Unwanted Events and Side Effects in Cognitive Behavior Therapy

Abstract

Side effects (SEs) are negative reactions to an appropriately delivered treatment, which must be discriminated from unwanted events (UEs) or consequences of inadequate treatment. One hundred CBT therapists were interviewed for UEs and SEs in one of their current outpatients. Therapists reported 372 UEs in 98 patients and SEs in 43 patients. Most frequent were "negative wellbeing/distress" (27% of patients), "worsening of symptoms" (9%), "strains in family relations" (6%); 21% of patients suffered from severe or very severe and 5% from persistent SEs. SEs are unavoidable and frequent also in well-delivered CBT. They include both symptoms and the impairment of social life. Knowledge about the side effect profile can improve early recognition of SEs, safeguard patients, and enhance therapy outcome.

Full text

Vol.:(0123456789)
1 3
Cognitive Therapy and Research (2018) 42:219–229
https://doi.org/10.1007/s10608-018-9904-y
ORIGINAL ARTICLE
Unwanted Events andSide Effects inCognitive Behavior Therapy
Marie‑LuiseSchermuly‑Haupt1· MichaelLinden1,2 · A.JohnRush3
Published online: 9 March 2018
© Springer Science+Business Media, LLC, part of Springer Nature 2018
Abstract
Side effects (SEs) are negative reactions to an appropriately delivered treatment, which must be discriminated from unwanted
events (UEs) or consequences of inadequate treatment. One hundred CBT therapists were interviewed for UEs and SEs in one
of their current outpatients. Therapists reported 372 UEs in 98 patients and SEs in 43 patients. Most frequent were "negative
wellbeing/distress" (27% of patients), "worsening of symptoms" (9%), "strains in family relations" (6%); 21% of patients
suffered from severe or very severe and 5% from persistent SEs. SEs are unavoidable and frequent also in well-delivered
CBT. They include both symptoms and the impairment of social life. Knowledge about the side effect profile can improve
early recognition of SEs, safeguard patients, and enhance therapy outcome.
Keywords Psychotherapy· Unwanted events· Side effects· Adverse treatment reactions· Quality assurance· Cognitive
behavior therapy· Deterioration
Introduction
Psychotherapy has both positive and negative effects (Bergin
1963). Global estimates for the rate of relevant side effects
(SEs) across different psychotherapies range from 5 to 20%
of patients (Barlow 2010; Berk and Parker 2009; Lilien-
feld 2007; Linden and Strauß 2013; Parker etal. 2013).
In an online survey, 13% of 1504 patients reported “bur-
dens caused by therapy” (Leitner etal. 2013). If psycho-
therapists are asked about their own treatments, 20–40%
remember harmful effects (Buckley etal. 1981; Macaskill
1992). The most frequent nominations were psychological
distress (29%) and marital or family stress (13%), although
some therapists thought this to be necessary for effective
treatment. Other examples of burdens caused by psycho-
therapy were patients feeling overwhelmed or being afraid
of the therapist, undermining of self-efficacy, deterioration
of symptoms, emergence of new symptoms, suicidality,
treatment failure, occupational problems, stigmatization,
strains in personal relationships, or changes in the social
network (Hoffmann etal. 2008).
Over the years, therapists and researchers have repeat-
edly pointed to the importance of side effects in psycho-
therapy, though little empirical data are available (Jonsson
etal. 2014). So far, there is no generally accepted definition
or assessment methodology for psychotherapy SEs. There is
the problem of differentiation between SEs on one side and
consequences of inappropriate or unethical behavior by the
therapists on the other. Also, it must be assumed that dif-
ferent treatments, different types of patients, different types
of patient therapist dyads, and even therapist training can
have their own side effects, so that detailed data are needed
to guide clinical practice (Henry etal. 1993). Finally, the
recognition of SEs poses special problems as therapists can
have difficulties accepting and recognizing negative conse-
quences of their work. Often they do not expect and there-
fore not recognize them (Hannan etal. 2005; Hatfield etal.
2010; Lambert 2010).
The assessment of side effects, be it in psychotherapy or
pharmacotherapy, follows a stepwise process, as shown in
Fig.1 (Linden 2013; Linden and Schermuly-Haupt 2014).
It starts with systematically recording unwanted events
(UEs), which occur parallel to treatment and which may
be related to the ongoing therapy or not (e.g. suicide). In a
second step a professional judgment is needed to determine
* Michael Linden
michael.linden@charite.de
1 Research Group Psychosomatic Rehabilitation
attheCharité University Medicine Berlin, CBF Hs II, E01,
Hindenburgdamm 30, 12200Berlin, Germany
2 Institute forBehavior Therapy Berlin, Berlin, Germany
3 Duke-National University ofSingapore, Singapore,
Singapore

220 Cognitive Therapy and Research (2018) 42:219–229
1 3
whether these UEs are adverse treatment reactions (ATRs),
i.e. related to treatment (has the suicide been caused by treat-
ment procedures?). Thirdly, it must be determined whether
they are caused by treatment according to professional stand-
ards (were the treatment procedures which caused suicide
following professional standards?). There may be unwanted
events during the course of therapy that just happen by coin-
cidence and there may be some unwanted events that are the
result of malpractice or even therapeutic unethical behavior.
None of those can be considered as side effects. This dis-
crimination helps therapists to accept that their work can
have not only beneficial but also negative consequences. The
distinction between UEs and ATRs makes clear that SEs
are often an unavoidable part of well-delivered treatment.
By expecting SEs, we can plan for their management and
mitigation.
As the incidence, nature, severity and duration of SE dif-
fer across different interventions, conditions and settings,
specific data are needed (Lambert and Ogles 2004; Parker
etal. 2013; Tarrier etal. 1999). The present study focuses
on cognitive behavior therapy (CBT). For this treatment
approach it has been reported that some patients with post-
traumatic stress disorder (PTSD) or anxiety disorders can
exhibit symptom worsening after imaginary exposure (Foa
etal. 2002; Mayou etal. 2000; Scheeringa etal. 2011). In
PTSD patients with hyper-arousal at baseline, emotional
debriefing (meaning single sessions of psychological assis-
tance shortly after the traumatic event where participants are
encouraged to talk about detailed aspects of the events and
their emotional reactions and thoughts) has resulted in sig-
nificantly more PTSD symptoms than educational debriefing
(single sessions with the focus on information on stress and
symptoms) or no debriefing (Rose etal. 2002; Sijbrandij etal.
