IJTAR 8: 1 Complete Issue

This study is the first of a series of seven, and belongs to the second Italian systematic replication of findings from two previous series (Widdowson 2012a, 2012b, 2012c, 2013; Benelli, 2016a, 2016b, 2016c) that investigated the effectiveness of a manualised transactional analysis treatment for depression through Hermeneutic SingleCase Efficacy Design (HSCED). The therapist was a white Italian woman with 8 years of clinical experience and the client, Anna, was a 33-year old white Italian woman who attended 16 sessions of transactional analysis psychotherapy. Anna satisfied DSM-5 criteria for mild persistent depressive disorder (dysthymia) with anxious distress. The conclusion of the judges was that this was a good-outcome case: the dysthymic symptoms improved over the course of the therapy and were maintained in the ‘healthy’ range at the 6-month followup, the client reported a positive experience of the therapy and described important changes in intrapsychic and interpersonal patterns. In this case study, transactional analysis treatment for depression has proven its efficacy in treating persistent depressive disorder.


Introduction
This study is the first of a series of seven, and belongs to the second Italian systematic replication of findings from two previous case series  and was conducted under the auspices of the European Association for Transactional Analysis (EATA) and the University of Padua.
Transactional analysis (TA) is a widely-practiced form of psyc are now what do hotherapy, supported with a vast literature (for a review see , but still it is under-recognised within the worldwide scientific community of psychotherapy. In order to define TA psychotherapy as an efficacious Empirically Supported Treatment (EST), its efficacy must have been established in at least one Randomised Clinical Trial (RCT) replicated by two independent research groups, or alternatively in at least three Single Case Experimental Design studies (SCED), replicated by at least two independent research groups, with each group conducting a case series of a minimum of three cases, without conflicting evidence . Recently, a wide community of researchers proposed that efficacy and effectiveness in psychotherapy are a complex object that cannot be adequately evaluated with either the experimental approach of RCT Westen, Novotny & Thomson-Brenner, 2004) or classical SCED (reverse or multiple baseline design) . Systematic case study research has been proposed as a viable alternative to RCT and SCED (Iwakabe & Gazzola, 2009). Considering that approaches without evidence from RCTs tend to be under-recognised, Stiles, Hill and  proposed collecting a series of mixed methods systematic single case studies as the first step toward recognition of marginalised and emerging models of psychotherapy.
Hermeneutic Single Case Efficacy Design (HSCED; Elliott, 2002;Elliott et al., 2009) is nowadays considered the most comprehensive set of methodological procedures for systematic case study research, and is a viable alternative to RCT and SCED in psychotherapy (McLeod, 2010). HSCED is gaining momentum with enhanced versions proposed by different research groups, to validate new psychotherapeutic approaches or extensions of previously validated psychotherapies for investigation into their effectiveness with other disorders (e.g. . Recently, a systematic review of all published HSCED studies found within English language peer-reviewed journals (Benelli, De Carlo, Biffi & McLeod, 2015) highlighted methodological issues related to different levels of stringency, offering solid alternatives to conducting sound research according to the available resources within practitioner research networks.
Systematic case study research has already been applied to investigate the effectiveness of TA for people with long term health conditions  and HSCED methodology has been successfully applied to TA and widely described in this Journal by . Recently, several HSCEDs supporting the effectiveness of TA treatment for depression  have been published, as was an additional adjudicated study which demonstrated effectiveness of TA for mixed depression and anxiety . Furthermore, a related study was published on TA for emetophobia (Kerr, 2013). The case series by Widdowson and Benelli have shown that TA can be an effective therapy for major depressive disorder when delivered in routine clinical practice, in private practice settings, with clients with mild to moderate impairment in functioning who actively sought out TA therapy and with white British and Italian therapist and client dyads.
The present study analyses the treatment of Anna, a 33year-old Italian woman who had been suffering from depressive symptoms for several years, worsening in the last few months. Approximately 3% to 6% of all adults in Western countries suffer from a form of depression that persists for at least two years during their lifetime (Kessler, Berglund, Demler, Jin & Merikangas, 2005). The Diagnostic & Statistical Manual of Mental Disorders 5 th Edition (DSM-5) (American Psychiatric Association, 2013) has introduced a new diagnostic category of persistent depressive disorder (PDD) that includes the first two of the following four subtypes of persistent forms of depression: (a) a continuing mild depressive mood (dysthymia); (b) a state meeting all criteria for major depression continuously (chronic major depression); (c) a recurrent major depression with incomplete remission between episodes; and (d) a superimposition of a major depressive episode on an antecedent dysthymia (double depression) (Klein, 2010).
The aim of this study was to investigate the effectiveness of the manualised TA treatment of depression  applied to a persistent depressive disorder (dysthymia). The primary target of the therapy was the depressive symptomatology, the secondary target symptoms were anxiety, global distress and severity of personality problems. Qualitative data was also collected from therapist and client on helpful aspects of the therapy and following change.

Ethical Considerations
The research protocol follows the requirements of the ethical code for Research in Psychotherapy of the Italian Association of Psychology (AIP, 2015), and the American Psychological Association guidelines on the "rights and confidentiality of research participants" (APA, 2010, p. 16). The research protocol has been approved by the Ethical Committee of the University of Padua. Before entering the treatment, the client received an information pack, including a detailed description of the research protocol, and gave an informed consent and written permission to include segments of disguised transcripts of sessions or interviews within scientific articles or for these to be presented at conferences. The client was informed that she would have received the therapy even if she decided not to participate in the research and that she was able to withdraw from the study at any moment without any negative impact on her therapy. All aspects of the case material were disguised, so that neither the client nor third parties are identifiable. All changes are made in such a way as to not lead the reader to draw false conclusions related to the described phenomena. The final article, in Italian language, was presented to the client, who confirmed that it was a true and accurate record of the therapy and gave her final written consent for its publication.

Method Inclusion and exclusion criteria
Psychotherapists participating in this case series were invited to include in their studies the first new client, with a disorder within the depressive spectrum as described in Persistent or Other Depressive Disorder), who agreed to participate in the research. Other current psychotherapy, active psychosis, domestic violence, bipolar disorder, antidepressant medication, alcohol or drug abuse were considered as exclusion criteria. As the overall aim of this project is to study the effectiveness of TA psychotherapy in routine clinical practice, comorbidity is normally accepted and both inclusion and exclusion criteria are evaluated case by case.

Client
Anna is a 33-year-old white Italian woman who lives alone in a large metropolitan area in Italy. She is a manager in a tour operator company and loves her job, she reports having a good relationship with her mother, who she described as "an angel" but has a difficult relationship with her father, who has been unable to demonstrate his affection for her in many life circumstances. She reports that her parents have been unable to protect her in some life decisions and situations. For this reason, nowadays she often appears to be very angry with them, especially for not understanding her feelings. She described that in her family everyone over-estimates her capacities to manage everything on her own. She has a younger sister and her parents pressed her to help her sister in finding a job, but she was not able to secure employment for her. The unemployment of her sister exacerbated her conflictual relationship with her family. In the past years, Anna described feeling responsible for her sister's situation, feeling guilty for achieving and having success in her life, and also feeling culpable for being incapable of doing more to help her sister. Anna felt lonely for her many bad relationship break-ups, which made her think, in the last several years, that there might be something wrong with her manner of relating with men. At the time of therapy, she was single. Anna reported often feeling extremely vulnerable, with periods of intense crying and stomachaches, and that in the last few years she had some difficulties in falling asleep. She stated she also feels anxious and disappointed in her relationships, and does not get the feedback she expects from her partners, which causes conflict and often in turn leads to men breaking-up with her. She is worried about her future, believes that she will not be able to create a family of her own, and feels that she is not important to anyone. She independently decided to seek therapy and asked a colleague to recommend a therapist.

Therapist
The psychotherapist is a 40-year-old white Italian woman with 8 years of clinical experience and who has a certification as Certified Transactional Analyst (Psychotherapy) (CTA-P). For this case, she received monthly supervision by a Teaching & Supervising Transactional Analyst (Psychotherapy) (TSTA-P) with 30 years of experience.

Intake sessions
Since the client had difficulties in paying for the therapy, the therapist proposed that Anna participated in the research protocol to access lower cost therapy. The client attended four pre-treatment sessions (0A, 0B, 0C, 0D), which were focused on explaining the research project, obtaining consensus, conducting a diagnostic evaluation according to DSM-5 criteria, developing a case formulation and a treatment plan, defining the problems she was seeking help for in therapy, as well as their duration and their severity (i.e. preparing the personal questionnaire, see later), and collecting a stable baseline of self-reported measures for primary (depression) and secondary (anxiety, global distress, personal problems) symtoms.

DSM-5 Diagnosis
During the diagnostic phase, Anna was assessed as meeting DSM-5 diagnostic criteria of mild persistent depressive disorder with mild anxious distress: she experienced depressed mood for more than two years (criterion A) insomnia (B2), reduced self-esteem (B4) and feelings of hopelessness (B6); she also felt excessively anxious (1) and worried (2). Knowing the level of an individual's personality functioning and personality traits provides the therapist with fundamental information for treatment planning. Therefore, a personality diagnosis using the alternative dimensional model developed for DSM-5 Section III was also conducted. This diagnosis allows for assessment of: 1) the level of impairment in personality functioning, and 2) personality traits. Anna showed impairment ranging from little to some in the level of organisation, and personality traits of depressivity, anxiousness, submissiveness, impulsivity, hostility and withdrawal. The therapist also rated the computerised   ) that supported the diagnosis of high level of functioning, with traits of depressive, histrionic and dependent personality types.

TA Diagnosis and Case formulation
Anna presented with Be Strong and Please Me drivers (Kahler, 1975) and the injunctions (Goulding & Goulding, 1976) 'Don't be important', 'Don't think', 'Don't be close', and 'Don't be yourself' (Don't be feminine). Anna's racket system (Erskine & Zalcman, 1979) showed beliefs such as 'Be compliant in order to obtain love'. Her script analysis involved substitute feelings (English, 1971) of sadness and anger, with somatisation as defense mechanisms. Interpersonally, Anna tended to alternate dramatic roles (Karpman, 1968) of Victim (when backing down without expressing her feelings), Rescuer (worrying about others, especially her sister), and Persecutor (during outbursts of hostility). Her life position (Ernst, 1971) was I'm Not OK, You're OK.

Treatment
The therapy followed the manualised therapy protocol of . The treatment plan focused on creating a therapeutic alliance, primarily providing permission (Crossman, 1966) congruent with the client's injunctions, namely: you can be important, think, be close, be yourself (feminine). The therapist offered Anna empathic listening, supporting her to feel and express her emotions, needs and wishes. During first sessions, the therapist also explained the ego state model, in order to give Anna some theoretical knowledge that might help her to better understand the emotional states she was experiencing and her behaviours. Then, the therapist focused on reinforcing self-esteem, supporting Anna's recognition of the importance of understanding her Child ego state needs for attention and being loved, exploring her experiences, and analysing her script  events such as the relationship with her father, which influences her actual relationships with men.

Analysis Team
The HSCED main investigator and first author of this paper is a Provisional Teaching and Supervising Transactional Analyst (Psychotherapy) (PTSTA-P) with 10 years of clinical experience, with a strong allegiance for TA. Despite recent literature suggesting that hermeneutic analysis should be carried out only by expert psychotherapists , we decided that when the research is investigating a new population or a therapy that lacks a research base, it is appropriate to follow , who proposed that analyses can be carried out by a team of 'reasonable persons ', not yet overly committed to any theoretical approach or professional role. The team comprised six postgraduate psychology students who were taught the principles of hermeneutic analysis by Professor John McLeod, in a course on case study research at the University of Padua. Following the indication of Elliott, Partyka, Wagner et al (2009), the students preferred to assume both affirmative and sceptic positions, and independently prepared their affirmative and sceptic cases. Then they met and merged their own cases, supervised by the main investigator, creating a consensual affirmative and sceptic brief and rebuttals.

Transparency statement
The research was conducted entirely independently of the previous case series (see . The last author, Mark Widdowson, was involved in checking that the research protocol and data analysis process was adhered to, in order to make the claim that this case series represents a valid replication of the initial study (with minor changes) and he was involved in the final preparations of this article.

Judges
The judges were three researchers in psychotherapy at the University of Padua and co-authors of this paper: Judge A, Vincenzo Calvo, clinical psychologist, psychotherapist trained in dynamic psychotherapy, PhD in development psychology, with expertise in attachment theory; Judge B, Stefania Mannarini, psychologist with experience in research methodology; and Judge C, Arianna Palmieri, neuropsychologist and psychotherapist with a training in dynamic psychotherapy. Judges A and C had some basic knowledge of TA but had never engaged in any official TA training, whereas Judge B has some clinical experience but no knowledge of TA.

