Unexpressed Wishes and Unmet Needs: a Mixed Methods Study of Youth Negative Experiences in Psychotherapy

Unlike best practice guidelines for detecting and addressing medication side effects, there are no widely adopted guidelines to help therapists identify and address the possible negative effects of psychotherapy. Psychotherapy is a first-line evidence-based intervention for youth depression and anxiety, but the literature contains few studies of psychotherapy’s potential negative effects, and even fewer direct youth reports of their negative psychotherapy experiences. A mixed-methods study design was used to study youth with self-reported depression and/or anxiety and measure negative experiences associated with psychotherapy treatment. Youth were recruited from a larger study. Qualitative and quantitative data were collected from 45 youth with psychotherapy experience. Study design was informed by feedback from youth with lived experience of mental health challenges. Youth completed standardized measures of psychopathology, treatment effects, and detailed qualitative interviews. All youth experienced at least one negative effect during psychotherapy, and 30 youth (66.7%) attributed a negative effect as ‘caused’ by psychotherapy. While this attribution did not significantly differ based on severity of underlying current psychopathology, effect size estimates suggest moderate to large group differences regarding internalizing issues and baseline coping skills. Four major qualitative themes were identified from the interviews, summarizing the negative experiences of psychotherapy: (1) Barriers Experienced in Psychotherapy; (2) Concerns about Therapist; (3) Concerns about Therapy Itself; (4) Negative Experiences as a part of the Therapeutic Process. No youth raised any of these concerns directly with their therapists. These data support the need for validated, systematic assessment of psychotherapy’s negative effects among youth, with the aim of improving engagement and treatment outcomes. This study is one of the first with direct reports from youth about their negative experiences in psychotherapy, which is a growing area of study in the psychotherapy literature. Youth have negative experiences in psychotherapy at high rates, but do not report them to their therapists. Negative experiences associated with psychotherapy need to be systematically anticipated, measured and addressed by therapists, or they may be interpreted by youth as symptom worsening or ineffective treatment, leading to treatment discontinuation. Using principles of “youth friendliness” to inform service provision could improve psychotherapy uptake and retention and decrease negative experiences. This study is one of the first with direct reports from youth about their negative experiences in psychotherapy, which is a growing area of study in the psychotherapy literature. Youth have negative experiences in psychotherapy at high rates, but do not report them to their therapists. Negative experiences associated with psychotherapy need to be systematically anticipated, measured and addressed by therapists, or they may be interpreted by youth as symptom worsening or ineffective treatment, leading to treatment discontinuation. Using principles of “youth friendliness” to inform service provision could improve psychotherapy uptake and retention and decrease negative experiences.