2006). Treatment of women with agoraphobia resulted in
psychological symptoms in their husbands (Hafner 1984;
Milton and Hafner 1979). Patients with generalized anxiety
disorder treated with behavior therapy that focused on life
problems instead of on worrying showed less remission than
patients who had attended unspecific group sessions (Fisher
and Durham 1990). SEs have also been reported with token
economy treatment for substance abuse (Rosen etal. 2010),
group psychotherapies (Roback 2000) and psychotherapies
for patients with a history of sexual abuse (Brainerd and
Reyna 2005; Lambert and Ogles 2004).
Given this background, the present study was designed
to assess the nature, frequency, severity and duration of
both UEs and SEs in outpatients who were receiving CBT.
Such knowledge about the routine SE profile of treatments
is needed to inform and safeguard patients. This paper uses
a conceptualization and in-depth assessment of SEs that rec-
ognizes the distinction between true side effects, malpractice,
clinical deterioration and other causes of negative outcomes
(Linden 2013). By including a large number of therapists and
outpatients, the SE profile of CBT under routine treatment
conditions is described (i.e., nature, frequency, severity and
duration of SEs). This is what therapists should know about
when informing their patients about the upcoming merits and
risks of treatment.
Fig. 1 Differentiation of
unwanted treatment events
UE
Unwanted
Event
ATR
AdverseTreatment
Reaction
Correct
Treatment
SE
Side effect
Incorrect
Treatment
Malpractice
Not Treatment
related
Natural
course
External
cause

221Cognitive Therapy and Research (2018) 42:219–229
1 3
Method
Setting andTherapists
A convenience sample of 100 psychotherapists from three
outpatient clinics for CBT were asked to participate in an
interview on UEs and to report about one current patient
who had received at least ten sessions of consecutive CBT,
as this duration is estimated to be long enough for therapeu-
tic effects and SEs to manifest (Anderson and Lambert 2001;
Harnett etal. 2010). The mean age of therapists was 32.2
(SD = 4.7) years, 78% were female, and 96% had a univer-
sity degree in psychology. Therapists had an average of 5.1
(SD = 3.4) years of professional experience.
All therapists were specifically trained in “cognitive
behavior therapy (CBT)”. According to German law and
health regulations this is specified by an official expert panel,
as basis for training and reimbursement. CBT is based on
behavioral analysis and refers to classical learning theory,
coping and social competency, dysfunctional cognitions, and
emotional regulation. All treatments in this study were reim-
bursed by the general health insurance. To ensure treatment
integrity, therapists were supervised after every fourth treat-
ment session using audio or video recordings that are rou-
tine in the clinics to ensure that treatments follow accepted
clinical standards. The recordings were not available for
research purposes. On average, therapists had undergone
7.1 (SD = 4.0) sessions of supervision in the course of the
treatment about which they reported.
As can be expected in a group of closely supervised thera-
pists, no indicators for gross violations of therapeutic rules
or inappropriate therapeutic behavior could be found. There-
fore, all ATRs in this report are subsequently called SEs.
The institutional boards of the clinics reviewed and
accepted the study protocol. After informing the therapists
about the study, written informed consent was obtained.
Patients were not personally involved, and data that could
identify individuals were not recorded.
Interview
A semi-standardized interview1 was conducted by the first
author, a state licensed clinical psychologist with compre-
hensive training as a cognitive behavior therapist. Thera-
pists were asked to give at least six examples of potential
side effects of psychotherapy that they had experienced or
could imagine. Thereafter, they were asked to report on
their last treatment case of that day or the last few days in
which the patient had attended at least 10 sessions. This
minimum number of sessions was required to make sure
that the treatment had a chance to show some effect. Socio-
demographic indicators and clinical and treatment data were
collected, including course of the illness, diagnoses, medica-
tion, number of sessions, etc.
Therapists were asked to rate the patient’s present state
of illness on the Fischer Symptom Checklist (FSCL), an
observer rating scale with 41 Items, covering common psy-
chiatric symptoms like mood impairment, anxiety, sleep,
thought process and content, or social activity. The Items
are scored on a 4 point scale from “not at all” to “severe”.
The FSCL has shown good reliability and validity and has
been used in many clinical trials (Fischer-Cornelssen and
Berchier 1982).
To systematically screen for SEs, the interview then fol-
lowed the Unwanted Events-Adverse Treatment Reactions
Checklist (UE-ATR Checklist; Linden 2013) that lists 17
domains of possible UEs (Fig.2) including, for example,
emergence of new symptoms or changes in family relations.
For each UE, it was determined whether the event was
an ATR by asking the therapists to describe the UE in detail
and explain what might have caused the event in his or her
view. The final rating on the “degree of relatedness to psy-
chotherapy” was rendered by the therapist using the UE-
ATR Checklist by taking into account the type of UE, the
relatedness to the psychotherapeutic processes, or its course
and timing. “Relatedness to therapy” was classified as (1)
“definitely unrelated”, (2) “rather unrelated”, (3) “rather
related”, (4) “most probably related”, or (5) “definitely
related”. Therapists were then asked for information on the
“duration” and “severity” of that specific UE. Categories of
“duration” were “hours”, “days”, “weeks”, “months”, and
“persistent”. “Severity” was rated on a 5-point Likert-scale
with respect to the impact on the patient from (1) “mild, no
negative consequences” to (5) “extremely severe”.
