Informing patients about possible negative effects of psychological treatment: A survey of Swedish clinical psychologists’ attitudes and practices

Abstract

Objective: The present study explored Swedish clinical psychologists’ attitudes and practices of informing patients about possible negative effects of psychological treatment. Method: An anonymous online survey with closed and open-ended questions was distributed via Facebook and two mailing lists. In total, 320 clinical psychologists (age M = 38.6, SD = 10.3; 76% women; 77.3% Cognitive Behavior Therapy) completed the survey. Responses were analyzed using χ², binominal regression analysis, and thematic analysis.

Results: No significant associations were identified between demographic variables selected a priori and informing patients about possible negative effects, i.e., years in practice, therapeutic orientation, age, and male gender. The thematic analysis resulted in several positive (e.g., creating realistic expectations of treatment and increasing resilience) and negative attitudes (e.g., causing excessive worry) of sharing patients with information about possible negative effects. It also revealed various factors that prevent an open discussion on the topic, despite being seen as important.

Conclusion: The need to offer patients information about possible negative effects should be included in ethical guidelines, codes of conduct, and taught during clinical training. Lack of knowledge, fear, lack of time, and patient characteristics however seem to prevent therapists from discussing the issue during the informed consent procedure.

Keywords:

• Negative effects
• attitudes
• practice
• thematic analysis
• online survey
• psychological treatment
• informed consent

Clinical or methodological significance of this article: The present study increases the current understanding about therapists’ attitudes and practices of informing their patients about possible negative effects of psychological treatment. It provides recommendations on how education, ethical guidelines, and codes of conduct could help therapists include information of negative effects during the informed consent procedure.

Introduction

Psychological treatment can help prevent and mitigate mental illness and provide effective care for those suffering from psychiatric disorders (c.f., Cuijpers et al., 2021, 2020; Van Dis et al., 2020). Yet,  … if a treatment—any treatment—produces beneficial results, it must also be capable of producing harmful effects (Hadley & Strupp, 1976, p. 1291). The definition of these “harmful effects” and what they might pertain have varied in the literature (Bergin, 1963; 1966; Hadley & Strupp, 1976; Mays & Frank, 1985), with a consensus on their characteristics still lacking (Herzog et al., 2019; Rozental et al., 2018). However, Linden (2013) proposed a comprehensive set of definitions in order to better differentiate between the various incidents that might occur. This includes an “unwanted event” (UE), i.e., a negative event that occur during or in the wake of treatment, “treatment-emergent reactions,” i.e., UE that are caused by treatment, “adverse treatment reactions,” i.e., UE that are probably caused by a correctly applied treatment, “malpractice reaction,” i.e., UE that are probably caused by an incorrectly applied treatment, “treatment non-response,” i.e., lack of improvement despite receiving treatment (which can be a result of either a correctly or incorrectly applied treatment), and “deterioration of illness,” i.e., worsening (which may or may not be an UE, and can be a result of either a correctly or incorrectly applied treatment). Linden (2013) also defined “therapeutic risk” as all adverse treatment reactions that are known and need to be shared with the patient during the informed consent process, and “contraindications” as conditions of the individual patient that can make adverse treatment reactions probable. In the present study, the term “negative effects” of psychological treatment is broadly applied to describe any incident of potential harmful character, regardless of whether or not it can be causally linked to treatment per se, thus corresponding with the description of an UE (Linden, 2013). This covers deterioration (“deterioration of illness”) and non-response (“treatment non-response”), but also incidents that might be unrelated to symptomatology, e.g., dependency (Hadley & Strupp, 1976). Moreover, the present study only focuses on possible negative effects of an appropriately applied psychological treatment that is provided by a trained therapist, in line with “adverse treatment reactions,” which differs from “malpractice reaction” that are composed of incidents stemming from so called “fringe psychotherapies” (Beyerstein, 2001), e.g., scared straight interventions for conduct problems in youths (Lilienfeld, 2007), and malpractice issues and unethical behavior (Lindgren & Rozental, 2022).

Informing About Possible Negative Effects

Therapists who provide psychological treatment often belong to a professional organization that underlines the importance of avoiding harm. In a similar fashion, the value of obtaining informed consent is also stressed, e.g., “psychologists inform clients/patients as early as is feasible in the therapeutic relationship about the nature and anticipated course of therapy … ” (American Psychological Association, 2016, p. 14). However, providing information about possible negative effects is seldom mentioned explicitly in ethical guidelines or codes of conduct, which might affect to what extent therapists are conscious of and willing to discuss this issue with patients. Responses from a survey of 500 randomly selected clinical psychologists within the American Psychological Association showed that 28% were completely unaware of, and 30% were mistaken about, the established deterioration rate of 5-10% (Boisvert & Faust, 2006). In addition, Bystedt et al. (2014) explored the attitudes of 74 Swedish Cognitive Behavior (CBT) therapists, indicating that 94.5% believed that negative effects exist and can cause problems, but that they were largely unaware of research findings and methods to identify these events. Meanwhile, a doctoral thesis by Sarkozy (2010) demonstrated that half of the 161 responding American clinical psychologists stated that they informed about the risk that treatment might not work or produce any benefits. Moreover, 34.1% described that they conveyed this information before the first session, 58% during the first session, and 6.5% when they felt so fit or not at all (1.4% choose not to respond at all). Sarkozy (2010) also demonstrated that 48.3% agreed that it is essential to address possible negative effects during the informed consent procedure, 68.6% concurred that addressing the issue does not harm the therapeutic alliance or treatment outcome, and 89.4% believed that a discussion about possible negative effects is ethically important. Interestingly, only one study has asked patients themselves about receiving information on the potential risks of psychological treatment, indicating that it was considered important and should be conveyed prior to commencing any interventions (Braaten & Handelsman, 1997).

