and Sheridan, 2001; Canter and Ioannou, 2004; of loneliness, and sometimes also sexual frustration.
Harmon et al., 1995; Mohandie et al., 2006; Sheridan and The initial motive is to establish contact, usually with a
Boon, 2002; Wright et al., 1996; Zona et al., 1993, 1998). The stranger encountered in a public place, such as a shop
typology employed here has obtained wide currency among or entertainment venue, in the hope this will lead to
clinicians (Mullen et al., 1999; Pinals, 2007). This has the friendship and/or a sexual relationship. The approaches
following five characteristic patterns which, though not tend to be crude and insistent, and may evoke fear or
entirely mutually exclusive, do provide a basis for under- revulsion. The stalker is blind, or indifferent, to the cool
standing and managing most cases. response and continues to pester even in the face of
1. The Rejected Stalker. In this case, stalking emerges in the obvious distress. The lack of a positive response coupled
with increasingly active resistance from the victim
context of the breakdown of a close relationship. The usually eventually makes it clear to the stalker that his/
pursuit of the former intimate is initially motivated by her efforts are fruitless, and they stop. Occasionally the
the search for reconciliation, or exacting revenge for more insensitive will persist for weeks, though most go
rejection, or a fluctuating mixture of both. The stalking away after a day or so. This type of stalking can be more
may persist way beyond a time when reconciliation or persistent when some semblance of a relationship,
revenge seem plausible motives, at least in part because however casual, preceded the unwanted approaches.
© 2014 by Taylor & Francis Group, LLC 375
Pathologies of passion and related antisocial behaviours
This type of stalker is socially incompetent, often likely to be in the context of assessing sex offenders
as a result of intellectual or interpersonal defects, but than from a referral specifically because of stalking.
occasionally just youth. The difference between the Serial rapists of the more organized type almost always
approaches of an overenthusiastic, gauche and inex- have histories of predatory stalking, not infrequently
perienced suitor and the stalking of the incompetent with some victims of the stalking escaping the ultimate
suitor is the indifference or blindness not just to the attack. The victims of a predatory stalker may be una-
lack of interest on the part of the target figure, but also ware of the unwanted attentions, but some will recount
to their actual distress. This margin may be narrow, unease and events which led them to suspect they were
but it is important. Spitzberg and Cupach (2001) have being followed and kept under observation.
described what they term ‘obsessional relational intru- 5. The Resentful Stalker. All the stalker types described
sions’. These, they suggest, are found among young so far are motivated by the desire to form, or reform,
people attempting unsuccessfully to negotiate their a relationship, however impoverished, one sided,
way through society’s courtship rituals. In unaccepta- and potentially destructive. Resentful stalkers are
ble levels of courtship persistence, the would-be suitor different. Their dramas are played out like all stalking,
ignores the signs of lack of interest in part because in a dyadic relationship, but their motivation is to
they misread the messages. This may be because they damage the other and not to relate to him/her. It
assume outdated cultural stereotypes, for example that might seem that a person from this group might be
apparent lack of interest is coquetry. The emphasis in equated with a vengeful ex-partner but, in practice,
their mode of acting is on reciprocity, and their tar- the latter, even if predominantly aiming to harm,
gets may actually provide inadvertent encouragement, has at least aspects of nostalgia for what was lost.
for example through misplaced politeness or because
they initially felt flattered. Employing a similar model, Resentful stalking emerges in the context of events
Sinclair and Frieze (2000, 2005) studied a large cohort which the stalker experiences as unjust, humiliating,
of students who had been either the subject, or object, and injurious. Such stalkers focus their animus either
of unrequited affections. They concluded that the on someone who they blame for a perceived injury, or
stalking behaviours which may emerge in such situa- on a representative of the group they hold responsible.
tions depend on this mix of failures to read correctly The initial motive is to strike back, often surreptitiously.
the intentions and responses of the other party and The stalking is sustained because of the sense of power
blindness to the distress caused. On the edges of the and control which comes from harassing the victim,
normal range, most of the clinically incompetent stalk- and from a self righteous conviction they are fighting
ers have an extreme level of inappropriate intrusiveness back against an oppressor. People from this group of
and of indifference to the victim’s responses. Unusual stalkers often have sensitive and self-referential person-
persistence, of longer duration than the typical brief alities, and/or fall into one of the paranoid spectrum
episodes of stalking of this type may occur because of disorders, though usually into well-organized types.
obsessional relational intrusions and/or prior acquaint-
anceship. Incompetent suitors often have personality Stalking Among Juveniles
vulnerabilities in the schizoid or antisocial spectrum,
with Asperger’s syndrome not infrequent. No reliable data exist on the frequency of stalking behav-
4. The Predatory Stalker emerges in the context of sexual iours by juveniles but existing evidence points to it being
deviation, usually related to fantasies of rape and substantial (Mullen et al., 2008). McCann (1998, 2000,
subjugation, but occasionally of child molestation. 2001) was the first investigator to suggest that stalking by
The initial motivation is related to the selection of a juveniles could be just as damaging to victims as that by
victim and the acquiring of information about that adults, with threats and assaults being at least as frequent.
victim preparatory to launching an attack. The victim Purcell and colleagues (2009) studied the records of some
is selected on the basis of fit with the predator’s 300 juveniles appearing before the children’s court follow-
desires and fantasies. The stalking is prolonged ing stalking behaviours. Most perpetrators were male (64%)
beyond the time necessary for selection, information and victims female (69%). The mean age of the stalkers
acquiring and planning, to obtain gratification from was some 15 years, but that for victims several years older.
the voyeurism, the rehearsing of the planned attack Almost all of these young people stalked someone they
in fantasy while watching the victim, and the sense knew (98%); ex-boyfriends/girlfriends made up a fifth of the
of power and control obtained from observing the victims but, unlike adults, same sex stalking was common
victim, who remains unaware of the oncoming danger. (57%). Threats and assaults were surprisingly frequent (75%
and 50% of cases respectively), though this may reflect the
Predatory stalkers are men, often with histories of court based data set.
deviant sexual behaviour and/or prior sexual offend-
ing. Clinical encounter with a predatory stalker is more Juvenile stalker typology has much in common with
its adult counterpart, but the commonest form among
376 © 2014 by Taylor & Francis Group, LLC
Stalking
young people appears to be an extension of bullying, police, who will generally be responsible for the first steps
with the purpose of taking the tormenting beyond the in applying them in practice.
school context into the victim’s home and social life.
The juvenile resentful stalker might be more accurately Cyberstalking
described as ‘retaliatory’, as, in the Purcell series, it usually
involved a more direct and rapid response to a supposed Cyberstalking has attracted great public interest, but little
insult or injury; this was the next most frequent (22%). systematic research. Definitions offered of this behaviour
Threats were common as were harassing phone calls and vary from the parsimonious: ‘the use of information com-
approaches but, like the adult equivalent, actual assault munication technologies to harass and intrude on others’
was infrequent. Among rejected stalkers (22%), the per- to the all-inclusive, which incorporate mention of a range of
petrator was usually male and assault common (44%) and sexual, aggressive, and even terrorist activities (Barak, 2005;
occasionally seriously injurious. A group not seen among Bocij and McFarlane, 2002, 2003; Ogilvie, 2000). Sheridan
adults was of youths who harassed multiple victims, often and Grant (2007) conducted a systematic examination of
spanning schoolmates, neighbours and acquaintances and the differences between cyberstalking and other forms of
frequently targeting adults. These were an unhappy, angry stalking. They reported that most cases of cyberstalking
and delinquent group of young people, often with histories formed one element among more familiar stalking behav-
of conduct disorder, and at war with their world (Mullen iours rather than a distinct type of activity.
et al., 2008). Infatuated stalkers, seeking intimacy, were
uncommon among youths. Cyberstalking behaviours include:
1. sending repeated, unwanted, and disturbing emails or
The impact of being stalked can be just as serious
among adolescent as adult victims. In fact, given its SMSs;
potential to disrupt social behaviour and academic perfor- 2. ordering goods and services on a victim’s behalf;
mance at critical moments in development, its long-term 3. publicizing private information of a potentially damag-
effects may be more damaging. Stalking among juveniles
needs more research, particularly research not starting ing or embarrassing nature;
from a premise that this is a normal developmental vari- 4. spreading false information;
ant of no great significance for perpetrator or victim. 5. information gathering online about the victim;
6. identity theft;
Stalking and The Law 7. encouraging others to harass the victim;
8. attacks against the victim’s computer and its data bases.
Stalking has been criminalized in most Western juris- The prevalence of cyberstalking is hard to estimate,
dictions. Stalking offences, however, present problems particularly as the various attempts have used widely
for the law. Usually each of the acts which make up the discrepant definitions and ascertainment methods. One
offence of stalking, such as phoning, sending letters or of the more informative approaches has been to study
approaching with banal requests, are, in themselves, the frequency of receipt of repeated harassing emails; this
legal. Furthermore, the intentions of the stalker are often experience has been reported by about 25% of stalking vic-
lawful – such as seeking reconciliation or attempting to tims who had been drawn from US college samples (Alexy
establish a relationship – and the stalker often regards et al. 2005; Fisher et al., 2000; Spitzberg and Hoobler, 2002).
his/her activities as offering friendship or love or other-
wise well intentioned. A critical element in the definition Stalking of Celebrities and Public Figures
of the offence is the victim’s statement as to how they
were affected by the behaviour, but some of those who The definition of stalking has to be modified when applying
are stalked have an obvious mental disorder by the time it to public figures. Superstars and the truly powerful are so
the case comes to the attention of the law. Those who well-screened by security services that they are often una-
draft anti-stalking statutes often struggle to frame the ware of being stalked. The fear criterion has, therefore, to
offence in a manner which deals with what is a criminal be transferred to the raising of significant concern for their
offence, partly victim-defined, in which guilty intent may protective services.
be absent, and which only becomes criminal by virtue of
the repetition of lawful acts. The result can be a law which A high public profile attracts unwanted attentions. A
either creates multiple loopholes for the well-represented Dutch study of public figures suggested a third had been
offender or, conversely, a breadth of coverage which stalked (Malsch et al., 2002). A study of television person-
places the harmlessly enthusiastic, or reasonably enquir- alities in Germany reported even higher rates (Hoffmann
ing, at risk of prosecution. It is perhaps most important, and Sheridan, 2005). For those in elevated political or
if such laws are to provide the protection victims of stalk- state positions, like the US President or the British Queen,
ing need, that they are understood and accepted by the stalkers are a constant problem ( James et al., 2008; Mullen
et al., 2009; Scalora et al., 2002a,b). Public figures attract
the unwanted attentions of intimacy seekers and the
resentful, but in addition they are a magnet for a mixed
© 2014 by Taylor & Francis Group, LLC 377
Pathologies of passion and related antisocial behaviours
bag of publicity seekers, campaigners and the deluded. It romance, or resentful stalkers, who believe they have been
is unsurprising that the famous and powerful can attract let down, mistreated, or humiliated by the professional and
attention and become the focus of a wide range of emo- are pursuing retribution.
tions from admiration, through emulation, to envy, and
detestation. Interest in, or even preoccupation with, a There are now a range of studies of the prevalence of
public figure only becomes problematic when it is either stalking of mental health professionals (Galeazzi, et al.,
an all absorbing fixation and/or motivates damaging 2005; Gentile et al., 2002; McKenna et al., 2003; Purcell
behaviours. These include: et al., 2005; Regehr and Glancy, 2011; Sandberg et al., 2002).
1. intimacy and intimacy seeking, with the desire for or Though the study methods varied, all reported high rates
of stalking victimization. Purcell and colleagues (2005), in
assertion of an amorous or advisory relationship; the methodologically most adequate study, reported that
2. unusually persistent petitioning, from belief in a nearly 20% of clinical psychologists had been stalked by
a patient. The highest rate was among those preparing
cause or a highly personal grievance; this group is reports for accident compensation or benefits, when they
characterized by an intense sense of entitlement, might have been acting for the agency rather than the
which easily translates into resentment when hopes are person who became a stalker. The male preponderance
disappointed; it overlaps with querulous complainants; among stalkers was about half that in general population
3. delusions about the right to the position or title occupied samples (37% : 63%). Resentful stalkers formed the largest
by their target, and a need to assert that (‘pretenders’); group (42%) followed by intimacy seekers (19%). Most
4. a perception of the public figure as the cause or solution clinicians were pursued for months, using a range of stalk-
of their problems, in the first case already ‘persecuted’ ing methods; a third were threatened and 10% assaulted.
and in the second, when disappointed in their hope of Perhaps equally troubling was that one-third of the stalk-
protection, shifting to a view of their one time saviour as ers made vexatious complaints to registration boards and
the prime persecutor; health boards.
5. the chaotic, for whom the motivation is felt as intense,
but remains unclear because of the disorganized and The Royal College of Psychiatrists has carried out
fluctuating nature of their mental state in general and a postal survey of its membership (Mullen et al., 2010;
their beliefs in particular. Whyte et al., 2008), using a conservative definition of
Those who stalk and intrude on public figures are far more stalking, involving extended pursuit and multiple intru-
likely to have serious mental illnesses than those who sions. About 10% of responding psychiatrists reported
plague more mundane targets. This may be because of having been stalked; with a broader definition, the level
the prominence of intimacy seekers, who have the highest was over 20%, and one in three reported harassment. Most
rate of mental illness among the various types of stalking, (71%) of the stalkers were patients, with the next largest
or because fame and power attracts those with a range of category being relatives or friends of the patients; just 5%
grandiose, and persecutory, and identity delusions who experienced stalking by partners/ex-partners and 5% by
would be unlikely to fixate on the less famous or powerful. colleagues. In more than half the cases (58%) the stalking
Some public figures attract literally hundreds of fixated went on for more than a year, and most (65%) experienced
people who make repeated attempts to communicate or significant anxiety or fear.
establish contact. Threatening behaviour is not uncom-
mon, but only a tiny minority will ever present a threat to Minimising stalking behaviours against clinicians
the safety of anyone, except themselves. This is little com- Studies to date thus suggest that being stalked is a signifi-
fort, however, without some way of identifying the danger- cant professional hazard. Falling victim to such behaviour
ous few and managing their risk effectively. Considerable is unrelated to experience or gender but mainly reflects the
efforts have been made to distinguish those who pose a patients encountered. Avoiding being stalked can only be
threat from the accompanying crowds of harmless nui- assured by not seeing patients. The impact may be mini-
sances (Dietz et al., 1991a,b; Fein and Vossekuil, 1995, 1998, mised, however, in several ways.
1999, 2003; Phillips, 2006; Meloy et al., 2008; Scalora et al., 1. Early recognition of being stalked. Some professionals
2002a,b; see also chapter 22).
seem to tolerate a range of inappropriate communica-
Stalking of Health Professionals tions and contacts outside designated appointment
times, for lengthy periods and without apparent con-
Health professionals in general, and mental health pro- cern, so stalking may be well established before even
fessionals in particular, are brought into daily contact being recognised.
with people who suffer distress and disorder, particularly 2. Discussion of concerns with colleagues at the earliest
the lonely, the angry, and the self-absorbed. Their stalk- moment. This allows checking the perceptions of being
ers, therefore, from this population, are usually intimacy stalked with that of more detached colleagues. Usually,
seekers, occasionally seeking parental caring rather than but not always, it will confirm the behaviour is stalking;
378 © 2014 by Taylor & Francis Group, LLC
Stalking
colleagues, including receptionists and security staff, with names of potential witnesses where possible. If
need to know about the stalking, first so they may assist, the case goes to court, such material is invaluable (see
and secondly to protect themselves. Pathé, 2002).
3. Telling family and friends. Again, they cannot help you 9. Take sensible security measures. Over-enthusiasm about
or protect themselves if they don’t know about the security can add little in the way of safety but adds
stalking. greatly to the victim’s own fears.
4. Transference of the care of the patient to a colleague.
It is important neither to abandon the patient nor to Evaluating and Managing Risk in Stalking
continue working personally with him or her in such
circumstances, so transfer to a colleague, preferably at Risk in stalking is usually considered only in terms of the
another clinic, is essential at the earliest opportunity. chances of escalation to violence to the victim. There is
The patient must be informed personally of the transfer, indeed, as described, some risk of this, but the stalking
ideally by the senior clinician, even if that clinician is victim faces other important risks too, especially those of
the victim. A colleague may be present if the stalker is significant psychological and social harms when stalking is
considered to pose a risk of violence, or even security persistent or recurrent.
personnel in extreme circumstances. In our experience
a non-confrontational approach is best, at least initially, The stalker also faces future hazards, which include
advising the patient that his/her communications continuance and/or escalation of their stalking until it
and unscheduled contacts have created anxiety; the undermines his/her social and psychological functioning
word ‘stalking’ should not be used with the patient, and/or incurs criminal sanctions.
for a number of reasons, but especially because it is
not therapeutic for the patient to see an anxious and Initial and somewhat tentative approaches have been
frightened therapist. This is not the time for discussion, made to apply actuarial, structured clinical judgment, and
debate or argument about possible gaps between classification and regression tree approaches to predict-
intention and reality (‘you may not have intended to ing violence by stalkers (Kropp et al., 2002; Meloy et al.,
frighten Dr X but you have’), but for firm action and 2001; Mullen et al., 2006; Rosenfeld and Lewis, 2005; see
clear transfer to a new treatment situation. Any refusal also chapter 22). None is entirely successful in the clinical
to accept this on the part of the patient should be met situation. The approach we employ involves:
with a courteous but unambiguous statement that 1. assigning stalkers and victims to high- or low-risk
continuing with the current arrangements is not an
option. Then the patient must be informed that there groups, applying broadly relevant risk factors such as
can be no further attempt to communicate with or prior relationship, presence of prior threats or violence,
contact the professional concerned and that breaking and whether the intensity of the stalking is increasing,
that rule could lead to legal consequences. decreasing or apparently stable;
2. monitoring, where possible, the level of fear produced
We owe our colleagues who are stalked the assis- in the victim; this may be a useful indicator of
tance of accepting the referrals of such patients. It is a escalation, and is certainly not a factor to be dismissed;
low-risk kindness as the stalker is unlikely to transfer 3. identifying current risk factors and future hazards
the unwanted attention to the new clinician. In over which are potentially remediable;
300 stalkers seen in the Melbourne service only one has 4. using the information derived from this process to
gone on to stalk a clinician in the service. inform management strategies.