The UE-ATR Checklist is an instrument for event sam-
pling in combination with a coding scheme. This is so far
the only instrument which explicitly discriminates between
unwanted events (UE), adverse treatment reactions (ATR)
and side effects (SE), and which systematically guides the
screening process. Reliability has been tested by having 623
reports of therapists rated by two raters using the 17 domains
listed in the UE-ATR Checklist as categories. Cohens Kappa
was 0.75 (p = .000), which is considered to be good agree-
ment (Wirtz and Caspar 2002). Of the reports, 92% could
be assigned to the domains of the checklist (8% were rated
as “others”). Furthermore, we tested interrater agreement
for the two scales “severity” and “relation to therapy” by
comparing therapist ratings with the interviewer judge-
ments. Unadjusted intra-class correlations (Shrout and Fleiss
1979) were 0.68 (p < .001) for severity and 0.84 (p < .001)
for relation.
1 The interview is available from michael.linden@charite.de.

222 Cognitive Therapy and Research (2018) 42:219–229
1 3
Results
Treatment Cases
Each of the 100 therapists reported on one treatment case (a
total of 100 patients). Of the patients, 51% were female, the
average age was 38.0 (SD = 11.3) years (range 19–68years),
53% were married or living in a relationship, and 16% had
a university degree.
The mean FSCL total score was 25.1 (SD = 14.2), indi-
cating a degree of psychopathology in the mild to moderate
range. The most frequent clinical diagnoses based on ICD-
10 criteria (Dilling 2011) were major depressive episode/
recurrent depression (36%), anxiety disorders (16%), and
personality disorders (15%). Forty-three percent of patients
had been in an inpatient psychiatric hospital at least once,
33% had at least one prior period of psychotherapy, and 42%
were receiving some psychotropic medication while CBT
was being delivered.
At the time of the interview, patients had completed
10–100 sessions of CBT with a mean of 28.5 (SD = 16.8)
sessions over 10.6 (SD = 7.0) months. Prior to conducting
the UE-ATR interview, each therapist was asked to make a
global judgment on the course of treatment. Five percent
expected full remission and 59% expected a good response
at the end of treatment.
Unwanted Events (UEs)
When asked for a spontaneous report, 74% of therapists
were not aware of any UE or SE in the treatment of their
patients. While answering the structured interview, which
systematically evokes domains of UEs, therapists reported a
total of 372 UEs (average of 3.7 (SD = 2.0) UEs per patient).
0102030405060
Abuse of therapy
Stigmatization
any change in the life circumstances of the patient
problems in the extended social net
sick leave of the patient
changes in work situation
strains in work relations
changes in family relations
strains in family relations
very good patient therapist relationship, dependency
strains in the patient therapist relationship
negative well being/ distress
deterioration of existing symptoms
emergence of new symptoms
non compliance of the patient
prolongation of treatment
lack of improvement or deterioration of illness
Percent of patients
definitely relatedmost probably related rather related
rather unrelated definitely unrelated
Fig. 2 Percentage of patients with different unwanted events and their relation to therapy

223Cognitive Therapy and Research (2018) 42:219–229
1 3
Only for two of the 100 patients no UEs were reported.
Three patients had a maximum of 9 UEs. Figure2 shows
the frequencies of the UEs for each of the 17 domains of
the UE-ATR Checklist. Most frequent UEs were: “negative
well-being/distress” (53% of patients) and “deterioration
of existing symptoms” (46% of patients). Least frequent
were “change in life circumstances” (6% of patients) and
“emergence of new symptoms” (10% of patients). The
mean “severity” was 2.13 (SD = 0.89) (i.e., 65.6% of UEs
were mild or moderate). Regarding the “duration” of the
UEs, 15.9% lasted only for hours, 14.0% for days, 26.1% for
weeks, 22.0% for months, and 16.7% were expected to be
persistent (Table1).
Side Effects (SEs)
Of the 372 UEs 35.8% were, according to professional
judgement of the interviewer, rated as “definitely unrelated”
to treatment, 19.9% as “rather unrelated”, 16.4% as “rather
related”, 9.4% as “most probably related” and 15.3% as “def-
initely related” to treatment. All events that were rated as
“definitely related” to treatment were considered to be SEs.
Of the 100 patients, 43% suffered from at least one SE
(average 0.57 (SD = 0.81) SEs per patient). Thirty-one
patients had only one SE reported, while one patient had the
maximum of 4 SEs reported. Figure2 shows the frequencies
of SEs for each of the 17 domains of the UE-ATR Checklist
and Table1 shows their “severity” and “duration”.
The most frequent SEs were “negative wellbeing/distress”
(7.2% of all UEs, 27% of patients), “deterioration of symp-
toms” (2.4% of UEs, 9% of patients) and “strains in family
relations” (1.6% of UEs, 6% of patients). Mean severity was
2.32 (SD = 0.78). Fourteen percent of the SEs were rated
as mild (in 9% of patients), 45.6% as moderate (in 25% of
patients), 35.1% as severe (in 18% of patients) and 5.3%
as very severe (in 3% of patients). No SEs were rated as
extremely severe. Regarding duration, 31.6% of SEs lasted
only for hours, 31.6% for days, 21.1% for weeks, 5.3% for
months and 8.8% were expected to be persistent (Table1).
To provide a better clinical understanding of the nature of
UEs and SEs, Table2 gives some qualitative reports of what
is meant by the different dimensions of the UE-ATR Check-
list. Examples describe severe and persisting side effects
such as suicidality, breakups, negative feedback from family
members, withdrawal from relatives, feelings of shame and
guilt, or intensive crying and emotional disturbance during
sessions.
The longer the treatment lasted, the more UEs, and to a
lesser extent SEs, were reported. The correlation between
the number of CBT sessions and the number of UEs was
r = .55 (p < .001), and the number of SEs r = .30 (p < .01).
There was no significant correlation between the FSCL
score and the number of UEs (r = .14, p > .05), but there
was a modest negative correlation between the FSCL score
and the number of SEs (r = − .22, p < .05). In this respect,
there was a medium-sized significant difference in regard to
symptom severity as measured with the FSCL [t(98) = 3.03,
p < .01, d = .65], with patients with SEs showing less symp-
tom severity than patients without SEs.