Quantitative Outcome Measures
Three standardised self-report outcome measures were selected to measure primary target symptoms (depression) and secondary symptoms (anxiety and global distress).
Patient Health Questionnaire 9-item for depression (PHQ-9;, which scores each of the nine DSM-5 criteria from 0 (not at all) to 3 (nearly every day), which has been validated for use in primary care . Total scores up to 4 are considered healthy, scores of 5, 10, 15 and 20 are taken respectively as the cut-off points for mild, moderate, moderately severe and severe depression. PHQ-9 score ≥10 has a sensitivity of 88% and a specificity of 88% for major depression (Kroenke, Spitzer, & Williams, 2001) and scores of <10 are considered subclinical.
Generalized Anxiety Disorder 7-item for anxiety (GAD-7;, which scores each of the seven DSM-5 criteria as 0 (not at all), 1 (several days), 2 (more than half the days), and 3 (nearly every day). Total scores of up to 4 are considered healthy, scores of 5, 10, and 15 are taken as the cut-off points for mild, moderate and severe anxiety respectively. Using the threshold score of 10, the GAD-7 has a sensitivity of 89% and a specificity of 82% for GAD and scores of <10 are considered subclinical. It is moderately good at screening three other common anxiety disorders -panic disorder (sensitivity 74%, specificity 81%), social anxiety disorder (sensitivity 72%, specificity 80%) and post-traumatic stress disorder (sensitivity 66%, specificity 81%) (Kroenke, Spitzer, Williams, et al, 2007).
Clinical Outcome for Routine Evaluation -Outcome Measure for global distress (CORE-OM) (Evans, Connell, Barkham, Margison, Mellor-Clark, McGrath, & Audin, 2002). Each of the 34 items is scored on a 5-point scale ranging from 0-4 (0 = not at all, 4 = most of the time). Total scores up to 5 are considered healthy, scores between 5 and up to 9 are considered low level (subclinical), and scores of 10, 15, 20 and 25 are taken as the cut-off point for mild, moderate, moderately severe and severe distress, respectively. The cut-off of 10 yields a sensitivity (true positive rate) of 87% and a specificity (true negative rate) of 88% for discriminating between members of the clinical and general populations. CORE-OM was used in assessment sessions, in sessions 8, 16 and follow ups, whereas CORE short form A and B were used in all other sessions (Barkham, Margison, Leach, Lucock, Mellor-Clark, Evans, McGrath et al, 2001).
All measures were evaluated according to Reliable and Clinical Significant Improvement (RCSI) (Jacobson & Truax, 1991). It is important to consider that even under the cut-off score there may be a subclinical disorder. To minimise Type I error (which occurs when cases with no meaningful symptom change are assumed to have improved) we employed also Reliable Change (RC) (Jacobson and Truax, 1991) to evaluate whether observed changes on a measure were statistically reliable and not due to chance. For example,  proposed that a minimum reduction of 6 points in the PHQ-9 would be indicative of reliable improvement. Transition from clinical to non-clinical population and reliable change combine to produce a Reliable and Clinically Significant Change Index (RCSI), as robust evidence of recovery in psychological interventions (Evans, Margison & Barkham, 1998;Delgadillo, McMillan, Leach, Lucock, Gilbody & Wood, 2012).
See Table 1 for Clinical Significance (CS) and Reliable Change (RC) values for each employed measure. All these measures were administered prior to the beginning of each session to measure the on-going process and to facilitate the identification of events in therapy that produced significant change.
Before each session, the client also rated the Personal Questionnaire (PQ) (Elliott, Shapiro, & Mack, 1999), a client-generated measure in which clients specify the problems they would like to address in their therapy and rate their problems according to how distressing they are finding each problem (1, not at all; 7, maximum possible). Scores up to 3 are considered subclinical. In this case series, for the PQ we adopted a more conservative RC of two points, rather than the RC of one point already used in the previous case series.
All of these measures were administered in the pretreatment phase in order to obtain a three-point baseline, and during the three follow-ups, except that in this case Anna's PQ score was not obtained from session 1.

Qualitative Outcome Measurement
The client was interviewed using the Change Interview protocol (CI) (Elliott, Slatick & Urman, 2001) one month after the conclusion of the therapy. The CI is a semistructured qualitative change measure which asks clients how they feel they have changed during the therapy and how they think these changes came about, what they felt was helpful or hindering in the therapy, and what changes they feel they still need to make. Clients are asked to identify key changes they made and to indicate on a five-point scale: 1) if they expected to change (1=expected; 5=surprising); 2) how likely these changes would have been without therapy (1=unlikely; 5=likely), and 3) how important they feel these changes to be (1=slightly; 5=extremely).
The client also completed the Helpful Aspects of Therapy form (HAT)  at the end of each session. The HAT allows the client to describe hindering or useful aspects of the session and to rate them on a nine-point scale (1=extremely hindering, 9=extremely useful).

Therapist Notes
A structured session notes form (Widdowson, 2012a, Appendix 6, p. 50-52) was completed by the therapist at the end of each session. In this form, the therapist provides a brief description of the session in which are identified key aspects of the therapy process, the theories and interventions used, and an indication of how helpful the therapist felt the session was for the client.

Adherence
The therapist, the supervisor, and the main researcher were all Transactional Analysts and they each independently evaluated the therapist's adherence to TA treatment of depression using the operationalised adherence checklist proposed by Widdowson (2012a, Appendix 7, p. 53-55) before agreeing on a final consensus rating. The conclusion of the three evaluators was that the treatment had been conducted coherently according to TA theory at a good to excellent level of application.

Affirmative Case
The affirmative position according to Elliott (2002) should locate evidence in the rich case record supporting the claim that the client has changed, and that the change is causally due to the therapy. A clear argument supporting the link between change and treatment must be established on the basis of at least two of the following five sources of evidence: 1. Changes in stable problems: client experiences changes in long-standing problems. The change should be replicated in both quantitative and qualitative measures. Change should be Clinically Significant (scores fall in the healthy range), Reliable (corrected for measure error) and Global (Reliable Change is replicated in at least two out of three measures); 2. Retrospective attribution: according to the client the changes are due to the therapy; 3. Outcome to process mapping: refers to the content of the post-therapy qualitative or quantitative changes that plausibly match specific events, aspects, or processes within therapy; 4. Event-shift sequences: links between client reliable gains in the PQ scores and significant within therapy events; 5. Within therapy process-outcome correlation: the correlation between the application of therapy principles (e.g. a measure of the adherence) and the variation in quantitative weekly measures of client's problem (e.g. PQ score).

Sceptic Case
A sceptic position requires a good-faith effort to find nontherapeutic processes that could account for an observed or reported client change. Elliott (2002) identified eight alternative explanations that the sceptic position may consider: four non-change explanations and four non-therapy explanations.
The four non-change explanations assume that change is really not present, and should consider: 1. Trivial or negative change which verifies the absence of a clear statement of change within qualitative outcome data (e.g. CI), and the absence of clinical significance and/or reliable change in quantitative outcome measures (e.g. PHQ9); 2. Statistical artefacts that analyse whether change is due to statistical error, such as measurement error, regression to the mean or experiment-wise error; 3. Relational artefacts that analyse whether change reflects attempts to please the therapist or the researcher; 4. Expectancy artefacts, analysing whether change reflects stereotyped expectations of therapy.
The four non-therapy explanations assume that the change is present, but is not due to the therapy, and should consider: 5. Self-correction which analyses whether change is due to self-help and/or self-limiting easing of a temporary problem or a return to baseline functioning; 6. Extra-therapy events that verify influences on change such as those due to a new relationship, work, or financial conditions; 7. Psychobiological causes which verify whether change is due to factors such as medication, herbal remedies, or recovery from medical illness; 8. Reactive effects of research, analysing the effect of change due to participating in research, such as generosity or goodwill towards the therapist.
The formulation of affirmative and sceptic interpretations of the case consists of a dialectical process, in which affirmative rebuttals to the sceptic position are constructed, along with sceptic rebuttals of the affirmative position.
Finally, each position is summarised in a narrative that offers a customised model of the change process that has been inferred, including therapeutic elements and an account of the chain of events from cause (therapy) to effect (outcome), including mediator and moderator variables.

Adjudication Procedure
Each single judge received the rich case record (session transcriptions, therapist and supervisor adherence forms and session notes, quantitative and qualitative data and also a transcript of the Change Interview) as well as the affirmative and sceptic cases and rebuttals by email, together with instructions. The judges were asked to examine the evidence and provide their verdict. They were required to establish: • If the case were a clearly good outcome case, a mixed outcome case, or a poor outcome case; • If the client had changed; • To what extent these changes had been due to the therapy; • Which aspects of the affirmative and sceptic arguments had informed their positions.
Furthermore, the judges had to observe which mediator factors in the therapy they considered to have been helpful and which characteristics about the client did they think had contributed to the changes as moderator factor(s).

Results
In earlier published HSCED's the rich case records, along with hermeneutic analysis and judges' opinions were often provided as online appendices (Benelli et al, 2015). Since all the material is in Italian language, we adopted here the solution of providing a summary of the main points, as proposed in MacLeod, Elliott and Rodger (2012). The complete material (session transcriptions, Change Interview, affirmative and sceptic briefs and rebuttal, judge opinions and comments) is available from the first author on request.

Quantitative Outcome Data
Anna's quantitative outcome data are presented in Table  1. The initial PHQ-9 score of 11 indicated a moderate level of depression. The GAD-7 score of 8.3 indicated a subclinical, mild level of anxiety. The CORE at 16.8 indicated a moderate level of global distress and functional impairment. The PQ at 6.2 indicated that the client perceived her problems as bothering her more than very considerably.
At session 8, (mid-therapy), all measures decreased. Depression passed into the subclinical mild range (6), anxiety remained in the mild range (6), global distress passed to subclinical range, with clinically significant and reliable improvement (7.4), and personal problems decreased to moderately bothering (4.3).
By the end of the therapy, the depressive score remained in the mild range (7), the anxiety reliably decreased to healthy range (4), the global distress returned within the mild range (12.1) with a lower score than pre-therapy, and the personal problems reliably decreased (3.7).
At the 1-month follow up, all measures except anxiety improved: depressive scores remained in the mild range (6), anxiety returned to mild (6), the global distress returned to a subclinical range (6.5) with clinically significant and reliable improvement, and personal problems remained reliably improved at moderately (3.5).
At the 3-month follow up, all measures improved: depression passed into the healthy range (3) with a clinically significant and reliable improvement, anxiety reliably decreased to the healthy range (4), global distress entered the healthy range (5) and personal problems were scored as bothering her only a little (3.2).
At the 6-month follow up all scores worsened: depression remained in the healthy range with clinically significant and reliable improvement compared to the pre-therapy (4); anxiety returned to the mild range, with a slight, nonreliable change compared to pre-therapy (6); global distress returned within the mild range (10.29), with a score lower than at the end of the therapy and reliably improved in respect to the beginning of the therapy; personal problems returned to moderately bothering (3.7), with a reliable improvement compared to the pretherapy score. Table 2 shows the 10 problems that the client identified in her PQ at the beginning of the therapy and their duration. Four problems were rated as maximum possible, five very considerably and one moderately bothering. Four problems lasted from more than 10 years, two from 6-10 years, three from 3-5 years and one from 1-2 years. Problems lasting for more than 10 years showed a reliable change at the end of the therapy and at the 6-month follow up (except item 8, anger toward parents). All problems lasting from 1-5 years showed an early reliable change within session 8, and of these, three out of four also showed a clinically significant change.
At the end of the therapy and at the 1-month follow up, 9 out of 10 problems showed a reliable change, and 3 of these showed also a clinically significant improvement (guilty, oppression, anger). At the 3-month follow up, all problems showed a reliable change, and 5 of these also a clinically significant change. At the 6-month follow up, all problems showed a reliable change (but 8, anger) and guilt, sleeping and feeling of oppression also showed a clinically significant improvement.  (Evans et al., 2002). PHQ-9 = Patient Health Questionnaire 9-item for depression ) GAD-7 = Generalized Anxiety Disorder 7-item . PQ = Personal Questionnaire (Elliott, Shapiro, & Mack, 1999). FU = follow-up.
Clinical cut-off points: CORE-OM ≥10; PHQ-9 ≥10; GAD-7 ≥10; PQ ≥3. Reliable Change Index values: CORE-OM improvement of five points, PHQ-9 improvement of six points, GAD-7 improvement of four points, PQ improvement of two points. a Mean value of pre-therapy assessment sessions. b First available score in session 2.

Qualitative Data
Anna compiled the HAT form at the end of every session (Table 3), reporting only positive/helpful events. All positive events were rated from 7 (moderately helpful) to 9 (extremely helpful). She reported helpful aspects on self esteem (HAT 1, appreciate myself; HAT 5, accept myself; HAT 8, faith in myself; HAT 12, reassuring myself); relationships (HAT 1, put boundaries; HAT 3, collect information, no expectations; HAT 9, time to understand; HAT 14, receive vs show off); symptoms (HAT 2, too responsible); emotions and inner experience (HAT 4, utter emotions, become aware; HAT 5, focus feelings; HAT 7 put out anger, awareness of feelings; HAT 8, confidence in myself; HAT 12, stop and think about feelings).
Anna participated in a Change Interview 1-month after the conclusion of the therapy. In this interview, she identified her main and significant changes (Table 4). Anna described her therapy as "very helpful" (Client line 31), "it helped me focus on how to protect myself" (C32). When Anna started the therapy, she felt "more vulnerable" (C35), whereas now she is not "throwing herself headlong in relationships", in fact she is "trying to create a more mature and equal way to relate with men" (C37). At the beginning of the therapy, she felt angry when thinking about people from her past, whereas now she is able to "distance them from my life" (C82). Before starting the therapy, Anna reported feeling guilty for the unemployment of her sister, whereas now she does not feel responsible any more: "Earlier I was focused on my guilt… I changed perspective… I don't feel guilty anymore… I only tried to help her… I did it in good faith" (C92).
Anna summarised four main areas of change. First, she observed an improvement in her way of relating with men. Anna stated that she expected such results, in fact that was her therapy goal (rated 2, somewhat expected), although that she believed that this outcome would have been unlikely without therapy (1) and was very important for her (4). The second change she identified was focusing the aspects she has to work on to improve her affective vulnerability, which she identified as somewhat expected (2) and that the change would have been unlikely to have happened (1) without therapy, rating it as extremely important for her (5). The third and fourth improvements were accepting her past and feeling calmer regarding familiar stress, both rated as neither expected nor unexpected (3), which would have been unlikely to have happened without the therapy (1) and considered respectively as extremely (5) (Elliott, Shapiro, & Mack, 1999). Clinical cut-off point: PQ ≥3. Reliable Change: PQ improvement of two points. +=indicates clinically significant change (CS). *=indicates reliable change (RC). The rating is on a scale from 1 to 7 and indicate how much each problem has bothered the client: 1 = not at all; 7 = maximum. m = months. y = year. FU= follow-up. a The first available score was in session 2.
. Note. The rating is on a scale from 1 to 9: 1 = extremely hindering, 5 = neutral, 9 = extremely helpful. HAT = Helpful Aspect of Therapy .

Change
How much expected change was (A) How likely change would have been without therapy (B) Importance of change (C) I feel capable to develop relationships 2 (somewhat expected) The rating is on a scale from 1 to 5; 1= expected, 3 = neither, 5 = surprising. b The rating is on a scale from 1 to 5; 1=unlikely, 3 = neither, 5 = likely. c The rating is on a scale from 1 to 5; 1 = slightly, 3 = moderately, 5 = extremely.
important. Anna also reported that a friend of hers realised that she was calmer (C41). "Once the therapist told me that it's ok to hit the rock, but afterwards you can raise back up… I feel better knowing this, because there were days in which I felt like a total failure" (C49). Anna in her CI did not report any negative, obstructive or unpleasant aspect of therapy. On the contrary, she felt that some sessions were "really painful, yet crucial to focus on my problems… I metabolised the pain… I have a stronger will to feel better" (C67-8).