5. Break all direct contact with the stalker. However The following is a list of empirically established and clini-
reasonable and inoffensive any further approaches or cally identifiable risk factors for the main problems.
queries from the patient may seem, the stalked clinician a. Persistence or recurrence: risk is associated with the
must cease all contact once the patient has been stalker being female, either rejected or intimacy
transferred to a colleague’s care. seeking in type, with a duration already exceeding 2
6. Resort to civil or criminal law if the stalking persists. weeks, and multiple methods of intruding, particularly
Restraining orders and other civil orders to not deter letters and unsolicited gifts (Hart et al. 1999; McEwan
most committed stalkers, but may stop the less dedicated. et al., 2007; Mohandie et al., 2006; Purcell et al., 2004).
The protections of the criminal laws, if available and if b. Psychological and social damage to the victim: risk is
properly administrated, are the most effective. increased by the stalker being male and the victim
7. Meticulous record keeping. Assume that the stalker will female, a prior intimate relationship between stalker and
make an official complaint at some stage and there will victim, a greater type and number of intrusions, threats
be an enquiry. and assault, victim withdrawal from social contact and
8. Retain evidence of all unwanted communications. Retain work roles in response to the pursuit, and victim failure
tapes of phone calls and record the unwanted contacts, to seek help from the law, stalker support groups, or
experienced therapists (Blaauw and Sheridan, 2002;
© 2014 by Taylor & Francis Group, LLC 379
Pathologies of passion and related antisocial behaviours
McEwan et al., 2007; Pathé and Mullen, 1997; Purcell Persistent complainants
et al., 2005). and vexatious litigants
c. Assault: risk is increased when stalker and victim are ex-
intimates, the stalker is of predatory type, has marked Morbid querulousness is a pattern of behaviour which
personality pathology and/or abuses substances, has is characterized by unusually persistent pursuit of a per-
breached restraining and other court orders, threats to sonal grievance, which becomes damaging to the person
assault have been made, there is escalating intrusive- in pursuit and damaging and disruptive to the pursued,
ness, and victim response includes counter-threats and to social structures such as courts or other agencies
and confrontation (Mohandie et al., 2006; Purcell seeking to resolve the dispute. There are three broad types
et al., 2002; Rosenfeld and Harmon, 2002; Rosenfeld and of querulous behaviour, which are not necessarily mutu-
Lewis, 2005). ally exclusive: the laying of persistent complaints without
Ideal management of stalking will provide treatment and recourse to the law; the persistent presentation of petitions,
risk reducing interventions for the stalker and support in which generally includes harassment of people perceived
various forms for the victim, usually through different clini- to be powerful; and seeking redress through the courts.
cal services or agencies. The victim needs practical advice, A wide variety of possibilities may form the focus of the
may need treatment for anxiety and depressive symptoms grievance, but clinicians are probably most familiar with
as they arise, and sometimes needs dissuading from self- hypochondriacal claimants, who particularly direct their
destructive responses (Mullen et al., 2008; Pathé, 2002). A resentment against doctors who they believe have failed to
first step for the stalker lies in establishing a relationship cure them, or even harmed them, by giving the wrong treat-
which is sufficiently secure to allow the clinician to lead ment or withholding the correct one. In common with the
him or her to accept s/he is stalking and damaging both other conditions considered in this chapter, the primary
the victim and self. In this context, it becomes possible to sufferer believes that s/he is entitled to something – here
establish treatments, as appropriate, for general factors some goods or rights rather than love or intimacy, not
which may be fuelling the stalking, including: having attained that entitlement; persistent complainers,
●● treatment of any mental disorder; petitioners or litigants then believe they must seek redress,
●● treatment of any substance misuse disorder; which they pursue relentlessly. From time to time, failure to
●● treatment for personality traits or social skills deficits; get the relief they seek boils over into violence against those
as well as interventions more specifically directed at the perceived to be being obstructive.
stalking behaviours, including:
●● forbidding/stopping contact between stalker and victim; Persistent Complainants
●● reduction of denial of the damaging nature of the
behaviour, perhaps supported by CBT; Many organizations, including health services, now encour-
●● assisting the intimacy seeking stalker to relinquish false age complaints, and have literature on systems for making
hopes of the relationship with the victim; them. Even so, a minority of complainants account for
●● focusing on the harm the behaviour is doing/will do to disproportionate use of the resources. In Australia, for
the stalker; example, various agencies of accountability estimated that
●● enhancing victim empathy; less than 1% of complaints were from such persistent com-
●● encouraging alternative outlets to meet desires and plainers, but they consumed up to 30% of resources (Lester
needs, including education, sports or other club et al., 2004). Very often their grievance seems to be about a
attendance. small matter, and certainly not of the order that would jus-
tify a sustained campaign. A case–control study of 52 unu-
Stalking: Summarizing sually persistent complainants found that the persistent
Knowledge and Progress complainers pursued their cases for longer and showed
characteristic patterns in the complaint material. The
Stalking is a problem behaviour which can result from a latter included its form, which showed multiple methods
range of social, interpersonal, psychopathological and cul- of emphasis, numerous foot or marginal notes, irrelevant
tural influences. Its recognition is increasing and, thanks to attachments of substantial length, and its content, which
new legal protections for victims, there is a real chance of included rambling discourse, misuse of legal and technical
reducing the damage stalking produces. Mental health pro- terms, inappropriately ingratiating statements, ultimatums
fessionals have a role in supporting and advising victims, and threats (Lester et al., 2004). Persistent complainants
for which task they need adequate knowledge about the differed most from controls, however, in seeking retribution
nature of stalking. They also have a role in both assessing not only against the person/people whom they believed
and managing stalkers to reduce the risks to the victim and had inflicted the initial hurt, but also people they believed
the damage to the stalker themselves. had been obstructing their route to justice. While seek-
ing to have individuals dismissed or prosecuted and/or
380 © 2014 by Taylor & Francis Group, LLC
Persistent complainants and vexatious litigants
institutions closed or subjected to punitive damage, they John Brown was living alone when, at 65, he had a heart
also differed in expecting public recognition for their cru- attack. He was admitted to hospital, high blood pressure
sading work. diagnosed, and medication prescribed for this – a beta-
blocker. He was apparently recovering, when he complained
Persistent Petitioners that he was having difficulty with his hearing. He attributed
this to the medication. His hearing was examined. It was
Less studied than other groups, persistent petitioners tend found that he did have some mild loss in one ear, but the
to be rather similar to persistent complainants in terms ear, nose and throat specialist reassured him first that
of the voluminous and repeated written communications his hearing was more or less within normal limits for his
the make. They do sometimes progress to contacting the age, secondly that there had been no previous reports of
influential figures they believe will take up their cause and, this medication affecting hearing, and thirdly that as the
although violence seems rare in this group, they have been hearing loss only affected one ear, it was extremely unlikely
responsible for many of the attacks on politicians in Western that any medication could be implicated.
governments over the last 20 years (Mullen and Lester, 2006). Mr Brown was not reassured. He asked for a second opin-
ion, which yielded the same advice.
Vexatious Litigants Mr Brown was again not reassured, particularly as he
now thought that maybe his eyesight was less good than
The main practical difference between persistent com- it had been. He wanted to see a neurologist. He did, and
plainants and vexatious litigants is that the latter have the same advice was forthcoming.
extended their complaining into the courts. Rowlands Mr Brown was still not reassured. He demanded, and got,
(1988) reviewed the development of legal restraints in further opinions, insisting that each new specialist should
England. First applied under the Vexatious Actions Act be wholly independent of the others.
1896, the situation now is that the Attorney General may Mr Brown was still not reassured, and so he acquired
make application in the High Court, under section 42 of many records at a number of hospitals, which, initially,
the Supreme Court Act 1981, prohibiting an individual were not connected. Finally, his general practitioner
from continuing or initiating legal actions. The names refused further referrals. Mr Brown was offended and
of such people are published in the London Gazette. This angry. He complained to the body which oversaw the
limitation is extremely rarely used; on average, only five or practice. This body found that Mr Brown had received
six orders per year are made. Most countries have some appropriate care.
legal measure whereby continuing action may be pro- Mr Brown was still not reassured. He began to complain
hibited (Freckleton, 1988), although how this affects the of professional protectionism and that all the healthcare
actual behaviours is less clear. organizations were in league with each other. He sought
legal advice about redress in law. His first lawyer got legal
In many cases, there is an understandable grievance at aid for him to pursue his case, and helped him to collate
the centre of the claims of the morbidly querulous, and it his by now voluminous records. Mr Brown, however, felt
is only with time that those trying to assist them begin to that the case was proceeding too slowly, so he changed his
suspect the presence of any pathology. Appropriate inquiry lawyer. His second lawyer followed much the same pat-
will elicit characteristic cognitive distortions: tern as the first, so Mr Brown changed his lawyer again
●● those who do not fully support their case are their and again. He began to realize that his failure to achieve
redress for all his suffering was part of a wider profes-
enemies; sional protectionism, and started petitioning his local
●● lack of progress is the product of malevolent member of parliament to help him expose this travesty.
Mr. Brown got no relief from this either, and he threat-
interference; ened, then attacked his latest lawyer.
●● any compromise is humiliating defeat; After the attack, he rushed away from the office. Given the
●● the grievance is the defining moment of their lives; extent of his documentation, and his own sense of being
●● because they are in the right, the outcomes they seek right, it was easy for the police to find him at home, but
less easy for them to arrest him. He had booby trapped the
must be not only possible but necessary. front door with electrical wires and a long-bore shotgun
Like the complainants, they may have voluminous docu- trained at it, to be triggered by a pulley system if the door
mentation of their case; they often represent themselves in was opened. Fortunately his mechanical skills were poor,
court, and may find themselves under contempt charges
from time to time as their passionate involvement in pro-
moting their rights may bring them into direct conflict with
the judge. As with the other conditions included here, there
is advantage for all parties in early recognition of the nature
of the problem, but this is difficult and, as the following
case illustrates, unchecked, the querulant state is progres-
sive and may end in violence.
© 2014 by Taylor & Francis Group, LLC 381
Pathologies of passion and related antisocial behaviours
so no other person was hurt. When asked, he explained 1977), there is a suggestion that pimozide might be help-
that he had set up the defences at home because it had ful (Ungvari, 1993), and a view that, as the condition
come to him that the reason why his efforts to get his dues appears to be founded in cognitive distortion, cognitive
were still being foiled was that his neighbours had joined behavioural therapy may help (Caduff, 1995). The age and
the others, and had been intercepting his mail to remove sparseness of the literature is more reflective of treatment
crucial documents. difficulties than useful guidance. In the Melbourne clinic,
low-dose atypical antipsychotic medication together with
Finally, Mr Brown was referred for the treatment he actu- support has been found to offer some relief, but progress
ally needed, but in a secure psychiatric unit. is very slow.
There is little evidence to support a particular way forward
for helping people like Mr Brown, partly because his state Conclusions
is unusual, and partly because of its nature – even if he
would consent to participation in research, how many These various disorders of passion and drive are difficult
researchers would want to engage him in experimental to diagnose early, because the ideas that drive them are
treatment? If recognized earlier, then both the person on a continuum between health and pathology and the
with such problems and the healthcare and legal systems individual sufferer often has initial competence in engag-
attempting to provide him with services would be likely to ing the sympathy of the person consulted, and yet there
be helped by referring him to a team experienced in man- are so many advantages for the sufferer and the system
aging such situations; a single worker, however experi- in identifying them at the earliest possible stage. Once
enced, would be likely to be overwhelmed in time by such established, any of them may finally drive the sufferer to
a case, even if such a patient would remain in a single 1:1 violence if the condition is not ameliorated and other
relationship. The patient will be to some extent reassured methods of attaining the object – be it love, retribution
by being referred to a team with specialist knowledge and or justice – are unsuccessful. A team with special inter-
experience of the problem. The tasks of resolving what can est in and experience of managing such difficulties has
be resolved, setting limits to further actions and offering an advantage; persistence with care, which may include
sympathy and support may pave the way to more specific antipsychotic medication, may produce some real change,
treatment attempts. Although efforts to treat have been but there is very little research into therapeutic outcome
reported as discouraging (e.g. Astrup, 1984; Winokur, to inform treatment strategies.
382 © 2014 by Taylor & Francis Group, LLC
16
Personality disorders
Edited by Written by
Pamela J Taylor Conor Duggan
Andrew Hider
Tony Maden
Estelle Moore
Pamela J Taylor
1st edition authors: Ron Blackburn, John Gunn, Jonathan Hill, David Mawson and
Paul Mullen
Doctors came to see her singly and in consultation, disorder and not given continued treatment or followed
talked much in French, German, and Latin, blamed up (Healthcare Inspectorate Wales, 2008; p.5, para 1.8).
one another, and prescribed a great variety of medi- At the other extreme, in England and Wales, it is arguable
cines for all the diseases known to them, but the simple that seeming political determination to provide for indefi-
idea never occurred to any of them that they could not nite secure hospital detention of a subset of people with
know the disease Natasha was suffering from, as no personality disorder, regardless of their treatability, has now
disease suffered by a live man can be known, for every largely succeeded in the form of the Mental Health Act 2007,
living person has his own peculiarities and always has which applies a wide definition of mental disorder and, with
his own peculiar, personal, novel, complicated disease, respect to treatment, the criterion only that it should be avail-
unknown to medicine. (Leo Tolstoy, ‘War and Peace’) able (see chapter 3). This, and the blossoming and waning of
special ‘dangerous and severe personality disorder’ (DSPD)
Concepts of personality disorder services, described further below, followed a single tragic
case in England. Here, the facts finally revealed in a report
The Legal and Political Context completed in 2000 but not published until 2006 (Francis
et al., 2006), hardly seemed to warrant such radical changes.
The personality disorders may simultaneously be the most Nevertheless, new and useful thought about personality dis-
over-used and under-considered collection of diagnoses order and treatment activity have emerged from the result-
in the practice of psychiatry and psychology – forensic or ant stimulation and from the English Department of Health’s
otherwise. This is unfortunate, since the application of such attempts to destigmatise the disorder (National Institute for
a diagnosis exerts a powerful effect on whether the person Mental Health in England, 2003; Duggan, 2011a,b).
attracting the label gains any access to treatment and,
whether in hospital or custody, probably also on his/her Concepts and Diagnoses
liability to preventive detention in the event of a criminal act.
At one extreme, the diagnosis may be misused as an excuse The task of diagnosing personality disorder is compli-
for doing nothing for a person who is suffering and asking cated by the nosological confusion around the whole
for help, sometimes with disastrous consequences for others concept (Mann and Moran, 2000). Buchanan (2005) pro-
as well as the primary sufferer. Ms A, in fact suffering from vided an introduction to problems in the use of cat-
schizophrenia, killed a fellow shopper who was a stranger egorical diagnosis of personality disorder as grounds for
to her: compulsory detention in hospital, together with a presci-
ent warning for clinicians involved in the care and treat-
Ms A’s contact with mental health services … can be ment of such patients in the current political climate:
divided into two periods: 1992–1998, when she was given
a diagnosis of schizophrenia and treated with anti- When doctors practice in politically contentious fields,
psychotics and followed up by services. And in 2003–2005 a robust nosology is one defence to the charge that their
when she was given a diagnosis of borderline personality motives are political.
© 2014 by Taylor & Francis Group, LLC 383
Personality disorders
It is arguable that, masquerading as descriptive psychiatry, ICD-10 (WHO, 1992), also makes the requirement that
personality disorder diagnoses reflect the categorical abso- personality disorders, as developmental disorders, must
lutism of nineteenth century ‘moral insanity’ (Prichard, have appeared in childhood or adolescence and persisted
1837; Maudsley, 1885). First steps seem to have been to into adulthood – any such disorders which arise later must
observe and consolidate a distinction from illness – hence be regarded as conditions of personality change. DSM-IV
the concept of ‘mania without delirium’ coined by Pinel (American Psychiatric Association [APA], 1994), in its multi-
in his 1801 Treatise on Insanity (translated Davies, 1806). axial approach to classification, formally places personality
This was followed by recognition of developmental impair- disorders on a separate axis – axis II – not only from illnesses,
ments, but in the absence of intellectual disability – in which generally represent a clear break from health, but also
Prichard’s terms ‘moral insanity’, a concept which per- such conditions as substance-related disorders, impulse
sisted into the twentieth century: control disorders and adjustment disorders. Insofar as such
distinctions are helpful, this hardly seems an advance on
a form of mental derangement in which the intellectual Jaspers’ (1923, translated by Hoenig and Hamilton, 1963)
faculties [are uninjured], while the disorder is manifested earlier, essentially multi-axial separation of organically based
principally or alone in the state of feelings, temper, or conditions (Group I), from the major psychoses (Group II),
habits… The moral … principles of the mind … are and from the personality disorders (Group III). In the lat-
depraved or perverted, the power of self-government is ter he included ‘isolated abnormal reactions that do not
lost or greatly impaired... (Prichard, 1837). arise on the basis of illness, neuroses and neurotic syn-
From the perspective of forensic mental health practice, dromes and abnormal personalities and their development’,
there have been particular questions relating to the con- while acknowledging that ‘in Group III the classifications
cept of agency and, by extension, culpability. Far from attempted by various investigators show the least agreement.’
being of purely scholastic concern, as categorical diagno-
ses, personality disorders would seem to be amenable to Later academics have gone further than acknowledging
the legal system’s habit of shoehorning criminal behaviour the struggle for agreement. Kendell (2002), for example,
into the categories of ‘guilty’ or ‘not guilty’. The presence concluded:
of a personality disorder, however, does not share the
same relationship with behaviour as, say, the presence of The historical reasons for regarding personality dis-
a leg fracture shares a relationship with the ability to walk orders as fundamentally different from illnesses are
unaided. Personality disorders, it is arguable, are emer- being undermined by both clinical and genetic evidence.
gent from normality rather than distinct from it. They are Millon (1996) views axis I disorders as being, essentially,
existentially elusive, being significantly dependent on the decompensated variants of axis II problems. Many current
context in which individuals live. They are interpersonal experts are resigned to the explanatory inadequacy of the
diagnoses in a way not comparable to other forms of atheoretical DSM-IV, and look forward to the inclusion of
psychiatric classification. Paraphrasing George Vaillant some kind of dimensional classification, allowing personal-
(1987), it is possible to imagine a human being alone on ity level dysfunction to be explained in terms of specific
a desert island and suffering from depression, or panic traits, such as impulsiveness, as well as in terms of categori-
disorder, or schizophrenia, but more difficult to imagine cal variables (Sperry, 2003). The actual proposals, however,
how it would be meaningful to ascribe to such a person are complex (Shedler et al., 2010 and below).