This study is one of the first with direct reports from youth about their negative experiences in psychotherapy, which is a growing area of study in the psychotherapy literature.
• Youth have negative experiences in psychotherapy at high rates, but do not report them to their therapists.
• Negative experiences associated with psychotherapy need to be systematically anticipated, measured and addressed by therapists, or they may be interpreted by youth as symptom worsening or ineffective treatment, leading to treatment discontinuation.
• Using principles of "youth friendliness" to inform service provision could improve psychotherapy uptake and retention and decrease negative experiences.
Depression is a serious mental illness affecting over 7% of youth (Leanne, 2017), and a leading cause of youth disability worldwide (Gore et al., 2011). Depression and anxiety disorders combined may affect over 13% of youth (Comeau et al., 2019). Along with being associated with suicide, which is the second leading cause of death in youth (Statistics Canada, 2015), youth depression often has lifelong negative health impacts, increasing the risk of adult mental illness and substance use disorders, problems in relationships, lower educational achievements, and increased co-morbid physical illness (Fletcher, 2010;Hammen et al., 2011;Hasler et al., 2005;Thapar et al., 2012). Effective interventions in youth depression and anxiety are therefore essential to address immediate suffering and to improve long-term health outcomes (Keenan-Miller et al., 2007).
Evidence-based treatments for youth depression and commonly co-occurring anxiety disorders include psychotherapy and antidepressant medications (Goodyer & Wilkinson, 2019), and psychotherapy can be a highly effective treatment. Unfortunately, effective youth psychotherapy remains hampered by issues of access, adherence and engagement. Premature drop-out rates have been measured as ranging from 28-75%, but there is scant research about why youth have such high rates of drop-out from psychotherapy (de Haan et al., 2013;O'Keeffe et al., 2019). Given the serious consequences of untreated mental illness in youth, more needs to be understood about youth's experiences of the therapy itself (Baruch et al., 2009).
Best practice guidelines for medication use to treat youth depression and anxiety indicate that practitioners should proactively discuss, identify and address medication side effects (Dodd et al., 2011;Gualtieri & Johnson, 2006;NICE, 2018;Psychiatrists TRAaN-ZCo, 2005), but there are currently no comparable expectations and guidelines for psychotherapy (Herzog et al., 2019;Ryan, 2005). Although the terminology of "side effects" is not typically associated with psychotherapy, it is nonetheless an active intervention which may produce both positive and negative effects. The "negative effects" of psychotherapy can be defined as those that cause an increase in symptoms or distress, or a deterioration in functioning (Linden, 2013). Measuring negative effects is complex, since they can be due not only to the treatment, but also to personal factors associated with the therapist, client or setting (Linden, 2013) and negative effects may arise even within a correct treatment (Linden & Schermuly-Haupt, 2014). Negative effects can also be difficult to distinguish from the typical variation seen in the index disorder over time. Negative effects may also be more likely in populations with more severe psychopathology, raising questions of whether the effects are in fact due to underlying illness issues such as negative cognitive biases or reduced distress tolerance, rather than to the psychotherapy treatment (Rheker et al., 2017). Due to all these factors, distinguishing whether a negative experience had by a client is in fact a negative effect caused by a problem with the treatment continues to be a central problem of this literature, and there is currently no consensus or widely-adopted framework (Herzog et al., 2019). Distinguishing between negative experiences (which might even be anticipated and therapeutic, such as transiently increased anxiety during exposure-based psychotherapy) and negative effects (where the psychotherapy process is causing harm to the client) is therefore highly complex, and currently understudied, particularly in youth.
Despite being included alongside medication in evidence-based treatment guidelines, much less has been studied about psychotherapy's possible negative effects (Barlow, 2010;Berk & Parker, 2009;Rozental et al., 2016). There have been growing calls in the adult psychotherapy literature for systematic assessment of potential negative effects, (Berk & Parker, 2009;Gualtieri & Johnson, 2006;Linden & Schermuly-Haupt, 2014) rather than relying on clients to discern and articulate these experiences. Raising concerns about negative effects is challenging due to clients' emotional vulnerability, the perceived power differential, and the concurrent social and cognitive impacts of mental illness. These challenges are likely heightened in youth, due to their developmental stage, which might inhibit them from confronting adult authority, or from easily articulating complex emotional experiences (Garber et al., 2016). Youth are also often closely embedded within their families, so psychotherapy's negative effects could also extend to familial tension about confidentiality or decisionmaking, feelings of guilt or shame among parents, or conflict between parents and clinicians (Jonsson et al., 2014).
Negative effects of psychotherapy for adults are reported at a prevalence ranging from 3-20% of clients (Berk & Parker, 2009;Linden, 2013;Mohr, 1995;Schermuly-Haupt et al., 2018). These include worsening or lack of improvement of symptoms, development of new symptoms, social difficulties such as increased family distress, stigma, and dependency on the therapist (Berk & Parker, 2009;Linden, 2013;Rozental et al., 2015;Schermuly-Haupt et al., 2018). These effects have been reported across a variety of disorders (Barkley, 2018;Moritz et al., 2015;Parker et al., 2013;Rozental et al., 2015) and for different modalities of therapy (Parker et al., 2013;Roback, 2000;Rozental et al., 2015;Schermuly-Haupt et al., 2018) with qualitative and quantitative findings. The youth psychotherapy literature contains only a few studies reporting on negative effects, without comparable detail and data sources and with no direct youth reports (Barkley, 2018;Jonsson et al., 2016;O'Keeffe et al., 2019;Warren et al., 2009). As clinician perspectives may not recognize all effects that youth experience (Jonsson et al., 2014;Schermuly-Haupt et al., 2018), an accurate and person-centered perspective on the negative effects of psychotherapy should include youth reports (Hawke et al., 2018).

Current Study
As a first step, and to generate hypotheses in this understudied area, the aims of the current study were to: 1. quantify the extent to which youth with depression and/or anxiety had negative experiences associated with psychotherapy and whether they believed them to be caused by the therapy, and 2. through qualitative interviews, understand more about the meanings and impacts on youth of these negative experiences. To achieve the study aims, an equal status mixed-methods design (sometimes referred to as interactive design) was applied (Greene et al., 1989;Schoonenboom & Johnson, 2017), whereby the quantitative and qualitative components were purposefully integrated to best understand the possible negative effects of psychotherapy among youth (Bryman, 2006;Greene, 2015). Using mixed methods also increases the likelihood of nuanced findings in this understudied area, expanding the findings beyond the scope of one measurement approach alone (Creswell, 2009).
Youth in the current study were recruited from a larger school-based cohort sample participating in the Research and Action for Teens (RAFT) project (Brownlie et al., 2019), a longitudinal study of the onset and trajectory of mental health issues among youth. The current study is distinct from the larger RAFT project, with a unique focus on psychotherapy experiences. The youth in the current study, although recruited from RAFT, completed new survey and interview questions independent from the RAFT measures, to create a unique dataset.