HSCED Analysis Affirmative Case
The affirmative team identified four lines of evidence supporting the claim that Anna changed and that the therapy had a causal role in this change.
Change in stable problems Quantitative data (Table 1) shows that there is an early improvement in primary outcome measure (depression) that is clinically significant since session 8 and with reliable and clinically significant improvement (RCSI) at 3-and 6-month follow up. Secondary outcome measures depict a reliable improvement in the initial subclinical score of anxiety (GAD-7) at the end of the therapy and at the 3-month follow-up. There is also an early change with RCSI for global distress (CORE) at session 8, maintained at 1-and 3-month follow up. In the PQ (Table 2), Anna identified 10 main problems at the beginning of the therapy that she was trying to solve, almost all rated as bothering her very considerably (6) to maximum possible (7), nine out of 10 standing from 3 to more than 10 years. All the problems referred to issues with self esteem, relationships, symptoms, emotions and inner experience. At the end of the therapy and at the 6-month follow up 9 problems out of 10 showed reliable change, and three problems also reached RCSI. Overall, there is support for a claim of global reliable change (reliable change in at least two out of three measures). Qualitative data supports this conclusion: in fact, in her Change Interview (CI) Anna reports as a main achievement in therapy her change in dealing with others, men, family and her past experiences, all problems rated in the PQ as long standing (more than 10 years). At the end of the therapy she also appears more capable of asserting herself (session 15, C33-35), that implies a change in self experience (vulnerable), another long bothering problem since more than 10 years. Since sessions 7 (C16-40) and 8 (C5-8) Anna showed up with a higher mood. Thus, we claim that Anna obtained a stable RCSI in persistent depressive disorder, and a reliable improvement in global distress and in long-standing problems.

Retrospective attribution
Anna identified in her Change Interview four important changes in different aspects of her life, all of which she attributed to therapy (Table 4). She considered her improvements very and extremely important, and stated that she believed all were unlikely to have occurred without therapy, with the first two changes expected and the last two neither expected nor unexpected. She recognised that the therapy allowed her to change different aspects of her way of relating with men (CI, C35), which was directly related to her therapy contract. The client asserted that the therapy was very useful to her, in particular for the kind of mature and equal relationships she feels she is now capable of establishing (CI, C35). She also affirmed that there were no negative aspects, obstacles or unhelpful aspects to her therapy. In session 16, Anna reported being sad about the ending of the therapy, because it had helped her to focus on her problems and learn what she needed to work on to change for the better (session 16, C155-156). Due to the new strategies she had been working on with her therapist and started to use in her everyday life (see Table 3), she had noticed positive changes.
Association between outcome and process (outcome to process mapping) The HAT completed at the end of each session provides us with regular and immediate reports of what Anna found helpful in each session. All reported events are considered moderately to extremely useful and are coherent with both the diagnosis and the interventions reported in the therapist's notes. One of the client's most important changes reported in the Change Interview refers to the ability to "focus on the aspects I need to work on" (Table 4, CI C31) that appear tied to the therapist's frequent interventions on the importance of Anna clearly and succinctly expressing what she feels and thinks. This is mirrored in the client's HAT 5 ("be able to focus better on what I feel"), 9 ("give me time to understand"), 12 ("stop and think"), 15 ("realise what I want") and 16 ("what I understand, what I want"). Also, the change "I was unable to develop relationships" appears tied to the therapist's interventions reported by the client in the HAT 1 ("to put boundaries in place"), HAT 3 ("do not interpret, but collect information"), 14 ("I show off") and 15 ("realise what I want in relationships"). The other change about family ("accept the past and feel calmer") appears connected to the HAT 2 ("talk and cry about painful family events"), 7 ("pull out the anger towards my father"), 13 ("talk about a trauma of the past").

Event-shift sequences (process to outcome mapping)
The PQ mean score shows a progressive decrease in severity of her problems from the initial score (6.2, more than very considerably) to the final score (3.7, less than moderately). The therapist's confrontation of the client's tendency to feel responsible for others, in particular her sister's employment problems (session 2), is reflected in the PQ item 4 (guilty), that decreased since session 3, achieved RCSI in session 9 and was maintained throughout the entire follow-up period. The interventions regarding her tendency to 'please others' in session 5 led the client to become aware of her anger and to use it for taking an assertive position with her family and colleagues (session 5, C 174). This was reflected in improved scores in PQ item 9 that reached RCSI since session 5 and was maintained at the 6-month follow up.
Sceptic Case 1.The apparent changes are negative (i.e. involved deterioration) or irrelevant (i.e. involve unimportant or trivial variables). The client entered the trial with moderate depression (PHQ-9, score 11), barely over the threshold for major depressive disorder. Considering the typical cyclical pattern of the diagnosed persistent depressive disorder, it is quite likely that a natural reversal may occur in the following months. Change on anxiety (GAD-7) is irrelevant since the initial score was subclinical and change is not maintained at the 6-month follow up. The global distress score (CORE) shows an inconsistent pattern, and remains in the clinical range at the end of the therapy and at the 6-month follow up. Reliable change is present in three measures out of four, and RCSI is present only for primary outcome, suggesting that a claim of Global Reliable Change is unwarranted. Also in qualitative data, we note evidence of inconsistent change: at session 15, Anna tells about an episode she had with some friends, in which they told her she is not improving in her way of relating to men. Furthermore, at the final session, she reports ruminating on whether she did the right thing with a man she liked. During the CI, the client states that she has not completely worked on her insecurity, and still feels frustrated when dealing with stressful people (like her boss). In the 3-month follow up, Anna still refers to feeling guilty about her sister's unemployment and that she sought explanations from the executive director at her company about why they did not offer her sister a job. Thus, change reported in quantitative self-reported measures does not appear to be supported by the client's statements. Thus, we conclude that the change observed in the primary outcome is more due to the typical pattern of persistent depressive disorder than to the therapy, and only a longer follow-up could determine the effect of the therapy.
2. The apparent changes are due to statistical artefacts or random errors, including measurement error, experiment-wise error from using multiple change measures, or regression to the mean. The pre-treatment baseline related to the PQ has not been collected due to technical problems, and the score is available only from the second session, making it difficult to draw any conclusions on change in relation to long-standing personal problems, due to missing a stable baseline which would enable clear comparison with subsequent scores. We also noticed that Anna's scores for the PQ item 7 ("I have difficulties in falling asleep") has correspondence to the GAD-7 item 4, in PHQ-9 item 3 and in CORE-B item 2, and they received different evaluation in the same session. For example, at session 13, the client scored this item 2 (sometimes) in the CORE-B, 1 (several days) in the PHQ-9, 2 (over half the days) in the GAD-7, and 5 (considerably) in the PQ. This suggests that the client might have some difficulties in relating her inner experience to scoring of individual items, thus introducing a possible inconsistency within quantitative results.
3. The apparent changes reflect relational artefacts such as global hello-goodbye effects on the part of a client expressing his or her liking for the therapist, wanting to make the therapist feel good, or trying to justify his or her ending therapy. In her CI, Anna reported only positive comments about the therapy and the therapist, and in her HAT forms she reported only positive/helpful events. Even session 11, after which she "forgot" to attend the following session without informing the therapist, is described as helpful, and the event is not mentioned in the CI. This suggests that CI and HAT may be biased by Anna's tendency to 'please others' and a desire to present a good image of her therapist to the researcher conducting the CI. Her 'please others' tendency is also in line with her diagnosis. Furthermore, Anna keeps asking the therapist whether she is doing the right thing or not (e.g. in session 14: C2, C45, C63), and comments that she was looking forward to the day of her session in order to talk to the therapist and ask her for advice (session 15, C 73), supporting the conclusion that she tends to depend on the therapist's advice and approval which could have affected her outcome measures.
4. The apparent changes are due to cultural or personal expectancy artefacts; that is, expectations or scripts for change in therapy. The sceptic team were not able to find any evidence within the rich case record that would support a claim that Anna's changes were associated with expectancy effects.
5. There is credible improvement, but it involves a temporary initial state of distress or dysfunction reverting to normal baseline via corrective or self-limiting processes unrelated to therapy. DSM-5 indicates that the typical pattern of persistent depressive disorder is likely to include a major depressive episode that may spontaneously revert to a subclinical level. The primary outcome measure could have captured this spontaneous cyclical pattern.
6. There is credible improvement, but it is due to extratherapy life events, such as changes in relationships or work. The sceptic team were not able to find any evidence within the rich case record which would support a claim that Anna's changes were associated with extra-therapy life events.
7. There is credible improvement, but it is due to psychobiological processes, such as psychopharmacological mediations, herbal remedies, or recovery of hormonal balance following biological insult. In the CI, Anna reports she has not been taking any kind of drugs (T7-C8). The sceptic team were not able to find any evidence within the rich case record which would support a claim that Anna's changes were associated with psychotropic medication or other herbal or similar kind of remedy.
8. There is credible improvement, but it is due to the reactive effects of being in research. Participating in the research implied a lower cost for the client, and this might have more or less unwittingly affected the rating scores, probably in interaction with the abovementioned 'please others' effect.

Affirmative Rebuttal
Global Reliable Change in the literature is referred to as a measure to control experiment-wise error, thus relying on reliable change and not on clinical significance (e.g., Elliott, 2002). Thus, we can claim that three out of four measures support a claim in favour of Global Reliable Change. Despite the typical cyclical patterns of the per-srsistent depressive disorder, the client identifies change in long-standing problems that were not resolved in previous years by the simple passage of time or natural course of the disorder. Inconsistency between client statements and outcome measures are evidence that stable change is a process achieved during therapy and gradually displayed after its end; the client reports being aware of still having issues to work on (session 15, C156); in the first follow-up the client expressed the desire to continue in the therapy (FU1, C55); and in the third follow-up she complained about not having been contacted to resume the treatment.
Thus, the deterioration observed in quantitative measure at the third follow-up, is not supported by the client's verbal reports on daily life, and may reflect her desire to appear as experiencing greater suffering and therefore be more needing of treatment in order to continue the therapy. Despite missing a clear baseline score for the PQ, we can assume that the score obtained in the second session is representative of the baseline score, since all problems were long standing in time. As for the supposed difficulty of the client in rating self-report measures, the validity of the outcome instruments is widely established and personal scores are corrected for measurement errors by the use of reliable change index. Regarding the 'please others' effect, in the HAT 15, in the CI and in the 1-month follow up, Anna says that some sessions had been painful for her (CI, C67; FU1, C83), showing an ability to critically appraise her own therapy.
During the Change Interview and the follow-ups, the client refers to her sense that sixteen sessions were not sufficient for achieving all of her desired results, and at the 1-month follow up she described herself as satisfied with how she had learnt to handle stressful and painful situations (C37), but asked to continue the therapy (without fee reduction) to allow her to work on more general problems related to her personality, supporting the claim that being included in research did not affect the outcome.

Sceptic Rebuttal
Despite lack of agreement within the literature on how to determine Global Reliable Change, it would be more conservative to claim Global Reliable Change only when both reliable change and clinical significance are achieved in two out of three measures. Within transcripts (session 11, 14 and 15) of the therapy it is possible to find evidence that although the client did show some signs of improvement, these were not of a sufficient magnitude to warrant a claim of 'Global Reliable Change of the client. Anna appears to recognise when she is making unfair expectations of men, yet she is still creating fantasies about them. On the other hand, it seems she is now able to feel free from her sense of guilt, but sometimes she acts in order to pacify her guilty feelings. The deterioration at the 6 month follow up suggests that the treatment did not obtain a stable change.

Affirmative Conclusion
Anna's dysthymia, anxiety, global distress and personal problems were related to difficulties in interpersonal patterns, in particular with men, and intrapsychic patterns and inner experience such as emotions and self-esteem. Since the beginning of therapy, the therapist created a climate where the client explored an ability to appreciate herself, expression of emotions such as guilt, sadness and anger, new behaviours such as putting boundaries in place with others, and achieving a new comprehension of her inner experience, thus allowing herself more time to reflect on her emotions and needs before acting. Furthermore, the client explored connections between present and past relational patterns, differentiating past and present. These experiences were reflected in changes in internal dialogues, interpersonal relationships, depressive symptoms, and personality traits of submissiveness, anxiety, hostility and impulsivity.

Sceptic conclusion
Anna asked for therapy during a deterioration in her otherwise subclinical or normal depressive symptomatology. During the therapy, the typical pattern of her persistent depressive disorder reverted to the normal range. Her personality traits (submissiveness, dependent) affected her relationships with the therapist and probably her outcome scores. Changes in intrapsychic and interpersonal patterns are therefore likely to be due to the spontaneous remission of symptoms and to the reassuring effect provided by the presence of the therapist on her personality traits.

Adjudication
Each judge examined the rich case record and hermeneutic analysis and independently prepared their opinions and ratings of the case ( Table 5). The judges overall conclusions are that this was a clearly good outcome case, that the client changed substantially, and that the changes are between substantially and completely due to the therapy.
Opinions about the treatment outcome (good, mixed, poor) Judge A (VC). This case appears to be a clearly good outcome (80% certainty) with some aspect of mixed outcome (20%). Quantitative data shows a reliable and clinically significant change on measures of primary outcome (PHQ-9) at 3-and 6-month follow-ups. Measures of secondary outcomes also improved and there is a Global Reliable Change with three out of four measures showing a reliable change. PQ scores and qualitative data supports the conclusion that a change in long-standing problems occurred: for example, relationships with partners, colleagues and family are fully described as improved.
Judge B (SM). This is a clearly good outcome (80% certainty) or a mixed outcome (20%). The primary outcome was that the client's depressive symptoms passed from moderate into the healthy range during the course of the treatment. The final sessions and the Change Interview report clear descriptions of change in the client's life.
Judge C (AP). This case is classifiable as a good outcome case (80%) to mixed case (20%). Considering quantitative primary and secondary outcomes, every measure improved at the end of the therapy, indicating a change in depression, general distress, anxiety and severity of personal problems.
Opinions about the degree of change Judge A. The client changed substantially (80% with 80% certainty). Quantitative measures support the claim that the client's PHQ9 shows a stable healthy score at 6month follow-up, indicating a change in persistent, longstanding depressive symptomatology. The problems reported at the beginning of the therapy in the PQ were almost all long-lasting problems, bothering her up to ten years, and almost all problems show a reliable decrease. In the Change Interview, the client described a clear change in self-representation (guilty vs no longer guilty, vulnerable vs no longer vulnerable) indicating deep changes in personality dimensions and not only symptomatic modifications.
Judge B. The client changed considerably (60% with a 100% certainty), above all in her relationships with others and her family, and reported a decrease in problems described in the PQ. Despite there being some doubt about a claim of Global Reliable Change, the dysthymic symptoms are still absent six months after the end of therapy, indicating a deep and stable change in symptoms and in depressive personality traits. The client reported detailed pre-post differences in relationships with her parents, sister, friends, and colleagues, and a different stance towards her own internal experience.
Judge C. The client showed a substantial change (80% with 80% of certainty) in quantitative and qualitative data. Changes in long standing interpersonal relationships (specifically her sister and parents) support the conclusion that a deep and stable change happened. A longer follow up could further explain the degree and stability of change.
Opinions about the causal role of the therapy in bringing the change Judge A. The observed change is substantially (80% with 80% of certainty) due to the therapy. Quantitative PQ scores change following interventions that are reported as very important and helpful both in the therapist's notes and in the client's HAT forms. The focus of the therapist on the past experience that still influences the present and the differentiation between present and past appear tied to relational change between sessions, as reported in verbatim transcriptions. Qualitative data (Change Interview) reports clearly retrospective attribution of the client's four main changes to the therapy. HAT forms (summarised in Table 3) are rich in information on what happened during the sessions, and they appear coherent with the change the client feels she has obtained and which she described in the Change Interview.
Judge B. The change is substantially (80% with 80% of certainty) due to the therapy. The client refers to several helpful aspects in her HATs, and clearly states that her main change would have been unlikely without therapy. She reports changes in relationships that appear clearly connected to interventions in psychotherapy (e.g., feeling guilty vs recognising her sister's responsibility for her own situation). In the CI the client clearly defined her changes as unlikely without therapy.
Judge C. The change appears completely due to the therapy (100% with 80% of certainty). Hermeneutic analyses provide a clear link between specific therapeutic foci and changes in PQ scores. It appears unlikely that the client could change her relational patterns without the interventions of the therapist, as described in the HAT forms.