a disorder of personality. Even the two main mental dis-
order classification systems – the American Diagnostic Such dilemmas are challenging for academics, but they
and Statistical Manual (DSM) and the International pose serious practical problems for practising forensic
Classification of Diseases (ICD) which still tend to treat clinicians:
these, as other conditions, categorically, emphasize this
interactive quality, the ICD also acknowledging a dimen- When relating to the Criminal Justice System, forensic…
sional conceptualisation: views need to be expressed in understandable clear
[personality disorders] comprise deeply ingrained and language, not theoretical ‘psychobabble’, the meaning of
enduring behaviour patterns, manifesting themselves which is only understood by the initiated (Williams, 1997).
as inflexible responses to a broad range of personal This is hard when the initiated do not agree the lingua
and social situations. They represent either extreme or franca. The criminal justice system expects its experts to
significant deviations from the way the average indi- know what they are talking about, so a retreat to nihilism
vidual in a given culture perceives, thinks, feels and is neither acceptable nor necessary. The diagnostic and
particularly relates to others… They are frequently, but classification systems are themselves explicit about the
not always, associated with various degrees of subjective need for caution and caveats in making reference to their
distress and problems in social functioning and per- guidance in court. DSM-IV, for example:
formance (World Health Organization (WHO), 1992a). When the DSM-IV categories, criteria, and tex-
tual descriptions are employed for forensic pur-
poses, there are significant risks that diagnostic
384 © 2014 by Taylor & Francis Group, LLC
Concepts of personality disorder
information will be misused or misunderstood. … measures (Eysenck and Eysenck, 1975; Weinryb et al.,
In most situations, the clinical diagnosis of a DSM- 1992), although others advocate an overlapping five factor
IV mental disorder is not sufficient to establish the model of personality: neuroticism–stability; extraversion–
existence for legal purposes of a ‘mental disorder’… introversion; agreeableness–antagonism; conscientious-
The pertinence of such caution becomes even more appar- ness–lack of self-discipline; openness to experience–rigidity
ent when considering individual personality disorders. Both (John, 1990; Costa and McCrae, 1992a; McCrae and Costa,
main classification systems go on to subdivide personality 1996). Some would argue further, however, that the hall-
disorder into a number of smaller categories, reflecting mark of disorder here is not necessarily extremes per se, but
the main areas of personal and social disruption. This has rather the inability to adapt adequately to the interpersonal
some face validity, but ‘diagnosis’ of an individual personal- environment (e.g. Leary, 1957; Blackburn 1998).
ity disorder tends to be less reliable than the more general
attribution of ‘personality disorder’ (Bronsich, 1992; Bronsich In work with offender patients, psychopathy has
and Mombour, 1994); even inter-rater reliability may be become a widely used concept (Hare, 1980, 2003), build-
barely satisfactory (mean kappa 0.56, range 0.26–0.75, APA, ing on characteristics common to the people in Cleckley’s
1980; mean kappa 0.41, range 0.00–0.49, Mellsop et al., 1982). (1976) clinical stories. Psychopathy in these terms overlaps,
Others, while acknowledging the limitations on reliability but is not precisely coterminous with, antisocial personal-
of diagnosis have also highlighted the problem of excessive ity disorder (Hart and Hare, 1989). While Hare and his team
focus on observable behaviour rather than core dysfunctions initially evidenced a two-factor model (Hare et al., 1990),
(e.g. Zimmerman, 1994). Then, too, according to Fleiss (1981) others have found a three-factor model better fits the pic-
and Cicchetti (1994), although the names of personality dis- ture (Cooke and Michie, 1997), and the Hare group has pro-
orders are similar between ICD and DSM, the concepts do gressed to a ‘2 factor, 4 facet’ model (Neumann et al., 2007).
not quite match. Furthermore, although it could be argued There is, however, agreement on the core factor being
that the extent of so-called comorbidity between these single affective and empathic impairment. Impulsive behavioural
categories calls into question their validity as discrete entities style is also broadly agreed, with Cooke and Michie adding
(see also below), others have argued that important aspects grandiose and deceitful style as the third factor, although
of personality pathology are ignored within the standard it may be these aspects which show more variation across
classificatory frameworks, for example passive–aggressive cultures (Cooke and Michie, 1999). The concept and meas-
or sadistic traits (Westen and Arkowitz-Westen, 1998; Clark, urement of psychopathy undoubtedly have some research
2007; Clark et al., 1997). A partial response to ‘comorbid- evidence basis, but the Cleckley–Hare list of items making
ity of personality disorders’, while retaining a categorical up PCL-R also reflects another important issue – the nature
approach, has been to consider them as belonging to clusters of the countertransference between the person with the
(Tyrer and Alexander, 1979; Cloninger, 1987). These are presumed disorder and the raters. A person described as
glib and superficially charming, with a grandiose sense of
●● Cluster A, a withdrawn group, avoidant of social contact self-worth, given to pathological lying, lacking sincerity,
including paranoid and schizoid personality disorders remorse or guilt for, say, promiscuous sexual behaviour and
(DSM adds the schizotypal); many types of offence – to take just seven of the poten-
tially 20 negative attributes – is unlikely to attract caring
●● Cluster B, a risk-taking irresponsible group, which clinicians. Bowers (2002) studied nursing staff attitudes
includes antisocial (DSM)/dissocial (ICD), borderline to working with people with personality disorder in a high
(DSM)/emotionally unstable (ICD) and histrionic security hospital. The characteristics of patients who they
personality disorders (DSM adds the narcissistic). labelled, in confidence, as ‘evil’ or ‘monstrous’ were: their
index offence had been serious violence against vulnerable
●● Cluster C, an anxious, avoidant, sometimes rigid group victims; it had been planned in advance and often involved
includes the avoidant (DSM)/anxious (ICD), obsessive– torture; they had not been abused as children; they showed
compulsive (DSM)/anankastic (ICD) and dependent no remorse and refused treatment, but they appeared to be
personality disorders. nice people. Bowers went on to show that it was not impos-
sible for staff to work effectively with such people, but that
Longitudinal, prospective studies provide limited endorse- ‘moral commitment’ and ‘[therapeutic] beliefs about the
ment for the validity of individual personality disorder disorder’ were essential.
types. Indeed, while borderline personality disorder
(BPD) may be a relatively robust concept, some dis- In clinical practice, the tendency towards categorical
orders, such as schizotypal, seem so closely to predict classification seems almost inescapable – even dimen-
onset of illness that it may be better to consider them sional measures, such as that of psychopathy, tend to be
almost as a forme fruste of the illness (McGlashan, 1983). used in a categorical way, so that, for example, the person
‘has psychopathy’, or not, as the case may be. According
An alternative approach is to consider personality dis- to Shedler et al. (2010), the 5th edition of the American
orders as statistical extremes on trait dimensions. Concepts DSM is proposing to try combining five clinical personality
of extraversion and neuroticism, and, to a lesser degree,
psychoticism have been shown to be robust dimensional
© 2014 by Taylor & Francis Group, LLC 385
Personality disorders
‘prototypes’ – antisocial/psychopathic, avoidant, border- Table 16.1 Secondary domains of personality
line, obsessive–compulsive and schizotypal – with six disorder and their components (the Mulder and
dimensions: (1) negative emotionality, which includes Joyce [1997] nomenclature is given in brackets).
‘facets’ of depression, anxiety, shame and guilt; (2) intro- (After Livesley, 2007a)
version, including withdrawal from social interaction; (3)
antagonism, including exaggerated sense of self-impor- Secondary domain Primary trait
tance; (4) disinhibition and impulsivity; (5) compulsivity, Dissocial behaviour
including perfectionism and rigidity; and (6) schizotypy, (Antisocial) Narcissism
which includes odd perceptions and beliefs; Livesley (2012) Exploitativeness
correctly notes a sixth prototype, narcissistic. Shedler and Emotional dysregulation Sadism
Westen (2004a) have argued for more prototypes and they (Asthenic) Conduct problems
propose that dimensional measures should be confined to Hostile-dominance
characteristics such as severity. In forensic mental health Inhibitedness (Asocial) Sensation seeking
practice, we generally practise in this way already – using Impulsivity
some simple shorthand category for brief communica- Compulsivity (Anankastic) Suspiciousness
tions, but in any serious assessment incorporating dimen- Egocentrism
sional approaches.
Anxiousness
The DSM-5 proposal has features in common with Emotional reactivity
Livesley’s (2007a) conceptualisation, but, as one comes Emotional intensity
to expect in this field, is not quite the same, and Livesley Pessimistic anhedonia
(2012) himself sets out proper concerns about the DSM-5 Submissiveness
proposal neither maintaining continuity with DSM-IV – Insecure attachment
the DSM-5 prototypes carry the same names as DSM-IV Social apprehensiveness
classes but are different – nor becoming evidence based. Need for approval
ICD-11 may do better. Duggan (2011b) highlights some Cognitive dysregulation
of the differences more specific to antisocial personality Oppositional
disorder (ASPD). Livesley suggests three general features Self-harming acts
which underpin all personality disorder, with both social Self-harming ideas
aetiological and treatment implications: (a) failure to
develop attachment or intimacy in interpersonal relation- Low affiliation
ships; (b) failure to behave pro-socially so that one can be Avoidant attachment
part of a group; and (c) failure to achieve a coherent sense Attachment need
of self. He also subscribes to four higher order factors, con- Inhibited sexuality
stituting the secondary domains. Mulder and Joyce (1997) Self-containment
named them as the ‘four As’: the antisocial (the equivalent Inhibited emotional expression
of cluster B), the asthenic (anxious dependent as in cluster Lack of empathy
C and those with emotional dysregulation as in borderline
personality disorder), the asocial (those prone to social Orderliness
withdrawal as in cluster C and A) and the anankastic Conscientiousness
(people with obsessive–compulsive traits as in cluster C,
although some would create an additional cluster, cluster over deviant acts and encourages flexibility in identifying
D for them). Livesley simply uses slightly different termi- disorders of personality through the two-level approach of
nology here, and regards these four factors, essentially first identifying – or not – the three general features and then
phenotypes, as the core types which are explained by screening for detail in the secondary domains.
genetic and organic factors (Livesley et al., 1998; Livesley,
2007b), although he also identifies 30 primary traits which Personality Disorder
make up the main four (Livesley, 2005). This system, which Assessment Tools
may sound rather complicated in narrative, is summarised
in table 16.1. Broadly, instruments used for the assessment of per-
sonality and its disorders are founded in two principal
This domain-based conceptualisation of personality dis- approaches – the structured clinical interview – with
order has a number of advantages for the clinician and for the the patient and/or with informants – and the self report
prospective patient. Empirically derived, they are more likely questionnaire. In line with the conceptual approaches to
to be valid than operational definitions drawn up by commit- personality disorder, the assessment instruments may
tee. It allows for potentially greater accuracy in identifying be further characterised by whether they map on to the
the core problems by emphasizing interpersonal difficulties descriptive classifications of DSM-IV-TR or ICD-10 or
whether they are trait-based and dimensional. Here, we
present a guide to some of the main assessment tools
rather than a comprehensive overview; further informa-
tion may be found in specialist books, such as Part 3
of Livesley’s (2001) Handbook of Personality Disorders.
386 © 2014 by Taylor & Francis Group, LLC
Personality Disorder Assessment Tools
Among clinicians, however, these conventional tools are using any instrument cited and have a good knowledge of
commonly seen as being of limited value in clinical prac- its reliability and validity.
tice; they perceive direct questions as being much less
useful in assessing personality disorder than either listen- The other important indicator of a test’s usefulness is
ing to patient narratives about their lives and drawing its validity – whether it measures that which it is purport-
inferences about repeating patterns and/or observing a ing to measure. Broadly, personality disorder interviews
patient’s behaviour with him/her in the consulting room show poor convergent validity (that is, it is not possible to
(Westen, 1997). Accordingly, Westen and others devised say with confidence that the same person would receive
a further method of systematising assessment, which the same diagnosis when assessed using different meas-
provides clinicians with 200 cards, each of which provides ures). A review of the literature in this area concludes that
a statement such as ‘living arrangements are chaotic problems of validity present an overwhelming foil to the
and unstable’ or ‘tends to be passive and unassertive’ use of a truly scientific nosology in the area of personality
which they have to sort according to their experience disorder.
of the patient (Westen and Shedler, 1999; Shedler and
Westen, 2004b; Westen and Muderrisoglu, 2006). The Until a revised conceptualization of personality disorders
Operationalized Psychodynamic Diagnostic system (OPD provides a firmer basis on which to develop convergent
Task Force, 2008), described more fully below, also empha- assessment instruments, personality disorder research
sises the importance of being able to assess relational will remain fragmentary because instrument-based find-
capacities and interactions. ings will be the rule rather than the exception (Clark and
Harrison, 2001).
Interview Schedules Yielding
a Diagnostic Perspective on Trait-Based Interviews of
Personality Disorder Pathological Personality
Diagnostically based interviews guide the clinician through Loranger (1999b) observed that the switch from categories
a structured set of questions designed to gather informa- to dimensions improved reliability of his measure. UK
tion from the patient, relevant to his or her concordance based clinicians also use Tyrer’s Personality Assessment
with the categorical criteria of DSM-IV-TR and/or ICD-10. Schedule (PAS, Tyrer, 2000), which yields scores on a num-
Such measures include the Structured Clinical Interview ber of pathological traits thought to be subordinate to the
for DSM-IV Axis II Personality Disorder (SCID-II, First DSM-IV and ICD-10 criteria. Cluster analytic techniques
et al., 1997) and the International Personality Disorder are used to reduce 24 traits into five higher order personal-
Examination (IPDE, Loranger et al., 1995, 1999a). The latter ity styles or dimensions. A specific advantage of the PAS
was developed at the request of the WHO and the USA is its ability to yield not only trait-based and dimensional
National Institute of Health to improve international reli- scores, but also to map such scores onto existing diagnostic
ability in diagnoses. Its items are distributed under six criteria. Further, the PAS also provides a measure of sever-
categories (work, self, interpersonal relationships, affects, ity as a function of the number of criteria the person meets.
reality testing and impulse control). It is designed to be
rated by experienced clinicians trained in its use, and to The advantage of dimensional measures is that they are
take account of information about age of onset and dura- generally more anchored in theory about the aetiology of
tion of behaviour from informants as well as the individual disorders of personality. This, coupled with their ability to
under assessment. Both inter-rater (agreement between delineate homogeneous trait structures, gives these meas-
two raters at any one time) and test–retest reliability (the ures greater utility in research, whether organically (e.g. see
extent to which the measure yields the same results over chapters 8 and 12) or behaviourally based (see also Gray
time, given the same level of disorder) vary with personal- et al., 2003). Nonetheless, clinicians should be aware that
ity disorder type, but for any specific personality disorder any attempt to isolate traits and measure them as if unpol-
are acceptable (kappas: 0.7 and 0.63 respectively; Loranger luted by other psychological constructs is problematic. In
et al., 1994, 1997). An important determinant of reliability the case of the PCL-R, for example, scores on the ‘affective/
lies in clinical experience and training, so, given the con- interpersonal dimension’ might, in some cases, be as well
siderable impact that the results of such assessments may explained by schizophrenia as psychopathy per se.
have on the present and future care of the person being
assessed, it is arguable that these measures should only Diagnostic Questionnaires for
be used by appropriately trained clinicians, assiduously Personality Disorders
following the guidance in the manuals. When presenting
evidence based on such measures in the court, clinicians The widespread use of questionnaires, often in the context of
should be prepared to set out their level of expertise for legal situations such as Mental Health Review Tribunals, has
tended to foster an acceptance of the reliability and validity
of these instruments which outstrips the scientific data.
This is unsurprising, given that they are easy to use and,
© 2014 by Taylor & Francis Group, LLC 387
Personality disorders
in providing a quickly obtainable quantitative estimate of by high scores on the various subscales of the MCMI-III
personality function or dysfunction, they seem to add objec- (Millon and Grossman, 2007).
tivity to assessment reports. Well known examples include
the Millon Clinical Multiaxial Inventory, 3rd Edition (MCMI- Trait-Based Non-Diagnostic
III™, Millon, 2009), the Minnesota Multiphasic Personality Questionnaires About Personality
Inventory Personality Disorder Scales (MMPI-II-PD, Morey
et al., 1985), the Personality Assessment Inventory (PAI, Experimental psychology has an extensive history of devel-
Morey, 1991) and the Personality Disorder Questionnaire 4+ oping measures to quantify ‘latent traits’ – underlying
(PDQ4+ Hyler, 1994; Hyler et al., 1990). Of these, the MCMI- psychological variables which are relatively fixed and which
III purports to assess both DSM-IV categories and also underpin behaviour, cognition and emotion, for example,
patients’ overall personality profile based on Millon’s own, the Eysenck Personality Questionnaire (EPQ, Eysenck and
evolutionary based, theory of general psychopathology. The Eysenck, 1969, 1975). This yields three dimensions (extra-
MMPI-PD and PAI are confined to the assessment of the version, neuroticism and psychoticism), but, according to
categorical descriptors of DSM-IV. The PDQ4+ is a useful current consensus, there are five superordinate personal-
screening tool among prisoners (Davidson et al., 2001). ity dimensions which act as a psychological substrate to
all human trait like variables (McCrae and Costa, 1987).
