Behaviour Research and Therapy 68 (2015) 70e75

Contents lists available at ScienceDirect

Behaviour Research and Therapy

Shorter communication

Cognitive-behavioural therapy for outpatients with eating disorders:

a r t i c l e i n f o a b s t r a c t

Article history: Whilst there is a growing evidence to support the impact of cognitive-behavioural therapy (CBT) in the
Received 17 November 2014 treatment of adults with eating disorders, much of this evidence comes from tightly controlled efficacy
Received in revised form trials. This study aimed to add to the evidence regarding the effectiveness of CBT when delivered in a
2 March 2015
routine clinical setting. The participants were 203 adults presenting with a range of eating disorder
Accepted 6 March 2015
Available online 7 March 2015
diagnoses, who were offered CBT in an out-patient community eating disorders service in the UK. Pa-
tients completed measures of eating disorder pathology at the start of treatment, following the sixth
session, and at the end of treatment. Symptoms of anxiety, depression, and psychosocial functioning
Keywords:
Eating disorders
were measured pre- and post-treatment. Approximately 55% of patients completed treatment, and there
Cognitive-behavioural therapy were no factors that predicted attrition. There were significant improvements in eating disorder psy-
Effectiveness chopathology, anxiety, depression and general functioning, with particular changes in eating attitudes in
Attrition the early part of therapy. Effect sizes were medium to large for both completer and intention to treat
analyses. These findings confirm that evidence-based forms of CBT can be delivered with strong out-
comes in routine clinical settings. Clinicians should be encouraged to deliver evidence-based treatments
when working in these settings.
© 2015 Elsevier Ltd. All rights reserved.

There is growing evidence to support the use of cognitive- with high levels of supervision. It is unclear as to whether similar
behavioural therapy (CBT) in the treatment of adults with eating outcomes can be obtained from effectiveness studies, where treat-
disorders. Whilst early trials demonstrated the impact of focused ments are delivered in routine clinical settings. In such conditions,
forms of CBT for the treatment of bulimia nervosa (e.g., Bulik, clinician adherence to protocols is less closely monitored and the
Sullivan, Carter, McIntosh, & Joyce, 1999; Fairburn et al., 1995), diversity of cases is likely to be greater (e.g., higher levels of
recent studies have demonstrated the efficacy of an enhanced form comorbidity).
of the treatment (CBT-E) that is suitable for a broader range of To date, very few studies have considered the effectiveness of
eating disorder presentations (e.g., Fairburn et al., 2009). CBT has CBT for the eating disorders in routine clinical settings. Byrne,
since been shown to be more effective than psychodynamic psy- Fursland, Allen, and Watson (2011) conducted an open trial of
chotherapy in the treatment of bulimia nervosa (Poulsen et al., CBT-E for patients presenting with a broad range of eating disor-
2014), and is also suitable for use with underweight patients ders, including patients with a body mass index (BMI) of 14þ. They
(Fairburn et al., 2013; Watson & Bulik, 2013; Zipfel et al., 2014). reported significant improvements in eating disorder and general
However, the majority of the evidence for CBT has come from ef- psychopathology, with changes in scores on a range of treatment
ficacy studies e well-controlled treatment studies that often have measures indicating medium to large effect sizes. Of those who
tight inclusion criteria and are delivered under strict conditions completed therapy, two thirds were in full or partial remission at
the end of treatment. In another effectiveness study of CBT for
bulimia nervosa and atypical cases, Waller, Gray, et al. (2014) re-
ported similar remission outcomes to those found in efficacy
* Corresponding author. Eating Disorders Service, Southern Health NHS Foun-
dation Trust, April House, 9 Bath Road, Bitterne, Southampton, SO19 5ES, UK.
studies, with approximately 50% of patients being in remission at
E-mail address: hannahturner0@gmail.com (H. Turner). the end of treatment. However, those effectiveness studies were

http://dx.doi.org/10.1016/j.brat.2015.03.001
0005-7967/© 2015 Elsevier Ltd. All rights reserved.
 H. Turner et al. / Behaviour Research and Therapy 68 (2015) 70e75 71