Mediator Factors
Judge A. A good therapeutic alliance and equal relationship appear important for the client's change in therapy. Explanation of the ego state model in the first sessions appears to have been a strong mediator of agreement on goals and alliance. The therapist focused the attention of the client during the sessions, modelling an ordered exploration of events rather than impressionist descriptions and impulsivity. The client's internal dialogue which generated feelings of guilt have been explored, examined, and reappraised. Behavioural submissiveness and a tendency to withdraw were challenged and reappraised.
Confrontation of maladaptive patterns, such as feeling guilty for others' failure, allowed change in depressive symptoms and personality traits. Differentiation between here and now and there and then emotional reactions to the stimuli allowed a change in interpersonal patterns.
Judge B. The client-therapist relationship is equal, with the therapist taking an active stance in the therapy, but without leading or suggesting. The therapist paid attention to helping the client to remain focused in the therapy and in defining vague and unspecific statements about events, feeling and behaviours, thus addressing personality traits of withdrawal, impulsivity and submissiveness. Systematic, early exploration of connections between present and past relationships appears tied to enhanced awareness and change in relationships. Focus on self-protection allowed the client to self-explore new behaviours in old relationships.
Judge C. In a manner which was coherent with the diagnosis of don't think and don't be important injunctions, the therapist focused on correspondent permissions, which supported the development of an early alliance. The therapist focused on promoting change in the client's interpersonal behaviours of submissiveness and withdrawal, and in supporting the exploration of alternative behaviours. The therapist focus on the difference between past and present relational experience supported the client in developing insight into attitudes learned in the past which were no longer appropriate in the present.

Moderator Factors
Judge A. The client appears able to immediately assume the 'client role'. She appears motivated, actively seeking therapy, with a high level of personality organisation and intelligence.
Judge B. The therapy was probably enhanced by moderator factors such as: the client's level of higher education, intelligence, and high level of personality functioning.
Judge C. The client was motivated, collaborative and willing to explore her inner world, and open to the therapist's interventions, and was searching for a caring relationship.

Discussion
This case aimed to investigate the effectiveness of a manualised TA treatment for depression in a client with moderate level of persistent depressive disorder (PDD). Primary target was depressive symptomatology, that showed a reliable change since session 8 and a clinically significant change since the 3-month follow up, maintained in the 6-month follow up. According to DSM-5, the course of PDD show a typical pattern with symptoms rising to the level of a major depressive episode, followed by a likely reversion to a lower level.
Symptoms in PDD are much more unlikely to resolve compared to a Major Depressive Disorder, and thus current clinical practice guidelines recommend the use of psychotherapeutic treatments for PDD (American Psychiatric Association, 2010;). The therapist conducted the treatment with a good to excellent adherence to the manual. Hermeneutic analysis pointed out changes in stable problem, retrospectively attributed to the psychotherapy, highlighting connections between outcome and process. The judges concluded that this is a clearly good outcome case, with a considerably to substantial degree of change, substantially to completely due to the therapy. These conclusions provide supporting evidence as to the effectiveness of manualised TA psychotherapy for depression, and provide evidence that the manual is suitable for use with persistent depressive disorder.
The therapeutic alliance appears to have been built on a non-directive but active style, focused on personality traits associated to symptoms and addressing their origin in the past. Specific TA techniques were: exploration of internal dialogue, developing the client's Nurturing Parent, exploration of the Be Strong and Please Others' drivers, racket analysis of guilt, sadness and hostility, disconnecting rubberbands (Kupfer & Haimowitz), game analysis (Berne) and analysis of drama triangle roles.

Limitations
The first author has a strong allegiance to TA, is a teacher of the members of the hermeneutic groups and a colleague of the three judges. The author was also funded for this research by TA institutions (see Funding below). Despite the reflective attitude adopted in this work, these factors may have influenced in subtle ways both the hermeneutic analysis and the judges' evaluations.

Conclusion
This case study provides evidence that the specified manualised TA treatment for depression  has been effective in treating a persistent depressive disorder in an Italian client-therapist dyad.
Despite results from a case study being difficult to generalise, this study adds evidence to the growing body of research supporting the efficacy and effectiveness of TA psychotherapy, and notably supports the effectiveness of manualised TA psychotherapy for depression as applied to persistent depressive disorder.

Funding
This study was supported by grants from the European Association for Transactional Analysis, as part of the project 'Transactional Analysis meets Academic Research in order to become an Empirically Supported

Introduction
This study is the second of a series of seven, and belongs to the second Italian systematic replication of findings from two previous case series  and was conducted under the auspices of the European Association for Transactional Analysis (EATA) and the University of Padua.
Transactional analysis (TA) is a widely-practiced form of psychotherapy, supported with a vast literature (for a review see , but still it is under-recognised within the worldwide scientific community of psychotherapy. In order to define TA psychotherapy as an efficacious Empirically Supported Treatment (EST), its efficacy must have been established in at least one Randomised Clinical Trial (RCT) replicated by two independent research groups, or alternatively in at least three Single Case Experimental Design studies (SCED), replicated by at least two independent research groups, with each group conducting a case series of a minimum of three cases, without conflicting evidence . Recently, a wide community of researchers proposed that efficacy and effectiveness in psychotherapy are a complex object that cannot be adequately evaluated with either the experimental approach of RCT Westen, Novotny & Thomson-Brenner, 2004) or classical SCED (reverse or multiple baseline design) . Systematic case study research has been proposed as a viable alternative to RCT and SCED (Iwakabe & Gazzola, 2009). Considering that approaches without evidence from RCTs tend to be under-recognised, Stiles, Hill and Elliott (2015) proposed collecting a series of mixed methods systematic single case studies as the first step toward recognition of marginalised and emerging models of psychotherapy.
Hermeneutic Single Case Efficacy Design (HSCED; Elliott, 2002;Elliott et al., 2009) is nowadays considered the most comprehensive set of methodological procedures for systematic case study research, and is a viable alternative to RCT and SCED in psychotherapy . HSCED is gaining momentum with enhanced versions proposed by different research groups, to validate new psychotherapeutic approaches or extensions of previously validated psychotherapies for investigation into their effectiveness with other disorders (e.g. . Recently, a systematic review of all published HSCED studies found within English language peer-reviewed journals (Benelli, De Carlo, Biffi & McLeod, 2015) highlighted methodological issues related to different levels of stringency, offering solid alternatives to conducting sound research according to the available resources within practitioner research networks.
Systematic case study research has already been applied to investigate the effectiveness of TA for people with long term health conditions  and HSCED methodology has been successfully applied to TA and widely described in this Journal by . Recently, several HSCEDs supporting the effectiveness of TA treatment for depression  have been published, as was an additional adjudicated study which demonstrated effectiveness of TA for mixed depression and anxiety . Furthermore, a related study was published on TA for emetophobia (Kerr, 2013). The case series by Widdowson and Benelli have shown that TA can be an effective therapy for major depressive disorder when delivered in routine clinical practice, in private practice settings, with clients with mild to moderate impairment in functioning who actively sought out TA therapy and with white British and Italian therapist and client dyads.
The present study analysed the treatment of 'Caterina', a 28-year-old Italian woman who had been suffering from depressive symptoms for more than ten years, worsening in the last year.
The aim of this study was to investigate the effectiveness of the manualised TA treatment of depression  applied to a major depressive disorder in comorbidity with general anxiety disorder. The primary target was the depressive symptomatology, with the secondary target symptoms of anxiety, global distress and severity of personality problems. Qualitative data was also collected from therapist and client on helpful aspects of the therapy and following change.

Ethical Considerations
The research protocol follows the requirements of the ethical code for Research in Psychotherapy of the Italian Association of Psychology (AIP, 2015), and the American Psychological Association guidelines on the "rights and confidentiality of research participants" (APA, 2010, p. 16). The research protocol has been approved by the Ethical Committee of the University of Padua. Before entering the treatment, the client received an information pack, including a detailed description of the research protocol, and gave an informed consent and written permission to include segments of disguised transcripts of sessions or interviews within scientific articles or for these to be presented at conferences. The client was informed that she would have received the therapy even if she decided not to participate in the research and that she could withdraw from the study at any moment without any negative impact on her therapy. All aspects of the case material were disguised, so that neither the client nor third parties are identifiable. All changes are made in such a way as to not lead the reader to draw false conclusions related to the described phenomena. The final article, in Italian language, was presented to the client, who confirmed that it was a true and accurate record of the therapy and gave her final written consent for its publication.

Method Inclusion and exclusion criteria
Psychotherapists participating in this case series were invited to include in their studies the first new client, with a disorder within the depressive spectrum as described in DSM-5 (Major, Persistent or Other Depressive Disorder), who agreed to participate in the research.
Other current psychotherapy, active psychosis, domestic violence, bipolar disorder, antidepressant medication, alcohol or drug abuse were considered as exclusion criteria. As the overall aim of this project is to study the effectiveness of TA psychotherapy in routine clinical practice, comorbidity is normally accepted and both inclusion and exclusion criteria are evaluated case by case.

Client
Caterina is a 28-year-old white Italian woman who lives with her mother and her younger sister in a metropolis in Italy. She works in a big company but she does not like her job. Her parents are divorced: her mother has dependent traits, whereas her father is a narcissistic ladies' man. She reports her parents as having being unable to put boundaries and protect her. Her younger sister is in therapy too. She feels frustrated and has many feelings of guilt both in her work and in relationships. She devaluates herself, feeling like she is not important for other, but especially for herself. When she was a little girl, if she expressed an opinion or taste that did not align with her mother's, she was frequently mocked by her. At the time of therapy, she did not have any kind of relationship. Two years earlier she had ended a four-year therapy, reporting no significant improvement. She decided to seek therapy again when she spoke to her sister's 'doctor', who recommended a therapist.

Therapist
The psychotherapist is a 43-year-old, white, Italian woman with 10 years of clinical experience and international certification as Provisional Teaching and Supervising Transactional Analyst (PTSTA-P). For this case, she received weekly supervision by a PTSTA-P with 15 years of experience.

Intake sessions
The client paid a normal fee for the therapy. She attended four pre-treatment sessions (0A, 0B, 0C, 0D), which were focused on explaining the research project, obtaining consensus, conducting a diagnostic evaluation according to DSM-5 criteria, developing a case formulation and a treatment plan, defining the problems she was seeking help for in therapy, as well as their duration and their severity (i.e. preparing the personal questionnaire, see later), and collecting a stable baseline of self-reported measures for primary (depression) and secondary (anxiety, global distress, personal problems) symptoms.

DSM-5 Diagnosis
During the diagnostic phase, Caterina was assessed as meeting DSM-5 diagnostic criteria of moderate major depressive disorder: she experienced depressed mood in daily activities for more than one year, most of the day, nearly every day (criterion A1), decreased pleasure in most activities (A2), hypersomnia (A4), feelings of worthlessness and inappropriate guilt (A7), diminished ability to think or concentrate (A8). She also met DSM-5 diagnostic criteria for generalized anxiety disorder: she experienced excessive anxiety and worry for more than one year (criterion A), she found it difficult to control the worry (B), she was easily fatigued (C2), she had difficulties in concentrating (C3) and she suffered sleep disturbance (C6). Knowing the level of an individual's personality functioning and personality traits provides the therapist with fundamental information for treatment planning. Therefore, a personality diagnosis was also conducted using the alternative dimensional model developed for DSM-5 Section III. This diagnosis allows assessment of: 1) the level of impairment in personality functioning, and 2) personality traits. Caterina showed impairment ranging from some to moderate in the level of organisation, and personality traits of depressivity, anxiousness, submissiveness, distractibility, emotional lability. The therapist also rated the computerised Shedler-Westen Assessment Procedure (SWAP-200; ) that supported the diagnosis of moderate level of functioning, with traits of depressive, dependent and histrionic personality.

TA Diagnosis and Case formulation
Caterina presented evidence of Please Me and Be Perfect drivers (Kahler, 1975) and the injunctions (Goulding & Goulding, 1976) Don't be important, Don't feel, Don't be close, and Don't be yourself. Caterina's racket system (Erskine & Zalcman, 1979) showed beliefs such as Compliance to obtain love. Her script ) analysis involved substitute feelings (English, 1971) of sadness. Interpersonally, Caterina tended to alternate dramatic roles (Karpman, 1968) of Victim (when backing down without expressing her feelings) and Rescuer (when worrying and helping others). Her life position was I'm Not OK, You're OK. (Ernst, 1971).