Reliability indicators for these tests are, in the main, Assessments of these are most commonly used in non-
measures of internal consistency (i.e. the strength of cor- clinical environments, such as in personnel recruitment,
relation between different items designed to measure the with the most commonly used measure being the NEO
same trait or category) rather than inter-rater or test–retest Five Factor Inventory (NEO FFI), the five factors being
reliability (see Rogers, 2001 for an extended discussion of neuroticism, extraversion, openness to experience, agreea-
such psychometric properties). This may account for the bleness and conscientiousness; the revised version is the
extent to which test scores vary over time for the same NEO PI-R (Costa and McCrae, 1992b), while dimensions
person. In the case of personality variables, a high level of on instruments such as the Omnibus Personality (OMNI)
temporal stability should be expected, given the definition and Personality Disorder (OMNI-IV) inventories (Loranger,
and theory of personality disorder as an enduring state. 2001) have been adapted for both non-clinical and clinical
purposes. Such inventories may be more widely used in the
In terms of validity, the self-report measures are as prone future, particularly with the move towards DSM-V combin-
to the problems just described in relation to structured inter- ing dimensional and categorical approaches (Widiger and
views and, indeed, have commonly been validated against Samuel, 2005).
them (e.g. Hyler et al., 1990). In general, however, self-report
measures show, at best, modest convergent validity between Assessments of Specific Domains
each other (Livesley, 2001). Further, their discriminant of Personality Pathology
validity tends to be poor, and they generate too many false
positives (Zimmerman, 1994 provides useful reflections on A number of assessment tools follow the principles of
these issues, and limitations of other approaches too). The structure or administration just described, but focus on
limitations of self-report tests are perhaps unsurprising as, specific domains of personality pathology, for example
after all, they are essentially measures of ‘meta-function’, an the PCL-R (Hare, 2003) or the Schizotypal Personality
individual’s own appraisal of his/her personality. They must Questionnaire (SPQ, Raine, 1991). In forensic mental health
assume that individuals are accurate in their self-perception, settings, such tools are extensively used in offender assess-
a situation ameliorated somewhat by the presence of scales ment as an aid to delineating social function and making
within the measures which purport to measure the extent to links between it and risk of further social breakdown,
which respondents adopt a self-abasing or ‘desirable’/‘faking including offending. The standardised Special Hospitals
good’ pattern of response. It is these ‘validity indices’ which Assessment of Personality and Socialisation battery
seem to have specific relevance to forensic practice, in par- (SHAPS, Blackburn, 1986), for instance, has been used
ticular in the detection of symptom dissimulation. There is a to correlate particular aspects of personality pathology
small research literature around the use of the MCMI-III for with re-offending risk (Craig et al., 2006), whereas the
this purpose (e.g. Schoenberg, 2003). PCL-R has been used as a risk assessment in itself (see
also chapter 22). The latter, in its short form, is even
For the clinician, the most important thing to remem- incorporated into the best known structured professional
ber is that, while self-report measures have their place, they judgment approach to risk assessment – the 20-item
function best as screening instruments, and should not be Historical/Clinical Risk Assessment (HCR-20, Webster
used in isolation as diagnostic instruments. Nevertheless, et al., 1997), although there is evidence that it adds little
they have an important role to play in the process of case to the risk assessment function here (Douglas et al., 1999).
formulation and, indeed, Millon has developed a ‘per-
sonalized psychotherapy’ based on appropriate clinical
responses to psychological presentations as characterized
388 © 2014 by Taylor & Francis Group, LLC
How common are disorders of personality?
Beyond Self-Assessment of Personality treatment aid. Preliminary ratings can be made on the basis
of a 1–2 hour guided interview along four psychodynamically
Klonsky et al. (2002), in a review of the literature examining relevant axes: axis I experience of illness and pre-requisites
concordance between self-report and informant report of for treatment; axis II interpersonal relations; axis III inner
personality disorder, found that concordance was highest conflicts; axis IV personality structure. A fifth axis maps on
in the case of those personality disorders with which peo- to ICD/DSM. Axis I allows for identification of personal and
ple most commonly present to forensic psychiatrists – the social network strengths as well as prioritising areas for treat-
cluster B disorders. Nevertheless, given the potential legal ment. The 2008 edition of the manual incorporates a spe-
and clinical implications of a diagnosis of personality dis- cific module for offender-patients. The interpersonal axis is
order, reliance on either a single clinical interview and/or exceptionally useful in clinical practice, for helping staff iden-
a self-rating schedule would be indefensible. Milton (2000), tify countertransference issues which could interfere with
reporting on the results of a postal survey into assess- good management or treatment, and to develop an effective
ment practice in inpatient forensic psychiatric facilities language for communicating the four perspectives of patient
in the UK, calls for the establishment of a common set of experience: how the patient habitually experiences him/
standardised personality assessment instruments to be herself, how s/he habitually experiences others, how others,
used within them, so that diagnostic inconsistency may including the therapist, habitually experience the patient
be minimised. While it would be difficult from a scientific and how the others, including the therapist, habitually
perspective to endorse a particular package of specific e xperience themselves when relating to the patient. The
instruments, it is surely a sound principle to have a pack- seven basic internal conflicts are harder to rate, as, almost
age which incorporates within it a mix of self-report, inter- by definition, an offender-patient has difficulty in internalis-
view data and external observations from a range of people ing conflict, and so there is a risk of ‘floor effects’. The seven
who have been socially and/or clinically involved with the are: dependence v. autonomy; submission v. control; desire
person over time. for care v. autarchy; conflicts of self-value; guilt conflicts;
oedipal sexual conflicts; identity conflicts. On the personal-
Measures of personality pathology which rely on inform- ity structure axis, four levels are recognised – from well to
ant information include the Standardized Assessment of poorly integrated – on six categories: self-perception; self-
Personality (SAP) (Pilgrim and Mann, 1990; Pilgrim et al., regulation; defence; object perception; communication; and
1993), the best known of these in the UK, drawing on bonding. Reliability is good under research conditions and
an interview with a relative or close friend of the per- satisfactory under clinical conditions; criterion, construct,
son being assessed. For offender-patients, the Chart of concordant and predictive validity have been found to be
Interpersonal Relations in Closed Living Environments acceptable in several studies (Cierpka et al., 2007).
(CIRCLE) (Blackburn and Renwick, 1996) was developed for
use in a hospital inpatient setting, and relies on staff percep- How common are disorders
tions of the patient. Both these instruments may be used to of personality?
create ICD/DSM style classifications. The CIRCLE may also
have a specific role in assisting assessment of risk of inpa- Personality disorders are common in the general popula-
tient violence (Doyle and Dolan, 2006a,b), and has, on occa- tion. To our knowledge, there are three community-based
sion, been used as a self-report measure (Milton et al., 2005). epidemiological surveys of the prevalence of the full range
of recognised personality disorders; the communities from
Such assessments lead to a more detailed considera- which the samples were drawn differ, and the prevalence fig-
tion of assessments of interpersonal function. Some are ures differ. In Norway, Torgersen et al. (2001) drew a random
explicitly derived from the original interpersonal theo- sample of 3,590 18–65 year olds from the National Register
ries of Stack Sullivan (1953) and Leary (1957) and, in a of Oslo. After attrition because of death, illness, refusal,
more recent development, from psychoanalytic theory. language problems and one or two other smaller groups of
They all yield a circumplex picture which allows clini- reasons for non-inclusion, 2,053 people were interviewed,
cians to describe the overall interpersonal functioning using the SCID (see above), leaving a sample slightly biased
of patients on more than one bipolar dimension (e.g. towards women, older people and outer city dwellers. A total
dominance–submissiveness), with structured narrative of 13.4% had at least one personality disorder of any kind;
formulations also encouraged in the psychodynamic the most common was avoidant personality disorder (5%),
model. Of the former, the Inventory of Interpersonal followed by paranoid personality disorder (2.4%). The other
Problems (IIP, Horowitz et al., 1988) may be particularly two studies were household surveys rather than true com-
useful in treatment planning for improving social func- munity samples, but substantial, and also used standardised
tioning. The CIRCLE has been widely used in this spirit interviews. In the USA, Samuels et al. (2002) used IPDE data
with offender patients in the UK. on a subset of 742 individuals aged 34–94 from the Baltimore
The Operational Psychodynamic Diagnostics system
(OPD, OPD Task-Force, 2001, 2008) measures more than per-
sonality and its disorders, and is, again, used particularly as a
© 2014 by Taylor & Francis Group, LLC 389
Personality disorders
Epidemiologic Catchment Area follow-up survey (Eaton of 4.8%, but with a considerable range; Merikangas and
et al., 1997). Sub-sampling reasons are not entirely clear, but Weissman (1986) put this between 0.5 and 9.4%. About
the sex and ethnic distribution of the group included in the half of the post DSM-III studies yielded a rate of under 1%,
study was apparently similar to the parent cohort, although with the highest rate given as 3.7%. Moran then goes on to
younger. Overall, 9% had any personality disorder, with the review prevalence figures in various treated populations,
most common being antisocial personality disorder (4.1%), from general practitioner surveys, where the prevalence is
and the next avoidant (1.8%). This study had the advantage higher, through general hospital settings, with a still higher
of informant interviews to supplement the primary partici- prevalence, to specialist forensic hospitals and prisons with
pant interviews. For England, Wales and Scotland, Coid et al. the highest prevalence figures.
(2006) drew from the British National Survey of Psychiatric
Morbidity (Singleton et al., 2001) of 8,886 people aged 16–74 Clinical assessment and
completing a computer assisted screening interview. The engagement in practice
sample included in the personality disorder study was of 626
people who had completed a face to face interview, using The Process of Clinical Assessment
the SCID-II, as well as the computer screening. Selection of Personality Disorder
was complex, but transparent, and differences from the par-
ent cohort were allowed for in analysis. Here, just 4.4% had Underpinning the assessment process in clinical forensic
any personality disorder; obsessive–compulsive personality practice is the principle that, using a combination of evi-
disorder was the most common of the individual disorders dence based findings and practical experience, the scien-
(1.9%), with no other single personality disorder exceeding a tist-practitioner is aiming to achieve the best fit between
prevalence of 1%. the needs of each individual who is referred and treatment.
Interventions are necessarily multi-modal in the case of
Questions have been raised as to whether preva- personality disorder (Taylor, 2006b), and thus the quality of
lence of personality disorder varies according to ethnic the final package is dependent on the quality of the assess-
group. McGilloway et al. (2010) conducted a systematic ment phase of intervention. This requires description of
literature review but found little information. As would the problems and their impacts, a formulation of what
be expected from the observations on the general state maintains or escalates them, and strategies for introducing
of the epidemiology, this is hardly surprising, but a change. The precise objectives of assessment will vary by
potentially interesting possibility emerged – that ‘black’, setting and purpose, but the ultimate goal is to gather good
but not Hispanic nor Asian groups, may be less likely enough information for a sound decision on readiness for
to attract a personality disorder diagnosis than ‘white’ treatment and prioritising its elements. As far as possible,
groups, mainly evident in UK studies, and mainly in those the process should leave the individual under assessment
which rely on case note data. The Baltimore, Samuels feeling satisfied if not better for the exchange, whatever
et al. (2002) study samples, described above, despite a the outcome. In some circumstances, even simply helping
roughly 60:40 white:black distribution, did not allow for such a person towards a sensitive understanding of his/her
ethnicity in their calculations. Not quite 3% of the sample difficulties may be an important step forward. An emer-
was non-white in the interview based Coid et al. (2006) gency room psychiatric team in New York described how
study in the UK, and no difference in personality disorder the provision of a framework for understanding chaotic
distribution was found. The authors argue that the appar- interpersonal lives enhanced the readiness of outpatients
ently lower prevalence of personality disorder among with personality disorder to comply with suggested inter-
black groups in some studies may be due to reliance on ventions (Sneed et al., 2003). Where treatment is likely to
case record analysis rather than structured interviews. follow, the assessment process should facilitate preliminary
They did not raise the more serious issue of the impact engagement and preparation for interventions which will
of immigration, either for its possible selection bias with increase wellbeing and reduce risk.
respect to mental disorders or, of particular relevance to
personality disorder diagnoses, the likelihood of being As just described, clinical and/or informant interview,
unable to obtain good informant histories. It would be scrutiny of records and self-report measures are all com-
interesting indeed if ‘black’ group membership were truly ponents of assessment, and early stages of assessment
a protective factor against personality disorder. may include overcoming the reluctance of the patient to
participate in one or more elements of the process. Tyrer
The epidemiology of antisocial personality disorder (2000a) has proposed the notion of severity of personal-
has been more extensively studied, and ably reviewed by ity disorder as useful in forensic psychiatry, because it
Moran (1999). His monograph shows how a range of fac- distinguishes between subgroups of people with disorders
tors influenced the community-based figures produced to with different levels of impact on self and others, and com-
that date, with higher figures for prevalence being found plementary differences in needs. The concept of severity
before the publication of DSM-III than after. Dohrenwend
and Dohrenwend (1982) calculated a median prevalence
390 © 2014 by Taylor & Francis Group, LLC
Clinical assessment and engagement in practice
also links to issues around comorbidity, whether in terms interview process in which the patient is actively enlisted
of meeting criteria for more than one personality disorder to participate, is likely to pay dividends in the direction of
or pathology in more than one trait, or in cases where solidifying commitment to suggested treatments.
the personality disorder is complicated by having other
developmental conditions or illnesses as well. Outcome in Structuring the Clinical Approach
personality disorder depends in part on severity: the more in Assessment Interviews for
ingrained and pervasive maladaptive traits are, the worse Personality Disorder
the prognosis (Stone, 1993). Tyrer and Johnson (1996)
delineate a procedure for reflecting levels of severity of As recommended in the first edition of this book, assess-
disorder categories, including: no personality abnormality; ment of personality disorder – and indeed other disorders
sub-threshold; ‘simple’ personality disorder (one or more too – in a forensic mental health context is more usefully
personality disorders, same cluster); complex personality concerned with symptoms and traits than with diagnostic
disorder (two or more, different clusters), and ‘severe’ (two categories. Schotte (2002) took up this theme when con-
or more personality disorders from different clusters creat- sidering borderline personality disorder, regarding the clas-
ing ‘gross societal disturbance’). Tyrer’s (2000b) criteria for sificatory system criteria as just the first tier of description.
gross societal disturbance are evidence of: (1) personality The second tier would then identify the relevant physical
disturbance that has influenced a wider group than family and psychological components, linking to a theoretical
and friends; (2) the creation of significant problems to at frame of reference and thus underpinning theoretically
least 50 other individuals; and (3) threat created by the pat- driven treatments. It is helpful to consider the symptom
tern of disturbance, typically through aggressive impulses, tier within five over-arching fields, all in the context of the
irresponsibility that puts others at risk, outbursts of anger interviewee’s life history, including:
and violence, and insensitivity to social norms. 1. Affect: what are the dominant emotions? What is the
Given that personality disorder is associated with rela- level of arousal in relation to events as told?
tionship breakdown that ranges from less than optimal to 2. Cognition: what thoughts/beliefs seem to be
abusive, highly toxic and overtly dangerous, it is not sur-
prising that assessment of personality disorder, dependent underpinning the actions described? What thinking
as it is on the formation of some kind of interpersonal styles are in evidence (e.g. justifications for offending)?
alliance or collaboration, poses particular challenges for What are the beliefs about the world, self and others,
clinicians undertaking this task. Clinical assessment is a and what rules has the person generated (consciously
two-way process, including how the (potential) patient or otherwise) to compensate for these enduring beliefs
experiences the assessor and how the assessor experi- (schema)?
ences the (potential) patient, and this is nested within the 3. Behavioural disorder: what is the range and scope of
wider culture and ideology of the service. It is always help- problem/offending behaviours? What is the occupational
ful to document the stated reasons for the assessment, history? Is impulsivity a problem? What seem to be the
and to re-visit these at the point of summary, formulation antecedents to significant events? What is the nature
and recommendation. of the discrepancies between the interviewee’s account
of events and that of others? Who were witnesses, and
Extreme poverty in level of collaboration is rare, even what, if any, might be their biases?
in forensic clinical services (Moore and Gudjonsson, 2002), 4. Interpersonal functioning: to whom does/has the
but occasionally the interpersonal difficulties even with interviewee relate(d) and with what success? (See
a highly trained clinician may be so challenging that it further below.) How does s/he view her/himself ? How
is impossible to get clarity on the nature of the disorder. does s/he view others? What is her/his capacity to
Factors which are associated with such difficulties include sustain relationships? What actions have been taken at
co-existing axis I disorder, or a history of previous assess- times of relationship crisis in the past?
ments of interventions that were experienced as alienating 5. Insight: to what extent (if at all) does the interviewee
and unhelpful. Probing questions or prompts to self-reflec- recognise that there are problems, and to what or to
tion may elicit hostility, evasion, sometimes departure, whom are they attributed?
and/or failure to elaborate beyond ‘don’t know’. This may Motivation for change is another important issue for early
be anticipated and the order of the problem reduced to assessment. It is not a static predictor of successful engage-
some extent if the clinician takes responsibility for trans- ment, or indeed, in its initial absence, of unsuccessful
parently explaining the context for the assessment and its engagement. Rather, it is best conceptualised as a poten-
possible outcomes, and if time can be set aside to allow the tially changeable state of readiness or eagerness to change
interviewee to respond at his/her own pace. A preliminary (Sainsbury et al., 2004), which may be affected by other
assessment of this kind may take 3 to 4 hours of 1:1 work, personal characteristics. Motivation to engage in interview
before starting with informant interviews and other data- (and subsequent treatment) is likely, for example, to be
gathering. ‘Collaborative assessment’ (Ben-Porath, 2004), an
© 2014 by Taylor & Francis Group, LLC 391
Personality disorders
influenced by ‘insight’. David (1990) describes three compo- ‘centralised’ or ‘distributed’ and less stable, distancing or
nents to insight, including: acknowledgement/recognition self-focused and inflexible.
of mental illness, treatment compliance and the readiness/
ability to re-label unusual mental events as pathological. A number of diagnostic systems, particularly those with
He generated his concepts through research with people a history of influence within the dynamic–analytic tradi-
who had schizophrenia, and they do not always map pre- tion, such as the OPD described above, employ the notion
cisely onto personality disturbance as the latter is, almost of stable patterns in relationships. Research indicates that
by definition, ‘egosyntonic’ (perceived as compatible with these are linked with (early) attachment experiences (Allen,
the self-image) rather than ‘dystonic’, or alien to it. Often, 2001). Patients with borderline personality disorder are
patients experience the system and the staff as biased and identified as insecure, preoccupied and fearful in their rela-
malevolent, and locate problems ‘in’ others rather than in tionships (Gunderson, 1996). It is essential, and inevitable,
their encounters, which renders others as difficult to trust. that the attachment relationship is stimulated in treat-
Patients with personality disorder, therefore, tend not to ment. When this occurs, the patient’s mental capacities are
share the view of (significant) others about how best to likely to become subsumed by over-arousal, and a reduc-
describe or categorise their experiences, or may be unable to tion in ‘mentalisation’ (see below) ensues. Thus, clinicians
conceptualise their problems due to cognitive constraints, are alerted to repetition of identified patterns through their
such as emotional processing deficits (Blair and Frith, 2000). own encounters with the patient. The task at the assess-
In these circumstances, the primary objective of the asses- ment stage is to describe and document such patterns. The
sor must shift from assessment of the personality disorder OPD system has refined a useful method for expressing
per se to exploration of this process, and subsequent formu- such patterns in a formal but coherent way. The assessor is
lation of what is likely to create future impasses, together asked to rate ‘time and again’:
with possible strategies for resolving these.