not conducted with the same rigor as existing efficacy studies (e.g., The Eating Disorders ExaminationeQuestionnaire (EDE-Q,
lack of follow-up). They varied substantially in attrition rates, with version 6; Fairburn & Beglin, 2008). The EDE-Q is a self-report
the Byrne et al. study having a higher rate than the Waller et al. questionnaire assessing key cognitive and behavioural aspects of
study, probably due to the presence of anorexia nervosa patients in eating disorders. It generates frequency ratings for key eating dis-
the former. order behaviours (e.g., objective binge-eating, self-induced vomit-
These preliminary studies of effectiveness indicate that CBT can ing, laxatives misuse, and excessive exercise), as well as the
be delivered with strong outcomes in routine clinical settings. following attitudinal subscales: dietary restraint, weight concerns,
However, it is well-established that clinicians routinely fail to use shape concerns, and eating concerns. A global attitudinal score can
CBT when working with the eating disorders (e.g., Tobin, Banker, be calculated by averaging the four subscales. The EDE-Q has good
Weisberg, & Bowers, 2007) or that they deliver it in sub-optimal psychometric properties and validity (e.g., Mond, Hay, Rodgers,
ways (Waller, Stringer, & Meyer, 2012), expressing concerns about Owen, & Beumont, 2004).
the use of core CBT techniques (e.g., Turner, Tatham, Lant, Clinical Impairment Assessment Questionnaire (CIA; Bohn &
Mountford, & Waller, 2014) and discounting the use of evidence- Fairburn, 2008). The CIA is a 16-item self-report questionnaire,
based manuals to support their work (e.g., Waller et al., 2013). assessing severity of psychosocial impairment due to eating dis-
Therefore, there is a need for further evidence from other routine order features. Respondents rate the impact that exercise, eating
clinical settings to demonstrate that CBT for the eating disorders is habits and feelings towards eating, shape and weight have on their
an effective treatment, which others can use in their own clinics. ability to function in the world. A higher total score indicates a
This study aims to build on previous work by testing the effec- greater level of clinical impairment. The CIA has good reliability and
tiveness of CBT in a further routine clinical setting. It reports clinical validity (Bohn et al., 2008).
outcomes for a large group of transdiagnostic patients who were Hospital Anxiety and Depression Scale (HADS; Zigmond &
offered CBT in a community eating disorders service in the UK. Snaith, 1983). The HADS has two seven-item subscales measuring
Unlike previous studies, there were very few exclusion criteria and anxiety and depression. Respondents rate their experiences over
no BMI cut-off. In this case, the variant of CBT used was based on a the past week. The following categories are used: 0e7 ¼ normal;
combination of elements from the relatively similar approaches of 8e10 ¼ mild; 11e15 ¼ moderate; and 16e21 ¼ severe. The HADS
Fairburn (2008) and Waller et al. (2007), as used by Byrne et al. has been shown to be suitable for use with eating disorder pop-
(2011) and Waller, Gray, et al. (2014) respectively. ulations (e.g., Padierna, Quintana, Arostegui, Gonzalez, & Horcajo,
2000; Seed et al., 2004).
1. Method Clinical Outcomes in Routine Evaluation-Outcome Measure
(CORE-OM; Barkham et al., 2001). The CORE-OM is a self-report
1.1. Participants questionnaire measuring general psychological problems
(including an assessment of risk) in those presenting for psycho-
The sample consisted of 203 patients (190 women and 13 men) logical therapy. It can be used as a measure of individual change
who had been referred to a specialist National Health Service eating over time, and hence clinical effectiveness. The CORE-OM has good
disorder service in the UK. Other referrals were not included psychometric properties (e.g., Barkham, Gilbert, Connell, Marshall,
because they did not meet criteria for an eating disorder. All of the & Twigg, 2005; Evans et al., 2002) and is suitable for use with
203 patients were offered a course of outpatient CBT between 2010 people with eating disorders (Jenkins & Turner, 2014).
and 2013. Each was assessed using the Eating Disorders Examina-
tion, version 16 (Fairburn, Cooper, & O'Connor, 2008) and was 1.3. Procedure
diagnosed using DSM-IV criteria (American Psychiatric Association,
1994). Of the 203 patients, 56 (28%) had a diagnosis of anorexia Participants completed the following measures at the start and
nervosa, 58 (29%) bulimia nervosa, and 89 (43%) eating disorder not end of therapy (EDE-Q, CIA, HADS & CORE-OM). They also
otherwise specified. The mean age of the sample was 27.6 years completed the EDE-Q after the sixth treatment session. These
(SD ¼ 9.2, range ¼ 17e59 years) and their mean BMI at the start of measures are administered as part of routine clinical practice, and
treatment was 21.0 (SD ¼ 6.8, range ¼ 12.6e59.4). aim to monitor early clinical change, as well as the overall effec-
tiveness of treatment. All patients gave consent for data collected as
1.2. Measures part of routine service evaluation to be used to monitor the prog-
ress and effectiveness of therapy.
Patients completed the Eating Disorders Examination (EDE,
Fairburn et al., 2008) at initial assessment, and measures of eating 1.3.1. Treatment
disorder pathology at the start of treatment, following the sixth The CBT delivered within this clinic followed that described in
session, and at the end of treatment. Anxiety, depression and published evidence-based manuals (Fairburn, 2008; Waller et al.,
psychosocial functioning were measured at the start and on 2007). It included key elements of evidence-based practice such
completion of CBT. These measures are administered routinely at as: engagement; psychoeducation; developing a formulation;
the clinic for all patients receiving outpatient psychological ther- keeping a food diary; weekly weighing; dietary change; exposure;
apy. As is common in routine settings, a small proportion of the surveys; and cognitive restructuring. The treatment aimed to
data were not collected, and therefore the numbers vary across normalise eating, and to reduce weight controlling behaviours,
some analyses (see Tables). abnormal eating attitudes, and body image concerns. Where
The Eating Disorder Examination (EDE, version 16, Fairburn necessary it also aimed to address broader psycho-emotional-social
et al., 2008). The EDE generates the following four subscales: di- functioning, including identifying and managing emotions,
etary restraint, weight concern, shape concern and eating concern, replacing the functions of illness with more adaptive means,
as well as frequency ratings for key eating disorder behaviours, improving self-esteem, reducing pathological perfectionism, and
including objective bulimic episodes, self-induced vomiting, laxa- reducing inter-personal difficulties. All clinicians had regular indi-
tive misuse and excessive exercise. It can be used to generate DSM- vidual supervision (frequency varied between weekly and monthly,
IV diagnoses and has good psychometric properties (e.g., Berg, and was determined by factors such as individual clinician need,
Peterson, Frazier, & Crow, 2012). level of experience and clinical outcomes). Trainees and newly
72 H. Turner et al. / Behaviour Research and Therapy 68 (2015) 70e75