Treatment
The therapy followed the manualised therapy protocol of Widdowson (2015). The treatment plan primarily focused on creating a therapeutic alliance, primarily providing Permission (Crossman, 1966) congruent with the client's injunctions, namely: be important, feel and be close. The therapist offered Caterina empathic listening, supporting her to feel and express her emotions, needs and wishes. During assessment sessions, the therapist also explained the ego state model, in order to give her some theoretical knowledge that might help her to better understand the emotional states she experiences and her behaviours. Then, the therapist focused on reinforcing self-esteem, supporting Caterina's recognition of the importance of understanding her Child ego state needs for attention and being loved, exploring her experiences and analysing her script events, such as the relationship she has with the parents, which influences her actual difficulties in being independent, and her feelings of being always judged by others. Caterina attended all 16 sessions, although she skipped, and made up in the following week, session 12. In fact, session 11 has been very intense for her. Caterina reported "I knew I had to come, then I suddenly forgot… probably another example of boycott… I felt so thrilled to come, it was in my mind till few hours earlier… I felt so upset, especially because I thought all week about the things I had to tell my Child" (S12, C3-5).

Analysis Team
The HSCED main investigator and first author of this paper is a Provisional Teaching and Supervising Transactional Analyst (Psychotherapy) (PTSTA-P) with 10 years of clinical experience, with a strong allegiance for TA. Despite recent literature suggesting that hermeneutic analysis should be carried out only by expert psychotherapists , we decided that when the research is investigating a new population or a therapy that lacks a research base, it is appropriate to follow , who proposed that analyses can be carried out by a team of 'reasonable persons', not yet overly committed to any theoretical approach or professional role. The team comprised six postgraduate psychology students who were taught the principles of hermeneutic analysis by Professor John McLeod, in a course on case study research at the University of Padua. Following the indication of Elliott, Partyka, Wagner et al (2009), the students preferred to assume both affirmative and sceptic positions, and independently prepared their affirmative and sceptic cases. Then they met and merged their own cases, supervised by the main investigator, creating a consensual affirmative and sceptic brief and rebuttals.

Transparency statement
The research was conducted entirely independently of the previous case series (see . The last author, Mark Widdowson, was involved in checking that the research protocol and data analysis process was adhered to, in order to make the claim that this case series represents a valid replication of the initial study (with minor changes) and he was involved in the final preparations of this article.

Judges
The judges were three researchers in psychotherapy at the University of Padua and co-authors of this paper: Judge A, Vincenzo Calvo, clinical psychologist, psychotherapist trained in dynamic psychotherapy, PhD in development psychology, with expertise in attachment theory; Judge B, Stefania Mannarini, psychologist with experience in research methodology; and Judge C, Arianna Palmieri, neuropsychologist and psychotherapist with a training in dynamic psychotherapy. Judges A and C had some basic knowledge of TA but had never engaged in any official TA training, whereas Judge B has some clinical experience but no knowledge of TA.

Quantitative Outcome Measures
Three standardised self-report outcome measures were selected to measure primary target symptoms (depression) and secondary symptoms (anxiety and global distress).
Patient Health Questionnaire 9-item for depression (PHQ-9; Spitzer, , which scores each of the nine DSM-5 criteria from 0 (not at all) to 3 (nearly every day), which has been validated for use in primary care . Total scores up to 4 are considered healthy, scores of 5, 10, 15 and 20 are taken respectively as the cut-off points for mild, moderate, moderately severe and severe depression. PHQ-9 score ≥10 has a sensitivity of 88% and a specificity of 88% for major depression (Kroenke, Spitzer, & Williams, 2001) and scores of <10 are considered subclinical.
Generalized Anxiety Disorder 7-item for anxiety (GAD-7; Spitzer, , which scores each of the seven DSM-5 criteria as 0 (not at all), 1 (several days), 2 (more than half the days), and 3 (nearly every day). Total scores of up to 4 are considered healthy, scores of 5, 10, and 15 are taken as the cut-off points for mild, moderate and severe anxiety respectively. Using the threshold score of 10, the GAD-7 has a sensitivity of 89% and a specificity of 82% for GAD and scores of <10 are considered subclinical. It is moderately good at screening three other common anxiety disorders -panic disorder (sensitivity 74%, specificity 81%), social anxiety disorder (sensitivity 72%, specificity 80%) and post-traumatic stress disorder (sensitivity 66%, specificity 81%) (Kroenke, Spitzer, Williams, et al, 2007).
Clinical Outcome for Routine Evaluation -Outcome Measure for global distress (CORE-OM) (Evans, Connell, Barkham, Margison, Mellor-Clark, McGrath, & Audin, 2002). Each of the 34 items is scored on a 5-point scale ranging from 0-4 (0 = not at all, 4 = most of the time). Total scores up to 5 are considered healthy, scores between 5 and up to 9 are considered low level (subclinical), and scores of 10, 15, 20 and 25 are taken as the cut-off point for mild, moderate, moderately severe and severe distress, respectively. The cut-off of 10 yields a sensitivity (true positive rate) of 87% and a specificity (true negative rate) of 88% for discriminating between members of the clinical and general populations. CORE-OM was used in assessment sessions, in sessions 8, 16 and follow-ups, whereas CORE short form A and B were used in all other sessions (Barkham, Margison, Leach, Lucock, Mellor-Clark, Evans, McGrath et al, 2001).
All measures were evaluated according to Reliable and Clinical Significant Improvement (RCSI) (Jacobson & Truax, 1991). It is important to consider that even under the cut-off score there may be a subclinical disorder. To minimise Type I error (which occurs when cases with no meaningful symptom change are assumed to have improved) we employed also Reliable Change (RC) (Jacobson and Truax, 1991) to evaluate whether observed changes on a measure were statistically reliable and not due to chance. For example,  proposed that a minimum reduction of 6 points in the PHQ-9 would be indicative of reliable improvement. Transition from clinical to non-clinical population and reliable change combine to produce a Reliable and Clinically Significant Change Index (RCSI), as robust evidence of recovery in psychological interventions (Evans, Margison & Barkham, 1998;Delgadillo, McMillan, Leach, Lucock, Gilbody & Wood, 2012).
See Table 1 for Clinical Significance (CS) and Reliable Change (RC) values for each employed measure. All these measures were administered prior to the beginning of each session to measure the on-going process and to facilitate the identification of events in therapy that produced significant change.
Before each session, the client also rated the Personal Questionnaire (PQ) (Elliott, Shapiro, & Mack, 1999), a client-generated measure in which clients specify the problems they would like to address in their therapy and rate their problems according to how distressing they are finding each problem (1, not at all; 7, maximum possible). Scores up to 3 are considered subclinical. In this case series, for the PQ we adopted a more conservative RC of two points, rather than the RC of one point already used in the previous case series.
All of these measures were administered in the pretreatment phase in order to obtain a three-point baseline, and during the three follow-ups, except that in this case Caterina's PQ score was not obtained from session 1.

Qualitative Outcome Measurement
The client was interviewed using the Change Interview protocol (CI) (Elliott, Slatick & Urman, 2001) one month after the conclusion of the therapy. The CI is a semistructured qualitative change measure which asks clients how they feel they have changed during the therapy and how they think these changes came about, what they felt was helpful or hindering in the therapy, and what changes they feel they still need to make. Clients are asked to identify key changes they made and to indicate on a five-point scale: 1) if they expected to change (1=expected; 5=surprising); 2) how likely these changes would have been without therapy (1=unlikely; 5=likely), and 3) how important they feel these changes to be (1=slightly; 5=extremely).
The client also completed the Helpful Aspects of Therapy form (HAT)  at the end of each session.
The HAT allows the client to describe hindering or useful aspects of the session and to rate them on a nine-point scale (1=extremely hindering, 9=extremely useful).

Therapist Notes
A structured session notes form (Widdowson, 2012a, Appendix 6, p. 50-52) was completed by the therapist at the end of each session. In this form, the therapist provides a brief description of the session in which are identified key aspects of the therapy process, the theories and interventions used, and an indication of how helpful the therapist felt the session was for the client.

Adherence
The therapist, the supervisor, and the main researcher were all Transactional Analysts and they each independently evaluated the therapist's adherence to TA treatment of depression using the operationalised adherence checklist proposed by Widdowson (2012a, Appendix 7, p. 53-55) before agreeing on a final consensus rating. The conclusion of the three evaluators was that the treatment had been conducted coherently according to TA theory at a good to excellent level of application.

HSCED Analysis Procedure Affirmative Case
The affirmative position according to Elliott (2002) should locate evidence in the rich case record supporting the claim that the client has changed, and that the change is causally due to the therapy. A clear argument supporting the link between change and treatment must be established on the basis of at least two of the following five sources of evidence: 1. Changes in stable problems: client experiences changes in long-standing problems. The change should be replicated in both quantitative and qualitative measures. Change should be Clinically Significant (scores fall in the healthy range), Reliable (corrected for measure error) and Global (Reliable Change is replicated in at least two out of three measures); 2. Retrospective attribution: according to the client the changes are due to the therapy; 3. Outcome to process mapping: refers to the content of the post-therapy qualitative or quantitative changes that plausibly match specific events, aspects, or processes within therapy; 4. Event-shift sequences: links between client reliable gains in the PQ scores and significant within therapy events; 5. Within therapy process-outcome correlation: the correlation between the application of therapy principles (e.g. a measure of the adherence) and the variation in quantitative weekly measures of client's problem (e.g. PQ score).

Sceptic Case
A sceptic position requires a good-faith effort to find nontherapeutic processes that could account for an observed or reported client change. Elliott (2002) identified eight alternative explanations that the sceptic position may consider: four non-change explanations and four non-therapy explanations.
The four non-change explanations assume that change is really not present, and should consider: 1. Trivial or negative change which verifies the absence of a clear statement of change within qualitative outcome data (e.g. CI), and the absence of clinical significance and/or reliable change in quantitative outcome measures (e.g. PHQ9); 2. Statistical artefacts that analyse whether change is due to statistical error, such as measurement error, regression to the mean or experiment-wise error; 3. Relational artefacts that analyse whether change reflects attempts to please the therapist or the researcher; 4. Expectancy artefacts, analysing whether change reflects stereotyped expectations of therapy.
The four non-therapy explanations assume that the change is present, but is not due to the therapy, and should consider: 5. Self-correction which analyses whether change is due to self-help and/or self-limiting easing of a temporary problem or a return to baseline functioning; 6. Extra-therapy events that verify influences on change such as those due to a new relationship, work, or financial conditions; 7. Psychobiological causes which verify whether change is due to factors such as medication, herbal remedies, or recovery from medical illness; 8. Reactive effects of research, analysing the effect of change due to participating in research, such as generosity or goodwill towards the therapist.
The formulation of affirmative and sceptic interpretations of the case consists of a dialectical process, in which affirmative rebuttals to the sceptic position are constructed, along with sceptic rebuttals of the affirmative position.
Finally, each position is summarised in a narrative that offers a customised model of the change process that has been inferred, including therapeutic elements and an account of the chain of events from cause (therapy) to effect (outcome), including mediator and moderator variables.

Adjudication Procedure
Each single judge received the rich case record (session transcriptions, therapist and supervisor adherence forms and session notes, quantitative and qualitative data and also a transcript of the Change Interview) as well as the affirmative and sceptic cases and rebuttals by email, together with instructions. The judges were asked to examine the evidence and provide their verdict. They were required to establish: • If the case were a clearly good outcome case, a mixed outcome case, or a poor outcome case; • If the client had changed; • To what extent these changes had been due to the therapy; • Which aspects of the affirmative and sceptic arguments had informed their positions.
Furthermore, the judges had to observe which mediator factors in the therapy they considered to have been helpful and which characteristics about the client did they think had contributed to the changes as moderator factor(s).

Results
In earlier published HSCED's the rich case records, along with hermeneutic analysis and judges' opinions were often provided as online appendices (Benelli et al, 2015). Since all the material is in Italian language, we adopted here the solution of providing a summary of the main points, as proposed in MacLeod, Elliott and Rodger (2012). The complete material (session transcriptions, Change Interview, affirmative and sceptic briefs and rebuttal, judge opinions and comments) is available from the first author on request.

Quantitative Outcome Data
Caterina's quantitative outcome data are presented in Table 1. The initial depressive score (PHQ-9, 14.3) indicated a moderate level of depression. The anxiety score (GAD-7, 15) indicated a severe level of anxiety. The global distress score (CORE, 19) indicated a moderate level of global distress and functional impairment. The severity score of personal problems (PQ, 6.5) indicated that the client perceived her problems as very considerably to maximum possible bothering.
At session 8, (mid-therapy), all measures decreased. Depression (5) and anxiety (6) passed to subclinical mild range, presenting a clinically significant and reliable improvement. Global distress (11.8) passed to mild range with reliable improvement, and personal problems decreased to moderately bothering (4), with reliable improvement.
By the end of the therapy, all measures presented clinical significance and reliable change. Both the depressive (9) and anxiety (5) scores remained in the subclinical mild range, the global distress (7.9) decreased to subclinical low level range, and the personal problems (2.3) were rated very little, subclinical, bothering.
At the 1-month follow-up, all measures maintained clinical and reliable change. Anxiety passed into the healthy range, whereas depression and global distress passed to subclinical range, and personal problems passed to subclinical little bothering.
At the 3-month follow-up, all measures maintained clinical significance and reliable change, with anxiety returned to subclinical mild range, whereas the other measures remained in the previous range.
At the 6-month follow-up, all measures maintained clinical significance and reliable change. Depression (0) passed to the healthy range, and personal problems (2) passed to the very little bothering range. Table 2 shows the 11 problems that the client identified in her PQ at the beginning of the therapy and their duration. 7 problems were rated as maximum possible bothering, 2 were rated very considerably and 2 considerably bothering. All problems but relationship at work (item 7, 3-5 years) were identified as bothering the client for more than 10 years. Problems are related to 5 main areas: symptoms (1, sadness; 5, concentrating), specific performances (9, late; 11 put off), relationships (7, take advantage; 8 over adapt), self-esteem (2, importance; 10, feeling less) and emotions/inner experience (3, oppressed; 4, frustrated; 6, bashful).
At the middle, 8 out of 11 problems showed a reliable change, and 3 of these also a clinically significant change. At the end of the therapy, all problems showed a reliable change, and 9 out of 11 also a clinically significant change. At the first follow-up, 10 problems maintained reliable change and 7 of these also a clinically significant change. At the second follow-up, 10 problems maintained reliable change and 9 of these also clinically significant change. At the third follow-up all problems lasting for more than 10 years showed a clinically significant and reliable change, and the only problem lasting from 3-5 years showed neither reliable nor clinical change.