(1) The patient experiences her/himself in such a way
The measurement of stage or readiness for change can that s/he is……;
be a useful proxy for treatability, particularly if resources
are scarce (Kosky and Thorne, 2001). If patients with (2) The assessor experiences the patient in such a way
personality disorder are not in a position to commit to a that s/he is……;
course of treatment it is likely that problems will escalate
if staff proceed regardless. Stone (2003) notes that, for peo- (3) The patient experiences others in such a way
ple with personality disorder, the particular combination that……;
of opposition to internal change with unrelenting insist-
ence on the need for ‘external change’ is typically fatal to (4) The assessor experiences him/herself in the
any hopes for psychotherapeutic benefit. relationship with the patient in such a way that……’.
In a clinical context, then, a basic aim of the diagnostic Thirty options are available for the assessor to fill those
evaluation is to obtain valid information that leads to a gaps and frame the experience (e.g. ‘much admiring and ide-
constructive conceptual framework, or case formulation, alising’, ‘rejecting’, ‘self-justifying’, ‘cutting him/herself off ’).
in which interventions are understood, selected and imple-
mented (Schotte, 2002). The impact of personality disorder Case Formulation
will require further articulation through specific review of
the interviewee’s interpersonal relationships over a reason- It could be argued that the most crucial component of
able period of time (Van-Velzen and Emmelkamp, 1996). the entire assessment process is the formulation of the
What is a reasonable period of time? Bateman and Fonagy problem(s) within a framework that makes sense to the
(2006) recommend 5 years. With whom are the important patient and referrer (court, clinical team, outside agency).
past and current relationships? What are the connections Lengthy psychiatric reports usually contain detailed
between these and the problems the interviewee or others description and may only include general recommenda-
describe? It is important, for example, to note if suicide tions during the early stages of alliance with the patient,
attempts are linked in time to relationship break-ups. because much has yet to unfold. Unless the information is
In their practical guide to Mentalization Based Therapy clearly summarised, however, and integrated at the end of
(MBT), Bateman and Fonagy (2006) suggest that clinicians such a report, the potential value of the assessment may be
characterise the interviewees’ relationships according to undermined, particularly as a ‘baseline’ document under-
their form, the interpersonal processes they entail, any pinning a pathway of treatment.
changes that are desired in the relationship, and the spe-
cific behaviours such changes may require. Underpinning Formulation constitutes the integration within a theo-
the questions is the framework that ‘normal’ relationship retical framework of information pertinent to the potential
representations are flexible, stable and balanced, distin- treatment of the problem. What factors (probably) gave
guishing them from less healthy relationships which are rise to the disorder? What factors are currently maintain-
ing it? – this is the opportunity to describe the adverse fac-
tors in the patient’s social context. What factors seem to
improve things for him/her? What factors exacerbate the
problems? What among such factors could be changed;
392 © 2014 by Taylor & Francis Group, LLC
Causes and explanations of personality disorders
how might this be evaluated, and to what extent might of ‘recognition’, response, adjustment, re-evaluation and
this be associated with a reduction in offending? It is at the re-adjustment. So long as an individual remains healthy,
formulation stage that the strengths of the individual, and even though this process may slow, it continues. In gen-
the frequency and duration of previous periods of com- eral, development is simultaneously at its most active
parative wellbeing/less toxic inter-relationships may be and most vulnerable during the earlier phases of life, and
highlighted as well as the difficulties. The former are likely this is certainly true of personality. Even newborn infants
to serve a protective function against deterioration and/or have been shown to have distinctive temperamental
future offending, and may present obvious starting places traits (Thomas et al., 1963; Thomas et al., 1970). Qualities
for interventions that will be most tolerable for the patient. in parenting have an effect on infants, but the infant’s
traits also have an effect on how the parents – and others
The outline of what such an assessment should achieve – respond (Thomas and Chess, 1984).
is summarised in table 16.2. This could be extended to take
a specific model of intervention into account. As a precur- The difficulties outlined in conceptualising and diagnos-
sor to dialectical behaviour therapy (DBT), for example, ing personality disorder also apply to efforts to understand
the therapist might focus on examples of interpersonal how abnormal personalities emerge. The literature, which
effectiveness, skills deficits, and emotion dysregulation in vacillates between categorical diagnoses and dimensional
more detail. For mentalisation based therapy (MBT), the traits and, indeed, antisocial behaviour not amounting to
emphasis would be on capacity for perceiving and inter- personality disorder, has barely tackled the complexity of
preting behaviour as conjoined with intentional mental combining the approaches in aetiological studies, although
states (Bateman and Fonagy, 2006). Livesley has been prominent in trying to formulate a way
to do so (Livesley et al., 1998; Livesley, 2005; Livesley, 2008).
Causes and explanations Most research which is relevant for forensic mental health
of personality disorders clinicians, however, focuses on antisocial personality disor-
der (ASPD) specifically, or referring only to personality dis-
Introduction to Genetic and order more broadly, or to psychopathy, a construct defined
Developmental Factors by a threshold score on the PCL-R, but with the chosen cut-
off varying in part according to whether studies are from
Our emphasis on the interpersonal and interactive in Europe or the USA. These various labels do not necessarily
assessing personality applies as strongly to considera- reflect states which are co-terminous or even meaningful.
tions of its development. The basic substrate of a person
may be a collection of genes and genetically programmed For an understanding of social and genetic influ-
physical structures but, before as well as after birth, the ences, we refer readers to chapters 7 and 8, although both
environment has an effect of varying degrees on whether, emphasise antisocial behaviour rather than antisocial
when and how physical characteristics present, and personality disorder per se. As the antisocial behaviour to
both primary and developing physical characteristics which they refer is a distinguishing feature of the people
determine aspects of whether, when and how the envi- under study, the antisocial characteristics of these people
ronment has its impact. Development is a constant cycle were, by definition, different in nature, degree or both
from those of the general population, but that does not
Table 16.2 Steps in forensic–clinical assessment of personality disorder
• Document reasons for referral
• Explain rationale and context for assessment
• Explain possible outcomes/implications of the findings
• Prepare from (medical) records: history of problems, their extent, range, duration
• List and date sources of information relating to historical events
• Diagnose axis I disorders (illnesses) where present
• Interview to describe impacts of axis II personality disorders according to: affect, cognition, actions, interpersonal functioning, insight, motivation (for
treatment), readiness for change
• I nterview to focus on relationship history: patterns in past and current relationships, describe attachments, significant events, critical periods
• Summarise and FORMULATE: history, personal background and antecedents (to offending), what motivates the person to behave pro/antisocially,
with/without regard for others, what appears to have moderated these actions in the past; what could bring about desired change (i.e. positive
increases in wellbeing and risk reduction)
• Recommend specific next steps/estimate a pathway of intervention
• Discuss the formulation with the patient and elicit feedback (which may lead to its revision)
© 2014 by Taylor & Francis Group, LLC 393
Personality disorders
necessarily mean they had antisocial personality disorder. therapies. Many people with personality disorder have, for
The association between antisocial personality disorder example, subtle difficulties with language and communica-
and antisocial behaviour of a nature or degree that leads tion, which are not immediately apparent in a structured
to imprisonment is very high (Singleton et al., 1998a), but interview or from their vocabulary. Some may be founded
not absolute. Thus, any numerical extrapolations from in such problems as theory of mind deficits, explored more
strength of relationship between genetic and environmen- fully below. Not only may these be of direct causative rel-
tal contribution to antisocial behaviour on to antisocial evance, but, again, they may exert their effect through cre-
personality disorder must be made with caution, but the ating adverse responses from others and, eventually, a toxic
principles drawn out in these chapters are sound. When social environment. They may hinder therapy if not recog-
disorders run in families, explanations must be sought nised and either ameliorated or allowance made for them.
variously in heritability, in shared family environment, Extensive damage or dysfunction is not necessarily a barrier
in interactions between these, and in events, like abuse, to psychological treatments but, if present, will mean that
which may not be shared but which nevertheless com- the work may need to be structured differently, perhaps
monly happen within the context of a dysfunctional envi- allowing for more than usual repetition, and will take longer.
ronment. Even the impact of seriously traumatic events
may be determined in part by a particular genetic her- Neurochemistry
itance (Caspi et al., 2002). This study of the male partici- Dysfunctions in the serotonergic, dopaminergic and
pants in the Dunedin birth cohort showed that, at age 26, noradrenergic neurotransmitter systems have each been
those having the X-linked gene for monoamine oxidase-A implicated as underpinning aspects of personality disorder
(MAO-A), conferring high levels of this neurotransmitter, (Bhagwager and Cowan, 2006). The story is essentially one
were less likely to develop antisocial behaviour problems of reduction in plasma and/or cerebrospinal fluid (CSF)
after abuse. Usually, where there are more directly herit- levels of the transmitters associated with impulsivity or
able factors, these will be for traits, like impulsivity, which aggression rather than personality disorder per se. The evi-
contribute substantially to the personality type or disor- dence is weak or mixed with respect to the noradrenergic
der and its impact rather than determining the disorder or dopaminergic systems, but quite strong with respect to
per se. Genetic influences may, however, have as great a the serotonergic system (see chapter 12 for more detail).
role as the environment on key comorbidities, such as Again, however, there is little here that would help with
alcohol or drug dependency disorders. When considering routine assessment. Methods of measurement of such neu-
complex disorders, it is unlikely that a single explanation rotransmitters tend to be intrusive – requiring at best blood
will suffice either for the population with the disorder or samples, but ideally CSF samples – and are just not practi-
individuals within it. The effect of neighbourhood envi- cal even for extended clinical evaluations. Their main value
ronment must also be considered. lies in their confirmation in principle that there is likely to
be an organic substrate for personality disorder.
Physical Routes to Disorders
of Personality Psychological Routes to Development
of Personality Disorder
Neuropsychiatry
Evaluations of brain structure and function in people with Theory of mind and empathy
personality disorder remain largely a matter for research. Theory of mind (ToM) refers to an ability to represent the
Personality change has been observed after cerebral insults mental states of others within one’s own mind, and distin-
of various kinds, including head injury, encephalitis and the guish them from one’s own mental state, allowing explana-
dementias. As with other psychiatric conditions, experience tion and prediction of the behaviour of others (Leslie, 1987).
with localising injuries has provided guidance on where to It is, essentially, a two-stage, largely cognitive process, first
look for damage or dysfunction among people with primary of recognition of the separate mental situation of the other;
personality disorder. The limbic system and frontal lobes and secondly, the processing of this information so that
are particularly implicated (see chapter 12), but the damage the other person’s perspective can be taken accurately. It
and/or dysfunction is usually subtle and not yet generally is thus related to role-taking (e.g. Chandler et al., 1974) and
susceptible to identification on investigations routinely perspective taking (e.g. Selman, 1976), and underpins the
available to clinicians, such as skull X-ray, standard electro- concept of empathy: ‘an affective response more appropri-
encephalogram or computerised tomography scan. ate to someone else’s situation than one’s own’ (Hoffman,
1987). Accurate empathy relies on the two elements of
The clinical implications of such damage or dysfunction theory of mind together with a capacity for emotional
are that preliminary work to correct any deficits, as far as it responsiveness (Feshbach, 1987). It has been postulated
is possible to do so, is likely to create a more advantageous that such capacities are linked with inhibition of aggressive
position for other therapies, particularly psychological
394 © 2014 by Taylor & Francis Group, LLC
Causes and explanations of personality disorders
and/or antisocial behaviour (Eisenberg, 1986; Feshbach, here for understanding the cause of one important type
1987), moral development (Kohlberg, 1981; Turiel, 1983) of personality dysfunction, but a case has been made for
and prosocial development (Underwood and Moore, 1982). further research rather than a definitive theory, or basis for
It follows that impairments in one or more of the elements treatment of people with personality disorder.
of theory of mind or emotional responsiveness might result
in pathological aggression and/or antisocial behaviour. Maladaptive learning and personality disorder
Learning takes place as qualities in the environment are
Blair (2005, 2008; Blair et al., 2005) has been chiefly perceived and accommodated by the individual, resulting
responsible for drawing out the complexities of the rela- in a change in one or more of the three components of
tionships between dysfunction with respect to theory of behaviour: thinking, feeling or action. Identification of what
mind, other aspects of empathy, psychopathy as a par- constitutes adaptive and maladaptive learning is, however,
ticular type of personality difficulty and other disorders far from straightforward. Those behaviours which would
in which aspects of empathy are impaired. His choice to be construed as maladaptive in a well ordered society may
focus on people who are given high scores on the PCL-R is also be behaviours which, at some stage in the life of an
based on an idea, which he shares with many others, that individual who has been subject to extremes of deprivation
the category of antisocial personality disorder, as defined or abuse, have constituted the only possible survival strate-
in the main diagnostic classification systems, is founded gies. As a result, they can be extremely resistant to change.
in a social deviance model while the concept of psychopa-
thy, which may or may not result in detected antisocial The origins of learning theory rest first in the idea
behaviour, indicates the primary personal developmental that behaviour can be learned or ‘conditioned’ by pairing
disorder. His research path started with work with children almost any stimulus with a particular outcome. The clas-
with autism. Such children have consistently been shown sic Pavlovian experiment with dogs coupled the sound of
to have theory of mind impairments (e.g. Baron-Cohen a bell with the arrival of food, so the experimental dogs
et al., 1985), which cannot be explained by language difficul- were, effectively, trained to salivate on the sound of the
ties (e.g. Leslie and Frith, 1988). They have also been shown bell; eventually, they did so regardless of whether the food
to have difficulties with recognising emotions in others was provided (Pavlov, 1927; a translation of all his lectures
(e.g. Tantam et al., 1989) and, as in adults, there is some evi- and a bibliography is available at http://psychclassics.yorku.
dence of association between autistic spectrum disorders co.pavlov/). The concept of operant conditioning followed
and violence (Scragg and Shah, 1994). Blair, however, found – that more complex behaviours are, essentially, shaped
that children with autism and theory of mind impairments by a system of ‘rewards’ and ‘punishments’, such that any
do respond to distress in others, finding it distressing (Blair, behaviour which changes an individual’s inner experience
1996) and arousing (Blair, 1999a). By contrast children with relatively positively will tend to be reinforced and any
‘psychopathic tendencies’ were found to have reduced elec- which changes it negatively will tend to be extinguished
trodermal responses to distress in others, but not to sense (Thorndike, 1931; Skinner, 1974). Thus, enduring traits may
of threat from them (Blair, 1999b). arise as an individual learns that a particular behaviour
or sequence of behaviours may relieve emotional pain,
While this work may be suggestive of a distinctive but, to the observer, initially including the therapist, the
pattern of deficits in empathic development, it was done behaviours may seem unpredictable because the negative
entirely with small numbers of children. A series of 25 affect is a conditioned response to something associ-
adult men crossing the PCL-R cut off and resident in high ated with the original trigger. Thus, an abused child may
security hospital had similar theory of mind function to become conditioned to associate a particular colour and/
healthy controls (Blair et al., 1996). In a later series, of 18 or smell with being beaten and, in turn, to respond to any
experimental men and 18 controls, such ‘psychopathic’ contact with such a trigger with a particular pattern of
adults were shown to have impaired responses to distress behaviours which have at least reduced the pain of this.
cues, but not to threat cues (Blair et al., 1997). Few others This might incorporate avoidance of authority figures –
have worked in this specific area, so the findings have to so never keeping outpatient appointments – and/or taking
be regarded with caution, however further indirect sup- alcohol or other drugs ‘to block things out’ and/or being
port is provided through neuroimaging studies. Amygdala aggressive or frankly violent to anyone in the vicinity at
dysfunction has been particularly associated with psy- such moments. To compound such difficulties, the learn-
chopathy and with impaired abilities to recognise facial ing of behavioural responses to situations through social
emotions in others, especially fear (Blair, 2008; Blair and modelling may be impaired (Bandura, 1969), especially in
Fowler, 2008). As Blair observes, however, theory of mind families and/or communities where there is a higher than
functions are not associated with the amygdala. He also average risk of abuse in any of its forms.
observes that the ventro-medial prefrontal cortex is com-
monly compromised in the presence of high PCL-R scores, The idea that maladaptive learning makes at least a
and theory of mind is associated with the medial frontal contribution to the development of disorders of personality
cortex (Fletcher et al., 1995). There is, then, the potential
© 2014 by Taylor & Francis Group, LLC 395
Personality disorders
is important because of the potential role it creates for signals from the child and ability to respond appropriately,
identifying and extinguishing the maladaptive patterns and and on the other from the child’s ability to seek their car-
engaging the individual in learning new strategies to cre- egiver, communicate distress and be sufficiently reassured
ate and maintain a sense of inner security – even, perhaps, to resume exploratory play. Next steps were to apply infant
more effectively than with the old strategies. More pro- models of attachment to adults (e.g. Main et al., 1985). This
social behaviours, which can provide for the individual’s is dealt with more fully in chapter 28, albeit there in relation
prospects of establishing him/herself in a safer external to post-traumatic stress disorder.