qualified staff received weekly supervision, whilst more experi- completed all measures at all time points. No data were substituted,
enced clinicians received fortnightly or monthly supervision (in so the N varies across measures (as shown in the tables, below). The
line with accreditation guidelines). The frequency of supervision data analyses in the tables are based on those numbers where there
was also reviewed in line with clinical outcomes, with additional is a complete data set for the relevant time points.
supervision being offered if necessary. A binary logistic regression was used to determine whether any
Supervision was provided by an external supervisor (GW), the pre-treatment variables predicted the patient terminating treat-
service clinical lead (HT), an accredited CBT supervisor, and another ment early (discounting those who left for reasons other than drop-
senior clinician in the team who has 10 years' experience of out). The variables used were the patients' ages and BMIs, their four
delivering and supervising CBT for eating disorders. Clinicians were individual EDE-Q attitudinal scales, the three EDE-Q behavioural
also encouraged to use evidence-based manuals (Fairburn, 2008; scales, HADS depression and anxiety scores, the CIA total score, and
Waller et al., 2007) to guide and inform the delivery of treatment. the CORE-OM total score. The overall model did not approach sig-
Manuals are readily available in the service and their use is sup- nificance (X2 ¼ 10.5; df ¼ 12; P ¼ .57), indicating that none of these
ported via supervision, with supervisors requiring clinicians to pre-treatment indices predicted loss to treatment. Furthermore,
demonstrate how their actions in therapy were related to the none of the individual variables approached significance (P > .10 in
relevant protocol and requiring them to read those manuals in all cases).
order to facilitate therapy. Treatment was delivered by 11 therapists
(three eating disorder therapists, four clinical psychologists, and 2.2. Remission rates
four trainee clinical psychologists). Two of the eating disorder
therapists were qualified mental health nurses and one was an A relatively strict definition of end of treatment remission was
accredited counsellor. All had attended training courses on deliv- used for all patients (BMI > 18.5; no reported objective binges,
ering CBT for eating disorders. vomiting or laxative use in the past 28 days; and EDE-Q total
Treatment length was typically 20 sessions, but that was score < 2.46 [under one SD above the community mean e Mond
shortened in the event of rapid change (to a minimum of 10 ses- et al., 2004). As a conservative strategy, all of the 179 patients were
sions) and extended for those patients with more significant co- included, with last available scores used to determine their outcome.
morbidity or a restrictive presentation (up to 40 sessions). Such Of the 179 patients, 34 (19%) achieved complete remission. However,
extension was on condition that the patient was actively engaged in that proportion differed across those who did and did not complete
therapy. Whilst a small number of patients had occasional dietetic treatment. Of those 100 who did complete treatment, 31 (31%)
reviews during the course of treatment, none had any other psy- achieved full remission. Of the 79 who did not complete treatment,
chological therapy or more intensive treatment (e.g., day care) only three (3.8%) achieved remission. This difference was significant
during the time they received CBT. (X2 ¼ 19.5; df ¼ 1; P < .001). There was no significant difference in
remission rates for anorexia nervosa (9/52 cases), bulimia nervosa (8/
1.4. Data analysis 51) or atypical cases (17/76) (X2 ¼ 1.02; df ¼ 2; P < .60).