Correlations between demographic variables and the practice and addressing possible negative effects during the informed consent procedure have also been studied. Sarkozy (2010) hypothesized that clinical experience (in years), older age, or having CBT as one’s therapeutic orientation would be positively correlated with sharing information about possible negative effects. The rationale behind this was that clinical psychologists with more extensive experience have been in contact with more patients and may therefore become more aware of ethical issues. Furthermore, as CBT tends to emphasize patient autonomy, this was considered to affect what is shared during the informed consent procedure. However, no support for these hypotheses was found, but more male than female clinical psychologists reported that they raised the issue with their patients.

As for what could prevent therapists from sharing information with patients about possible negative effects, this remains largely unexplored. The patient's symptomatology, personality, and general functioning seem to influence whether or not this is conveyed, with Sarkozy (2010) reporting that 39.5% agreed to these factors impacting the decision to address possible negative effects. Lack of knowledge and training about negative effects in psychological treatment have also been found to inhibit therapists from discussing the issue (Bystedt et al., 2014). Similarly, Sarkozy (2010) showed that those clinical psychologists who had undergone specific teaching activities in ethics and the informed consent procedure were more likely to talk with patients about this topic. However, more research on the factors that might prevent and help therapists discuss possible negative effects is warranted.

The Present Study

There is currently scarce research on the topic of informing patients about possible negative effects of psychological treatment. However, given the increased attention during the last decade concerning potentially harmful events in psychological treatment (Rozental et al., 2018), therapists might be more open to talk about possible negative effects today. The present study thus seeks to explore this issue using an anonymous online survey distributed to practicing clinical psychologists in Sweden. The purpose is to further the understanding of therapists’ attitudes and practices with regard to possible negative effects of psychological treatment. More specifically, the aims of the present study are to: (1) explore the proportion of therapists who are aware that possible negative effects occur, (2) the proportion of therapist who inform their patients about possible negative effects, (3) whether any demographic variables are related to providing this information to patients, (4) positive and negative ideas of informing patients, and (5) factors that may prevent the provision of such information.

Method

Sample

The present study involved practicing Swedish clinical psychologists. In Sweden, this is a professional title regulated by the National Board of Health and Welfare [Socialstyrelsen]. In Swedish healthcare, only people with such a license or undergoing practical training to become a psychologist [Praktisk Tjänstgöring för Psykologer (PTP)] can call themselves clinical, i.e., licensed psychologists (Sveriges Psykologförbund, n.d.). All Swedish clinical psychologists are obliged to follow the ethical guidelines in Sweden (Sveriges Psykologförbund, 1998). All clinical psychologists working within Swedish healthcare are also required to follow the Swedish National Healthcare Law (Patientlagen, 2014, p. 821), which stipulates that information about expected results and possible negative effects should be shared with a patient before initiating treatment, thus making it relevant population to explore in the present study.

In total, 320 clinical psychologists participated. Most respondents were women (76.2%), which is close to the gender distribution (73%) among Swedish clinical psychologists reported by the Swedish Psychological Association. The age of the sample ranged from 24 to 72 years (M = 38; SD = 10.3). For an overview of the demographics, see Table 1. When asked about their therapeutic orientation, 214 (77.3%) choose to define themselves as CBT, 38 (13.7%) Psychodynamic Psychotherapy (PP), 2 (0.7%) Humanistic/Existential Psychotherapy, and 23 (8.3%) the category Other. There are no records in Sweden with regard to the distribution between therapeutic orientations, although CBT and PP have historically been the two dominating modalities.

Table 1. Demographic information of the respondents.