Participants
Participant inclusion criteria were a history of mood or anxiety issues on self-report standardized instruments and at least one previous experience of psychotherapy. To be eligible, youth had to have a previous score of 1 or higher on the Global Appraisal of Individual Needs-Short Screener (GAIN-SS) Internalizing Disorder subscale (i.e., moderate likelihood of meeting criteria for a diagnosis in the internalizing domain), or a score of 12 or higher on the the Center for Epidemiologic Diseases Depression Scale-12 (CESD-12; i.e., meeting screening criteria for depression). 105 symptomatic youth were recruited, and of these 53 had received previous psychotherapy and were included in the study and administered the self-report questionnaires. Approximately 10% (n = 5) did not answer any survey items, and an additional 5% (n = 3) were removed due to answering "I prefer not to answer" on more than 50% of the items. Thus, the final quantitative sample consisted of 45 youth. The youth were 19.13 years old on average (SD = 1.44, ranging from 16-22 years), including 7 boys/young men, 36 girls/young women, and 1 young person with a transgender/non-binary gender identity (Table 1). 35 youth were Caucasian and 10 were racialized.
Based on qualitative research standards, a sample size of 20-30 youth was considered sufficient to gather representative qualitative information (Mason, 2010). Of the 105 youth with mood/anxiety issues who provided quantitative data, 60 volunteered to participate in interviews, of whom 33 had experienced psychotherapy. One interviewed youth subsequently reported no negative experiences in psychotherapy, so the remaining 32 interviews were analyzed, and saturation was reached.

Procedure
Participants were recruited by email from a pool of previous RAFT study participants who had consented to be contacted about future research. Interested participants then received a phone call in which study staff described the study purpose and procedures. Written informed consent was obtained from all participants, in accordance with Research Ethics Board guidelines. More details on study procedures are provided in Mehra et al. (2021).
In addition to a demographic questionnaire, a battery of self-report questionnaires comprised of standardized measures was administered using the online survey platform REDCap. A subset of those who answered the online survey participated in an in-depth semi-structured telephone interview to obtain qualitative data. The online survey and interview were administered independently to maintain confidentiality. The study was approved by the research ethics board of the Center for Addiction and Mental Health.
Mixed methods (quantitative and qualitative measures) were used to capture negative effects, such that the quantitative survey data identified the frequency and type of negative effects experienced by youth, and the qualitative data revealed more about the meaning and impact on youth of these experiences.

Quantitative instruments
Negative effects The Negative Effects Questionnaire (NEQ) is a 32-item scale validated in adults to capture quantitative data about the negative effects of psychotherapy (Rozental et al., 2016;Rozental et al., 2019). Respondents of the NEQ endorse specific negative effects and whether the effect was attributed to "the treatment I received" or attributed to "other circumstances". For each negative effect endorsed, respondents also rate its severity on a four-point Likert scale of "how negatively it affected me". The authors of the NEQ also conducted a factor analysis of its items, identifying six subscales of possible negative experiences associated with psychotherapy treatment (symptoms, quality, dependency, stigma, hopelessness, and failure (Rozental et al., 2016). The NEQ authors also note that these six subscales can be further conceptualized into two broad categories: individual-level effects and therapy-level effects. Individual-level effects are defined as mental health symptoms that worsen or emerge during treatment, and therapy-level effects are defined as pertaining to the treatment itself. In the current sample, Cronbach's alpha, a measure of scale internal consistency, was 0.91 suggesting the scale items aggregate as expected and therefore can be interpreted with confidence.
Mental health and substance use Youth completed the Global Appraisal of Individual Needs-Short Screener (GAIN-SS; (Dennis et al., 2006), a validated self-report screener assessing mental health and substance use concerns. There are four GAIN-SS subscales, each measuring a different facet of psychopathology, including: internalizing scale, externalizing scale, substance use scale, and crime/violence subscale. Clinical cut-off scores are calculated for each subscale. The current study reports the proportion of youth who meet the clinical cut-off for each scale. The Cronbach's alpha for the GAIN-SS subscales range from 0.60 to 0.80, suggesting moderate to good scale reliability across the measure.
Youth depression The Center for Epidemiologic Diseases Depression Scale-12 (CESD-12 (Poulin et al., 2005), is a validated measure of youth depression. In the current sample, the scale had good reliability with a Cronbach's alpha of 0.83.

Emotion regulation
The coping strategies subscale of the Difficulties in Emotion Regulation Scale (DERS; (Gratz & Roemer, 2004), is a validated youth measure of Notes. a Positive Cohen's d indicates youth who reported negative effects had more characteristics that were problematic (e.g., not in school) or uncommon b Chi-2 analysis conducted only on boys/young men and girls/young women cells due to small sample size c One cell with expected cell count less than 5, therefore must interpret with caution emotion regulation difficulties. In the current sample, the subscale had very good reliability with a Cronbach's alpha value of 0.90.

Qualitative interview
The interview protocol was developed by the research team and is available upon request. The interview included open-ended questions about the types of psychotherapy, number and frequency of sessions, confidentiality, stigma, relationship with therapist, activities or homework assigned, and positive and negative experiences of psychotherapy. In the interviews, "negative effects" were any statements where youth explicitly identified an experience as both negative and directly related to their psychotherapy. Examples include statements where the participants were adversely affected by the therapist's approach, or type of therapy; experienced a worsening of their symptoms; and/or a decline in adherence to treatment or attendance in therapy.