The impact of start of treatment features on attrition was tested 2.3. Treatment outcomes for eating pathology
using binary logistic regression. As not all data were normally
distributed, change in symptoms were tested using non-parametric Tables 1 and 2 shows eating disorder attitudes (EDE-Q scores)
tests. Completer and intention to treat analyses are presented, with and behaviours (objective bulimic episodes, self-induced vomiting
the latter involving the carrying forward of the last available data and laxative misuse) at the three time points, for the completer and
point. In 11 cases, the participant completed treatment, but the end the intention to treat analyses.
of treatment measures were not completed. These individuals were
included in the ‘Completer’ group, carrying forward their last 2.3.1. Completer analysis
available data point. This was adopted as a relatively conservative Considering only those who completed therapy (Table 1), there
approach, as it reduces the chances of finding an effect of therapy, were significant overall reductions in eating disorder attitudes and
while reflecting the true therapy retention rate. Eating disorder behaviours. The EDE-Q attitudinal scales showed early improve-
symptom change was measured comparing scores across baseline, ments, as well as longer-term improvement. In contrast, the be-
session six, and end of treatment, using Friedman tests and post- haviours changed more slowly, with no significant change by
hoc multiple comparisons (Wilcoxon tests). Given evidence of the session six. The effect sizes (tau) for these changes between session
importance of early symptom change (e.g., Agras et al., 2000; 1 and the end of therapy were large for all variables with the
Wilson et al., 1999), the inclusion of session six scores allows for exception of vomiting and laxative misuse, which showed medium
identification of early change. Change in other features (HADS, CIA effects. For those under a BMI of 17.5 at the start of therapy, the
and CORE-OM scores) from baseline to end of treatment was tested Table shows that their BMI rose significantly across the three time
using Wilcoxon tests. Finally, for underweight patients (BMI < 17.5) points, with strong effect sizes for each pairwise comparison.
only, BMI levels were compared between the start of therapy,
session 6 and the end of therapy, using Friedman and post-hoc 2.3.2. Intention to treat analysis
Wilcoxon tests. Effect sizes (tau) were calculated for each signifi- There was less change across therapy in this analysis (Table 2),
cant pairwise difference. as would be expected. However, the pattern of significant change
was almost identical to that in the completer analysis, as was the
2. Results pattern and general strength of effect sizes.

2.1. Attrition from treatment 2.4. Treatment outcomes for non-eating measures

Of the 203 patients who started CBT for their eating disorder, 24 Table 3 shows the start and end of treatment scores for anxiety,
left therapy for system-level reasons (transfer to another area, etc.). depression, functional impairment and general pathology, showing
Of the remaining 179, 100 completed treatment and 79 dropped both the completer and intention to treat analyses. Both analyses
out. This yields an attrition rate of 44.1%. However, not all patients show a significant improvements in all those domains, with strong
 H. Turner et al. / Behaviour Research and Therapy 68 (2015) 70e75 73

Table 1
Change in eating disorder pathology (EDE-Q scores) during CBT for eating disorders in a routine clinical setting (Completer analysis).

EDE-Q measure N Measurement point Friedman's test (df ¼ 2) Effect size (Tau)

Session 1 (S1) Session 6 (S6) End of CBT (ET) X2 P Wilcoxon MC S1eS6 S6-ET S1-ET
tests (P < .05)
M (SD) M (SD) M (SD)

Attitudes
Total 80 4.20 (1.21) 3.17 (1.38) 2.07 (1.42) 117.0 0.001 S1 > S6 > ET 0.56 0.64 1.19
Restraint 80 3.87 (1.51) 2.33 (1.48) 1.32 (1.43) 137.8 0.001 S1 > S6 > ET 0.54 0.75 1.29
Eating 76 3.69 (1.30) 2.82 (1.46) 1.66 (1.41) 83.2 0.001 S1 > S6 > ET 0.48 0.52 1.01
Weight 80 4.39 (1.40) 3.46 (1.71) 2.27 (1.71) 83.1 0.001 S1 > S6 > ET 0.36 0.62 0.98
Shape 80 4.82 (1.28) 4.12 (1.58) 3.04 (1.79) 66.1 0.001 S1 > S6 > ET 0.31 0.56 0.87
Behaviours over 28 days
Objective binges 80 8.28 (12.1) 6.40 (11.2) 3.30 (10.0) 22.7 0.001 S1 ¼ S6 > ET e 0.34 0.53
Vomiting 80 7.32 (14.3) 5.33 (13.0) 1.81 (9.53) 35.6 0.001 S1 ¼ S6 > ET e 0.23 0.40
Laxatives 78 4.41 (9.37) 4.58 (23.2) 0.49 (2.51) 19.5 0.001 S1 > ET e e 0.24
BMI (<¼17.5) 18 15.78 (1.08) 16.36 (1.43) 18.43 (2.24) 15.2 0.001 S1 < S6 < ET 0.59 0.80 0.81

Table 2
Change in eating disorder pathology (EDE-Q scores) during CBT for eating disorders in a routine clinical setting (ITT analysis).