Gender: n (%)	Female	211 (76.2)
	Male	64 (23.1)
	Other	2 (0.7)
Age (years): M (SD)		38.6 (10.3)
Therapeutic orientation: n (%)	Cognitive Behavior Therapy	214 (77.3)
	Psychodynamic Psychotherapy	38 (13.7)
	Humanistic/Existential Psychotherapy	2 (0.7)
	Other	23 (8.3)
Years of experience as clinical psychologist^a: M (SD)		8.71 (8.36)
University for degree as clinical psychologist^b: n (%)	Gothenburg University	31 (11.2)
	Karlstad University	6 (2.2)
	Karolinska Institutet	11(4.0)
	Linköping University	23 (8.3)
	Linnaeus University	5 (1.8)
	Lund University	34 (12.3)
	Mid Sweden University	14 (5.0)
	Stockholm University	31(11.2)
	Umeå University	32 (11.5)
	Uppsala University	55 (19.9)
	Örebro University	23 (8.3)
	Foreign degree	12 (4.3)
Year of one's university degree: n (%)	1981–1985	4 (1.5)
	1986–1990	3 (1.1)
	1991–1995	7 (2.5)
	1996–2000	12 (4.3)
	2001–2005	17 (6.1)
	2006–2010	31 (11.2)
	2011–2015	59 (21.3)
	2016–2020	99 (35.7)
	> 2021	45 (16.3)
Further education^c: n (%)	Psychotherapist	46 (16.6)
	Specialist degree^d	32 (11.6)
	Doctoral degree	6 (2.2)
	Other	33 (12)
	None	41 (14.3)
Workplace: n (%)	Adult psychiatry	80 (28.9)
	Child and adolescent psychiatry	34 (12.3)
	Private clinic	45 (16.2)
	Habilitation	9 (3.2)
	Occupational health	11 (4)
	Primary care	65 (23.5)
	Somatic care	8 (2.9)
	Inpatient care	2 (0.7)
	Other	23 (8.3)
Number of patients per week: M (SD)		14.9 (9.11)
Informing about possible negative effects: n (%)	Yes	171 (74)
	No	60 (26)

Note. In total, 277 respondents completed most demographic questions, and 43 chose not to respond to these. A total of 287 respondents completed the question about further education, but 33 chose not to complete this question.

^a Including practical training.

^b The listed universities provide the educational program to become a clinical psychologist in Sweden.

^c Several options were available, the total percentage is therefore not 100%. A total of 287 respondents completed this question.

^d Specialist degree is a title issued by the Swedish Psychological Assocation following several years of further education and a research thesis.

Instrument

A questionnaire was created with closed and open-ended questions (see Appendix). Respondents entered a website with information about ethics, data management, and the contact details to the principal investigator (AR). However, the purpose of the present study was not fully disclosed in order to prevent social desirability. Instead, the purpose was described as furthering the understanding of how clinical psychologists discuss the effects of psychological treatment in general with their patients. Informed consent was provided online before the respondents completed the 22 questions that were included in the questionnaire. The first nine questions covered demographic information and the remaining 13 questions concerned treatment effects, which consisted of closed questions (i.e., yes/no) that were followed by open-ended questions. However, only open-ended questions regarding possible negative effects of psychological treatment were used in the present study, the rest were included to avert priming effects (e.g., asking about how one discusses positive effects). The web host Squarespace (2022) was used to create the website and the online survey was created using Limesurvey (2022). A pilot was performed with a convenience sample of 20 psychology students and clinical psychologists in Sweden to ensure the quality and clarity of the survey, but these responses were not included in the final dataset.

Data Collection

A short description of the present study and a link to the website featuring the online survey was shared via two Swedish mailing lists, one belonging to the clinical division of the Swedish Psychological Association [Sveriges Kliniska Psykologers Förening] (approximately 200 members), and the Swedish Association of Behaviour Therapy [Beteendeterapeutiska föreningen] (approximately 800 members), and one Facebook group for psychologists in Sweden (approximately 9000 members). The conditions for participation included (a) being a clinical psychologist in Sweden (undergoing practical training or licensed), and (b) being engaged in clinical practice. The respondents also received information that they could, having completed the online survey, enter a lottery to win a self-help book on perfectionism by the principal investigator (AR). The survey was accessible for six weeks (12 September to 24 October 2022). Six reminders of the present study were published in the Facebook group. The survey was not time restricted, and the average time for completing the survey was 15 min.

Statistical Analysis

Quantitative data from the current study was presented descriptively and analyzed using χ² and binominal logistic regression. Four hypotheses were formulated a priori in accordance with Sarkozy (2010): it was assumed that more years in practice, having CBT as one’s therapeutic orientation, being older, and identifying oneself as male gender would be associated with informing patients about possible negative effects of psychological treatment. Additional analyses of relationships were made exploratively. All statistical analyses were conducted using jamovi (2023).

Thematic analysis was performed on the responses to the three open-ended questions that concerned possible negative effects of psychological treatment. The analysis aimed to be a detailed and realistic description of the data set. Themes were coded semantically and inductively due to the lack of theoretical framework and research on the subject of informing patients about possible negative effects. The analysis was performed in accordance with the phases and criteria by Braun and Clarke (2021). First, all of the responses were copied into Excel and read multiple times. Second, initial codes were generated. Third and last, themes were identified, reviewed, defined, and named. Frequencies and labels for the themes were named in accordance with Hill et al. (2005). Frequency was calculated based on how many respondents endorsed the themes within the survey question where the theme was most frequently occurring. The first author (FW) organized the responses and performed the thematic analysis under the supervision of the principal investigator (AR). The first author had just completed her clinical training in CBT. The second author had 12 years of experience working with CBT, had written a doctoral dissertation on negative effects of psychological treatment, and had researched the topic for ten years.

Results

Descriptives

Demographic information of the respondents and the rate of informing patients about possible negative effects of psychological treatment is presented in Table 1.