Youth feedback on measures
Three Youth Engagement Facilitators were consulted . These are young people with lived experience of mental health challenges who regularly advise on research processes. They suggested revisions to clarify instructions without changing the validated questionnaires; re-worded the demographic form to minimize the potential for youth to feel judged for disclosing sensitive information (Hawke et al., 2018); and gave feedback about the phrasing of interview questions. The interviewer also took part in a training on using youth-friendly language and approaches to conducting interviews.

Quantitative analysis
NEQ item-level (frequency, mean severity) and factorlevel data were analyzed. Youth who attributed any negative effects to their psychotherapy were compared to those who did not attribute negative effects to psychotherapy, across all measures of psychopathology. Standard independent sample t-tests were conducted to compare groups across the CESD-12 and the DERS measures. GAIN-SS scores were dichotomized according to the clinical cut-off score for each subscale, and chisquare statistics were conducted to compare groups. We also calculated Cohen's d to examine the magnitude of difference (i.e., effect size) between the groups (Cohen, 1988). While the statistical comparisons tell us whether our groups significantly differ, they do not speak to the magnitude of difference between our two groups. Given our smaller sample size, it is possible the statistical comparisons may not reach significance, while our groups still differ in terms of degree of psychopathology experienced. As such, we calculated effect sizes which speak to the size of difference between our groups, independent of statistical significance. Effect sizes were calculated so that a negative d-value indicates worse psychopathology on the part of youth who attributed negative effects to psychotherapy. As a general rule of interpretation, an absolute value of Cohen's d from 0.0 to 0.29 is considered negligible, 0.30 to 0.49 is considered small, 0.5 to 0.79 is considered moderate, and greater than 0.8 is considered large (Cohen, 1988).

Qualitative analysis
The qualitative analysis was informed by an interpretive phenomenological approach (Braun & Clarke, 2006), which uses multiple forms of data collection, such as our use of surveys and interviews, to help describe and interpret the meanings participants ascribe to specific experiences. The phenomenon being analyzed in this study are youth's perceived negative experiences of psychotherapy.
Thirty-three youth interviews were recorded and transcribed. Using NVivo 12 Pro, three study authors conducted a deductive thematic analysis to identify patterns, develop codes and form themes (Braun & Clarke, 2006). To establish reliability and a multi-coder analysis agreement, two primary coders (E.L. and M.C.) worked independently to generate initial codes, then reviewed their coding together with the first author (P.W.) to establish a code book and receive interpretive support. An audit trail was kept to document the rationale for any coding changes (e.g., code reassignment, regrouping, or renaming). Coders used reflective diaries to ensure that assumptions underpinning the reading of data were acknowledged. Quotes from participants describing negative experiences were coded verbatim as descriptive initial codes, and subsequently merged under "mid-level" codes that best summarized the concepts (Braun & Clarke, 2006). Insights from quantitative data were used to inform the grouping of overarching qualitative themes (Braun & Clarke, 2006).

Quantitative Results
Negative effects of psychotherapy Sample demographics are reported in Table 1. All 45 youth reported at least one negative effect, and 30 youth (66.7%) attributed at least one negative effect specifically to their psychotherapy. Although increased stress, anxiety, unpleasant memories, and fear that people would find out about treatment were the most frequently reported individual-level negative effects (Table 2), only increased stress and unpleasant memories were most frequently attributed to therapy, along with increased perception that the problem could not get better and decreased selfesteem. For therapy-level negative effects, the most frequently endorsed were also the most frequently attributed to therapy problems: treatment did not produce results, did not have confidence in treatment, treatment expectations were not fulfilled, and did not understand the treatment (Table 2).
In terms of NEQ factor scores (Table 3), among those youth who attributed negative effects to their psychotherapy (n = 30), the "symptoms" factor was most endorsed, but the mean severity was highest for the "hopelessness" factor. Thus, while fewer youth endorsed items related to hopelessness, they experienced hopelessness as more severe. Within therapy-level effects, the "quality of treatment" factor was most endorsed, and has the highest mean severity rating, indicating that youth found treatment quality to be of the greatest concern.

Association with psychopathology
Youth who attributed negative effects to therapy were compared to youth who did not attribute negative effects to therapy, across all measures of psychopathology (Table 4). There were no significant differences between groups of youth on the CESD-12 depression measure, the coping DERS measure, or GAIN-SS score.
Overall, the pattern of effect size findings was variable. A small positive effect size (d = 0.36) was observed for depression severity. In contrast, youth who attributed negative effects to therapy reported worse psychopathology when it came to internalizing issues with a moderate effect size of d = −0.54. This suggests a moderate degree of difference in the magnitude of internalizing issues between the groups, with a greater proportion of youth who attributed negative effects to psychotherapy meeting the clinical internalizing cut-off. There was a large degree of difference in the magnitude of coping ability (d = −0.93) between the two groups, with youth who attributed negative effects to psychotherapy scoring worse.