EDE-Q measure N Measurement point Friedman's test (df ¼ 2) Effect size (Tau)
2
Session 1 (S1) Session 6 (S6) End of CBT (ET) X P Wilcoxon MC S1eS6 S6-ET S1-ET
tests (P < .01)
M (SD) M (SD) M (SD)

Attitudes
Total 120 4.17 (1.29) 3.40 (1.39) 2.92 (1.68) 117.8 0.001 S1 > S6 > ET 0.51 0.43 0.95
Restraint 120 3.87 (1.65) 2.70 (1.54) 2.28 (1.82) 136.2 0.001 S1 > S6 > ET 0.63 0.35 0.98
Eating 117 3.71 (1.39) 3.05 (1.50) 2.57 (1.72) 74.7 0.001 S1 > S6 > ET 0.41 0.36 0.74
Weight 120 4.32 (1.49) 3.63 (1.67) 3.09 (1.90) 63.9 0.001 S1 > S6 > ET 0.39 0.40 0.79
Shape 119 4.75 (1.37) 4.22 (1.55) 3.75 (1.85) 63.9 0.001 S1 > S6 > ET 0.30 0.49 0.69
Behaviours over 28 days
Objective binges 119 8.62 (12.1) 7.37 (13.1) 5.41 (11.6) 20.1 0.001 S1 ¼ S6 > ET e 0.26 0.30
Vomiting 119 9.59 (18.0) 7.26 (15.5) 5.42 (13.2) 23.3 0.001 S1 > ET e e 0.29
Laxatives 116 4.79 (14.4) 4.03 (19.0) 2.12 (7.19) 29.4 0.001 S1 > ET e e 0.23
BMI <¼ 17.5 31 15.68 (1.25) 16.30 (1.71) 17.67 (2.27) 16.6 0.001 S1 < S6 < ET 0.49 0.68 0.79

effect sizes. their level of experience, all received regular clinical supervision
and all were actively encouraged to use treatment manuals to guide
3. Discussion the delivery of treatment. This study differs from those of Byrne
et al. (2011) and Waller, Gray, et al. (2014) in two important ways.
This study reports the effectiveness of CBT for eating disorders - First, it included patients with lower BMIs than either of those
outcomes when delivered in a routine clinical setting, where there studies. Second, the therapy used was an amalgam of the two
are few exclusion criteria and where adherence to evidence-based versions of CBT that those two studies employed (Fairburn, 2008;
protocols is less intensively monitored. Whilst therapists varied in Waller et al., 2007).
The findings were broadly comparable to previously published
effectiveness studies. First, the attrition rate was similar to that
Table 3
Change in mood, anxiety and functional impairment from the beginning to the end
reported in other studies (e.g., Byrne et al., 2011; Campbell, 2009).
of CBT for eating disorders in a routine clinical setting (Completer and ITT analyses). Second, there was a substantial reduction in eating attitudes and
behaviours, with the change in global EDE-Q mirroring that re-
Measure N Measurement point Wilcoxon test ES
ported by Byrne et al. (2011). Showing comparability with existing
Session 1 End of Z P Tau effectiveness and efficacy studies (Byrne et al., 2011; Fairburn et al.,
treatment
2009), the element of eating attitudes that showed the least
Completer analysis improvement by the end of treatment was the EDE-Q shape con-
HADS depression 93 M 8.92 4.79 7.17 0.001 0.74
cerns scale. There was also a significant reduction in anxiety and
(SD) (4.23) (3.83)
HADS anxiety 93 M 12.7 9.66 6.10 0.001 0.63 depression, as well as an overall improvement in general psycho-
(SD) (4.03) (3.99) logical functioning. These broader improvements in general mental
CIA total 94 M 31.4 15.5 7.66 0.001 0.79 health and quality of life reflect those found in previous studies
(SD) (10.2) (11.9) (Byrne et al., 2011; Waller, Gray, et al., 2014). Furthermore, those
CORE-OM 88 M 3.49 1.67 7.19 0.001 0.77
patients who had an AN-spectrum presentation also benefitted.
(SD) (1.40) (1.53)
Intention to treat Their mean change in BMI (2.6, SD ¼ 1.8) compares favourably with
HADS depression 116 M 9.19 6.97 6.87 0.001 0.64 that reported in previous research trials (e.g., Dare, Eisler, Russell,
(SD) (4.32) (4.73) Treasure, & Dodge, 2001; McIntosh et al., 2005), and at a similar
HADS anxiety 116 M 12.9 11.3 5.75 0.001 0.53
level to that found by Fairburn et al. (2013).
(SD) (4.18) (4.51)
CIA total 117 M 32.4 23.3 7.86 0.001 0.73 The temporal pattern of change is also relevant, given the
(SD) (10.6) (14.8) importance of early change in predicting treatment outcome in
CORE-OM 110 M 3.22 2.39 6.65 0.001 0.63 eating disorders (e.g., Agras et al., 2000; Raykos, Watson, Fursland,
(SD) (1.36) (1.68) Byrne, & Nathan, 2013; Wilson et al., 1999). The present findings
74 H. Turner et al. / Behaviour Research and Therapy 68 (2015) 70e75