Quantitative Analyses

The four hypotheses formulated a priori were tested, but were not related to informing patients of possible negative effects of psychological treatment; more years in practice (OR = 1.04, 95% Confidence Interval (CI) [0.99, 1.09]), therapeutic orientation, χ2(1; n = 205) = 3.11 (p = .078), age (OR = 1.02, 95% CI [0.98, 1.05]), and gender, χ2(1; n = 201) = < 0.00 (p = .98), albeit with potential trends pointing towards more years in practice and those practicing PP being more likely to inform patients about possible negative effects.

Post-hoc analyses concerning the other demographic variables were not significant, although a potential trend was observed for having received information about negative effects during one's clinical training χ²(1; 205) = 2.76 (p = .096).

Lastly, the respondents’ understanding of current research findings on rates of deterioration, non-response, and improvement were also explored (i.e., questions 18–20 in the Appendix). The reported estimates in the current study were 12.9% (SD = 8.07%), 29.8% (SD = 12.9%), and 58.4% (SD  = 13.0%).

Qualitative Analysis

In total, 170 respondents completed the question about positive expectations of addressing the negative effects of psychological treatment, and 127 answered the corresponding question about negative expectations. A total of 82 respondents completed the question about barriers of informing their patients about negative effects. A total of 379 responses were included in the thematic analysis.

The thematic analysis yielded two domains. The first domain, Expectations, consisted of four themes and ten sub-themes, see Table 2. This included both positive and negative expectations of informing about negative effects of psychological treatment; (1) Familiarity with the treatment process, (2) Increased focus on negative effects, (3) Improved collaboration, and (4) Professional and ethical responsibility. The second domain, Hindrances, included four themes; (1) Lack of knowledge, (2) Fear, (3) Lack of time, and (4) Patient characteristics, see Table 3.

Table 2. Themes and sub-themes of the domain Expectations.

Themes	Sub-themes	Examples	Frequency^a	Label
1. Familiarity with the treatment process	1.1 Realistic expectations	“ … trying to nuance the idea that a treatment will solve ‘everything’.”	43	Variant
	1.2 Increased participation	“ … include the patient in his care —> increases the patient's sense of self-efficacy? Make the patient an active participant in the treatment.”	11	Variant
	1.3 Mental preparedness and resilience	“Joint working alliance, motivation and resilience against potential aspects of deterioration.”	62	Variant
	1.4 Normalizing initial negative emotions and experiences	“Just as with the introduction of SSRI, it is important to convey that a temporary increase in symptoms, behavioral changes, that exposure is usually difficult, can make you feel worse before you feel better.”	27	Variant
	1.5 Decreased motivation and trust in treatment	“Reduced commitment, reduced trust in the treatment, feeling of hopelessness or helplessness.”	99	Typical
2. Increased focus on negative effects	2.1 Detecting negative effects	“It is good that the person is aware of possible risks and can recognize himself if he were to suffer any negative effect.”	5	Variant
	2.2 Excessive worry	“That they exaggerate the negative effects, for example due to an anxiety problem, which is what they are seeking help for.”	22	Variant
	2.3 Negative impact on expectancy	“Of course, in some cases this could make the patient to look for symptoms of deterioration + nocebo.”	21	Variant
3. Improved collaboration	3.1 Increased trust in the therapist	“Being open about the fact that a treatment is not suitable for everyone, that it is not certain that it will help, to precede and to have talked about what we do then - instills trust and increases trust.”	27	Variant
	3.2 Increased reporting and handling of negative effects	“ … That they are more sincere in verbal reports and estimates, because they know that we are trying to manage and make a plan together.”	35	Variant
4. Professional and ethical responsibility		“So that patients can make an informed choice as to whether they want to start treatment or not, so that the timing is right and you have good conditions to cope with treatment."	56	Variant

Note.^a The total number of responses composing the sub-theme.

Table 3. Themes and sub-themes of the domain Hindrances.

Theme	Examples	Frequency^a	Label
1. Lack of knowledge	“I feel that much of my work is from what I learned during my educational program and a lot of the books I bought then are the same ones I use now, or later updated copies. These lack information about negative treatment effects … ”	28	Variant
2. Fear	“Fear in me of dropout … ”	41	Typical
3. Lack of time	“The time limitation prevents me from always being able to prioritize discussing the negative effects of treatment … ”	13	Variant
4. Patient characteristics	“In a patient group with severe symptoms and high co-morbidity, it is sometimes difficult to work according to a set method. Also difficult with diagnostics. This makes it difficult to talk about treatment effects … ”	14	Variant

Note. ^aThe total number of responses composing the sub-theme.

Domain: expectations

1. Familiarity with the treatment process.

The theme Familiarity with the treatment process was related to how respondents believed information about possible negative effects could help familiarize the patient with the entire treatment process. This was mostly associated with positive aspects, such as creating more realistic expectations and normalizing initial negative reactions to treatment, but some respondents were fearful this could also affect trust and motivation in treatment negatively.