To explore whether one could predict which patients
would develop SEs, a binary logistic regression analysis
was calculated with patient and therapist characteristics as
predictors and group (patients with or without SEs) as a
dichotomous independent variable. The set of potential pre-
dictor variables focused on variables found to impact out-
come in other studies such as severity of disease, experience
Table 1 Frequencies of “severity” and “duration” for unwanted events and side effects
a 5.4% missing values in duration of unwanted events; 2 1.8% missing values in duration of side effects
Severity Mild Moderate Severe Very severe Extremely
severe
Unwanted events (n)102 142 107 20 1
Percent (n = 372) (%) 27.4 38.2 28.8 5.4 0.3
% of patients 69 94 70 18 1
Side effects (n) 8 26 20 3 –
Percent (n = 57) (%) 14.0 45.6 35.1 5.3 –
% of patients 9 25 18 3 0
DurationaHours Days Weeks Months Persisting
Unwanted events (n) 59 52 97 82 62
Percent (n = 372) (%) 15.9 14.0 26.1 22.0 16.7
% of patients 51 40 62 47 40
Side effects (n) 18 18 12 3 5
Percent (n = 57) (%) 31.6 31.6 21.1 5.3 8.8
% of patients 19 17 0 3 5

224 Cognitive Therapy and Research (2018) 42:219–229
1 3
Table 2 Examples of severe or very severe and persistent side effects
UE-ATR domain Patient characteristics Sessions Narrative description Severity Duration
Deterioration of existing symptoms 45 year-old male with phobias, panic disorder, trauma and sub-
stance abuse in history
53 Suicidality after intense exposure Severe Days
37 year-old male with narcissistic personality disorder and depres-
sion
78 Increased feelings of hopelessness, decline of happiness, increas-
ing social isolation after confrontation with diagnosis and social
costs
Severe Weeks
33 year-old female with generalized anxiety disorder and personal-
ity disorders
75 Increased anxiety and insecurity, emotional outbursts after reflect-
ing her problems, awareness made her feel worse
Severe Days
27 year-old female with personality disorders, depression, social
phobia and obsessive–compulsive disorder
42 Felt destabilized right at the beginning of therapy, crying and
ruminating more often when talking about her problems
Severe Weeks
35 year-old female with post-traumatic stress disorder and depres-
sion
100 Increased symptoms of dissociation, dissociation in the domes-
tic environment, which had not happened before, patient felt
stressed by talking about what happened, insufficient coping
strategies and feelings of overburden
Severe Weeks
Negative wellbeing, distress 47 year-old male with alcohol dependence 18 Felt under pressure and stressed by the question of his therapist,
felt angry and tense during sessions
Severe Hours
37 year-old male with narcissistic personality disorder and depres-
sion
78 Feelings of surprise, shame and guilt after video feedback and
discussion of degrading offenses against his therapist
Severe Days
39 year-old male with agoraphobia 26 Intensive crying during session when talking about family conflicts Severe Days
35 year-old female patient with depression and agoraphobia 12 While preparing confrontation therapy sessions the patient felt
very uneasy and anxious, but did not dare to talk about it in the
beginning, confrontation therapy had to be postponed
Severe Days
26 year-old male with recurrent depression 30 Felt very uncomfortable during role plays, thought them embar-
rassing
Severe Hours
40 year-old female with recurrent depression and phobia 30 Burdened by exposure therapy, intense fear, blushing, trembling,
felt forced by her therapist
Severe Hours
68 year-old female with recurrent depression 24 After role play intensive crying, agitated and desolate at the end of
session, could not be contained because of time, refused further
role plays
Severe Weeks
50 year-old female with recurrent depression, somatization disor-
der and with a history of post-traumatic stress disorder
21 Negative emotions during meditation and relaxation exercises,
feelings of weariness of life through unhappy memories
Severe Days
33 year-old female with social phobia and adjustment difficulties
after cancer diagnosis
18 Discomposed just by talking to her therapist, blushing, nearly
crying during sessions, rigid posture, feels so exhausted that ses-
sions must be planned in the afternoons
Severe Months
24 year-old female with depression 40 Some issues are very hard for her to discuss, during a family ses-
sion she felt rejected and criticizes, she cried and blushed and
stammered, made her realize that the relationship to her father
is not good
Severe Hours
Emergence of new symptoms 33 year-old male with pathological gambling 46 After talking about biographical background the patient experi-
enced strong feelings of guilt and shame towards his mother, that
he had never felt before
Very severe Days
Very good patient therapist relationship,
dependency
45 year-old male with phobias, panic disorder, trauma and sub-
stance abuse in history
Patient will have problems to go on without therapy, relationship
to therapist is grown and strong, patient finds therapy supportive
and is afraid to struggle on his own, may need help in the future
as well
Mild Persistent

225Cognitive Therapy and Research (2018) 42:219–229
1 3
Table 2 (continued)
UE-ATR domain Patient characteristics Sessions Narrative description Severity Duration
Strains in family relations 50 year-old female with obsessive–compulsive disorder and
agoraphobia
16 Concerned by feedback from family members, who find her irri-
tated and behaving unusual when she tried to deal with negative
emotions
Severe Weeks
43 year-old female with dysthymia and panic disorder 23 Concerned by feedback of her husband, who finds her more indi-
viduating, taking more time for herself and less caressing
Severe Weeks
27 year-old female with personality disorders, depression, social
phobia and obsessive–compulsive disorder
42 Tries to individuate from her family, in order to work on her social
anxiety she tries to distance from her mother, agreed not to talk
to her every day and to not see her three times a week, patient
feels guilty and cries sometimes
Severe Hours
54 year-old male with depression and pedophilia 35 After the discussion about how social support prohibits offenses,
outed himself with his wife, confessed possession of illegal
material, his wife was shocked, the couple discussed separation
Severe Persistent
33 year-old female with social phobia and adjustment difficulties
after cancer diagnosis
18 Critical view on her parents, questions their way, relates them to
her problems
Very severe Persistent
Changes in family relations 27 year-old female with personality disorders, depression, social
phobia and obsessive–compulsive disorder
42 Broke up with her boyfriend after discussing the disadvantages of
an exploitive relationship in therapy, felt very sad about it but
hoped for reward in the future
Very severe Weeks
35 year-old female with post-traumatic stress disorder and depres-
sion
100 Patient decided to end the relationship to her mother after years of
problems and conflicts, that was her wish when starting therapy
and what she had planned, but she needed support and was sad
when she did it
Mild Persistent
Non-compliance of the patient 51 year-old female with dysthymia and obsessive–compulsive
personality disorder
12 Neglects homework, refuses exercises, criticizes interventions,
anxious that her therapist might be into Buddhistic religion,
complaining about poor match with her therapist, speculated
about drop out of therapy
Severe Weeks
42 year-old female with generalized anxiety disorder and somatiza-
tion disorder
26 Patient is not doing her homework or only half of it, she is not
working for her goals, it seems we’re not working on something
important for her, therapy goes in the wrong direction
Moderate Persistent

226 Cognitive Therapy and Research (2018) 42:219–229
1 3
of therapist or match of gender (Crits-Christoph etal. 1991;
Hamilton and Dobson 2002; Lambert 2011; Leitner etal.