Qualitative Data
Caterina compiled the HAT form at the end of every session, reporting positive/helpful events and one hindering event. All positive events were rated from 8 (greatly helpful) to 9 (extremely helpful) and are reported in Table 3. The hindering event was reported in session 9 and rated 3 (moderately hindering): "I got here earlier believing I was late, I went away and then I got back (late), forgetting the money to pay the session. It has been hindering because this made me feel very anxious, which created in me this succession of events completely out of my control, which added up with other events that happened throughout my whole day".
She reported a rich description of therapeutic process, related to all five main areas reported in the PQ.
therapy, she felt "so exhausted" (C83) and she "would have liked to exchange my life with any other one" (C18), whereas now she reports taking her life back (C82).
Caterina summarised six main areas of change. First, she observed an improvement in her way of giving importance to her life. Caterina referred to being surprised by such a result (rated 5, very much surprised), unlikely without therapy (1) and extremely important (5). The second and the third changes she identified were the decrease of her senses of oppression and frustration, with both as somewhat surprised (4) and that the changes would have unlikely happened (1) without therapy, rating them as extremely important for her (5). The fourth improvement was her "increase of selfesteem" (5), which would have unlikely happened without the therapy (1) and considered as extremely important (5). The last two changes were "greater respect at work" and "less devaluation of important things", identifying them somewhat surprising (4), somewhat unlikely without the therapy (2) and very important (4). Caterina also reported that some friends of hers told her she is now a better person (C28-29). Caterina felt that some sessions were "really painful, but were those that allowed me to go on" (C21).

HSCED Analysis Affirmative Case
The affirmative team identified four lines of evidence supporting the claim that Caterina changed and that the therapy had a causal role in this change.
Change in stable problems Quantitative data (Table 1) show that there is a significant improvement in primary outcome measure (depression) that is clinically significant and reliable since the middle of the therapy and is maintained at the end and at 1-, 3-and 6-month follow-up, with a solid Reliable and Clinically Significant Improvement (RCSI). Secondary outcome measures depict an early RCSI in the anxiety (GAD-7) score, maintained throughout the follow-ups. At the end of the therapy there is also an RCSI for global distress (CORE), maintained at 1-, 3-and 6-month follow-up.
In the PQ (Table 2), Caterina identified 11 main problems at the beginning of the therapy that she was trying to solve, almost all rated as bothering her maximum possible (7). All problems standing from more than 10 years showed a RCSI at the 6-month follow-up. For these reasons, there is claim for a stable global reliable   Realizing I felt the need of protection (and that a man should have protected me). Identifying the paradox between the search of an ongoing autonomy and the frustration of not being able to guarantee it. Or having to suffer "anything" to guarantee it to me. Recognizing the 8 year old girl inside me, who hasn't received the possibility to "fly high", seeking something that she thought was the greatest expression of herself, an idea of happiness.
I understood I can't ignore my child side, if it remembers me my/its neglected needs.
Identify sadness as an ongoing mood. It remembered me a book I'm reading. The point is that more or less we are always sad. In the end sadness is not recognized anymore. And so you are not sad.

9 (extremely)
Understanding the mechanism that makes me entrust to others, judges about myself and of who I am.
It has been important to understand that according to this mechanism I AM NOT, if others don't say what I am. I gained an emptiness to fill: I, independently from others.
Giving credit where credit is due to the two parts of me that are still dealing with my ex.

9 (extremely)
It has been explained to me that there's a middle zone between the pedestal of perfection and the deep of devaluation, where it is possible to live a good life.
It has been important and reassuring "focusing" this mechanism. I've found it very liberating and it seems to me that I now have a clearer goal.
Connecting the dynamics of the pedestal and the deep in my relationships. Saving me from "raping" myself.

9 (extremely)
Finding out how, inside me, the importance of the inside and the substance of the appearance coexist.
I understood which are the origins of the war inside me. I understood why I act in a certain way, aiming at seduction and appearance.
Seeing the Parent, the Adult, the Child and myself. It gave me a lot of optimism because no matter how long the path might be, it's not so uneven as I thought.
Identify the practical aspects like a sensation that, until today I felt like generalized apprehension, and find concrete answers that allow me to go over my obstacles, making it as a duty for myself (as a person with some value, who has necessities that deserve to be listened to).

9 (extremely)
When the therapist connected all my improvements to my giving more importance to myself as a person It has been important because I understood I possess a strong base upon which I can build anything. Or create a solid base to sustain everything else. Recognizing the value and dignity of a person as his/her needs and wishes.
Realising the aptitude to consider sentimentally people that until recently I would have considered out of reach, without any possibility. Maybe because it leads to observe other as people (like myself) that live in this world like myself, and that can consider me as I consider them. See me as I see them. Not considering myself invisible before me and before others.

Tabl e 3: Cat erina's helpf ul as pect of therapy (HAT for ms)
Note. The rating is on a scale from 1 to 9: 1 = extremely hindering, 5 = neutral, 9 = extremely helpful. HAT = Helpful Aspect of Therapy .
change (reliable change in at least two out of three measures) in quantitative outcome measures. Qualitative data support this conclusion: in fact, in her Change Interview (CI) Caterina reports as a main achievement in therapy giving importance to her life, a long-standing problem (more than 10 years). She also reports that she changed her way in approaching life (CI, C13), in relating with others and her availability in opening up to others (CI, C26). She reports that friends saw her as a "better person" (CI, C28-C29).
Reading the session's transcriptions, from session 12 Caterina showed up with a higher mood, that is reflected in the scores of the outcome measures. In fact, in session 11, they worked on Caterina's tendency to suppress herself as a defence mechanism (Table 3, HAT 11), originated when she was a child and her mother made fun of her. She understood she needs to feel OK and love herself as her mother didn't do. This very intense session lead Caterina to skip the following one, breaking the alliance with the therapist. Nevertheless, this helped Caterina rebuild the therapeutic alliance, triggering a new way of relating with others and giving importance to herself and her life. Thus, we claim that Caterina obtained a stable RCSI in major depressive disorder, in general anxiety disorder, in global distress and in long standing personal problems, in main areas such as symptoms, relationships, specific performances, selfesteem, emotion and inner experience.

Retrospective attribution
Caterina recognised in her Change Interview six important changes in different aspects of her life, which she attributes to therapy (Table 4). She also re-examines all PQ items, scoring for each one its improvement and importance. All her improvements are considered very or extremely important, all surprising or almost surprising and all unlikely or quite unlikely without the therapy. She recognised that therapy allowed her to give more importance to her life and taking her life back, which was her therapy contract. Before beginning the therapy she would have given up her life for any other one, whereas now she understands that she is able to "fix" her life (CI, C18). The client asserts that the therapy was very useful to her and that it was exactly what she needed to get better (CI, C8): "now it's difficult for me to feel so bad like before starting the therapy" (CI, C14). In fact, she did not expect all these improvements for so long standing problems (CI, C36-C37), that without therapy would have been impossible to happen (CI, C82). She also affirms that the most painful sessions were the ones that allowed her to move on and work on herself (CI, C21). Previously, Caterina had been in therapy for four years, referring that "in four years I have never felt such big changes as I did in such a short time in this one" (CI, C85). From session 11, when the therapist asked her how she would like others to describe herself, she noticed that everything she underlined already belonged to her. This achievement is recalled in session 15, when speaking about her problems, Caterina realised that she reached different goals throughout the therapy (Table 3, HAT 11,15). For these reasons, we claim that the therapy had a causal role in Caterina's change.
Association between outcome and process (outcome to process mapping) The HAT completed at the end of each session provides us with regular and immediate reports of what Caterina found helpful in each session. All reported positive events are considered greatly or extremely useful and are coherent with both the diagnosis, the treatment plan and the interventions reported in the therapist's notes. In particular, it is important to notice the therapeutic focus since the first session on applying in daily life the achievement; an attitude that is maintained throughout the therapy. Thanks to the therapist's work, some items of the PQ (Table 2, item 1, 2, 7 and 8) show a clinically significant and a reliable change from session 12, maintained throughout the follow-ups, demonstrating an improvement in old aspects in her interpersonal life (Table 3, HAT 12,13,14,16). In fact, in the HAT Caterina writes about these mechanisms used throughout the session (Table 3, HAT 2,5,9). Her work on her first two main changes ("Decrease sense of oppression" and "Decrease sense of frustration") can be seen since HAT 4 and again in HAT 13; "I give importance to my life" and her feeling of having "greater respect at work" has been Change How much expected change was (a) How likely change would have been without therapy (b) Importance of change ( The rating is on a scale from 1 to 5; 1= expected, 3 = neither, 5 = surprising. b The rating is on a scale from 1 to 5; 1=unlikely, 3 = neither, 5 = likely. c The rating is on a scale from 1 to 5; 1 = slightly, 3 = moderately, 5 = extremely. focused specifically in HAT 6 ("It is useful to focus on what I don't want for myself" and "It has been useful to share […] the mechanisms that keep me tied to my job (which completely absorbs my life)"). Again, "I give importance to my life" and her "Increase of self-esteem" can be seen specifically in HAT 16 ("When the therapist referred me to all my improvements, that was giving me more value as a person") Event-shift sequences (process to outcome mapping) The PQ mean score shows a progressive decrease of problems' severity from the initial score (5.7, very considerably) to the final score (2.3, very little). The therapist's confrontation of the client's tendency to not give value to her life and feeling that others are more important than her (session 1), reflected respectively in the PQ item 1 and 2, that decreased since session two, became RCSI in session five and maintained through the follow-ups. Self-report data also shows a substantial change starting from session 11, thanks to the use of the rechilding technique (S11, C24), which allowed Caterina to recognise her anger (Table 3, HAT 11).
Sceptic Case 1.The apparent changes are negative (i.e. involved deterioration) or irrelevant (i.e. involve unimportant or trivial variables). According to quantitative data, Caterina's depression reached an early RCSI, maintained at the end of the therapy and throughout the follow-ups. Despite it, in session 16 she reports feeling still "depressed in specific contests" (C16). For this reason, the changes reported in quantitative self-reported measures appear not supported by client's statements. In the Change Interview, she also reports some changes that she feels being negative for her, like feeling "less responsible, […] less disposed to be always available, day and night, for anything" (C35). She also reports "I should have more concentration, I should better optimise my time, I should have a schematic control of time and things, which I still don't have, because I'm always late, I lose myself, I'm distracted, so I don't believe I should allow myself to tone down my sense of duty and my responsibilities" (C35). Regarding her problems of relationships, in session 14 she reports having troubles in creating new relationships (C40). Furthermore, any positive change can be attributed to her past four years of therapy. Even if quantitative data support a positive change, it is highly improbable that such an improvement could have happened in only 16 weeks of therapy.
2. The apparent changes are due to statistical artefacts or random errors, including measurement error, experiment-wise error from using multiple change measures, or regression to the mean. The sceptic team were not able to find any evidence within the rich case record which would support a claim that Caterina's changes were associated with statistical artefacts or random errors.
3. The apparent changes reflect relational artefacts such as global hello-goodbye effects on the part of a client expressing his or her liking for the therapist, wanting to make the therapist feel good, or trying to justify his or her ending therapy. Even if Caterina in her CI and in her HAT forms did not report only positive comments/helpful events about the therapy and the therapist, (see Table 3, session 9), the sceptic team believes that Caterina's improvement may be biased by her tendency to Please Others, in line with her dependent personality and submissiveness traits and over-adjustment. In fact, at the end of the therapy, the item 7 of her PQ ("In relationships I over-adapt") is still scored 4 (moderately bothering).
4. The apparent changes are due to cultural or personal expectancy artefacts; that is, expectations or scripts for change in therapy.
Having been in therapy for four years and having her younger sister in therapy too might have unconsciously led Caterina into expecting something would have changed in a short time.
5. There is credible improvement, but it involves a temporary initial state of distress or dysfunction reverting to normal baseline via corrective or self-limiting processes unrelated to therapy. The sceptic team were not able to find any evidence within the rich case record which would support a claim that Caterina's changes were associated with a reversion to normal baseline via corrective or self-limiting processes unrelated to therapy.
6. There is credible improvement, but it is due to extratherapy life events, such as changes in relationships or work. When Caterina went on vacation (between session 7 and 8), all her scores dropped (PHQ-9 dropped from 12 to 5, reaching RCSI; GAD-7 from 12 to 6, also with RCSI; CORE from 20,6 to 11,8, with reliable change; and PQ from 4.6 to 4), but in session 9, all four measures returned to their previous score (PHQ-9 to 11; GAD-7 to 12; CORE to 18,9; and PQ went to 4.91, higher than before her holiday). Thus, the early change claimed by the affirmative team appears tied to vacation, rather than therapeutic effect. Furthermore, at the end of the therapy, she says that she was thrilled to participate in a formation program where she wanted to propose some innovations inside her company (S16, C29-30). As holidays helped her to get better, this event might have led her to feel better.
7. There is credible improvement, but it is due to psychobiological processes, such as psychopharmacological mediations, herbal remedies, or recovery of hormonal balance following biological insult. The sceptic team were not able to find any evidence within the rich case record which would support a claim that Caterina's changes were associated with psychobiological processes.
8. There is credible improvement, but it is due to the reactive effects of being in research. In the Change Interview, Caterina reported that the research initially blocked her, making her feel the therapy was mechanical and difficult (C11), especially for being forced to make every single aspect of her life fit into a categorical definition (C86).

Affirmative Rebuttal
We can claim that all four measures support a Global Reliable Change. In only 16 sessions, Caterina made great improvements, reporting that she did not expect such a big change in a so short a period of time (CI, C37). Her being late and distracted is a passive-aggressive and oppositional defiant trait of her personality and changing difficult personality traits is a work that cannot be fully accomplished in only sixteen sessions. Even if she reports feeling still a little depressed, quantitative data show that there is a global and stable change in Caterina's depression, to a score of 0 at the 6-month follow-up. Furthermore, in session 16, speaking about the formation program in her work place, she was willing to propose a continuing education course for more efficient communication (C30), showing that she wanted to improve this aspect in which she feels she lacks. In fact, in the CI she reports a change in her way of approaching others and to life (C13). Also, according to what Caterina said, she found this actual therapy to be more efficient than the previous one because she felt the therapist was more empathic (CI, C83), declaring that "comparing these two therapies, this one is better" (CI, C85). Besides, she never speaks of the previous therapy, whereas she reports gaining more benefit from this one. Caterina also reports feeling better only after painful sessions (CI, C21). If she was complaisant towards the therapist, she would not have said she suffered. About extra-therapy events, there is no evidence that reports an improvement due to her participation in the formation program. Finally, her difficulties in dealing with self-report are only present at the beginning of the therapy, in fact she says: "it wasn't so difficult after all, and slowly it became natural and I didn't feel it so difficult […], it was just an initial block" (CI, C11).