1.1 Realistic expectations.

Informing about possible negative effects was suggested to make the patient develop more nuanced and reasonable expectations of treatment, for instance by making it easier to discuss its unpredictability and limitations:  … trying to nuance the idea that a treatment will solve “everything.” [Respondent 146]

1.2 Increased participation.

Presenting information about possible negative effects was also thought to emphasize the importance of the patient’s active participation in treatment and how this in turn may influence what outcomes to expect:  … include the patient in his care -> increases the patient's sense of self-efficacy? Make the patient an active participant in the treatment. [Respondent 214]

1.3 Mental preparedness and resilience.

Providing information about possible negative effects was also perceived to make the patient less surprised and more mentally prepared for any eventual negative incident that might arise, as well as making the patient more resilient: Joint working alliance, motivation and resilience against potential aspects of deterioration. [Respondent 205]

1.4 Normalizing initial negative emotions and experiences.

Many respondents argued that informing about possible negative effects may normalize initial negative emotions and experiences, such as increased anxiety, and that patients might feel worse before feeling better: Just as with the introduction of SSRI, it is important to convey that a temporary increase in symptoms, behavioral changes, that exposure is usually difficult, can make you feel worse before you feel better. [Respondent 33]

1.5 Decreased motivation and trust in treatment.

Other respondents however believed that presenting information about possible negative effects would decrease the patient’s motivation, hope, compliance, and trust in the treatment, which in turn could lead to dropout: It might prevent motivation to reach that optimal level if you do no not have confidence in treatment. [Respondent 297]

2. Increased focus on negative effects.

The theme Increased focus on negative effects concerned some of the reservations respondents had about informing patients about possible negative effects, particularly the assumption that it would lead to unwelcome attention of certain incidents in treatment.

2.1 Detecting negative effects.

Albeit relatively rare, a few respondents believed that presenting information about possible negative effects could increase monitoring and thus the likelihood of the patient noticing adverse and unwanted events during the treatment process: It is good if the person is made aware of the potential risks involved and understands if they have been negatively affected. [Respondent 28]

2.2 Excessive worry.

It was expected that sharing information about possible negative effects might increase the patient’s worry about experiencing such incidents, especially those with anxiety-related problems: That they exaggerate the negative effects, for example due to an anxiety problem, which is what they are seeking help for. [Respondent 239]

2.3 Negative impact on expectancy.

Several respondents also thought that increased knowledge about possible negative effects could turn out to be a nocebo (i.e., the expectation by a patient that treatment may produce more negative than positive results): Of course, in some cases this could make the patient to look for symptoms of deterioration + nocebo. [Respondent 232]

3. Improved collaboration.

The theme Improved collaboration was referred to by many respondents as the advantage of sharing information about possible negative effects, which would improve collaboration between the patient and therapist.

3.1 Increased trust in the therapist.

Transparent and honest sharing of information about possible negative effects was expected to increase the patient’s trust in the therapist and strengthen the therapeutic alliance: Being open about the fact that a treatment is not suitable for everyone, that it is not certain that it will help, to prepare for and to have talked about what we do then - instills trust and increases trust. [Respondent 49]

3.2 Increased reporting and handling of negative effects.

It was also suggested that talking about possible negative effects could increase the patient’s willingness to honestly report such incidents. Moreover, it was suggested that presenting information about possible negative effects would help patients and therapists collaboratively plan for and deal with adverse and unwanted events, should they occur.  … that they are more sincere in verbal reports and estimates, because they know that we are trying to manage and make a plan together. [Respondent 5]

4. Professional and ethical responsibility.

The theme Professional and ethical responsibility concerns the idea of sharing information about possible negative effects as one’s duty and obligation as a therapist. Respondents described that patients could only make a fully informed decision about commencing treatment if they have received adequate information about what to expect. This was often compared to informing about possible side effects of pharmacological treatment: So that patients can make an informed choice as to whether they want to start treatment or not, so that the timing is right and you have good conditions to cope with treatment. [Respondent 40]

Domain: hindrances

1. Lack of knowledge.

Several respondents referred to a lack of knowledge and training in informing about possible negative effects when discussing potential obstacles to raise the issue with patients. They felt uncertain how information about such incidents should be presented and that it was hard to find a balance between instilling hope and being completely transparent. A few respondents also described a lack of clarity in the research on possible negative effects, as well as a lack of information about the issue in many clinical textbooks: I feel that much of my work is from what I learned during my educational program and a lot of the books I bought then are the same ones I use now, or later updated copies. These lack information about negative treatment effects …  [Respondent 307]

2. Fear.

Many respondents felt that discussing possible negative effects with patients was associated with negative emotional states, such as anxiety, fear, and being uncomfortable. Several respondents were also fearful of causing the patient more suffering, should the issue be brought up. Respondents tried to deal with their fears differently, some by avoiding the topic altogether. Fear in me of dropout …  [Respondent 14]

3. Lack of time.

A few respondents mentioned a lack of time as preventing them from providing information about possible negative effects. A high workload with little room for lengthy conversations about what to expect from treatment refrained them from raising the issue: The time limitation prevents me from always being able to prioritize discussing the negative effects of treatment …  [Respondent 301]