indicated a significant early reduction in EDE-Q attitudes, reflecting the outcomes of treatment delivered in rigorously controlled
the pattern of change shown by Raykos et al. (2013). Raykos et al. treatment trials and those delivered in routine clinical settings.
have defined early rapid response as a change of 1.52 or greater on Comparing the current study and other effectiveness studies (e.g.,
the global EDE-Q over the first 3e6 sessions. This level of change Byrne et al., 2011) with efficacy studies (e.g., Fairburn et al., 1995;
was achieved by 30% of the sample in the present study, which is 2009), the most important focus is no longer whether the treat-
similar to the 34% reported by Raykos et al. (2013). However, the ment works in clinical settings, but the need to enhance retention
transdiagnostic nature of the sample means that early changes in in routine clinical settings, to ensure that more patients can benefit
behaviours cannot be meaningfully compared to the findings of from proven treatments.
other papers (e.g., Agras et al., 2000; Wilson et al., 1999; Waller,
Gray, et al., 2014), as a proportion of patients in this study were Conflict of interest
not engaging in binge eating or self-induced vomiting at baseline.
This study has confirmed the previous contention that CBT for The authors have no conflict of interest to declare.
eating disorders can be effective in routine clinical practice, based
on a relatively large sample size and the use of well-validated
References
measures. However, the study also has a number of the limita-
tions that are inherent in such clinic-based work. In particular, Agras, W. S., Crow, S. J., Halmi, K. A., Mitchell, J. E., Wilson, G. T., & Kraemer, H. C.
although therapists were encouraged to follow evidence-based (2000). Outcome predictors for the cognitive behavior treatment of bulimia
manuals and were guided in the delivery of CBT via supervision, nervosa: data from a multisite study. American Journal of Psychiatry, 157,
1302e1308.
there were no external check that the treatment delivered was American Psychiatric Association. (1994). Diagnostic and statistical manual of mental
actually the intended amalgam of evidence-based forms of CBT. It is disorders: DSM- IV (4th ed.). Washington, DC: American Psychiatric Association.
therefore possible that the interventions were not as theoretically Barkham, M., Gilbert, N., Connell, J., Marshall, C., & Twigg, E. (2005). Suitability and
utility of the CORE-OM and CORE-A for assessing severity of presenting prob-
or therapeutically ‘pure’ as those in efficacy studies. The different lems in psychological therapy services based in primary and secondary care
levels of training and experience of the therapists might have settings. British Journal of Psychiatry, 186, 239e246.
impacted on outcomes, and this would merit further study. Simi- Barkham, M., Margison, F., Leach, C., Lucock, M., Mellor-Clark, J., Evans, C., et al.
(2001). Service profiling and outcomes benchmarking using the CORE-OM:
larly, the impact of therapists' attitudes to and use of manuals to toward practice-based evidence in the psychological therapies. Journal of
support treatment for eating disorders (e.g., Wallace & von Ranson, Consulting and Clinical Psychology, 69, 184e196.
2011; Waller et al., 2013) requires further research, to clarify Berg, K. C., Peterson, C. B., Frazier, P., & Crow, S. J. (2012). Psychometric evaluation of
the eating disorder examination and eating disorder examination-
whether such manuals facilitate the dissemination of evidence-
questionnaire: a systematic review of the literature. International Journal of
based interventions. Furthermore, the lack of follow-up data Eating Disorders, 45, 428e438.
means it is not known whether the changes made during treatment Black, C., & Wilson, G. T. (1996). Assessment of eating disorders: Interview versus
were sustained long-term. Another methodological issue is that the questionnaire. International Journal of Eating Disorders, 20, 43e50.
Bohn, K., Doll, H. A., Cooper, Z., O'Connor, M. E., Palmer, R. L., & Fairburn, C. G. (2008).
recording of eating disorder behaviours and weight was taken from The measurement of impairment due to eating disorder psychopathology.
the EDE-Q, rather than from weekly symptom monitoring records. Behaviour Research and Therapy, 46, 1105e1110.
This use of EDE-Q behavioural measures might mean that there was Bohn, K., & Fairburn, C. G. (2008). Clinical Impairment Assessment questionnaire
(CIA 3.0). In C. G. Fairburn (Ed.), Cognitive behavior therapy and eating disorders
over-reporting of binge-eating in this study (e.g., Black & Wilson, (pp. 315e317). New York, NY: Guilford.