environment, may also be developed. Almost by defini-
tion, however, if persistent adversity or abuse within the Evidence for a relationship between insecure attach-
relatively consistent environment of family, school and local ments and personality disorder is partly inferred from
community have provided any of the triggers or reinforcers observations of the conditions of separation, deprivation
of habitual behavioural presentation, the usual environ- and abuse that underpin such pathological attachments
ment of the individual is unlikely to be congenial or safe in being particularly evident in the early lives of people who
a way which would be recognisable to most therapists, and develop those personality disorders which are so promi-
there is a resultant risk that changes in behaviour which nent in offender populations, including borderline per-
are advantageous outside the home community are risky sonality disorder (Ogata et al., 1990; Brown and Anderson,
within it. Therapies which yield such changes thus, in effect, 1991; McClellan et al., 1995) and antisocial personality
have potentially serious side-effects. This sort of issue was, disorder (Luntz and Widom, 1994). Although there is some
for example, thought to be a key explanation for the finding more direct evidence of measured insecure attachment
that young male offenders in the experimental treatment among people with personality disorder (Patrick et al.,
group in a US custodial setting who initially appeared to do 1994; Fonagy et al., 1995; van Ijzendoorn and Bakermans-
significantly better than the controls in a standard custo- Kranenburg, 1997; Frodi et al., 2001), sample sizes tend to
dial setting were apparently doing much less well at long- be small, and personality disorder groups to be proportion-
term follow-up; they had gained skills which were useful ately more likely to show insecure attachments rather than
in the therapeutic community and while supported in the there being any absolute relationship. Nevertheless, this
community, but less so when left to cope for themselves conceptualisation of the pathway to personality disorder
back home (McCord and Sanchez, 1982). Recognition of a also provides for theoretical models of intervention. First,
need to take a whole systems approach is well established it paves the way – or should do so – for understanding
outside offender/offender-patient settings (Mikesell et al., how it is that people with personality disorder have such
1995; Magnavita, 2000; McGoldrick and Gerson, 1985; Paris, difficulty in committing themselves to treatment – in effect
1996), and perhaps needs more emphasis within them. attaching themselves to therapists or services. Secondly, it
suggests the importance of tackling emotional regulation
Attachment theory in the short term, perhaps with medication, to facilitate
Bowlby (1944) was probably the first to propose a relation- improvements in reflective functions (Fonagy et al., 2002).
ship between early disorders of attachment and crime,
founded in his observations of the family life of young Trauma as a precursor to personality disorder
thieves attending a child guidance clinic. He went on A distinction between the traumatic origins of personal-
to extensive observations of infants with their primary ity disorder and an understanding of the emergence of
caretakers, most often their mothers. He postulated that the disorder as a form of chronic attachment disorder is
attachment bonds were formed during the first year of to some extent an artificial one, since early trauma com-
life, healthy attachments serving to protect the vulner- monly has the effect of disrupting attachment. Insofar as
able infant and safeguard its development and survival adult experienced trauma is associated with chronic dys-
(Bowlby, 1969, 1973, 1980). Disruption of these bonds for function in interpersonal relationships, the disorder classi-
any reason would leave the infant in a painful affective state fication systems, somewhat artificially, label the resultant
which could predispose to later sustained psychopathol- disorder as a personality change rather than personality
ogy, in particular sustained difficulties in regulating affect. disorder. The importance of bearing this aetiological path-
Ainsworth extended ethological study of children with way in mind may lie less in the style of therapy to be offered
their parents, drawing attention to the importance both of as in recognition of the risk of re-traumatising patterns in
the main caregiver’s capacities and patterns in the child’s the individual’s life-course. These may affect treatment
attachment behaviours to the development of secure bond- relationships as well as ordinary social ones, and increase
ing (e.g. Ainsworth and Wittig, 1969). Contributions to the the risk of avoidant behaviours, including avoidance of
attachment process thus come on the one hand from the therapy. Thus, work will have to be expended on sustain-
main caregiver’s capacity for internalization of the mental ing the therapeutic relationship as well as resolving the
state of the child, and thus their sensitivity to emotional particular traumas which may have contributed to the
shaping of the individual.
396 © 2014 by Taylor & Francis Group, LLC
Causes and explanations of personality disorders
Comorbidity was indicative that these conditions were not independ-
ent in the first place. Fyer et al. (1988) found that only 8%
This brief review of explanations of how personality disor- of their series of 180 cases of borderline personality disor-
der may emerge raises questions as to how far the disorders der had ‘pure’ disorder, however, this was a treated sample,
observed in clinical practice are correctly designated as so may have been biased towards inclusion of more com-
personality disorders alone, and how far there are, in effect, plex cases. It makes some sense to be as parsimonious as
multiple disorders to be dealt with – perhaps multiple possible with such diagnoses, because variety in symptom
disorders of personality, but also other specific develop- presentation may only indicate different manifestations or
mental disorders, such as attention deficit hyperactivity phases of the same disorder. Furthermore, both main diag-
disorder, other traumatic/attachment related disorders, nostic classificatory systems offer an option for diagnosing
such as post-traumatic stress disorder, and illnesses, to mixed personality disorder, ICD (WHO, 1992a), explicitly
which perhaps, people with personality disorder may be with this name:
disproportionately vulnerable. It has, for example, been
shown that attention deficit hyperactivity disorder as a With features of several of the disorders in F60. – but
child is not only a risk factor for antisocial behaviour and without a predominant set of symptoms that would allow
alcohol abuse in adult life (Loeber et al., 2003; Langley for a more specific diagnosis.
et al., 2010) but also may persist into adult life and com- Apparent co-occurrence with a personality disorder of
plicate the presentation of personality disorder and other other developmental disorders also poses a theoretical
conditions (Young et al., 2003; Young and Toone, 2000). challenge in that, like other personality disorders, they
too may lack real independence from the main disorder of
The term comorbidity was probably coined by Feinstein interest. Generalised or specific learning disabilities may,
(1970) but, as a concept, has a longer history. It has always for example, appear to co-occur with personality disorder,
created philosophical, theoretical and practical difficul- but to what extent, when they do so, are all these devel-
ties, but has a certain practical value too, so the concept opmental disorders, including the personality disorder,
is retained. It indicates that two or more conditions may multiple manifestations of a single pathology? To what
occur together, perhaps concurrently throughout their extent may one have a powerful causative effect in the
course, perhaps with one being more intermittent than the emergence of another or others? Here, however, there is a
other. Jaspers (1923/1963) recognised the co-occurrence practical advantage in recognising the multiple presenta-
of different clinical phenomena, but considered that the tions, because of implications for treatment. One or more
principles of medical diagnosis meant that a single, primary of the apparently co-occurring conditions may have to be
disease entity should be diagnosed, with the remaining treated effectively to enable the individual to engage in the
phenomena being regarded as secondary or accidental. more active requirements of treatment of the personality
Clarkin and Kendall (1992), by contrast, were comfortable disorder per se.
with the notion that two or more diagnoses could co-occur,
but took a strict view of comorbidity within the framework Injury or illness, even in the presence of personality
of disease classifications. Still others have extended the disorder, may at first sight seem to pose a sufficiently clear
concept to include a range of behaviours beyond diagnosis, break in relative health that it seems obvious that they
such as violence or antisocial behaviour (e.g. McCord and should be regarded as comorbid conditions, although this
Enslinger, 1997; Slomkowski et al., 1997). has not always been accepted (Hare and Cox, 1978). Now,
both the multi-axial structure of the DSM system, which
The diagnostic systems, if anything, confuse the pic- places personality disorder on a different axis from illnesses,
ture further, in that a multi-axial classification system impulse-control disorders and adjustment disorders and an
such as the American DSM system would seem to encour- explicit statement in ICD-10 (p.201) seem to endorse this:
age recognition of comorbidities, perhaps artificially
inflating their rate of occurrence; however, also in such If a personality disorder precedes or follows a time-limited
systems, some categories of disorder are defined explicitly or chronic psychiatric disorder, both should be diagnosed.
by the absence of others, thus perhaps artificially lowering Even this is not, however, straightforward, as the tran-
their prevalence. The tensions relating to comorbidity are sition between some disorders of personality and later
particularly strong with respect to personality disorder onset of illness is so consistent that the recommenda-
per se. Some clinical researchers are adamant that several tion is to regard them as manifestations of the same con-
personality disorders may co-occur (e.g. Pfohl et al., 1986; dition, for example schizotypal personality disorder and
Coid, 1992), while others indicate the absurdity of this schizophrenia (McGlashan, 1983). Nevertheless, there
position, arguing that the high rates of co-occurrence are now several published studies in substantial gen-
simply indicate that the disorders are not independent of eral psychiatric samples (e.g. Moran et al., 2003), foren-
each other; Tyrer and Stein (1993) are among the latter, sic psychiatric samples (e.g. Taylor et al., 1998; Blackburn
suggesting that their finding of a co-occurrence of 46% et al., 2003; Putkonen et al., 2004) and among prisoners
between borderline and histrionic personality disorders
© 2014 by Taylor & Francis Group, LLC 397
Personality disorders
(e.g. Singleton et al., 1998a) which report findings of exten- the main requirement. It is, after all, impossible to measure
sive personality disorder–illness comorbidity. These are all, outcome adequately unless the desired outcome is clear.
however, cross-sectional studies, and the parallel courses of Setting aside for a moment the potential role for resolving
the disorders over time are not at all clear. Neither are the comorbid conditions, which may be the critical factors in
implications for treatment of personality disorder. There are episodes of decompensation, ought the function of inter-
many studies which illustrate the extent to which the pres- vention for personality factors per se be to achieve the mini-
ence of personality disorder may complicate the treatment mal sufficient adjustments needed for slightly improved
of comorbid illnesses (e.g. Reich and Green, 1991; Alnaes and social integration or should they be aimed at more radical
Torgersen, 1997), although Mulder (2002), in a systematic personality change? Many who suffer with personality dis-
review of the treatment of major depression in the context order are as likely as those who do not to equate their per-
of personality disorder, found a main problem that, even in sonality with their sense of self and being an independent
such trials, people with personality disorder were less likely human being, so they are understandably both offended and
to receive adequate treatment for their depression. What, frightened by the idea that, in effect, they must be changed.
however, is the effect of comorbid illness on the treatment Perhaps they would be reassured by the suggestion:
of personality disorder? One study in a high security hospital
hints at a lower success rate (Reiss et al., 1996), but substan- Psychotherapists do not aim to make radical changes
tive studies are urgently needed in this area. The implications in a patient’s personality, but rather to smooth down the
for the structure of treatment programmes are potentially rough edges with fine sandpaper – to make the abrasive
substantial, not only in a hospital setting but also, given the person more polite, the impulsive person more restrained,
high rates of comorbidity among prisoners presented in the and so on (Stone viii, 2006a).
Singleton et al. (1998) England and Wales national prisoner A third major challenge lies in finding the optimal timing
survey, within prisons. for interventions. It seems logical to postulate that inter-
ventions should be offered as early as possible, preferably
Treatment of personality as preventive measures for children and adolescents at
disorder risk of developing personality disorders. This makes sense,
because traits and risk factors which are associated with
The Challenge of Providing Treatment personality disorder in an adult are, by definition, already
for People With Personality Disorder evident in children and/or adolescents and they pose con-
siderable risks to the health of the individual concerned,
There is no difficulty in listing potential barriers to deliv- including accidental death and suicide (Robins and Rutter,
ering treatment for offenders with personality disorder, 1990), as well as risks to others and substantial costs to
but four are especially salient. First, the evidence base for society (Scott et al., 2001; Walsh, 2001). Furthermore, given
the effectiveness of any currently recognised interven- the nature of the pathologically intrusive disorders of
tion is weak. While the nihilist might interpret this as interpersonal function which are at the heart of personal-
‘nothing ought to be done’, in practice it often has the ity disorder, it is arguable that the counter-transferences
opposite meaning, since the results of treatment trials engendered over time, and perhaps other environmental
generally reflect small effects from each treatment and/ influences, will serve to reinforce the pathological traits
or little difference between them in their impact, rather (Maughan and Rutter, 2001). Thus, the longer treatment is
than an absence of effect or harmful impact. Thus, a more deferred, the harder it will be to get a good outcome. On
appropriate interpretation of the findings may be that any the other hand, there is a view that, insofar as there are
of a wide range of treatments may be tried, since, from more organic explanations for the difficulties, interventions
dynamic psychotherapy on the one hand to antipsychotic which have a fundamentally holding function may be suf-
medication on the other, the weakness of the evidence is ficient to keep the individual safe while natural maturation
on what works best for whom. processes take place (Stone, 1990).
One reason why evaluation of treatments yields less A final problem for consideration here, although this
information than the practitioner would like lies in the is by no means an exhaustive list, is that even among the
second major challenge – that there is disagreement on the subgroup of people with personality disorder who become
purpose(s) of interventions for offenders with personality offenders, a very large number need help with personal-
disorder. To put the dilemma in its crudest form – is the ity disorder, but resources for providing this are scarce
main aim to reduce the rate of re-offending or to improve – both in terms of the cash needed to fund services and
psychological wellbeing, or both – perhaps achieving the the expertise needed even when funds are available. Thus,
former through the latter? Underpinning this confusion is many people languish in prison or in the community with
a further question about the nature of the psychological substantial unmet need, surrounded by a largely untrained
improvement to be expected even if it accepted that this is workforce that ought to offer some treatment, but cannot
398 © 2014 by Taylor & Francis Group, LLC
Treatment of personality disorder
do so. The magnitude of the resultant task facing those focused – for alleviating aspects of the child’s environment
trying to develop a sensible and sustainable policy is clear. which would put him/her at increased risk of developing
The issue of resources also poses limits on what can be personality disorder. With respect to the latter, preven-
achieved through preventive programmes. tion and amelioration of the effects of child abuse are
particularly important, through early health visitor atten-
Faced with these challenges, there is a temptation to tion (Olds et al., 1998) or other home visit programmes
be disheartened. If, for instance, practitioners cannot agree designed to improve parental safety skills and knowledge
on such fundamentals as the nature and categorisation (MacMillan et al., 2009). The difficulties in sustaining such
of personality disorder, the novice might be tempted to programmes are, however, considerable. Most people
abandon the field altogether and focus only on people with struggle to maintain beneficial health strategies which will
axis I conditions where the terrain appears more certain. prevent disease and other forms of harm at some notional
Succumbing to this temptation is, we believe, a mistake time far in the future – hence the difficulties in establish-
as there is now sufficient agreement on essentials that ing healthy eating or drinking programmes and preventing
sensible and informed plans can be developed for offend- smoking (e.g. Connor and Norman, 1996). Mere knowledge
ers with personality disorder, potentially saving lives and of the issues is necessary, but insufficient, and outcomes
money. In fact, for a set of disorders in which there is inher- influenced by factors such as perceived ability to influence
ent possibility of maturation over time, the prognosis in one’s own life and, in effect, capacity for delayed gratifica-
appropriately treated cases may be much better than that tion. The concept of promoting the possible absence of
for people with the chronic deteriorating conditions which some unwanted problem at some vague time in the future
constitute the more severe forms of schizophrenia and at the cost of denying oneself the more immediate and
other functional psychosis. obvious rewards of a clearly pleasant or wanted experi-
ence is hard for any of us. Such problems are multiplied
Prevention of Personality Disorder when the primary beneficiaries, almost by definition,
lack motivation for prevention programmes. Further, in
Harrington and Bailey (2003, 2004a) summarised the dif- economically strained times, the community too must be
ferent approaches to prevention of antisocial personality motivated to maintain financial commitment to relevant
disorder and reviewed evidence of their effectiveness. programmes (Offord and Bennett, 2002).
Broadly, there are two main approaches to primary pre-
vention, aimed at reducing the incidence of adult antiso- The third major approach lies in, effectively, secondary
cial personality disorder – the application of ‘universal’, prevention and actual treatment of the child or adolescent
school-based programmes and selective prevention. The at risk, and/or his/her family.
former is aimed at improving relevant aspects of the
environment for everyone, and particularly targets aggres- Paving the Way to Effective
sive behaviours, academic failure and low commitment Treatment of Personality Disorder
to school. Thus, interventions include reduction in class
size, improved organisation and oversight of classroom Recognising the needs of the staff
behaviour, behaviour management strategies and good Clinicians justifiably complain about a disconnection
citizenship skill promotion (Hawkins and Herrenkohl, between the careful assessment and diagnosis of personal-
2003). Bullying has received specific attention, and effec- ity disorder and deciding on the most appropriate treat-
tive programmes reduce this by at least 50% (e.g. Olweus, ment. Problems are as likely to lie with the staff as with
1994). A criticism of this type of approach with respect to the prospective patients. For staff, some difficulties arise
prevention of disorder is that, because it has to be relevant in their reading of the literature on effectiveness of treat-
to everyone, it is too dilute to be of much use to those who ments, which we will explore more fully below, but there
are most at risk. are likely to be even more fundamental difficulties in get-
ting started. A qualitative evaluation of narratives on the
Selective prevention, however, requires identification concerns of psychiatrists in general about working with
of those children thought to be most at risk, which in itself people who have a personality disorder revealed a core cat-
requires resources, but also raises concerns in the minds egory of ‘echoing the pathology’, which indicated that the
of many workers about the risk of stigmatising children, complex pathologies of these disorders tend to be reflected
and that the resultant sense of otherness and/or open in the behaviours and attitudes of professionals ( Jones,
acknowledgement of the difficulties of these children 2011). This was underpinned by five conceptual categories
could add to their burden. Interventions here may be vul- of ‘failure to pin down the concept’, ‘disruption of role iden-
nerability focused – for some personal problem which may tity’, ‘projection of blame’, ‘absence of buy-in’ and ‘intra/
increase the risk of personality disorder, such as attention
deficit hyperactivity disorder – or they may be situation
© 2014 by Taylor & Francis Group, LLC 399
Personality disorders
inter-professional turmoil’. The theory generated is that Promoting patient engagement
work with people with personality disorder is a process of during the assessment process through
working with this echo of pathology. The process entails a collaborative approach
movement along a continuum between simply echoing
the pathology at one end, through recognition and man- With expert support, in a treatment unit in which the
agement of this, to being able to use personal experience patient can gain a sense of personal safety, offender-
to help both the patient and colleagues. Factors likely to patients with personality disorder may be helped towards
provide barriers to this process include perception of the engagement with treatments designed to meet their
wide array of treatments available, each needing its own needs (Sainsbury et al., 2004; Cordess, 2006). How might
set of skills, unfamiliarity with most of those skills and with the foundation for such a climate be created? It must start
the kind of intensively supportive team work which must early in the assessment process, so that, from the outset,
accompany them. All of this may engender a sense of help- the patient has an opportunity to feel like a partner in
lessness. Factors facilitating treatment include provision of the decisions to be made about treatment. For offender-
appropriate training (Miller and Davenport, 1996; Krawitz, patients under compulsory hospital detention and/or a
2004; Commons Treloar and Lewis, 2008), s upervision and treatment contract under a community sentence, those
reflective practice (Moore, 2012). aspects of management which relate to security may be
non-negotiable, but with respect to treatment all the
Setting boundaries usual clinical principles of information sharing and real
If there is something in this concept of working with ‘ech- consent should apply. By definition, however, most people
oes of pathology’ – an echo in itself of psychoanalytic con- with personality disorder will have difficulties in form-
cepts of transference and countertransference – it follows ing a healthy therapeutic alliance. Common to several
that boundary recognition and setting are important steps models of personality disorder is the expectation that
in preparing for treatment. Boundaries may be physical – acts of sabotage are likely to threaten the alliance and
the most obvious being the physical structures inherent therefore continuity of the intervention (Bender, 2005).
in secure hospitals or prisons – but they must also be psy- These issues are therefore taken up at the outset, using
chological. Walls and locks are an aid, not a substitute for non-punitive interpretation and collaborative assignment
well-structured and maintained clinical teams, members of responsibility for preservation of the treatment with
of which meet regularly, communicate clearly, have explicit the patient, (Plakun, 1994). Derksen (1995), Millon (1999)
lines of responsibility, and keep their work under constant and Bender (2005), amongst others, have provided sum-
review. These teams need to be embedded in a supportive, maries of characteristic ways of relating allied to each
wider management structure, in which, in turn, there is cluster (A, B and C) of personality disorders which have
honesty about what may go wrong as well as achievements specific value in framing appropriate responses for effec-
expected. As part of a process of enhancing service quality tively involving and retaining the patient in treatment.
and keeping adverse events to a minimum, resources are These are very broadly summarised in table 16.3, but we
necessary, but administrators can only provide these if they recommend that readers seek out the original texts for full
are adequately justified and if all parties are working in a context and further information.
climate in which good practice is at least as well marked as
any poor practice, and there can be full and frank inquiry Ben-Porath (2004) notes that the first five sessions
into the adverse events which will inevitably occur from with a prospective patient with personality disorder may
time to time, however good the practice. hold particular significance, as an alliance is being estab-
lished. It has been argued that factors such as the indif-
ference and inexperience of some ‘therapists’ contributes
Table 16.3 Characteristics that complicate the process of engagement in forensic clinical assessment,
according to personality types. (Sources: Derksen, 1995; Bender, 2005; Millon, 1999.)