4. Patient characteristics.

Several patient characteristics were brought up as factors preventing respondents from informing about possible negative effects, such as younger patients, cognitive difficulties, short attention span, greater tendency to worry, low motivation, ambivalence, and having a tendency to think in black-or-white. Additionally, patients with more severe conditions, co-morbidity, autism, and high risk of suicide also prevented some respondents from sharing information about the topic: In a patient group with severe symptoms and high co-morbidity, it is sometimes difficult to work according to a set method. Also difficult with diagnostics. This makes it difficult to talk about treatment effects …  [Respondent 289]

Discussion

The present study surveyed practicing clinical psychologists in Sweden with regard to their attitudes and practices of informing patients about possible negative effects of psychological treatment. Most of the respondents (74%) reported that they do provide such information, a number much greater than the estimate by Sarkozy (2010), where only half of the American clinical psychologists endorsed a similar question. Issues concerning clinical training, practice, as well as differing ethical guidelines and codes of conduct might explain this difference. However, it could also be an effect of the topic of possible negative effects of psychological treatment gaining ground during the last decade (Rozental et al., 2018). Additional research on the practices surrounding what information is disclosed to patients during the informed consent procedure should be made to understand this process better, including whether or not this is also affected by social desirability effects and norms which might differ between countries.

None of the four hypotheses formulated a priori or the post-hoc analyses revealed a significant relationship between demographic variables and informing patients about possible negative effects. This goes in line with prior findings (Sarkozy, 2010), with the exception of gender, where male therapists were previously found to affect the tendency to provide such information to patients. More research is needed to verify these results, but it might be the case that such aspects as years in practice, therapeutic orientation, age, and gender are less relevant when it comes to sharing this information. Meanwhile, other factors may play a greater role and need to be studied further, such as personality traits, which in some studies have been shown to have a small yet significant impact on treatment processes and outcomes (Fletcher & Delgadillo, 2022). The present study did however find a few trends that might facilitate additional research. First, more years in practice seem to be positively correlated with addressing the issue with patients, indicating that greater clinical experience makes therapists better at discussing what patients might expect of psychological treatment. However, studies on the long-term development of therapists implies that increased clinical experience is not always associated with better outcomes for their patients (Goldberg et al., 2016), suggesting that it might be the quality rather than quantity of clinical experience that is most important for one’s continued development as therapist. Whether this is true also for addressing possible negative effects is unknown and warrants further study. Second, practicing PP might be related to informing patients about possible negative effects to a greater degree than CBT, but this should be interpreted with some caution. The number of therapists in the present study practicing CBT far outnumbered PP, despite several attempts at reaching out to Swedish organizations that provide clinical training, supervision, and continued education for the latter. Hence, those identifying themselves as PP in the present study may not be representative for therapists of this modality in general. Lastly, having received information oneself about negative effects during clinical training may be associated with a greater tendency to discuss this with patients. This is consistent with Sarkozy (2010), who found that those who underwent specific training in ethics and informed consent were more likely to address negative effects prior to commencing psychological treatment. This implies that research and practices surrounding possible negative effects should be an integral part of clinical training.

To compare the respondents’ understanding of the rates of deterioration, non-response, and improvement, three questions probed for their estimates. This illustrated a relatively good approximation on the number of patients faring worse (12.9%), close to the 5–10% often referred to in the literature (Boisvert & Faust, 2003; Hatfield et al., 2010; Lambert, 2010, 2013). For patients not responding, the estimate in the present study was however in the lower range (29.8%), where research suggest that 30–50% are expected to remain unchanged (Lambert, 2010; Loerinc et al., 2015; Rozental, Andersson, et al., 2019). This indicates an overestimation when it comes to the improvement rate (58.4%). Sarkozy (2010) posed a different set of questions in relation to this, where 30.3% disagreed with the statement that 10% of patients fare worse. Nevertheless, therapists do seem to vary with regard to their understanding of the benefits and limitations of psychological treatment. Whether it is important to know the exact numbers is unclear, but it could be argued that a more accurate idea of how effective psychological treatment is may influence how aware therapists also are of its possible negative effects.

The thematic analysis resulted in two domains; Expectations and Hindrances. In terms of the former, respondents believed that discussing possible negative effects of psychological treatment could make patients more accustomed to and familiar with the treatment process, which in turn may lead to more nuanced and realistic expectations of what it pertains, in line with the findings by Bystedt et al. (2014). Addressing this issue with patients was also perceived as positive for engagement and becoming mentally prepared, potentially increasing resilience should possible negative effects occur. In a similar vein, sharing information about incidents that might be interpreted as negative by the patient was assumed to normalize emotional responses to and experiences made during psychological treatment, such as discussing the possibility of feeling worse before feeling better. This could be important to address, since events that are conceived as negative, albeit temporary, might still be unpleasant, and in turn affect adherence and lead to worse treatment outcome and dropout (Crawford et al., 2016).