Qualitative Results
There were four major themes that summarized the negative effects of psychotherapy as conveyed by youth: (1) Barriers Experienced in Psychotherapy; (2) Concerns about Therapist; (3) Concerns about Therapy Sessions; (4) Negative Experiences as a part of the Therapeutic Process. All themes developed are reported in the results section, and the qualitative data of all 32 youths were included in one or more theme. Most youth could not identify what specific therapy modality they had received (26/32), though three youth identified Cognitive Behavioral Therapy, two identified Dialectic Behavior Therapy and one youth identified family therapy.
The number of youth who contributed to each overall theme and subthemes were as follows: Concerns about therapist (n = 30), Communication/rapport problems (n = 27), Therapist's Stance (n = 20), Therapist's Identity (n = 6); Concerns about therapy itself (n = 26), Therapy process (n = 26), Clinical environment (n = 5); Barriers experienced in psychotherapy (n = 20) Access (n = 17), Stigma (n = 17); and Negative experiences as a part of the therapeutic process (n = 20), Tolerating distress with a desire to recover (n = 10), Improvement over time (n = 8), Strong therapeutic alliance (n = 6). Concerns about therapist was the most discussed theme, with 274 references made. Concerns about therapy itself and Barriers experienced in psychotherapy were referenced nearly equally with 123 and 116, respectively. While the fourth theme, Negative experiences as a part of the therapeutic process, was referenced the least by youth (35 references), it was emphasized by youth in lengthy and detailed quotes and therefore retained in the reporting (Greenhalgh & Taylor, 1997).

Barriers experienced in psychotherapy
The first major theme, Barriers Experienced in Therapy, was comprised of two 'mid-level' codes or subthemes: Access and Stigma.
Access Access problems were not restricted to finding therapy, but also to maintaining therapy once it was underway. Youth described significant challenges including ongoing costs, difficulty making appointments/ finding a therapist, limited hours or scope of services, and travel and wait times. Financial and time resources presented barriers for families too, as one participant explained: "I was worried about putting my parents out, either financially, or just causing them just one more thing to deal with…I was worried about it being a burden".
Age was also a noted barrier, with numerous youth noting that age-restricted services impacted their course of care: "It's not as easy as people make it seem-lots of places only take youth. I only got in because I was under 18. Within the year I was going to turn 18, so they got me in pretty quick, but I only had a year and then they shipped me on off to my family doctor". These factors were calculated using the scoring sheet associated with the NEQ tool (http://neqscale.com/negative-effects-questionnaire/dow nload/ These barriers were impactful throughout the course of therapy, affecting treatment engagement and effective service provision.
Stigma The second sub-theme regarding barriers had to do with the stigma associated with receiving therapy. Youth feared being seen as "delicate", "dependent" or "weak", or that others would "pity" them. Youth also reported concerns that family, friends, or school staff would become "overly supportive", "nosey", "judgmental", or dismiss their problems as "attention-seeking", making seeking psychotherapy more difficult.

Concerns about therapist
The second major theme reflected problems in the therapeutic alliance. Youth struggled to relate to their therapist's perceived identity; noted communication/rapport problems; and expressed unmet expectations and needs arising from the therapist stance. These all compromised the alliance.
Therapist identity Youth described challenging perceptions of therapist identity and values. One common preference expressed was for having a female therapist over a male, explained by some youth as feeling "safer" and more "comfortable", or a "mother-type model". Perceived differences in cultural background were also noted to decrease the engagement between therapist and client: "I just feel like we didn't really have that much in common […]. And at our cultural standpoint I guess maybe certain things I'd want to talk about, but I couldn't because there's that fear of them not understanding".
Communication problems Communication issues were the most frequently reported difficulty. Youth described communication omissions, such as therapists not explaining therapy homework's purpose, benefits, and applicability to real-life problems. These concerns around the framing of homework extended to the therapy itself, which youth often noted was not well-explained or rationalized. Youth described pressure to meet these unclear expectations despite competing time constraints and not understanding their purpose: "My counselor-she kind of did this thing where like I'd do some work and then I'd have to bring it in to show her-so that I could show her that I completed it or whatever…I felt like I had to have something done when I went in to see her, like it wasn't a relaxing experience. It felt kind of tense". A perception that therapists had rigid expectations around homework was a noted source of distress and disconnection, and reinforcing these expectations was interpreted as a lack of understanding or of caring by the therapist: "I think she was just really not-she was really not compassionate. (…) For someone who works in mental health, she was like 'You should be doing your homework earlier, or you should be going to bed earlier.' And it's like well, yeah, I'd be doing that if I didn't have mental health issues. So she didn't seem to get it or didn't care".
Negative therapist responses to incomplete therapy homework were described by various youth as "judgmental", "uncaring", "pushy" and "unsupportive". All 32 interviewees reported that their therapists never asked what type of clinical approach or therapy model might be a good fit for them. Youth conveyed how therapy became less effective as it became clear that these unexpressed wishes would remain unmet, and treatments were not adjusted in response to their elicited needs and concerns.
Therapist stance Youth described difficulties not just with specific therapy expectations, but also with the way in which therapists interacted with them, captured in the subtheme of Therapist Stance. Problematic therapist stances were "too scientific", "too textbook" or "too vague and not individualized". A therapist's clinical approach was unhelpful if it was too generic: "The counselor I saw, she honestly just sounded like she was regurgitating self-help sites on-line, like she was nothing of substance". Attunement problems arose when the therapist's affect was misaligned with the youth's: "She was just too happy all the time and she put on this mask like she was just happy-golucky, and I didn't really like that because she was happy when I was not, and I just felt like I couldn't talk to someone who was like that".
Youth variously perceived problematic therapist stances as "cold", "uninviting", "unengaging", "judgmental", and "patronizing". When therapists were experienced in these negative ways, youth reported that they disengaged from therapy because they were unable to open up, began to feel negative about themselves, or felt "anxious and sad" talking to their therapist in particular.