2013; Sotsky etal. 1991). A test of the model against a
constant-only model was statistically significant, indicat-
ing that the predictors distinguished between patients with
and without SEs (chi square = 18.417, p < .05 with df = 9).
Table3 shows the variables and their coefficients. With a
Nagelkerke´s R2 of .23, the effect size was very small, result-
ing in a successful prediction of group in only 63% of cases
(instead of 57%).
Further analyses showed that the FSCL score was the
only variable that could by itself significantly predict the
classification of patients with or without SEs. Patients with
a lower FSCL score tended to be patients with SEs. No sig-
nificant prediction was found with respect to number of ses-
sions, number of diagnoses, age of patient, age of therapist,
years of experience of the therapist, gender of therapist,
gender of patient or match of gender.
Discussion
To our knowledge, this is the first study to attempt to sepa-
rate unwanted events (UE), adverse treatment reactions
(ATR), and side effects (SE) in a large sample of psycho-
therapy patients. It demonstrates that one can make this
distinction, and it confirms the concept and the tool. Look-
ing for unwanted events independent of any early judg-
ment as to causality helps to overcome the non-recognition
bias of therapists. We found that 74% of therapists were
not aware of any side effects in their treatment before the
systematic evaluation began. The present findings suggest
that the use of structured assessment methods like the UE-
ATR Checklist can improve the recognition of side effects.
A separate judgment about whether an unwanted event
is an adverse treatment reaction helps to avoid incorrectly
inflating the rate of side effects. In our sample, 98% of
patients had experienced some unwanted events, but only
43% were found to suffer from adverse treatment reactions
or side effects. It is an open question who is in the best
position to report side effects: the patient, the therapist or
other professionals. The patient is positioned to identify
negative feelings or outcomes, but does not have the pro-
fessional knowledge or perspective to say whether these
are side effects. The therapist can make a judgment about
side effects, especially those of clinical relevance, but may
have a recognition or judgement bias. Our method of hav-
ing an independent professional interview the therapist is
perhaps best, though it might be improved if a separate
interview of the patient had been possible. Future research
should combine both perspectives.
We found 372 unwanted events in 98 of 100 patients.
Unwanted events that were “definitely related” to treatment
were found in 43% of cases. This is a conservative estimate
of side effects. Had we included those unwanted events
that were rated as “most probably” and “rather related”
to the treatment, 63% of cases would have been classified
as suffering from side effects. This rate of side effects is
somewhat higher than in most other studies (Barlow 2010;
Berk and Parker 2009; Leitner etal. 2013; Lilienfeld 2007;
Linden and Strauß 2013; Parker etal. 2013), which can be
explained by the use of the structured UE-ATR procedure.
A point of discussion is whether ordinary reactions
to CBT which may be indispensable for the success of
treatment, such as distress during exposure treatment
should be called “side effects”. We argue that they are
side effects although they may be unavoidable, justified, or
even needed and intended. If there were an equally effec-
tive treatment that did not promote anxiety in the patient,
the present form of exposure treatment would become
unethical as it is a burden to the patient. It is important
to make a distinction between unavoidable and possibly
even intended negative effects on one side, and desired
ones on the other. This is a general rule in medicine, like
in surgery, where in earlier times it may have been neces-
sary, unavoidable and intended to remove a breast to fight
cancer. But still, it was not desired and so surgeons devel-
oped new treatments without this burden to the patient.
Unavoidable and intended negative effects are burdens
to the patient and therefore undesired. To acknowledge
this is important for the improvement of psychotherapy
in the individual case as this can help to avoid unnec-
essary distress for the patient and select the best treat-
ment option. It is also important for the development of
Table 3 Logistic regression analysis of group membership (patients
with or without SEs)
The dependent variable in this analysis is group coded so that
0 = patients without SEs and 1 = patients with SEs
FSCLFischer Symptom Checklist; SEs Side effects
Independent variable b SE Wald Probability Odds
Number of sessions .018 .014 1.680 .195 1.018
Number of diagnoses .208 .252 .683 .409 1.232
Sum FSCL − .054 .019 8.039 .005 .948
1.041
Age of patient .040 .021 3.603 .058 1.041
Age of therapist .001 .061 .000 .990 1.001
Years of experience
of the therapist
− .005 .086 .003 .955 .995
Gender of therapist .226 .599 .143 .706 1.254
Gender of patient .423 .472 .805 .370 1.527
Match of gender .261 0.303 0.740 .390 1.298
Model χ2 = 18.417 p < .05
Pseudo R2 = .226
n = 100

227Cognitive Therapy and Research (2018) 42:219–229
1 3
treatment alternatives that are better tolerated, similar to
the development of strategies in surgery which allow to
keep a breast.
Most of the side effects in this study were rated as mild
or moderate (59.6%) and transient (89.6%). However, more
than 40% of side effects were rated as severe (i.e., counter-
measures are necessary) or very severe (i.e. enduring nega-
tive consequences) and 8.8% as persistent. Psychotherapy
is not harmless.
Apart from prevalence, our data also show that side
effects are very multi-facetted in regard to type and content.