Cluster A: Schizotypal, schizoid, paranoid personality disorders
Key issue: Profound impairments in interpersonal relationships
Strategies: Seek to establish trust, build alliances slowly over time, respect the need for distance
Cluster B: ‘Dramatic’: antisocial, borderline, histrionic, narcissistic personality disorders
Key issues: Limit pushing and impulsivity; emotion dysregulation; discontinuities in the sense of self; relationship breakdown; grandiosity
Strategies: Collaborate, contract, formulate, set limits, repair ruptures, manage countertransference
Cluster C: ‘Anxious/fearful’: avoidant, dependent, obsessive–compulsive personality disorders
Key issues: Emotional inhibition; aversion to interpersonal conflict
Strategies: Note instances of feeling guilty and internalisation of blame; empower to reduce over-compliance; monitor treatment objectives
particularly in relation to the termination of contact
400 © 2014 by Taylor & Francis Group, LLC
Treatment of personality disorder
to treatment failures with offenders with psychopathy enhanced it, with patients spontaneously contributing to
(Martens, 2004). In addition to the general clinical ethic the process of the description of their problems.
of collaboration, the use of contracts, goal-setting and
techniques of validation, including overt listening to and As Schotte (2002) has suggested, descriptive diagnostic
understanding the patient (Linehan, 1997), may assist assessment of personality disorder, based on a biopsycho-
with engagement. What have offender-patients them- social model, involves feedback to the patient in the clini-
selves reported to be helpful? Detained patients with cal context. There is some small scale research evidence
personality disorder in high security preferred a ‘firm but (D’Silva and Duggan, 2002) that patients with personality
fair’ staff attitude (Ryan et al., 2002). In addition, clini- disorder in forensic settings remain less likely to ‘know’
cians have been asked to share the basic issues without their diagnosis than those with other mental health prob-
jargon, ask relevant questions, and be flexible about time; lems, and more likely to have undertaken their own search
this last refers to the tension between boundaried and for this type of information by reading their clinical records.
containing practice and the experience of some in secure Building on the usefulness of a psycho-educational pack-
conditions that it is difficult to convey adequately the age, Banerjee and colleagues (2006) have demonstrated
complexity of their histories within an hour on a once a positive impacts on the therapeutic alliance following four
week basis (Denborough, 1996). sessions of information sharing in which the patterns asso-
ciated with specific personality traits are articulated, and
In an interesting study with people with drug depend- areas for change are identified by the patient. If this type of
ence histories in residential treatment under a criminal feedback and engagement were routinely applied, or even
justice mandate, Sung et al. (2001) observed that incidents adapted as a groupwork package which incorporates a sup-
of rule violation were common. They added, however, that, portive function, talking about personality disorder would
aside from a minority who never engaged with treatment, no longer be done in the patient’s absence, and active
the more minor episodes of non-compliance seemed not inclusion would become a more richly embedded clinical
to pose serious obstacles to eventual recovery. Reports in practice. Such processes are used in the English high secu-
the literature thus seem unanimous in highlighting two rity hospitals (e.g. Perkins et al., 2007; Tennant and Howells,
essentials for successful assessment of people with complex/ 2010; Willmot and Gordon, 2011),
severe personality disorder:
●● continuing attention to the potential need for repair- Setting the goals of treatment
Cure is a false hope in many conditions in all fields of medi-
ing ruptures in the patient–doctor/therapist alliance, cine. ‘Cure’, meaning complete resolution of the condition,
and an ability to do so, especially during early meetings is desirable, and may be achievable, but it is often an unre-
(Plakun, 1994); alistic goal. This is also true in respect of work with people
●● flexibility of approach within clearly boundaried inter- who have a personality disorder but, also as in other fields
changes (Meux and Taylor, 2006). The therapeutic in medicine, there are other acceptable alternative goals.
community model of treatment includes patient These include improvement, holding without deterioration
involvement from the outset, and staff structures that and even palliation, in the form of some relief of the maxi-
are sufficiently flexible to avoid unnecessarily hierar- mum distress to self or others without real change. Until
chical and authoritarian styles that can undermine the Mental Health Act 2007, for England and Wales, most of
engagement (Stern et al., 1986; Norton, 1992; Warren the goals from this range were even explicit when consider-
and Dolan, 1996). ing compulsory detention for treatment.
Sharing the diagnosis of personality The process of assessment just described allows for set-
disorder with patients ting attainable goals for treatment with the patient, and for
Fears of somehow compromising the therapeutic alliance their prioritisation. At best, this leads to an explicit state-
through mention of the ‘label’ of personality disorder to the ment of a series of desired outcomes, such that the patient
patient appear in the clinical literature of the early 1990s and the therapists all have clear markers for success,
(Davidson, 2000). Every patient is, however, entitled to and thus can recognise progress when it is being made.
know and understand his/her diagnosis. In England and The creation of stepwise goals is important – achieving
Wales this is even now stipulated in the Patient’s Charter change in personality, or even personality traits, is a large
for Mental Health Services (DOH, 1997a), perhaps reflect- and complex task and, as an entirety, too hard for most
ing the extent to which mutual denial of the condition people, including staff, to contemplate. Modest increments
may become a problem. The essential question becomes: of change are more attainable, and markers of success
how might the diagnosis of personality disorder be most along the way provide a framework which better engages
constructively shared with the recipient? In a short paper the patient and therapist alike. At first, the increments of
on the benefits of feedback, Tyrer (1998) observed that, change must be small, so that they are fairly quickly achiev-
rather than diminishing collaboration, feedback typically able, and reinforce the idea for all parties that change is
© 2014 by Taylor & Francis Group, LLC 401
Personality disorders
possible. Completion of a complex assessment process is in and finally victim empathy at the highest level of the third.
itself an indication of progress for many people. Everyone A variant of this model is shown in figure 16.1. Setting out
needs evidence of progress to stay motivated. with a clear structure of this kind, while retaining the capac-
ity for flexibility as the treatment is implemented, is the key
Structuring the programme with achievable to successful engagement and retention of the patient. An
goals which reward patient and staff alike attentive therapist will see when this more-or-less sequen-
Completion of an extended assessment is in itself an tial approach needs adjustment, and, for example, be able
achievement, which needs to be acknowledged by staff and to help an individual work on self-integration (phase 3)
patients alike. Next steps are commonly around treating while also working on relationships with others (phase 1).
comorbid illness, offering treatment or training in coping
strategies for certain developmental and/or educational Evidence for the Effectiveness of
deficits and perhaps offering medication to reduce the Treatments of Personality Disorders
impact of specific enduring traits. Patients for whom a
particular trait, such as impulsivity, is a prominent part of Measurement of the effects of treatment on personality dis-
presentation may not be able to engage in psychological order is extremely difficult. The general principles for doing
therapies until they are able to get some relief from this so are that the condition to be treated, the treatment itself
through medication. Failure to cover this ground may limit and the increments of change which are indicative of suc-
capacity to engage. In this context, even comorbid psycho- cess should all be clearly definable. As we have already dis-
sis, so long as it is well controlled, is not necessarily a bar cussed, while there is still some concern over the reliability
to treatment of personality disorder. The sort of develop- of diagnosing individual personality disorders, uncertainty
mental issues which are critical here relate to speech and about the life course of personality disorders is even more
language. It is not uncommon for people with personality substantial. Moran (1999, pp.43–50) purports to show
disorder to have language skills in the borderline impaired the ‘natural course’ of antisocial personality disorder/psy-
range, and yet most treatment opportunities depend on chopathy, but the studies he lists are, in fact, of people with
being competent enough for quite sophisticated psycho- whom health and criminal justice services have intervened
therapies; some of the language difficulties may be quite a good deal, albeit perhaps not therapeutically. Others have
subtle, such as difficulties in understanding and dealing followed people with borderline personality disorder for
with irony or certain forms of humour, like satire; so, test- up to 27 years but, again, people identified by service use,
ing to ensure recognition of such difficulties and teaching and at least intermittently using services (e.g. Stone, 1990;
relevant coping strategies may be vital. Paris and Zweig-Frank, 2001; Zanarini et al., 2010). True
natural course studies are, realistically, unlikely among peo-
Thereafter, a programme of therapeutic work may be ple who are such heavy service users, and would probably
mapped out according to need and developmental stage be unethical. Interpretations of the outcomes of trials of
in recovery. In one comprehensive model for working with treatment are, as for most conditions, further bedeviled by
personality disorder in a high security hospital (Newrith the fact that the most valued form of evaluation – the ran-
et al., 2006) the post-stabilising programme of psychological domised controlled trial – is invariably completed with the
work included structured, task oriented streams of group most altruistic and co-operative people, often with disorder
work and an unstructured more psychodynamic stream, which is in the middle range of seriousness. Altruistic, co-
each delivered in an acknowledged hierarchy, such that, operative people are, by definition, unusual among those
with progress, patients ‘graduated’ from a lower level group with serious personality disorder, so it is arguable that
to a higher level group. True psychodynamic/analytic ther- results of such trials should be regarded with more than
apy would invariably be the pinnacle of active psychological usual caution in this field. Measurement of change within
treatment with, as an inpatient, the task also of beginning to treatment programmes, particularly when that is within a
prepare for separation from the institution and away from therapeutic institution of some kind, may be further con-
perhaps the only healthy attachments many of the patients founded by the difficulty in these circumstances of knowing
had so far achieved. The task oriented stream, largely when change is sufficiently extensive and robust that it is
based on a cognitive behavioural method of working, but generalisable to other environments or whether the meas-
including drama work as a means of accessing affect and ured changes better reflect the institution’s capacity to find
empathic responses, included social, sexual and self and a good fit with the individual’s needs. If the latter, this could
sensory awareness streams. The hierarchy of tasks within mean that as long as s/he remains there, s/he functions
such streams was from social attitudes and skills, through optimally, but once back in his/her natural environment,
anger management to assertiveness and social coping in the pathological traits and coping mechanisms will come
the first, sexual knowledge, interpersonal relationships and to the fore again. That said, it would be unthinkable to dis-
then sexual relationships within those in the second, and charge a person with psychotic illness from inpatient treat-
self and sensory awareness, followed by family awareness ment without an adequate package of ongoing treatment
402 © 2014 by Taylor & Francis Group, LLC
Treatment of personality disorder
Phase 1 Phase 2 Phase 3
Establishing therapeutic Managing dissocial Integration of
engagement behaviour self-states
Containment Maintaining engagement Preparing for transition
Main Therapeutic Schema
activities Alliance Formulation
Initial Assessment Medication Social Skills Psychodynamic
CBT Schema focused
CBT
Anxiety level Unpredictable Low Cognitive analytic therapy
of staff
Significant Minimal High
Therapeutic Training Training
expertise Significant
Training
Figure 16.1 A stepped approach to the treatment of personality disorder.
and reviews. There is, however, a tendency not to structure except when treating comorbid conditions. This does not
transitional and discharge arrangements with the same mean that drugs cannot be used, but rather that they should
rigour for people who have personality disorder. Small be used with caution and with particularly systematic moni-
wonder, then, that from an early general overview of the toring of any positive and/or negative effects.
evidence on treatment of personality disorder (Dolan and
Coid, 1993) to more recent systematic reviews (see below), Stein (1993), in what is not styled as a systematic
the overall answer to questions about the effectiveness of review, but appears to be so, examined the evidence for the
treatment for personality disorder is equivocal. effectiveness of the range of substances which had been
studied in this context up to that date – treatment with low
Drug treatments for personality disorders doses of neuroleptics, tricyclic antidepressants for people
Medication is commonly prescribed for people with per- with borderline personality disorder, monoamine oxidase
sonality disorder, at least at times. Early estimates were that inhibitors (MAOIs), lithium, benzodiazepines, anticonvul-
from about half (Soloff, 1981) to more than three-quarters sants, psychostimulants and even electroconvulsive treat-
(Andrulonis et al., 1982) of people presenting for treatment ment. He was positive about progress in the field, since he
would be prescribed something. There is no specific ‘person- considered that both medications available and evidence
ality disorder drug’, so an ever present question is whether on their effects had improved so much during the 1980s,
any medication is really treating aspects of the personality albeit starting from a state of almost no knowledge at all of
disorder per se or, rather, helping with comorbid conditions this particular aspect of treatment. Most of the evidence,
which may have been the trigger to a period of decompen- however, appeared to relate to people with borderline
sation, and, in the presence of personality disorder, may not personality disorder, and a range of methods of study had
always present in classical form. Perhaps this doesn’t matter been applied, including simple observational accounts of
as long as the patient improves, and, in this context, if it can naturalistic studies. He discussed the difficulties in com-
be said that any medication appears to be helpful, then it pleting trials with people who have such problems, not
may at least not be contraindicated. In the UK, however, no least because of difficulties, inherent to the disorders, with
drug has been licensed as a specific treatment for personality drug compliance and retaining people in trials, and because
disorder, so drug use must still be regarded as experimental some drugs which appeared to confer benefit might be just
too risky; an example of the latter would be the MAOIs,
which for the patient’s physical safety require abstinence
© 2014 by Taylor & Francis Group, LLC 403
Personality disorders
from alcohol, and a good many other substances – legal disorder; again, trials were most commonly with people who
and illegal. Nevertheless, even by 1983, several randomised had borderline personality disorder. Multiple meta-analyses
controlled trials (RCTs) had been completed. were performed with the data to take account of the nine
classes of drugs trialed and the range of behavioural features
By 2006, then with the aid of 15 electronic databases, measured. The latter were cognitive perceptual symptoms,
35 RCTs of pharmacological treatments for personality affective dysregulation (including depression, anxiety, anger
disorder could be located (Duggan et al., 2008). The range and hostility), impulsive behavioural dyscontrol, global func-
of treatments tried showed little difference from the Stein tioning – this area including perhaps the most typical of per-
study; atypical neuroleptics were added, electroconvulsive sonality disorder features in interpersonal symptoms/signs
treatment not mentioned, and one experiment had been – physical function, and leaving the study early. Reduction
conducted each with a hypotensive agent (clonidine) and of aggression in the context of anticonvulsant prescription
a dietary supplement (omega-3 fatty acids). The studies and modulation of cognitive perceptual and other subjec-
included are shown in table 16.4. Probably the most striking tive mental state disturbance with anti-psychotic medi-
features are the small sample sizes in most studies and also cation were the only two domains of significant success.