Concerns about therapy itself
The third major theme across the interviews was Concerns about therapy itself, containing two mid-level codes of Therapy processes and Clinical environment. Negative experiences of the expectations and course of therapy were tied to reported early drop-out and reluctance to re-engage in later treatment.
Therapy processes Some youth-reported stressors arose from the inherent demands of therapy itself, rather than any particular treatment deficit. These challenges made up a prevalent mid-level code among participants, including experiences of vulnerability when sharing and reflecting upon the psychological distress that brought them to therapy in the first place: "I think a lot of time how mental health issues start is you either, you know, subconsciously or consciously put stuff out of your mind that you don't want to think about-obviously when you bottle things up it's not good, but the therapist kind of made me lay it out in the open and that's what's challenging and scary".
Youth reported that disclosing sensitive details caused increased anxiety, self-critique, or feeling worse after sharing their problems; these negative experiences led at times to negative coping after therapy sessions: "I used to drink a lot after appointments. I would sometimes feel worse, just because it brought up all these things and then I would cope with it by going out or something, so-I think that was negative".
Youth noted that sharing in therapy could be "daunting", "uncomfortable", "awkward", and "stressful". When youth felt worse after sharing, they interpreted this as an indicator that the therapy itself was a problem Clinical environment Negative experiences also arose from the treatment setting and personnel, as described by one youth who called to cancel an appointment: "I called the day of and the secretary was just like really, really, really, rude to me and abrupt and almost seemed pissed off that I didn't cancel the appointment in the set time… that just spiked my anxiety, and I was like counseling is supposed to help my anxiety, but it's making me feel worse because I have to deal with all that".
The physical environment was also a source of concern; youth reported being "apprehensive" of sessions occurring in rooms with "no natural lighting", "no windows", and walls that were "plain", "bare" or "made you feel trapped".

Negative experiences as a part of therapeutic process
The fourth overarching qualitative theme contained a more positive understanding of how negative experiences might unfold. Although negative experiences might make therapy unhelpful and prompt discontinuation for some youth, multiple others noted that they came to understand these experiences as integral to the therapeutic process and recovery. These youth rationalized their negative experiences in three subthemes: Tolerating distress with a desire to recover, Strong Therapeutic Alliance, and Improvements over Time.
Tolerating distress with a desire to recover Here therapy was described as a "good struggle" and a "struggle that is worth it" to grow and heal towards a "great future": "At the beginning it might seem tough, but at the end of the counseling or at the end of sessions, even, you did feel a lot better, like that's how I felt anyway. I just felt more accomplished somehow and moreyeah more positive.
[…] I think for me, the biggest issue was I wanted to get better and I wanted to have a great future with nothing releasing me of my abilities, so I wanted to be able to seek help and get the help I need … so that I would be able to push myself further and achieve what I always really wanted…" Strong therapeutic alliance Youth relayed how their problems were eased by a positive view of the therapist and therapeutic alliance. Even when it "felt weird venting and unloading stress onto someone else", youth expressed how it "ultimately felt good" when a therapist was able to convey genuine concern and understanding: "It was not easy at all, only because some of the things were very private and were not mentioned ever-but [the therapist] was very calm and soothing, I guess in a way, and understanding. So it made that part easier, and he seemed concerned, so that made me feel better, because it was really genuine that way". Improvements over time Despite high rates of initial hesitancy, some youth noted that the relationship with their therapist improved over time, with the most acute challenges being "at the beginning, not really knowing what to expect going in there, and talking to someone for the very first time was uncomfortable, and then afterwards it got better".