Deterioration of symptoms can be caused by increased feel-
ings of hopelessness or despair when looking at existing
problems. Feelings of dependency and lack of self-efficacy
can be caused by a very close and supportive therapeutic
relation. Break up of relations with partners or parents,
or avoidance of work can result from explicit or implicit
incrimination of living situations. Problems in therapy
cooperation can follow feelings of shame after not doing
homework assignments. Those reports have to be interpreted
carefully and may not be applicable to other samples and set-
tings. Still, these examples show that regular and appropriate
therapeutic interventions can have negative consequences,
which may even be enduring like quitting a job, getting into
trouble with close persons, undermining of self-confidence,
or aggravation of problem perception.
The positive correlation between number of sessions and
number of unwanted events and side effects gives validity
to the assessment, as more unwanted events should emerge
over the course of time. An interesting and somewhat unex-
pected finding was that patients with side effects had lower
scores on the FSCL, which suggests that side effects are
not an expression of severity of illness. It can be assumed
that mild side effects are better recognized in less severe
cases. The FSCL score was the only significant predictor of
side effects. Other variables, such as number of diagnoses or
experience of therapists, did not contribute to the classifica-
tion of patients with or without SEs. This finding may be due
to a relatively small sample size concerning the number of
variables in the model. Further research is needed to identify
patients at risk for developing side effects and to replicate
the association between symptom severity and side effects.
Our study has several implications. First, therapists are
often not aware of side effects in their treatment as they
primarily focus on positive changes (Tomba etal. 2017).
An awareness and recognition of unwanted events and side
effects in all therapies will benefit patients, improve therapy
or reduce attrition, analogous to the benefit of measurement-
based monitoring of treatment progress (Castonguay etal.
2010). It is probably not realistic to expect that therapists
use checklists in their daily practice, as is similar with other
clinical scales. Still, it may be helpful to know the dimen-
sions of the UE-ATR Checklist, as this can help therapists to
not forget important aspects in the clinical assessment. The
scale could and should be used in the training of therapists,
so that they learn what a side effect is, how to recognize side
effects, and how to ask patients about them (Nolan etal.
2004). The therapist bias of not seeing side effects must be
overcome during training. Finally, the checklist may be used
in research and especially in clinical trials, where there is
often a lack of focus on side effects (Jonsson etal. 2014).
The research question is not only to describe side effects
but also to develop strategies on what to do should they
occur, to identify patients at risk, and to develop therapies
without certain side effects (Creed etal. 2014). Finally, the
data describe the SE profile of CBT under routine clinical
conditions in a heterogenous patient population. This is what
can occur with some probability in patients undergoing this
mode of treatment. Therapist can use this knowledge when
informing their patients about the treatment and also for risk
monitoring in the course of treatment.
Limitations
The identification of SEs in this report depends on our judg-
ment as to the appropriateness of the therapy and the unob-
served but presumably ethical behavior of the therapists.
We can say that all therapists were trained in CBT, but that
does not necessarily verify treatment integrity and compe-
tency (Lambert 2011). However, since the therapists were
closely supervised, gross malpractice is very unlikely. The
cross-sectional study design has limitations, including that
of recall. Prospective, longitudinal data would have allowed
a more fine-grained analysis and potentially a more accu-
rate assessment of the events during the course of therapy.
Another limitation is that we do not have data on the rela-
tionship between SEs and the outcome of therapy. This
report focused only on the nature, frequency, severity and
duration of UEs, but not their consequences. Since about
half of the patients were receiving additional psychotropic
medication, we cannot attribute the presence of SEs to the
psychotherapy alone. The instructions for the therapists were
very clear as to report unwanted events and subsequently rate
whether those events were caused by psychotherapy and psy-
chotherapy only. But since those ratings can be flawed some
SEs may have been related to the pharmacotherapy itself
or an interaction between the pharmacotherapy and CBT.
To use therapists ratings only, limits the generalizability of
the results per se. It may be possible that patients would
consider the same events noted by therapists as wanted and
beneficial instead of unwanted (e.g. a break up) or not caused
by psychotherapy but some aspects outside of therapy (e.g.
job loss). A combination of the patient, therapist and a sepa-
rate interviewer may indeed be the best way to approach
this problem (Smith etal. 2003). Since we investigated a

228 Cognitive Therapy and Research (2018) 42:219–229
1 3
heterogeneous sample of patients with different conditions
treated by a range of therapists with a range of skills and
experiences and different methods of CBT, the results could
be different in different samples of patients, conditions, ther-
apists and even types of therapy.
Conclusions
A structured assessment, using the UE-ATR-Checklist
allows to measure side effects of psychotherapy (SE), and
discriminate between SEs and other unwanted events (UEs).
SEs include symptoms, but also additional problems in many
areas of life. They are common even in the course of rig-
orously supervised CBT. The data describe the SE profile
of CBT under routine clinical conditions in a heterogenous
patient population. Such information is needed in clinical
practice, training, and research.
Acknowledgements The authors thank the Institute for Behavior Ther-
apy Berlin, the Centre for Psychotherapy at the Humboldt-University
Berlin and the Institute for Behavior Therapy Brandenburg for partici-
pating in the study. We would also like to acknowledge the editorial
assistance of Jon Kilner, MS, MA (Pittsburgh, PA).
Compliance with Ethical Standards
Conflict of Interest Marie-Luise Schermuly-Haupt, Michael Linden
and A. John Rush declare that they have no conflict of interest.
Human and Animal Rights All procedures performed in studies involv-
ing human participants were in accordance with the ethical standards of
the institutional and/or national research committee and with the 1964
Helsinki declaration and its later amendments or comparable ethical
standards.
Informed Consent Informed consent was obtained from all individual
participants included in the study.
References
Anderson, E. M., & Lambert, M. J. (2001). A survival analysis of clini-
cally significant change in outpatient psychotherapy. Journal of
Clinical Psychology, 57(7), 875–888.
Barlow, D. H. (2010). Negative Effects from psychological treatments.
A perspective. American Psychologist, 65(1), 13–20. https ://doi.
org/10.1037/a0015 643.