the brevity of treatment in the majority. A few of the stud-
ies had targeted a particular behaviour, such as substance Lieb et al. (2010) set out to update the review specific
misuse or suicidal acts, but among people with personality to treatment of borderline personality disorder. They found
Table 16.4 Summary of completed randomised controlled trials of pharmacological treatments for
people with personality disorder
Study Sample Drug tested Usable outcomes1 Duration of Authors’ claims
trial
* Arndt et al.,1992 29 men with ASPD Desimipramine Days with psychological 12 weeks2 Those with ASPD made few gains
among 59 substance v. placebo, both problems, opiate use, with desimipramine or placebo
* Battaglia misusers with standard or medical problems 6 months
et al., 1999 methadone Marked reduction in self-harm in
* Bogenschutz and 32 men 25 women treatment Leaving the study early; 12 weeks both groups
Nurnberg, 2004 repeated self-harm Low dose v. suicidal behaviour
* Coccaro and with PD ‘ultra-low’ dose 12 weeks Olanzapine superior on BPD
Kavoussi, 1997 15 men, fluphenazine Leaving the study early measures on BPD-clinical global
25 women with Olanzapine v. 32 days impressions scale
* de la Fuente and BPD placebo Leaving early; side- 12 week trial + Sustained reduction in irritability
Lotstra, 1994 28 men effects; quality of 12 week open and global rating of improvement
Zanarini et al., 12 women, any PD Fluoxetine v. life; mental state3; label with fluoxetine, regardless of
2011 with aggression and placebo aggression/irritability depression, anxiety, or alcohol
irritability 12 week trial; use
Schulz et al., 2008 Carbamazepine v. Leaving early; 12 week open Carbamazepine: no significant
6 men placebo behavioural dyscontrol label extension positive effects
* Frankenburg and 14 women BPD 2.5 mg Mental state, self-harm, 6 months Higher dose olanzapine superior
Zanarini, 2002 119 men olanzapine v. core BPD symptoms on ZAN-BPD total score, but not
* Goldberg 322 women 5–10 mg (Zanarini scale) 12 weeks depression nor self-harm; 2.5 mg
et al., 1986 olanzapine v. superior on self-harm and identity
91 men placebo Mental state, self-harm, disturbance only
223 women with Flexible dose core BPD symptoms Olanzapine superior on
BPD olanzapine v. (Zanarini scale) aggression and irritability
placebo measures, but not core BPD
27 women with Leaving early, mental measures
BPD Divalproex sodium state, behaviour, side- Divalproex superior in reducing
v. placebo effects anger and interpersonal sensitivity
21 men Leaving early
29 women with Thiothixine v. Significant thiothixene effect on
BPD and/or placebo psychotic, obsessive–compulsive
schizotypal PD and phobic anxiety symptoms,
‘more than an antipsychotic effect’
404 © 2014 by Taylor & Francis Group, LLC
Treatment of personality disorder
Table 16.4 (Continued)
Study Sample Drug tested Usable outcomes1 Duration of Authors’ claims
trial
Hallahan et al., 7 men Omega-3 fatty Depression, self-harm, 12 weeks Omega-3 fatty acids superior
2007 15 women acid v. placebo, impulsivity, aggression, on depression, self-harm and
both with standard daily stresses 10 weeks daily stresses, not impulsivity or
* Hollander 10 men psychiatric care aggression
et al., 2001 11 women BPD Divalproex v. Leaving early; global 12 weeks Divalproex superior to placebo
placebo state; mental state; 6 weeks + 6 for impulsive aggression, irritability
* Hollander 91 (63% men) behaviour months follow- and cluster BPD global rating
et al., 20034 Divalproex v. Leaving early; up Divalproex superior for impulsive
* Joyce et al., 2003 40 men placebo behaviour 9 weeks aggression
43 women, 30 Fluoxetine v. Leaving early; mental Poor outcome for depressed
* Koenigsberg BPD, 53 other PD, nortryptyline state patients with BPD on nortryptiline
et al., 2003 all major depression
19 men Low dose Leaving early, global Risperidone superior on PANNS
4 women with risperidone v. state, mental state, negative and general scale by
schizotypal PD placebo schizotypal symptoms week 3 and positive symptoms
scale by week 7
* Leal et al., 1994 11 men Amantidine v. Leaving early; money 12 weeks ASPD poor prognostic indicator;
8 women ASPD + desimipramine per week spent on 6 weeks medication no advantage
* Leone, 1982 cocaine abuse v. placebo, all + cocaine 10 weeks
standard care 6 months Both drugs improved symptoms
* Loew et al., 32 men Loxapine succinate Symptom changes; with loxapine being superior, side-
2006 48 women with v. chlorpromazine leaving early; side- 8 weeks effects occurred in a third
* Montgomery BPD effects Significant advantage for
et al., 1983 56 women with Topiramate v. Leaving early, side- topiramate on mental state and
BPD placebo effects, mental state, PD measures
* Nickel et al., quality of life No improvement; no significant
2006 20 men Mianserin v. Suicide attempts difference between groups
38 women with placebo
BPD and/or Mental state (SCL-90); Aripripazole superior on most
histrionic PD + self- Aripripazole v. self-injurious outcomes mental state scores and state-trait
harm placebo anger but not self-harm
9 men Topiramate superior in treating
43 women BPD anger in men with BPD
Topiramate superior in treating
* Nickel et al. 44 men with BPD Topiramate v. Leaving early, mental 8 weeks anger in women with BPD
2005 placebo state CBT superior (but for social
* Nickel et al., 31 women with Topiramate v. Leaving early, mental 8 weeks phobia rather than PD)
2004 BPD placebo state
* Oosterbaan 48 men Moclobemide v. Leaving early, side- 15 weeks + 2 No significant advantage for
et al., 2001 34 women, 50% CBT v. placebo effects and 15 month ziprasidone
with avoidant PD, all follow-up
Pascual et al., 2008 with social phobias Ziprasidone v. Clinical Global No significant advantage for
60 patients with placebo Impression Scale for 12 weeks naloxone, but the more BPD
* Philipsen BPD BPD criteria the better the naloxone
et al. 2004a Naloxone v. Dissociative symptoms Given twice in response
9 women with BPD placebo 15 min before and after double blind
* Philipsen et al. naloxone/placebo cross-over Acute inner tension, dissociation,
2004b 14 women with Clonidine 75 µg design during suicide related behaviours
BPD v. clonidine 150 µg None 6–35 days reduced in both groups (within
4–16 days the hour)
(Continued)
© 2014 by Taylor & Francis Group, LLC 405
Personality disorders
Table 16.4 (Continued)
Study Sample Drug tested Usable outcomes1 Duration of Authors’ claims
trial
* Powell et al., 65 men ASPD + Nortryptiline v. Leaving early; alcohol Only significant advantage of
1995 alcohol dependence bromocriptine v. abstinence 6 months (post- active drugs was with nortryptiline
placebo detox) in the ASPD group
* Rinne et al., 38 women with Leaving early, side- Fluvoxamine superior for
2002 BPD Fluvoxamine v. effects, mental state, 6 weeks double sustained reduction in rapid
placebo + behaviour blind; 6 weeks mood shifts, but not impulsivity or
single blind; 12 aggression
* Salzman 8 men Fluoxetine v. Mental state weeks follow-up Fluoxetrine superior in reducing
et al., 1995 14 women with placebo 12 weeks anger
BPD/BPD traits
* Serban and Thiothixene v. Leaving early, side- 3 months 84% patients markedly improved
Siegal, 1984 36 men haloperidol effects at 3 months in both groups
16 women BPD
* Simpson and/or schizotypal Fluoxetine v. Leaving early; global 12 weeks No added benefit for fluoxetine
et al., 2004 placebo, both + state; mental state;
20 women with DBT impulsive aggression,
BPD suicide related
Olanzapine v. behaviours, ER visits
* Soler et al., 2005 8 men placebo, both with 12 weeks Improvement in both groups;
52 women with adapted BPD Leaving early, global olanzapine superior for
BPD state, mental state 5 weeks depression, anxiety and impulsive
5 weeks aggression
* Soloff et al., 26 men Haloperidol v. Leaving early, global
1993 82 women with phenelzine state, mental state 6 months Phenelzine superior for anger and
BPD hostility; no other drug advantages
* Soloff et al., Haloperidol v. Leaving early, global 8 weeks
1989 22 men amitryptiline v. state, mental state, Up to 52 Haloperidol superior to placebo
68 women with placebo impulsivity weeks in depression, hostility, schizotypal
BPD 6 weeks symptoms, impulsivity global
8 weeks function; amitriptyline only
Steiner et al., 2008 24 women with Olanzapine v. Standard measures or 8 weeks superior for depression
BPD placebo, both with irritability, aggression,
DBT self-harm and More rapid reduction of irritability
* Tritt et al., 2005 24 women with depression with olanzapine; self-injury tended
BPD Lamotrigine v. to decrease more with DBT
placebo Leaving early, mental alone
state
Lamotrigine superior for all
* Verkes et al., 37 men Paroxetine v. Leaving early aspects of anger except internally
1998 54 women, recent placebo directed anger
suicide attempt,
* Zanarini and 84 with PD Olanzapine v. Leaving early Paroxetine superior in reducing
Frankenburg, 2001 placebo suicide related behaviours; not
28 women with Leaving early, mental other behaviours/symptoms
BPD Omega-3 fatty state, behaviour
acids v. placebo Olanzapine superior in rate of
* Zanarini and 30 women with Leaving early; side- improvement in all areas except
Olanzapine (o) effects depression
Frankenburg, 2003 BPD v. fluoxetine (f) v.
o + f Omega-3 fatty acids superior
* Zanarini 45 women with in reducing aggression and
et al., 2004 BPD depression
All three conditions effective, but
fluoxetine alone least so
1 Usable outcomes refers to all those outcomes measured systematically and reported sufficiently transparently to allow for meta-analysis;
2 Times all refer to periods of active treatment; many studies explicitly added baseline periods;
3 Reference to mental state as an outcome means that a systematic measure was used;
4 A later publication (Hollander et al., 2005) is not included here, it overlapped with other Hollander studies. It analysed borderline personality
disorder data separately, suggesting divalproex superior for impulsive aggression.
*Included in Duggan et al. (2008) meta-analyses.
ASPD, antisocial personality disorder; BPD, borderline personality disorder; DBT, dialectical behaviour therapy; ER, emergency room; PANNS, positive
and negative symptoms of schizophrenia; PD, personality disorder; SCL-90, symptom checklist 90 items.
406 © 2014 by Taylor & Francis Group, LLC
Treatment of personality disorder
just four new trials, two of olanzapine (one of these a drug are dealing with such complex presentations that similarly
company trial of its own product), one an additional study complex interventions are needed in response. Reconciling
of omega-3 fatty acids, and a study of ziprasidone. These this tension between simple, brief interventions available
have been added to table 16.4. There was really no update of to many versus complex interventions available only to the
conclusions possible, and the article drew somewhat acid few is the leitmotif of current service provision.
observations from Kendall et al. (2010). It is hard to see why
this update justified a major publication; what is needed There are, however, many types of brief psychological
is a simple system of notes to provide clinicians with interventions. Given the current state of the evidence on
periodic updates. For the USA, the American Psychiatric effectiveness, how does one choose? Frank et al. (1991) and
Association (APA) does this in the form of a ‘guideline Wampold (2001), among others, have observed that many
watch’ (APA, 2005), which updates its position on the specific types of psychotherapeutic treatments achieve
treatment of borderline personality disorder (APA, 2001). virtually the same effects, largely because of a common set
This is the only APA guidance on treatment of personality of curative processes. While few disagree that this is true,
disorder. In England, the National Institute of Health and many, nonetheless, feel uncomfortable with its implica-
Clinical Excellence (NICE) guidelines on treatment of anti- tions. Does this emphasis on the non-specificity of any one
social personality disorder and borderline personality dis- treatment, not discredit that treatment’s effectiveness?
order are more recent (NICE, 2009a,b), with their cautious Glass (2001), for example, wrote:
recommendations that for antisocial personality disorder
‘prescribers will use a drug’s summary of product charac- There are those health policy analysts who argue that any
teristics to inform their decisions for each person’ and the therapy that uses non-specific diagnoses and non-specific
slightly more specific, additional advice with respect to treatments is somehow bogus witchcraft lacking indica-
borderline personality disorder on use of drugs to manage tions of when to begin and when to end, and its appli-
crises, comorbid conditions and insomnia. cation should be excluded from third-party coverage.
Time limited psychological treatments for Rachman and Wilson (1980), drawing on Alice in
offenders with personality disorder Wonderland and the Dodo Bird’s verdict on the caucus
race that ‘everyone has won and all must have prizes’, wrote
A major consideration influencing the take-up of an trenchantly on results of clinical trials in the field, which
effective treatment is the amount of effort and skill usually arrive at the conclusion that there is little difference
needed from care providers and the amount of tech- between the various psychological treatments:
nical back-up that they require (Marks, 1991).
Brief, simple and effective treatments will have a much If the indiscriminate distribution of prizes carried true con-
greater impact in the population overall than those that viction … we end up with the same advice … ‘Regardless of
require a great deal of technical knowledge and expertise. the nature of your problem seek any form of psychothera-
Most treatment offered to offenders with personality dis- py.’ This is absurd. We doubt even the strongest advocates
order – if they are fortunate enough to be offered any at of the Dodo Bird argument dispense this advice (p.167).
all – is likely to be of this brief type, although the national
guidance for England on treatment both of antisocial per- How, then, does one reconcile the research evidence sup-
sonality disorder (NICE, 2009a) and borderline personality porting importance of common factors on the one hand
disorder (NICE, 2009b) both express caution about the with the clinical reality that practitioners choose specific
need for a full and appropriate clinical framework when- therapies for certain specific problems? This dilemma may
ever brief psychological treatments are given; such guid- be resolved at two levels. Frank and Frank (1991), who were
ance draws on advice from an international collaboration among the first to develop the common factor position,
of researchers from elsewhere in Europe, Canada, New deal with one of these levels:
Zealand and the USA (http://www.agreecollaboration.org).
The idea, however, of time limited treatment for personality [our] position is not that technique is irrelevant to out-
disorder may read like a contradiction in terms. After all, come. Rather, as developed in the text, the success of all
surely people with a personality disorder are, by definition, techniques depends on the patient’s sense of alliance with
suffering from a long-term disturbance, so how could, say, the actual or symbolic healer. This position implies that
10–20 sessions of treatment be adequate? Could it even be ideally therapists should select for each patient the ther-
harmful? Further, if people with a personality disorder are apy that accords, or can be brought to accord, with the
also offenders, they have not been found to be especially patient’s personal characteristics and view of the prob-
responsive to interventions of any type (Woody et al., 1985; lem. Also implied is that therapists should seek to learn
Huband et al., 2007). It could be argued that forensic mental as many approaches as they find congenial and convinc-
health services in particular, and prisons more generally, ing. Creating a good therapeutic match may involve both
educating the patient about the therapist’s conceptual
scheme and, if necessary, modifying the scheme to take
into account the concepts the patient brings to therapy.
Given that the therapist’s person is, in part, the therapeu-
tic tool, a partial solution to this dilemma that differing
© 2014 by Taylor & Francis Group, LLC 407
Personality disorders
psychological therapies may be equally effective is to accept rates, with patient c haracteristics and need together with
that treatments are only likely to be effective when used by some environmental factors influencing this (McMurran
a practitioner who finds that therapy congenial, believes in et al., 2010). The four studies that investigated the relation-
it and practises it consistently. Hence, the choice of therapy ship between non-completion and treatment outcome
has to fit the therapist as well as the patient for it to be showed adverse outcomes for non-completers. One even
effective. The wide range of psychological therapies avail- showed a higher rate of re-offending among those who dis-
able for personality disorder may appropriately reflect the engaged than among those who were never offered treat-
wide range of practitioners in the field. ment in the first place (McMurran and Theodosi, 2007).
Hence, there is an ethical obligation to consider the risks of
A further important issue along these lines, which is disengagement, and what can be done to minimise it.
almost never considered in the research literature, is the fit
not only between therapist and therapy, but also between Motivational interviewing (Rollnick et al., 2008) is one
therapist and patient. Early observations suggested that way of reducing the risk of dropout from treatment, psych-
this was critical for people with schizophrenia – that oeducation (Livesley, 2001) is another, either way helping
therapists with certain personality styles did not or could the patient into a position whereby s/he can make accurate
not engage such patients and treat them successfully observations about her/his own behaviour rather than
(Whitehorn and Betz, 1954, 1960). Similar work would this coming more directly from the therapist and carry-
be important to explore fit between personal style of the ing a sense of threat. Methods have also been developed
therapist and patients with the main types of personality specifically to encourage these observations and help co-
disorder presenting for treatment. operation over prioritising the individual’s difficulties so
that they seem less overwhelming (e.g. D’Silva and Duggan,
A second possible solution to this dilemma over postu- 2002). Such interviewing styles also help to direct the
lated lack of specificity of treatments is that, while a range therapist’s attention as to where possible fault lines in the
of therapies may be applied, each will only work when therapeutic alliance are likely to develop, so that these can
there is a clear rationale for its use – both for those who be anticipated and worked through. Thus, for instance, the
deliver the treatment and for the patient making use of it. therapist might say to someone with a paranoid personality
Most forensic psychiatrists will have limited opportunity disorder: ‘I realise that it is part of your personality to feel
for consistent, long-term, individual psychotherapeutic suspicious of strangers, and perhaps that they might wish to
contact with their patients, indeed it may be inappropriate take advantage of you – so, you may from time to time feel
for the same person to have the overarching management like that about me too. This is something that we both need
responsibility in the case and be the therapist (see also to be aware of, and we need to be able to work together on
chapters 23 and 26). Thus, forensic psychiatrists require finding ways to increase your confidence in our working
an understanding of the strengths and weaknesses of the relationship so that your concerns don’t get in the way of
various approaches and the strengths of their colleagues in treatment.’ Preventive measures are not invariably success-
this particular respect, in order to be able to integrate this ful, however, so it is important, as well, to have strategies for
information into the decisions on therapy. Thus, specific dealing with therapeutic ruptures if they occur. Safran and
psychological treatments may be provided as the mainstay Muran (2000) offer suggestions for identifying early ‘rupture
of short-term intervention or, perhaps more commonly, as markers’, which may include increased irritability or other
an element of a treatment package, as already described. change within or outside the sessions, or decreased involve-
ment in homework, and note that they must immediately
Another important argument for short-term treatment be actively explored in the here and now. This is often
is that people with personality disorder struggle to complete daunting for both patient and therapist, as the tendency
treatment, although the data are largely restricted to those for both is to ignore the rupture, so the therapeutic alli-
with borderline personality disorder. Figures as high as 60% ance needs constant monitoring. Confrontation and inter-
early withdrawal rates have been reported in some trials pretation should be used sparingly at this stage, as these
(Waldinger and Gunderson, 1984; Gunderson et al., 1989), approaches provoke anxiety and are likely to be interpreted
although a systematic review of 41 studies of treatment as attacks, increasing the risk of withdrawal. Such anxiety is
completion specifically relating to people with borderline better contained by helping the individual to observe and
personality disorder was more encouraging (Barnicot et al., explore his/her own behaviour. Giving feedback to a patient
2011). For interventions of less than 12 months, they found with clinically significant narcissistic traits is particularly
that a random effects meta-analysis gave an overall com- difficult; in one case, one of us (CD) found that no matter
pletion rate of 75% (95% confidence interval (CI): 68–82%); what was said, the patient interpreted it as a ‘put down’,
it was barely different for longer interventions (71%, CI: and so became extremely angry and sullen for the rest of
65–76%). Factors predicting dropout were low commitment the day. He was helped by encouraging him to examine this
to change, poor therapeutic relationship and high impulsiv- recurrent pattern for himself. He then started to recognise
ity; sociodemographics were not relevant. Another meta- his need for unequivocal admiration, and that no matter
analysis of non-completion for people with personality
disorder more generally yielded slightly worse completion
408 © 2014 by Taylor & Francis Group, LLC