At the same time, respondents raised a concern of increased focus on possible negative effects if this was discussed together with patients. This was believed to affect both motivation and trust, which might cause some to refrain from commencing psychological treatment. Likewise, excessive worry about possible negative effects and expecting to experience such events were also raised, as in patients becoming more inclined to look for and notice incidents during the treatment period. In particular, respondents considered this to be a problem for patients with anxiety disorders, who may overestimate the risks of certain interventions and dismiss their benefits. One way of overcoming these issues would be to implement expectation-management strategies, as in a study on oncological treatment by Nestoriuc and Gerke (2021). Informing patients about framing effects, explaining the concepts of placebo and nocebo, and developing individual coping strategies on how to manage an unwanted event helped generate more realistic and positive expectations of treatment.

Respondents also stressed that sharing information about possible negative effects may lead to improved collaboration between the patient and the therapist. Among other things, it was perceived to increase trust and improve the therapeutic alliance, as well as help the patient detect, report, and handle possible negative effects when they occur. A study by Muschalla et al. (2022) gives some credence to this notion, indicating that monitoring and discussing possible negative effects of psychological treatment with patients did not jeopardize the therapeutic alliance, but could rather improve it. Similar findings have been demonstrated in relation to pharmacological treatment (Jose & AlHajri, 2018), where information did not affect the occurrence or reporting negative effects. This also corresponds to prior findings of patients being positive towards receiving information about the potential risks of psychological treatment (Braaten & Handelsman, 1997), suggesting there might be more advantages than disadvantages of conveying this during the informed consent procedure. Respondents in the present study believed this could help them set up a plan with patients to manage an unwanted event, thereby making them prepared for any uncomfortable experiences that could arise during psychological treatment. However, it should be noted that there are no general rules on when to present this information to a patient. Instead, this might have to be determined on a case-by-case basis and depending on the patient’s presenting problems.

In addition, many respondents conceived sharing information about possible negative effects as one’s professional and ethical responsibility. This was often compared to pharmacological treatment, where disclosure of side-effects is commonplace. This also goes in line with the Swedish National Healthcare Law (Patientlagen, 2014, p. 821), postulating that patients should understand the risks and benefits involved before initiating any type of treatment. This highlights the importance of providing instructions and tutoring regarding the informed consent procedure and what it should include already during clinical training. Moreover, ethical guidelines and codes of conduct might also have to be more explicit on this issue as negative effects of psychological treatments are seldom mentioned specifically. For example, the American Psychological Association (2016) currently mention that patients should be informed about “the potential risks involved … ,” but only in relation to treatments that are not yet established or undergoing development. In order to make this an integral part of the informed consent procedure, this should however be provided for any type of treatment, also those who are considered evidence-based.

As for the second domain, Hindrances, respondents referred to lack of knowledge, fear, lack of time, and the patient’s characteristics as factors preventing them from addressing possible negative effects of psychological treatment. Similar to Bystedt et al. (2014), lack of clinical training with regard to the topic seems to be an essential issue. This is not only important for universities providing educational programs in clinical psychology, but also emphasizes the importance of keeping oneself up to date with regard to research findings and pursuing deliberate self-practice long after one’s clinical training (Chow et al., 2015). Meanwhile, fear was mentioned by many respondents and seems to be related to the therapist’s own beliefs of how discussing possible negative effects would be received by the patient. This leads to negative emotional experiences on the part of the therapist, who then might choose to avoid addressing the topic altogether. A similar type of experiential avoidance can also be seen in CBT for obsessive-compulsive disorder, where the therapist’s own fear and avoidance behavior have been shown to impact the delivery of exposure and response prevention negatively (Scherr et al., 2015). Again, this points to the importance of including information on how to discuss possible negative effects during clinical training, but also to practice self-reflection in order to identify and work with challenging emotional responses (Chigwedere et al., 2021). As for lack of time, some of the respondents raised the issue of seeing too many patients, thus not having any room for a lengthy conversation on what to expect in terms of possible positive and negative effects. This seems to convey more general problems of how psychological treatment is being administered in many sectors. Exactly how to deal with the demands imposed on many therapists is a difficult question, but reveals a conflict between delivering adequate care and the priorities set up by healthcare providers. Lastly, the patients’ characteristics were raised by some of the respondents as a factor precluding them from addressing possible negative effects. This included such issues as comorbidity, anxiety disorders, and cognitive impairment, which was also raised by Bystedt et al. (2014). However, whether or not these aspects would actually affect the patient’s ability to understand and make sense of this information remains an empirical question. So far, no study has demonstrated such a relationship, but further research is needed.

Clinical Implications

The present study suggests there are both positive and negative expectations about discussing possible negative effects, including factors that seem to refrain therapist from raising the topic in psychological treatment. Overall, there appears to be many benefits in having an open and balanced conversation with patients about what to expect, creating an opportunity to discuss any potential misconceptions. Sharing this information might also have a positive impact on trust and the therapeutic alliance, increase resilience and prepare the patient for unpleasant experiences, and make the patient better at detecting and reporting possible negative effects that may otherwise lead to dropout and worse treatment outcome. Meanwhile, most of the negative expectations and hindrances seem to revolve around beliefs about how the patient will react. Understanding the occurrence and characteristics of possible negative effects could therefore be an important during clinical training, as this would make therapists more correctly informed and perhaps less reluctant to share this information with patients. Clinical training should however not only convey basic estimates, but also include practice on how to discuss and manage questions regarding possible negative effects during the informed consent process, such as via role-playing. Several self-report measures probing for possible negative effects also exist and could help therapist address the issue with patients by administering them prior to and during the treatment period, e.g., the Negative Effects Questionnaire (Rozental et al., 2016, 2019). Lastly, addressing possible negative effects of psychological treatment should be included in professional policy documents, such as ethical guidelines and codes of conduct for clinical psychologists, as this would stress the importance of the topic to practicing therapists.