Sceptic Rebuttal
The sceptic team believes that Caterina's change is principally due to her previous therapy and that she needed this second one only to resume and fix the previous therapy work. If she will not continue with the therapy after the 6-month follow-up, she will inevitably return to her previously dysfunctional state of depression and anxiety.

Affirmative Conclusion
Caterina's depression, anxiety, global distress and personal problems were tied to childhood experiences of being devaluated when she was taking decisions, which led the client to have many difficulties in interpersonal patterns and intrapsychic patterns relating to inner experience, emotions, self-esteem. The therapist created from the beginning a climate where the client explored appreciations of herself, expression of emotions such as sadness and anger, and achieved a new comprehension of her inner experience, allowing herself to relate with others and give value to her life. Furthermore, the therapist focused on Caterina's selfcritical ego state internal dialogue, self-esteem, sense of identity, with regressive techniques. These experiences were reflected in changes in internal dialogues, interpersonal relationships, depressive symptoms, and personality traits of depressiveness, submissiveness, anxiety. Caterina's drop out between session 11 and 12 helped her to create a stronger alliance with the therapist, which affected her way of relating with others.

Sceptic conclusion
Caterina asked for therapy after a two-years suspension of a four-years therapy, consequent to her sister's doctor's advice. Her trait of personality (submissiveness, dependent) affected her relationships with the therapist and probably her outcome scores. Changes in intrapsychic and interpersonal patterns are probably due to the previous therapy and to the reassuring effect provided only by the presence of the therapist on her personality traits.

Adjudication
Each judge examined the rich case record and hermeneutic analysis and independently prepared their opinions and ratings of the case ( Table 5). The judges' overall conclusions are that this was an outstanding clearly good outcome case, that the client made substantially to completely changes, and that the changes are substantially to completely due to the therapy.
Opinions about the treatment outcome (good, mixed, poor) Judge A (VC). This case appears to be a clearly good outcome (100% certainty). Quantitative data show a reliable and clinically significant change on all measures of primary outcome (PHQ-9) and secondary outcome (GAD-7, CORE, PQ) at the end and through 1-, 3-and 6month follow-ups. Personal problems rated as lasting for more than 10 years present a clinical and reliable change, maintained through the follow-ups. It appears evident that there is a Global Reliable Change. Qualitative data from Change Interview clearly support such conclusion.
Judge B (SM). This is a clearly good outcome (80% certainty). Despite outstanding evidences of good change on quantitative measures, qualitative reports of the client support the conclusion that quantitative scores may be biased by personality traits.
Judge C (AP) This case is classifiable as good outcome case (100%). This opinion is based on quantitative measures and qualitative data that are coherent in indicating a stable global change in long-standing problems.
Opinions about the degree of change Judge A. The client changed substantially (80% with 100% certainty). Quantitative measures support the claim that depressive symptoms are in the healthy range six months after the conclusion of the therapy, indicating a change in persistent, long standing depressive symptomatology. The clear improvement in anxiety symptoms, global distress and long standing personal problems suggest that the therapy, despite focused on depression, deeply changed personality traits.
Judge B. The client changed substantially (80% with an 80% certainty). Qualitative data suggest that in daily life the client experienced new ways to relate with others and a renewed self-esteem and inner experience.
Judge C. The client showed a complete change (100% with 80% of certainty), as showed in quantitative and qualitative data. With respect to the beginning, there is a global change in symptoms, relationships, perception of self. Hermeneutic analysis illustrated deep change in daily life that are beyond those expected in a short-term psychotherapy.
Opinions about the causal role of the therapy in bringing the change Judge A. The observed change is substantially (80% with 100% of certainty) due to the therapy. HAT and Change Interview present rich descriptions of change in the client's life and their connections with the therapist's interventions. Specific homework addressed the main daily difficulties of the client and were discussed with great attention to the therapeutic alliance. The therapist tends often to connect the experiences outside the therapy to what is happening within the session, allowing the client to experiment with change in maladaptive patterns within the secure therapeutic relationship, and then fostering the generalisation of the change within relationships outside the therapy.
Judge B. Change is substantially (80% with 80% of certainty) due to the therapy. There are clear statements in the Change Interview where the client affirms that the changes in her Personal Questionnaire were due to the therapy, and unlikely without it. The client presents a rate of change that is not usual in a short-term psychotherapy, probably due to the previous experience of psychotherapy, that acted as a solid base for the actual change.
Judge C. The change appears completely due to the therapy (100% with 100% of certainty). The client refers in her Change Interview to many important changes, unexpected and unlikely without therapy. We have no information about the previous therapy, but in the session transcription it appears that the actual change is not due to past or present external factors.

Mediator Factors
Judge A. Techniques such as regression to archaic relational episodes appear tied to deep and stable change in self-perception and relational patterns. The therapist explained the ego state model in early sessions and the client used often the specific language of the model, suggesting that the comprehension of what is going on may improve therapeutic alliance and psychotherapy process.
Judge B. The therapist challenged in an active way the beliefs and behaviours of the client, supporting imagination of what could happen from changing her way to think and to stay in relationship. The therapist also fostered the application of the new comprehension of self in the real relationship, accelerating the process of change. There is some doubt about the missed appointment after session 11, which may suggest an excessive burden of active interventions. Despite it, the therapist used the event for strengthening the therapeutic alliance, allowing the client to express her fantasies and emotion on the event. Judge B. the therapist appears able to create an affective climate that can hold rupture of therapeutic alliance.
Judge C. The client appears able to explore immediately, since the first sessions, the inner world, probably due to the previous therapy.

Discussion
This case aimed to investigate the effectiveness of a manualised TA treatment for depression in a client with moderate level of major depressive disorder (MDD) and general anxiety disorder. Primary outcome was depressive symptomatology, that showed an early reliable and clinically significant change since session 8, maintained at the end of the therapy and through the 1-, 3-, and 6 month follow-ups. Secondary outcomes were that anxiety, global distress and severity of personal problems all showed a reliable and clinically significant change at the end of the therapy and through the three follow-ups. The therapist conducted the treatment with a good to excellent adherence to the manual. Hermeneutic analysis pointed out changes in stable problem, retrospectively attributed to the psychotherapy, highlighting connections between outcome and process. The judges concluded that this is a clearly good outcome case, with a substantial to complete degree of change, substantially to completely due to the therapy. The treatment appears to be effective also for anxiety symptoms, suggesting that common mental health disorders such as depression and anxiety might share a common etiopathogenetic mechanism.
The therapeutic alliance appears to have been built on an active style, focused on personality traits associated to symptoms, transference and countertransference analysis. Specific TA techniques were: early sharing of the ego state model, exploration of inner dialogue, developing of Nurturing Parent, exploration of drivers Be Perfect and Please Me, racket analysis of guilt and sadness. This result appears partially moderated by previous treatment, that probably facilitated therapeutic alliance and early, deep exploration of interpersonal and intrapsychic maladaptive patterns.

Limitations
The first author has a strong allegiance to TA, is a teacher of the members of the hermeneutic groups and a colleague of the three judges. The author was also funded for this research by TA institutions (see Funding below). Despite the reflective attitude adopted in this work, these factors may have influenced in subtle ways both the hermeneutic analysis and the judges' evaluations.

Conclusion
This case study provides evidence that the specified manualised TA treatment for depression  has been effective in treating a major depressive disorder associated with generalised anxiety in an Italian client-therapist dyad. Despite results from a case study being difficult to generalise, this study adds evidence to the growing body of research supporting the efficacy and effectiveness of TA psychotherapy, and notably supports the effectiveness of manualised TA psychotherapy for depression as applied to persistent depressive disorder.

Introduction
This study is the third of a series of seven, and belongs to the second Italian systematic replication of findings from two previous case series  and was conducted under the auspices of the European Association for Transactional Analysis (EATA) and the University of Padua.
Major Depressive Disorder (MDD) affects all age groups, and is considered the fourth leading cause of disability in Europe and North America, calculated by Disability Adjusted Live Years (DALYs) (Murray, Vos, Lazano et al., 2012). Depression in childhood and adolescence has an estimated prevalence of 2,8% amongst children and 5,9% amongst adolescents (Costello, Erkanli & Angold, 2006). This figure rises to almost 25% of adolescents attending primary care settings, where it is possible to find depressive symptoms below the diagnostic threshold for MDD (but which cause significant distress and impairment in functioning), a condition termed Subthreshold Depression (SD) that is considered on a continuum of severity with MDD. Indeed, SD is often considered to be a precursor of MDD (Wesselhoeft, Pedersen, Mortensen et al., 2015). In childhood and adolescence, it is also common to see a clinical presentation of comorbid anxiety and affective disorders, with some evidence that the former anxiety precedes, and could cause the latter affective disorder (Seligman & Ollendick, 1998). Considering that SD seems to precede MDD and that pre-pubertal onset of MDD leads to worse outcome than onset in adulthood (Van Noorden, Van Fenema, Van der Wee, Zitman & Giltay, 2012), it appears appropriate to develop standardised interventions targeting SD and MDD in childhood and adolescence.
Transactional analysis (TA) is a widely-practiced form of psychotherapy, supported with a vast literature (for a review see , but still it is under-recognised within the worldwide scientific community of psychotherapy. In order to define TA psychotherapy as an efficacious Empirically Supported Treatment (EST), its efficacy must have been established in at least one Randomised Clinical Trials (RCT) replicated by two independent research groups, or alternatively in at least three Single Case Experimental Design studies (SCED), replicated by at least two  or three  independent research groups, with each group conducting a case series of a minimum of three cases, without conflicting evidence. Recently, a wide community of researchers proposed that efficacy and effectiveness in psychotherapy are a complex object that cannot be adequately evaluated with either the experimental approach of RCT Westen, Novotny & Thomson-Brenner, 2004) or classical SCED such as reverse design . Systematic case study research has been proposed as a viable alternative to RCT and SCED (Iwakabe & Gazzola, 2009). Hermeneutic Single Case Efficacy Design (HSCED; Elliott, 2002;Elliott et al., 2009) is nowadays considered the most comprehensive set of methodological procedures for systematic case study research, and is a viable alternative to SCED in psychotherapy . HSCED is gaining momentum and enhanced versions have been proposed by different research groups, both to validate new psychotherapeutic approaches and to extend a previously validated psychotherapy to new disorders (e.g., . Recently, a systematic review of all published HSCED studies found within English language peer-reviewed journals (Benelli, De Carlo, Biffi & McLeod, 2015) highlighted methodological issues related to different levels of stringency, offering solid alternatives for conducting sound research according to the available resources within practitioner research networks.
Considering that approaches without evidence from RCTs tend to be under recognised, Stiles, Hill and  proposed collecting a series of mixed methods systematic case studies as the first step toward recognition of marginalised and emerging models of psychotherapy. Systematic case study research has already been applied to investigate the effectiveness of TA for people with long term health conditions  and HSCED methodology has been successfully applied to TA and widely described in this Journal by . Recently, several HSCEDs supporting the effectiveness of TA treatment for depression  have been published, as was an additional adjudicated study, which demonstrated effectiveness of TA for mixed depression and anxiety . Furthermore, a related study was published on TA for emetophobia (Kerr, 2013). These case series have shown that TA can be an effective therapy for MDD when delivered in routine clinical practice, in private practice settings, with clients with mild to moderate impairment in functioning who actively sought out TA therapy and with white British and Italian therapist and client dyads. Currently, no systematic research has been conducted on TA treatment for depression in childhood and adolescence.
The aim of this study was to investigate the effectiveness of the manualised TA treatment for depression in adults  for use with adolescents who present with depression. Reviewing the literature, we would expect MDD in adolescence to be characterised by a progression along a continuum beginning with SD and comorbid Generalized Anxiety Disorder (GAD). The present study analyses the treatment of 'Deborah', a 15year-old Italian young girl who had been suffering from depressive and anxiety symptoms, with a personal history of self-harm, cannabis use and self-induced vomiting, which had emerged in the previous year and had been steadily getting worse in the last few months. The quantitative primary outcomes investigated were depressive and anxious symptomatology; the secondary outcomes were client-generated personal problems and behavioural problems. Quantitative and qualitative analyses were conducted.

Ethical Considerations
The research protocol follows the requirements of the ethical code for Research in Psychotherapy of the Italian Association of Psychology (AIP, 2015), and the American Psychological Association guidelines on the "rights and confidentiality of research participants" (APA, 2010, p. 16). The research protocol has been approved by the Ethical Committee of the University of Padua. Before entering the treatment, the client and her parents received an information pack, including a detailed description of the research protocol, and they gave a signed informed consent and written permission to include segments of disguised transcripts of sessions or interviews within scientific articles or conference presentations. They were informed that the therapy would be provided even if they decided not to participate in the research and that they were able to withdraw from the study at any point, without any negative impact on their therapy. All aspects of the case material were disguised, so that neither the client nor third parties are identifiable. All changes are made in such a way that does not lead the reader to draw false conclusions related to the described clinical phenomena. Finally, as a member checking procedure, the final article, in Italian language, was presented to the client and her parents, who read the manuscript and confirmed that it was a true and accurate record of the therapy and gave their final written consent for its publication.

Method Inclusion and exclusion criteria
Psychotherapists participating in this case series were invited to include in their studies the first new client, with a disorder within the depressive spectrum as described in DSM-5 (Major, Persistent or Other Depressive Disorder), who agreed to participate in the research. Other current psychotherapy, active psychosis, domestic violence, bipolar disorder, antidepressant medication, alcohol or drug abuse were considered as exclusion criteria. As the overall aim of this project is to study the effectiveness of TA psychotherapy in routine clinical practice, comorbidity is normally accepted and both inclusion and exclusion criteria are evaluated case by case. In this study, the occasional recreational use of marijuana was not considered an exclusion criterion, considering its high diffusion amongst adolescents and that there was no evidence that this habitual use was indicative of a significant physical addiction which would require specialist medical intervention.