Discussion
In this mixed-methods study, which integrated quantitative and qualitative data to understand the possible negative effects of psychotherapy, all youth endorsed negative experiences in psychotherapy, and two-thirds attributed at least one negative experience as caused by the psychotherapy itself. Qualitative data show that these attributions are prime reasons for discontinuation of treatment and symptom exacerbations (e.g., hopelessness). Distinguishing whether a negative psychotherapy experience is in fact a negative effect due to poor treatment or an unavoidable challenge inherent in the treatment process continues to be a central problem of this literature. There is currently no consensus on how this should be measured, and minimal studies on this question in youth (Herzog et al., 2019). Even if these negative experiences are not side effects due to poor treatment per se, the attribution made by youth is key to understanding their impacts. If a youth believes the negative experience is a sign that therapy is not working or is harmful, then this attribution will affect treatment success (Baruch et al., 2009;O'Keeffe et al., 2019). The qualitative data shows how the attribution of negative experiences impacts youth across the therapy timeline, spanning from pre-treatment engagement to in-session struggles to early discontinuation. Youth reported hesitancy to begin therapy due to concerns of whether the therapist would be nonjudgmental, trustworthy, and skilled. Similar youththerapist tensions arose during treatment and were ultimately cited as reasons for discontinuing therapy. Further, our qualitative data show that this negative attribution is not shared and processed with the therapist (32/32 did not share). When these effects are left undetected and unaddressed, youth may quit therapy or feel hopeless. The literature on youth mental health literacy shows that youth often are not equipped to understand the expectations and course of treatments (Gulliver et al., 2010;Radez et al., 2021). Few youth participants were able to identify the specific psychotherapy modality they received, and they noted that therapists often did not explain the purpose of specific treatment steps. Therefore, any negative experiences, even those deemed by the therapist as essential aspects of the therapy, need to be contextualized and explained in a dialogue between youth and therapist, at all stages of therapy, in order to prevent outcomes such as early termination or deterioration. While negative experiences may be unavoidable, attributing them to poor treatment and therefore dropping out, is not.
This attribution to treatment did not differ significantly between youth with higher or lower levels of psychopathology, indicating that attributing negative effects to therapy was not necessarily due to negative cognitive biases or low distress tolerance associated with low mood or anxiety, as has been indicated in the adult literature (Herzog et al., 2019;Rheker et al., 2017). However, moderate to large negative effect sizes were found for internalizing issues, externalizing issues, and coping skills, showing that youth who attributed negative experiences to psychotherapy had greater psychopathology on these measures. Variation in levels of psychopathology might therefore have influenced whether youth attributed a negative effect as being caused by the psychotherapy. However, other interpretations are possible. For instance, it is possible that youth who had negative experiences with psychotherapy may have terminated their therapy too soon, and the greater psychopathology observed reflects a worsening over time since therapy discontinuation. Establishing the mechanism between psychopathology and attributing negative effects to psychotherapy was beyond the scope of the current paper, and to our knowledge there are currently no published studies examining this question in youth. Our data identify this as an important avenue for future research.
These rates of negative experiences attributed to psychotherapy were higher than those reported in the adult literature, and are associated with concerning symptoms such as increased hopelessness. Importantly, youth did not attribute these effects to treatment equally, distinguishing certain ones as more likely due to the treatment itself, such as increased unpleasant memories. This finding makes sense, since therapy itself often requires the recall and discussion of previous unpleasant experiences; and it shows that youth could differentiate between negative experiences that arose in therapy, and negative experiences that arose because of therapy. For instance, youth attributed their increased difficult feelings to therapy at much lower rates than they attributed difficult memories. One notable exception, however, were the negative experiences of increased hopelessness: 20% of youth endorsed increased hopelessness, and 31% reported an increased perception that their problems could not get better, after the therapy process. When these items are combined in the hopelessness factor, almost all youth (n = 13/14) attributed this worsening specifically to the therapy they had received. As hopelessness is a predictor of recurrent and more severe depression in depressed youth and adults (Horwitz et al., 2017), this finding deserves special attention. One of psychotherapy's primary positive effects on depression may be the instillation of hope (Irving et al., 2004), so measuring and understanding this experience is essential. Intervention at a young age may improve longterm depression outcomes (Neufeld et al., 2017), but if youth feel more hopeless about their problems after therapy, and attribute this hopelessness to the therapy itself, then they are at increased risk for a worsened course of illness and possibly also less likely to seek further treatment. The hopelessness shown in the quantitative data is echoed in the qualitative themes of Communication Problems and Therapist Stance, where youth noted that a therapist unresponsive to their struggles with the therapy left them feeling that they "couldn't talk to someone who was like that".