Limitations

When reviewing the results of the present study, several limitations need to be addressed. First, no definition of possible negative effects was presented in the online survey. Consequently, the respondents may have had a different conception of what is meant by possible negative effects when completing the questionnaire. However, providing a definition was deemed to impact the responses to the open-ended questions and was therefore left out. Second, social desirability could have influenced the responses, particularly with regard to such an important issue as possible negative effects. Given that AR has researched the topic for many years and discussed it in forums run by the Swedish Psychological Association, this may have affected the respondents’ perception of possible negative effects. Third, recruitment was made via advertisements, and although the number of respondents was greater than prior studies (Sarkozy, 2010), more therapists practicing CBT than PP were included, limiting the generalizability of the results. Furthermore, most quantitative questions had a response rate of 277, as compared to only 82 for some of the open-ended questions. This might be due to declining motivation and interest in the online survey as the questionnaire required written responses, but it could also be attributed to some of the open-ended questions being perceived as too similar and thus repetitive. Lastly, the thematic analysis was mostly performed by FW, which may have affected the results. Including several researchers in the coding process and subsequent thematization is sometimes used to get a richer and more nuanced analysis (Braun & Clarke, 2021), but not a requisite. Instead, the concept of trustworthiness is more important to consider. In the present study, including information about recruitment, providing the open-ended questions, and presenting quotes increase the findings credibility and transferability. As for dependability and confirmability, relating the results to prior research and being transparent about the research process also help to make the present study trustworthy. However, additional steps would have been to triangulate the findings by, for instance, exploring patients’ conception of possible negative effects.

Ethical Statement

The present study included an anonymous online survey and did not collect any sensitive or personal information. The respondents provided informed consent online prior to responding to any questions.

Acknowledgements

The authors would like to thank George Vlaescu for his help with the online survey, and the Swedish Psychological Assocation for providing demographic information about their members. Both the clinical division of the Swedish Psychological Association [Sveriges Kliniska Psykologers Förening], and the Swedish Association of Behaviour Therapy [Beteendeterapeutiska föreningen] are thanked for their help in advertising the present study. Author contributions: The idea behind the present study was drafted by AR, who also acted as principal investigator. FW and AR developed the survey. FW advertised the study and managed the data collection. Quantitative analyses were made by AR, while FW conducted the qualitative analyses under supervision of AR. Both FW and AR authored the manuscript.

Disclosure Statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

The present study did not receive any funding.

Appendix

1. Gender (n = 277)

2. Age (n = 277)

3. Therapeutic orientation (n = 277)

4. Years of experience as a clinical psychologist (including practical training) (n = 277)

5. At what university did you get clinical training? (n = 277)

6. What year did you complete your clinical training? (n = 277)

7. Have you undergone any further education? (n = 277)

8. What is your primary workplace? (n = 277)

9. Approximately, how many patients do you see for psychological treatment in a typical work week? (n = 277)

10. When starting psychological treatment with a patient, do you inform about the possible results of the treatment? (n = 248) In what way? (n = 174)

11. Do you inform your patients about the possible positive outcomes of treatment, such as symptom improvement? (n = 234) In what way? (n = 172)

12. Do you inform your patients about the possible negative outcomes of treatment, such as worsening of symptoms? (n = 231) In what way? (n = 151)

13. Do you think there could be any advantages to informing your patients about the possible positive effects of the treatment? (n = 216) What are these? (n = 182)

14. Do you think there could be any disadvantages to informing your patients about the possible positive effects of the treatment? (n = 216) What are these? (n = 127)

15. Do you think there could be any advantages to informing your patients about the possible negative effects of the treatment? (n = 216) What are these? (n = 172)

16. Do you think there could be any disadvantages to informing your patients about the possible negative effects of the treatment? (n = 216) What are these? (n = 126)

17. Do you think something is preventing you from discussing possible negative effects of treatment with your patients in connection to starting treatment? (n = 216) What might this be? (n = 82)

18. Approximately, what proportion (0-100%) of all patients undergoing psychological treatment do you think improve? (n = 211)

19. By estimate, what proportion (0-100%) of all patients undergoing psychological treatment do you think deteriorate? (n = 211)

20. Approximately, what percentage (0-100%) of all patients who undergo psychological treatment do you think remain unchanged? (n = 211)

21. During your clinical training (the psychology program or equivalent), have you learned about the occurrence and characteristics of negative effects of psychological treatment? (n = 210) In what way? (n = 60)

22. Is there anything you would like to add? Please tell us or press the “Submit” button below to save your answers and complete the study. (n = 42)