1996). It will also be important for future research to consider Bulik, C. M., Sullivan, P. F., Carter, F. A., McIntosh, V. V., & Joyce, P. R. (1999). Pre-
symptom change on a session-by-session basis (e.g., Waller, dictors of rapid and sustained response to cognitive-behavioral therapy for
bulimia nervosa. International Journal of Eating Disorders, 26, 137e144.
Taltham, Turner, Mountford & Tritt, 2014), in order to understand Byrne, S. M., Fursland, A., Allen, K. L., & Watson, H. (2011). The effectiveness of
the importance of early and sudden change on clinical outcomes in enhanced cognitive behavioural therapy for eating disorders: an open trial.
more detail. Behaviour Research and Therapy, 49, 219e226.
Campbell, M. (2009). Drop-out from treatment for the eating disorders: a problem
These findings have clear clinical implications. First, CBT's evi-
for clinicians and researchers. European Eating Disorders Review, 17, 239e242.
dence base for use in routine clinical settings is enhanced. Clini- Dare, C., Eisler, I., Russell, G., Treasure, J., & Dodge, L. (2001). Psychological therapies
cians' own concerns about the delivery of CBT (Turner et al., 2014) for adults with anorexia nervosa: randomised controlled trial of out-patient
treatments. British Journal of Psychiatry, 178, 216e221.
should not be used to justify avoiding its use. Nor can clinicians
Evans, C., Connell, J., Barkham, M., Margison, F., McGrath, G., Mellor-Clark, J., et al.
simply argue for adopting a more eclectic approach (Tobin et al., (2002). Towards a standardised brief outcome measure: psychometric proper-
2007) or omitting key elements of CBT (Waller et al., 2012) on the ties and utility of the CORE-OM. British Journal of Psychiatry, 180, 51e60.
grounds that CBT is ‘not suitable for our patients in real-life set- Fairburn, C. G. (2008). Cognitive behavior therapy and eating disorders. New York, NY:
Guilford.
tings’. Second, it will be important to ensure the delivery of Fairburn, C. G., & Beglin, S. J. (2008). Eating disorder examination questionnaire
evidence-based training, supervision and outcome monitoring, in (6.0). In C. G. Fairburn (Ed.), Cognitive behavior therapy and eating disorders (pp.
order to ensure the development of clinical skills and competence 309e313). New York, NY: Guilford.
Fairburn, C. G., & Cooper, Z. (2011). Therapist competence, therapy quality, and
(Fairburn & Cooper, 2011) and to promote protocol adherence (e.g., therapist training. Behaviour Research and Therapy, 49, 373e378.
Hogue et al., 2008). Such training should stress the importance of Fairburn, C. G., Cooper, Z., Doll, H. A., O'Connor, M. E., Bohn, K., Hawker, D. M., et al.
tracking early intervention response to inform clinical decision- (2009). Transdiagnostic cognitive-behavioral therapy for patients with eating
disorders: a two-site trial with 60-week follow-up. American Journal of Psy-
making (i.e., to continue, augment, or change strategies), given chiatry, 166, 311e319.
the documented prognostic value of early change in eating disorder Fairburn, C. G., Cooper, Z., Doll, H. A., O'Connor, M. E., Palmer, R. L., & Dalle Grave, R.
treatment. That focus is particularly important given the strong (2013). Enhanced cognitive behaviour therapy for adults with anorexia nervosa:
a UK-Italy study. Behaviour Research and Therapy, 51, R2eR8.
links between early change and later treatment outcome, over and Fairburn, C. G., Cooper, Z., & O'Connor, M. (2008). Eating disorder examination
above the impact of other therapeutic elements such as the work- (16.0D). In C. G. Fairburn (Ed.), Cognitive behavior therapy and eating disorders
ing alliance (Raykos et al., 2014). (pp. 270e306). New York, NY: Guilford.
Fairburn, C., Norman, P., Welch, S., O'Connor, M., Doll, H., & Peveler, R. (1995).
Overall the findings from the present study demonstrate that
A prospective study of outcome in bulimia nervosa and the long-term effects of
evidence-based forms of CBT can be delivered with strong out- three psychological treatments. Archives of General Psychiatry, 52, 304e311.
comes in routine clinical settings. It is important that clinicians Hogue, A., Henderson, C., Dauber, S., Barajas, P., Fried, A., & Liddle, H. (2008).
should deliver those evidence-based treatments and monitor out- Treatment adherence, competence, and outcome in individual and family
therapy for adolescent behavior problems. Journal of Consulting and Clinical
comes as a matter of course, to ensure maximum available benefit Psychology, 76, 544e555.
to patients. Such practice should allow us to close any gaps between Jenkins, P., & Turner, H. M. (2014). An investigation into the psychometric properties
 H. Turner et al. / Behaviour Research and Therapy 68 (2015) 70e75 75