Practical Implications
Sharing the burden of raising negative experiences Not one of the youth interviewed raised their concerns directly with their therapists, and so their experiences remained unshared but influential. Without explicit discussion there is no opportunity for these treatment dyads to reframe the challenges, clarify any confusion, and improve the treatment. However, requiring youth to bear the entire burden of raising these concerns is unreasonable, given the developmental challenges of articulating sensitive experiences to an adult who may be perceived as having greater power or expertise (Bear et al., 2019). Knowing how to understand and communicate a negative psychotherapy experience is additionally challenging due to the low mental health literacy found among many youth (Gulliver et al., 2010;Radez et al., 2021). Therapists must therefore be proactive in raising and discussing possible negative experiences early and throughout treatment, to ensure that they are processed and understood as part of a healing trajectory, rather than a negative effect of the therapy (O'Keeffe et al., 2019). This practical approach aligns with evidence-based standards of care in the administration of pharmacological treatments (NICE, 2018). To hold the same standards for psychotherapy treatments could improve high rates of drop-out and low rates of access and engagement in youth psychotherapy, and further increase its effectiveness.

Purposeful measurement of negative experiences
The ubiquity of reported negative experiences across youth suggests therapists must anticipate and elicit them from youth in an ongoing way. While many therapists believe that they do routinely detect and address their clients' negative experiences in therapy, numerous studies have shown they are often unaware of negative changes occurring (Hatfield et al., 2010) and may overestimate the benefits of the treatments they are providing (Walfish et al., 2012), thereby missing the reasons for drop-out and disengagement (Chen et al., 2017). Further, therapists may not be providing treatments with fidelity, or selecting treatments according to evidence-based guidelines (Watson et al., 2019). Therapists report that client symptom worsening is their most common cue for deterioration, but systematic and validated measures for detecting deterioration are not used in typical clinical practice (Barlow, 2010;Cuijpers et al., 2018).
In light of these findings, there have been increased calls for measurement-based care and continuous outcome monitoring (Clark et al., 2018); to meet this standard of care, therapists need to implement a deliberate practice to detect and address negative experiences (Chow et al., 2015), rather than relying on clinical judgment alone. Unfortunately, the youth psychotherapy literature on negative experiences and therapist assessment is even more sparse than the adult literature, so development of measures and standards to improve youth psychotherapy outcomes is urgently needed. In the interim however, a practical approach may be to administer the NEQ with youth on a semi-regular basis, using the results to assess and recalibrate the therapeutic approach to improve outcomes.

Creating youth-friendly spaces
Youth reported concerns regarding inflexible hours, agelimited services and setting concerns like unfriendly staff, inaccessible locations and "bleak" décor. Increasing youthfriendliness to improve service uptake and retention has become an international priority, including a proposed World Health Organization framework (Organization, 2012). Ensuring that treatment settings and routines adhere to the principles of youth friendliness can increase the likelihood of youth retention .

Strengths and Limitations
This study had several strengths and limitations, and points the way to future research questions. These data show that negative experiences are prevalent and impactful, providing initial data in an under-studied but highly important area of youth psychotherapy treatment. As the study was designed to be a first step in investigating this issue, the inclusion criteria were broad, admitting youth with anxious or depressive symptoms and any previous experience of psychotherapy. Therefore, it was not designed to analyze all factors which might contribute to negative experiences, such as variations across therapy modalities, fidelity in treatment provision, time elapsed since treatment, or treatment adherence by youth. As a retrospective study about subjective therapy experiences, the findings are also subject to recall bias. Verifying certain data, such as modality of therapy received or fidelity of the treatment delivered was therefore not possible. The small sample size limited the quantitative analyses, as it restricted the statistical power needed to sufficiently examine questions such as whether the attribution of negative effects might vary by variables such as gender, age or therapy type. There is currently no validated screening instrument for assessment of psychotherapy negative effects in youth, so no estimates of reliability and validity for the measurement tool are available for this population.
Strengths of this study were that it used a mixed methods approach to capture detailed data about an under-researched and poorly understood area of youth psychotherapy, with design input from youth with lived experience. Mixed methods are useful when trying to capture nuanced data about subjective experiences, and can strengthen the interpretations when the quantitative measure hasn't been validated in youth, as the qualitative data help provide a fuller understanding of the phenomena being described. To our knowledge, this is the first study of its kind capturing youth reports of negative experiences of psychotherapy, and it provides initial evidence that these experiences are an important concern among depressed and anxious youth, and merit further study.

Future Directions
Future directions in research should include the development of a validated screening instrument for use in studies of youth psychotherapy experiences. Future research could build on the current findings to conduct a larger examination of variables that may influence the onset and effects of negative experiences. Larger prospective studies might recruit participants with single mental disorders and no comorbidities, in order to evaluate whether negative effects vary by disorder. Prospective studies could also measure and track treatment fidelity and youth adherence, to examine questions of whether negative experiences vary by treatment type, and to ensure that the treatment being delivered is evidence-based. This would help distinguish whether the negative experiences are indeed adverse and unanticipated effects of therapy, or instead are difficult experiences that could be contextualized and addressed. Prospective designs would also allow for repeated tracking of negative experiences over the course of treatment, and measurement of whether explicitly addressing these experiences improves youths' treatment experiences, and ultimately their clinical outcomes.
Publisher's note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons. org/licenses/by/4.0/.