Bergin, A. E. (1963). The effects of psychotherapy: Negative results
revisited. Journal of Consulting Psychology, 10, 244–250.
Berk, M., & Parker, G. (2009). The elephant on the couch: Side effects
of psychotherapy. Australian & New Zealand Journal of Psychia-
try, 43(9), 787–794. https ://doi.org/10.1080/00048 67090 31075 59.
Brainerd, C. J., & Reyna, V. F. (2005). The science of false memory.
New York: Oxford University Press.
Buckley, P., Karasu, T. B., & Charles, E. (1981). Psychotherapists view
their personal therapy. Psychotherapy: Theory, Research & Prac-
tice, 18(3), 299–305. https ://doi.org/10.1037/h0088 377.
Castonguay, L. G., Boswell, J. F., Constantino, M. J., Goldfried, M. R.,
& Hill, C. E. (2010). Training implications of harmful effects of
psychological treatments. American Psychologist, 65(1), 34–49.
https ://doi.org/10.1037/a0017 330.
Creed, T. A., Wolk, C. B., Feinberg, B., Evans, A. C., & Beck, A.
T. (2014). Beyond the Label: Relationship Between Community
Therapists’ Self-Report of a Cognitive Behavioral Therapy Orien-
tation and Observed Skills. Administration and Policy in Mental
Health and Mental Health Services Research, 43(1), 1–8. https ://
doi.org/10.1007/s1048 8-014-0618-5.
Crits-Christoph, P., Baranackie, K., Kurcias, J., Beck, A., Carroll, K.,
Perry, K., Luborsky, L., McLellan, A., Woody, G., Thompson,
L., Gallagher, D., & Zitrin, C. (1991). Meta-analysis of thera-
pist effects in psychotherapy outcome studies. Psychotherapy
Research, 1(2), 81–91.
Dilling, H. (2011). Internationale Klassifikation psychischer Störun-
gen: ICD-10 Kapitel V (F); klinisch-diagnostische Leitlinien (8.,
überarb. Aufl. unter Berücksichtigung der Änderungen entspre-
chend ICD-10-GM 2011.). Bern: Huber.
Fischer-Cornelssen, K. A., & Berchier, P. (1982). Validität und Relia-
bilität einer Symptom-Checkliste (FSCL): Anwendung in der
Psychogeriatrie. Zeitschrift für Gerontologie und Geriatrie, 15,
31–37.
Fisher, P. L., & Durham, R. C. (1990). Recovery rates in generalized
anxiety disorder following psychological therapy: An analysis of
clinically significant change in the STAI-T across outcome studies
since 1990. Psychological Medicine, 29(6), 1425–1434.
Foa, E. B., Zoellner, L. A., Feeny, N. C., Hembree, E. A., & Alvarez-
Conrad, J. (2002). Does imaginal exposure exacerbate PTSD
symptoms? Journal of Consulting Clinical Psychology, 70(4),
1022–1028.
Hafner, R. J. (1984). The marital repercussions of behavior therapy for
agoraphobia. Psychotherapy: Theory, Research, Practice, Train-
ing, 21(4), 530–542. https ://doi.org/10.1037/h0085 999.
Hamilton, K. E., & Dobson, K. S. (2002). Cognitive therapy of depres-
sion: Pretreatment patient predictors of outcome. Clinical Psy-
chology Review, 22(6), 875–893.
Hannan, C., Lambert, M. J., Harmon, C., Nielsen, S. L., Smart, D.
W., Shimokawa, K., & Sutton, S. W. (2005). A lab test and algo-
rithms for identifying clients at risk for treatment failure. Journal
of Clinical Psychology, 61(2), 155–163. https ://doi.org/10.1002/
jclp.20108 .
Harnett, P., O´Donovan, A., & Lambert, M. J. (2010). The dose
response relationship in psychotherapy: Implications for
social policy. Clinical Psychology, 14(2), 39–44. https ://doi.
org/10.1080/13284 207.2010.50030 9.
Hatfield, D., McCullough, L., Frantz, S. H., & Krieger, K. (2010).
Do we know when our clients get worse? An investigation of
therapists’ ability to detect negative client change. Clinical Psy-
chology and Psychotherapy, 17(1), 25–32. https ://doi.org/10.1002/
cpp.656.
Henry, W. P., Strupp, H. H., Butler, S. F., Schacht, T. E., & Binder, J. L.
(1993). Effects of training in time-limited dynamic psychotherapy:
Changes in therapist behavior. Journal of consulting and clinical
psychology, 61(3), 434–440.
Hoffmann, S. O., Rudolf, G., & Strauß, B. (2008). Unerwünschte und
schädliche Wirkungen von Psychotherapie. Der Psychotherapeut,
53, 4–16.
Jonsson, U., Alaie, I., Parling, T., & Arnberg, F. K. (2014). Reporting
of harms in randomized controlled trials of psychological inter-
ventions for mental and behavioral disorders: A review of current
practice. Contemporary Clinical Trials, 38(1), 1–8. https ://doi.
org/10.1016/j.cct.2014.02.005.
Lambert, M. J. (2010). Predicting negative treatment outcome: Meth-
ods and estimates of accuracy. In M. J. Lambert (Ed.), Preven-
tion of treatment failure: The use of measuring, monitoring, and

229Cognitive Therapy and Research (2018) 42:219–229
1 3
feedback in clinical practice (pp.83–105). Washington, DC:
American Psychological Association.
Lambert, M. J. (2011). What have we learned about treatment failure in
empirically supported treatments? Some suggestions for practice.
Cognitive and Behavioral Practice, 18(3), 413–420. https ://doi.
org/10.1016/j.cbpra .2011.02.002.
Lambert, M. J., & Ogles, B. M. (2004). The efficacy and effective-
ness of psychotherapy. In M. J. Lambert (Ed.), Bergin & Gar-
field’s handbook of psychotherapy and behavior change (5thedn.,
pp.139–193). New York: Wiley.