of the CORE-OM in patients with eating disorders. Counselling and Psychother- Wallace, L. M., & von Ranson, K. M. (2011). Treatment manuals: use in the treatment
apy Research, 14, 102e110. of bulimia nervosa. Behaviour Research and Therapy, 49, 815e820.
McIntosh, V. V. W., Jordan, J., Carter, F. A., Luty, S. E., McKenzie, J. M., Bulik, C. M., Waller, G., Corstorphine, E., Cordery, H., Hinrichsen, H., Lawson, R., Mountford, V.,
et al. (2005). Three psychotherapies for anorexia nervosa: a randomized, et al. (2007). Cognitive-behavioral therapy for the eating disorders: A compre-
controlled trial. American Journal of Psychiatry, 162, 741e747. hensive treatment guide. Cambridge, UK: Cambridge University Press.
Mond, J. M., Hay, P. J., Rodgers, B., Owen, C., & Beumont, P. J. V. (2004). Validity of the Waller, G., Gray, E., Hinrichsen, H., Mountford, V., Lawson, R., & Patient, E. (2014).
Eating Disorder Examination Questionnaire (EDE-Q) in screening for eating Cognitive-behavioral therapy for bulimia nervosa and atypical bulimic nervosa:
disorders in community samples. Behaviour Research and Therapy, 42, 551e567. effectiveness in clinical settings. International Journal of Eating Disorders, 47,
Padierna, A., Quintana, J. M., Arostegui, I., Gonzalez, N., & Horcajo, M. J. (2000). The 13e17.
health-related quality of life in eating disorders. Quality of Life Research, 9, Waller, G., Mountford, V. A., Tatham, M., Turner, H., Gabriel, C., & Webber, R. (2013).
667e674. Attitudes towards psychotherapy manuals among clinicians treating eating
Poulsen, S., Lunn, S., Daniel, S. I. F., Folke, S., Mathiesen, B. B., Katznelson, H., et al. disorders. Behaviour Research and Therapy, 51, 840e844.
(2014). A randomized controlled trial of psychoanalytic psychotherapy versus Waller, G., Stringer, H., & Meyer, C. (2012). What cognitive-behavioral techniques do
cognitive behavior therapy for bulimia nervosa. American Journal of Psychiatry, therapists report using when delivering cognitive-behavioral therapy for the
171, 109e116. eating disorders? Journal of Consulting and Clinical Psychology, 80, 171e175.
Raykos, B. C., McEvoy, P. M., Erceg-Hurn, D., Byrne, S. M., Fursland, A., & Nathan, P. Waller, G., Taltham, M., Turner, H., Mountford, V., & Tritt, A. (2014 October). The ED-
(2014). Therapeutic alliance in enhanced cognitive behavioural therapy for 10 questionnaire: Development, psychometric properties and preliminary clinical
bulimia nervosa: probably necessary but definitely insufficient. Behaviour validation of a brief session-by-session measure of eating cognitions and behav-
Research and Therapy, 57, 65e71. iours. San Diego, CA: Poster presented at the Eating Disorders Research Society.
Raykos, B. C., Watson, H. J., Fursland, A., Byrne, S. M., & Nathan, P. (2013). Prognostic Watson, H. J., & Bulik, C. M. (2013). Update on the treatment of anorexia nervosa:
value of rapid response to enhanced cognitive behavioral therapy in a routine review of clinical trials, practice guidelines and emerging interventions. Psy-
clinic sample of eating disorder outpatients. International Journal of Eating chological Medicine, 43, 2477e2500.
Disorders, 46, 764e770. Wilson, G. T., Loeb, K. L., Walsh, B. T., Labouvie, E., Petkova, E., Lui, X., et al. (1999).
Seed, J. A., Dahabra, S., Heffernan, T., Robertson, B., Foster, K., Venn, H., et al. (2004). Psychological versus pharmacological treatments of bulimia nervosa: pre-
Everyday memory and related processes in patients with eating disorders. dictors and processes of change. Journal of Consulting and Clinical Psychology, 67,
Clinical Effectiveness in Nursing, 8, 176e188. 451e459.
Tobin, D. L., Banker, J. D., Weisberg, L., & Bowers, W. (2007). I know what you did last Zigmond, A. S., & Snaith, R. P. (1983). The hospital anxiety and depression scale. Acta
summer (and it was not CBT): a factor analytic model of international psy- Psychiatrica Scandinavica, 67, 361e370.
chotherapeutic practice in the eating disorders. International Journal of Eating Zipfel, S., Wild, B., Groß, G., Friederich, H. C., Teufel, M., Schellberg, D., et al. (2014).
Disorders, 40, 754e757. Focal psychodynamic therapy, cognitive behaviour therapy, and optimised
Turner, H., Tatham, M., Lant, M., Mountford, V. A., & Waller, G. (2014). Clinicians' treatment as usual in outpatients with anorexia nervosa (ANTOP study):
concerns about delivering cognitive-behavioural therapy for eating disorders. randomised controlled trial. The Lancet, 383, 127e137.
Behaviour Research and Therapy, 57, 38e42.