Leitner, A., Märtens, M., Koschier, A., Gerlich, K., Liegl, G., Hinter-
wallner, H., & Schnyder, U. (2013). Patients’ perceptions of risky
developments during psychotherapy. Journal of Contemporary
Psychotherapy, 43(2), 95–105. https ://doi.org/10.1007/s1087
9-012-9215-7.
Lilienfeld, S. O. (2007). Psychological treatments that cause harm.
Perspectives on Psychological Science, 2(1), 53–70. https ://doi.
org/10.1111/j.1745-6916.2007.00029 .x.
Linden, M. (2013). How to define, find and classify side effects in psy-
chotherapy: From unwanted events to adverse treatment reactions.
Clinical Psychology and Psychotherapy, 20(4), 286–296. https ://
doi.org/10.1002/cpp.1765.
Linden, M., & Schermuly-Haupt, M. L. (2014). Definition, assessment
and rate of psychotherapy side effects. World Psychiatry, 13(3),
306–309. https ://doi.org/10.1002/wps.20153 .
Linden, M., & Strauß, B. (2013). Risiken und Nebenwirkungen von
Psychotherapie. Erfassung, Bewältigung, Risikovermeidung. Ber-
lin: Medizinisch-Wissenschaftliche Verlagsgesellschaft.
Macaskill, N. (1992). Psychotherapists-in-training evaluate their per-
sonal therapy: Results of a UK Survey. British Journal of Psycho-
therapy, 9(2), 133–138. https ://doi.org/10.1111/j.1752-0118.1992.
tb012 11.x.
Mayou, R. A., Ehlers, A., & Hobbs, M. (2000). Psychological debrief-
ing for road traffic accident victims: Three-year follow-up of a
randomized controlled trial. British Journal of Psychiatry, 176(6),
589–593. https ://doi.org/10.1192/bjp.176.6.589.
Milton, F., & Hafner, R. J. (1979). The outcome of behavior therapy for
agoraphobia in relation to marital adjustment. Archives of General
Psychiatry, 36(7), 807–811.
Nolan, S. A., Strassle, C. G., Roback, H. B., & Binder, J. L. (2004).
Negative treatment effects in dyadic psychotherapy: A focus on
prevention and intervention strategies. Journal of Contemporary
Psychotherapy, 34(4), 311–330. https ://doi.org/10.1007/s1087
9-004-2526-6.
Parker, G., Fletcher, K., Berk, M., & Paterson, A. (2013). Development
of a measure quantifying adverse psychotherapeutic ingredients:
The Experiences of Therapy Questionnaire (ETQ). Psychiatry
Research, 206(2–3), 293–301. https ://doi.org/10.1016/j.psych
res.2012.11.026.
Roback, H. B. (2000). Adverse outcomes in group psychotherapy. Risk
factors, prevention, and research directions. Journal of Psycho-
therapy Practice Research, 9(3), 113–122.
Rose, S. C., Bisson, J., Churchill, R., & Wessely, S. (2002). Psycho-
logical debriefing for preventing post traumatic stress disorder
(PTSD). The Cochrane database of systematic reviews, 2002(2),
1–47. https ://doi.org/10.1002/14651 858.CD000 560.
Rosen, M. I., Rounsaville, B. J., Ablondi, K., Black, A. C., & Rosen-
heck, R. A. (2010). Advisor-teller money manager (ATM) therapy
for substance use disorders. Psychiatry Services, 61(7), 707–713.
https ://doi.org/10.1176/appi.ps.61.7.707.
Scheeringa, M. S., Weems, C. F., Cohen, J. A., Amaya-Jackson, L.,
& Guthrie, D. (2011). Trauma-focused cognitive-behavioral
therapy for posttraumatic stress disorder in three-through six
year-old children: A randomized clinical trial. Journal of Child
Psychology and Psychiatry, 52(8), 853–860. https ://doi.org/10.1
111/j.1469-7610.2010.02354 .x.
Shrout, P. E., & Fleiss, J. L. (1979). Intraclass correlations: Uses in
assessing rater reliability. Psychology Bulletin, 86(2), 420–428.
Sijbrandij, M., Olff, M., Reitsma, J. B., Carlier, I. V. E., & Gersons, B.
P. R. (2006). Emotional or educational debriefing after psycholog-
ical trauma. Randomised controlled trial. British Journal of Psy-
chiatry, 189, 150–155. https ://doi.org/10.1192/bjp.bp.105.02112 1.
Smith, S. R., Hilsenroth, M. J., Baity, M. R., & Knowles, E. S. (2003).
Assessment of patient and therapist perspectives of process: A
revision of the Vanderbilt Psychotherapy Process Scale. American
Journal of Psychotherapy, 57(2), 195–205.
Sotsky, S. M., Glass, D. R., Shea, M. T., Pilkonis, P. A., Collins, J. F.,
Elkin, I., Watkins, J. T., Imber, S. D., Leber, W. R., Moyer, J., & Oli-
veri, M. E. (1991). Patient predictors of response to psychotherapy
and pharmacotherapy: Findings in the NIMH Treatment of Depres-
sion Collaborative Research Program. American Journal of Psy-
chiatry, 148(8), 997–1008. https ://doi.org/10.1176/ajp.148.8.997.
Tarrier, N., Pilgrim, H., Sommerfield, C., Faragher, B., Reynolds,
M., Graham, E., & Barrowclough, C. (1999). A randomized trial
of cognitive therapy and imaginal exposure in the treatment of
chronic posttraumatic stress disorder. Journal of Consulting and
Clinical Psychology, 67(1), 13–18.
Tomba, E., Tecuta, L., Schumann, R., & Balladini, D. (2017). Does
psychological well-being change following treatment? An
exploratory study on outpatients with eating disorders. Compre-
hensive Psychiatry, 74, 61–69. https ://doi.org/10.1016/j.compp
sych.2017.01.001.
Wirtz, M. A., & Caspar, F. (2002). Beurteilerübereinstimmung und
Beurteilerreliabilität: Methoden zur Bestimmung und Verbesse-
rung der Zuverlässigkeit von Einschätzungen mittels Kategorien-
systemen und Ratingskalen. Göttingen: Hogrefe.