Client
Deborah was a 15-year-old white Italian female adolescent, who lived with her parents and two younger siblings in a large city in north Italy. She was an intelligent, curious, altruistic and resourceful teenager, with many positive values, and who had good selfreflective and evaluative capacities. At the age of 12 she noted some strange symptoms and self-diagnosed (through symptom-checking on the Internet) a metabolic disorder, which was confirmed by subsequent medical investigation. She gained much weight and followed a very strict self-imposed diet, frequently not eating and sometimes inducing vomiting. Deborah attended the second year of high school, with a wide network of relationships but recently had become withdrawn. She had a tendency to easily forgive others, even when they hurt her, and felt "evil and guilty" when she did not. She often self-inflicted injuries on her wrists and ankles, and pulled her hair and picked at her fingernails until they were sore. She had a boyfriend, whom she described as critical and devaluing. She ended this relationship around one month after the onset of therapy. Her mother suffered from depressive disorder. Her father had many episodes of alcohol and drug abuse and was afflicted with a cancer for which he was being treated with a cycle of chemotherapy. Deborah reported having difficulties in her relationship with her parents, especially that she felt unable to communicate with her father. She felt that she was not accepted, not understood, and felt attacked and criticised, and also felt that she was lacking guidance and protection from her parents, and felt strongly guilty for disappointing them.
She independently decided to seek therapy, showing a strong motivation for treatment. In the first appointment, which her parents also attended, the therapist noted that Deborah's parents appeared relieved that they could delegate their daughter's cure to somebody else.

Therapist
The psychotherapist was a 54-year-old, white, Italian woman with 20 years of clinical experience, and who is a Provisional Teaching and Supervising Transactional Analyst (PTSTA-P). For this case, she received monthly supervision by a Certified Transactional Analyst Trainer (CTA Trainer) with 30 years of experience.

Intake sessions
The therapy was conducted in a public clinic, once a week and free of charge. The client attended two pretreatment sessions along with her parents and three individual pre-treatment sessions (0A, 0B, 0C), which were focused on explaining the research project, obtaining consensus, conducting a diagnostic evaluation according to DSM-5 criteria (American Psychiatric Association, 2013), defining the problems she was seeking help for in therapy along with their duration and severity, developing a case formulation including TA diagnosis, treatment plan and contract, and collecting a stable baseline of self-reported measures for primary (depression and anxiety) and secondary (personal problems and problematic behaviours) outcomes.
Deborah asked to learn how to protect herself, how to express her needs, thoughts and emotions to others, especially her parents, how to regulate her intense and overwhelming emotions and stop self-injuring behaviours.

DSM-5 Diagnosis
The initial diagnostic phase identified the client's primary diagnosis. Deborah was assessed as meeting DSM 5 diagnostic criteria for moderate MDD: she experienced depressed mood in daily activities for more than one year, most of the day, nearly every day (criterion A1), decreased pleasure in most activities (A2), weight change (A3), restlessness regulated with self-harm (A5) overwhelming feelings of worthlessness and inappropriate guilt (A7), diminished ability to think and concentrate (A8). She also met criteria for moderate GAD: excessive anxiety and worry (criterion A), that were uncontrollable (B), with restlessness (C1), difficulty in concentration (C3) and irritability (C4).

TA Diagnosis and Case formulation
Case formulation was conducted according the TA diagnostic categories presented in the treatment manual. Deborah assumed a life position (Ernst, 1971;Berne 1972) I'm Not OK, You're OK, that interacted with her stroke economy (Steiner, 1974), which was characterised by an absence of positive strokes and abundance of negative strokes. This in turn led to internalisation of an under-active and under-functioning internal Nurturing Parent and an over-active internal Critical Parent, which activated intense self-critical internal dialogues (Kapur, 1987). Furthermore, the underlying Injunctions (Goulding & Goulding, 1976;McNeel, 2010): Don't be a child (be adult and take care of your parents); Don't think (avoid problem solving) Don't be important (the other's needs are more important than yours), and Don't feel (especially angry when abused) were also identified. These led to the observable drivers (Kahler, 1975) of Try Hard and Please Others and the assumption of drama triangle roles (Karpman, 1968) such as Rescuer with parents and when forgiving friends, Victim when feeling helpless, frail and unable to protect herself, and Persecutor when her active Critical Parent was externalised in conflictual transactions with parents. Script conclusions and decisions (Berne, 1961) were observable through script beliefs and contaminations (Berne, 1961;Stewart & Joines, 1987 such as: "I am wrong" "others are more important than me", "I cannot be angry with others", "I must forgive and please others", "I must take care of and support my parents", "my body is not pretty therefore I don't deserve love", "I am evil because I disappoint my parents", "others cannot understand what I mean", "when you have a problem let the time pass away, wait and do not attend to it". The script system (Erskine & Zalcman, 1979;Erskine, 2010) involved all of the above-mentioned thoughts and behavioural manifestations, as well as repressed primary anger when she receives abuse or is not loved and considered by others, which was covered by secondary sadness, helplessness, feelings of being unprotected and unlovable, with worry and restlessness, self-injury behaviours, which in turn triggered the memory recall of episodes of criticism and neglect.

Treatment
The therapy followed the manualised therapy protocol of , including the 12 Key tasks and the research-based principles. Throughout the whole treatment, the therapist focused on 1) building the therapeutic alliance providing empathic listening, 2) giving strong support to the client's self-esteem and recognising her resources and positive strengths; 3) developing the observing self and TA problem solving protocol, in order to enhance Adult functioning, and 4) permeating the sessions with permissions (Crossman, 1966), especially those congruent with the client's injunctions, namely: be and express your needs as adolescent, be important, think (about the consequences on your future), feel and express all your primary emotions, especially anger when you are abused, thus providing systematic implicit experiential disconfirmation and modelling a positive and potent Nurturing Parent. In the first phase (sessions 1-6) the focus was on the recognition and decontamination of script beliefs, emotional literacy and emotion regulation. In the second phase (sessions 7-12) the therapist focused on changing internal dialogue from critical to nurturing and enhancing the client's internal Nurturing Parent. In the third phase (sessions 13-16), the focus was on problem solving strategies, and emotional and behavioural regulation in daily situations.

Analysis Team
The HSCED main investigator and first author of this paper is a Provisional Teaching and Supervising Transactional Analyst (Psychotherapy) (PTSTA-P) with 10 years of clinical experience, with a strong allegiance for TA. Despite recent literature suggesting that hermeneutic analysis should be carried out only by expert psychotherapists , we decided that when the research is investigating a new population or a therapy that lacks a research base, it is appropriate to follow , who proposed that analyses can be carried out by a team of 'reasonable persons', not yet overly committed to any theoretical approach or professional role. The team comprised six postgraduate psychology students who were taught the principles of hermeneutic analysis by Professor John McLeod, in a course on case study research at the University of Padua. Following the indication of Elliott et al (2009), the students preferred to assume both affirmative and sceptic positions, and independently prepared their affirmative and sceptic cases. Then they met and merged their own cases, supervised by the main investigator, creating a consensual affirmative and sceptic brief and rebuttals.

Transparency statement
The research was conducted entirely independently of the previous case series (see . The last author, Mark Widdowson, was involved in checking that the research protocol and data analysis process was adhered to, in order to make the claim that this case series represents a valid replication of the initial study (with minor changes) and he was involved in the final preparations of this article.

Judges
The judges were three researchers in psychotherapy at the University of Padua and co-authors of this paper: Judge A, Vincenzo Calvo, clinical psychologist, psychotherapist trained in dynamic psychotherapy, PhD in development psychology, with expertise in attachment theory; Judge B, Stefania Mannarini, psychologist with experience in research methodology; and Judge C, Arianna Palmieri, neuropsychologist and psychotherapist with a training in dynamic psychotherapy. Judges A and C had some basic knowledge of TA but had never engaged in any official TA training, whereas Judge B has some clinical experience but no knowledge of TA.

Quantitative Outcome Measures
Four standardised self-report outcome measures were selected to measure primary target symptoms (depression) and secondary symptoms (anxiety and global distress).
Patient Health Questionnaire 9-item for depression (PHQ-9;, which scores each of the nine DSM-5 criteria from 0 (not at all) to 3 (nearly every day), which has been validated for use in primary care . Total scores up to 4 are considered healthy, scores of 5, 10, 15 and 20 are taken respectively as the cut-off points for mild, moderate, moderately severe and severe depression. PHQ-9 score ≥10 has a sensitivity of 88% and a specificity of 88% for major depression (Kroenke, Spitzer, & Williams, 2001) and scores of <10 are considered subclinical. A change of at least 6 points on PHQ-9 score is considered to assess a reliable improvement or deterioration (RCI).
Generalized Anxiety Disorder 7-item for anxiety (GAD-7;, which scores each of the seven DSM-5 criteria as 0 (not at all), 1 (several days), 2 (more than half the days), and 3 (nearly every day). Total scores of up to 4 are considered healthy, scores of 5, 10, and 15 are taken as the cut-off points for mild, moderate and severe anxiety respectively. Using the threshold score of 10, the GAD-7 has a sensitivity of 89% and a specificity of 82% for GAD and scores of <10 are considered subclinical. It is moderately good at screening three other common anxiety disorders -panic disorder (sensitivity 74%, specificity 81%), social anxiety disorder (sensitivity 72%, specificity 80%) and post-traumatic stress disorder (sensitivity 66%, specificity 81%) (Kroenke, Spitzer, Williams, et al, 2007). A change of at least 4 points on GAD-7 score is required in order to assess a reliable improvement or deterioration (RCI).
Youth-Self Report (YSR) (Achenbach, 1991;Achenbach and Rescorla, 2001) was rated at session 1, 8, 16, and at 6-month follow-up. YSR is a self-descriptive measure which investigates social competencies and behavioural problems in 11-18 year olds. There are 112 items, coded from 0 (not true in the last six months) to 2 (very often or often true in the last six months). Items measure eight subscale symptoms: Withdrawal, Somatic complaints, Anxiety and depression, Social problems, Thought problems, Attention problems, Aggressive behaviours, and Rule-breaking behaviours (Achenbach, 1991). The first three subscales are grouped into the Internalizing scale, the last two into the Externalizing scale, and the remaining three scales are categorised as Other problems. Overall behavioural and emotional functioning is measured by the Total problems scale. The sum of the scores for each scale and sub-scale may be converted to T-scores for which the manual gives the cut-offs for the clinical and borderline range for boys and girls. A change of at least 6.2 points on YSR Total problems score indicates a reliable improvement or deterioration (RCI).
The Personal Questionnaire (PQ) (Elliott, Shapiro, & Mack, 1999;Elliott, Wagner, Sales, Rodgers, Alves & Café, 2016) is a client-generated measure in which clients specify the problems they would like to address in their therapy and rate their problems according to how distressing they are finding each problem (1, not at all; 7, maximum possible). Scores up to 3.25 are considered subclinical. In this case series, missing the Italian normative score, for the PQ we adopted a more conservative RCI of two points, rather than the RCI of 1.67 recently proposed by Elliott et al. (2016). The PQ procedure suggests including problems from five areas: symptoms, specific performance or activity (e.g., work), relationships, mood/emotions and self-esteem/internal experience.
All quantitative outcome measures were evaluated according to Reliable and Clinically Significant Change (RCSC) (Jacobson & Truax, 1991). Clinical significance (CS) is obtained when the observed score on an outcome measure drops under a cut-off score that discriminates clinical and non-clinical populations. For example, the PHQ-9 considers a score of ≥10 as an indicator of current moderate major depression (Kroenke, Spitzer & Williams, 2001). It is important to consider that even under the cut-off score there may be a subclinical disorder. For example, the PHQ-9 considers a score between 0 and 4 an indication of healthy condition, and a score between 5 and 9 as an indicator of mild (subclinical) depression. Reliable Change Index (RCI) is a statistic that enables the determination of the magnitude of change score necessary to consider a statistically reliable change on an outcome measure (Jacobson and Truax, 1991). In particular, it is helpful in minimising Type I errors which occur when cases with no meaningful symptom change are assumed to have improved. For example,  proposed that a reduction of at least 6 points in the PHQ-9 score would be indicative of a reliable improvement. Only when we observe the presence of both CS and RCI, we have a RCSC, which is considered a robust method for assessing recovery in psychological interventions (Evans, Margison & Barkham, 1998;Delgadillo, McMillan, Leach, Lucock, Gilbody & Wood, 2014). To control experiment error which occurs when multiple significance tests are conducted on change measures, we consider that a RCSC is required in at least two out of three outcome measures, thus demonstrating a Global Reliable Change (GRC) .
All of these measures were administered in the pretreatment phase in order to obtain a three-point baseline, and during the three follow-ups. Deborah's quantitative data are presented in Table 1. Since the client expressed some discomfort in completing self-report measures, we chose to reduce the number of questionnaires the client was required to fill in prior to any session. Thus, we have scores of the GAD-7 only in the assessment phase and at the 1-month and 3-month follow-ups. The first PQ score was available in session 1.

Qualitative Outcome Measurement
The client was interviewed using the Change Interview protocol (CI) (Elliott, Slatick & Urman, 2001) one month after the conclusion of the therapy. The CI is a semistructured qualitative change measure which asks clients how they feel they have changed during the therapy and how they think these changes came about, what they felt was helpful or hindering in the therapy, and what changes they feel they still need to make. Clients are asked to identify key changes they made and to indicate on a five-point scale: 1) if they expected to change (1=expected; 5=surprising); 2) how likely these changes would have been without therapy (1=unlikely; 5=likely), and 3) how important they feel these changes to be (1=slightly; 5=extremely).
The client also completed the Helpful Aspects of Therapy form (HAT)  at the end of each session. The HAT allows the client to describe hindering or useful aspects of the session and to rate them on a nine-point scale (1=extremely hindering, 9=extremely useful).

Therapist Notes
A structured session notes form (Widdowson, 2012a, Appendix 6, p. 50-52) was completed by the therapist at the end of each session. In this form, the therapist provides a brief description of the session in which are identified key aspects of the therapy process, the theories and interventions used, and an indication of how helpful the therapist felt the session was for the client.

Adherence
The therapist, the supervisor, and the main researcher were all Transactional Analysts and they each independently evaluated the therapist's adherence to TA treatment of depression using the operationalised adherence checklist proposed by Widdowson (2012a, Appendix 7, p. 53-55) before agreeing on a final consensus rating. The conclusion of the three evaluators was that the treatment had been conducted coherently according to TA theory at a good to excellent level of application.

HSCED Analysis Procedure Affirmative Case
The affirmative position according to Elliott (2002) should locate evidence in the rich case record supporting the claim that the client has changed, and that the change is causally due to the therapy. A clear argument supporting the link between change and treatment must be established on the basis of at least two of the following five sources of evidence: 3. Outcome to process mapping: refers to the content of the post-therapy qualitative or quantitative changes that plausibly match specific events, aspects, or processes within therapy; 4. Event-shift sequences: links between client reliable gains in the PQ scores and significant within therapy events; 5. Within therapy process-outcome correlation: the correlation between the application of therapy principles (e.g. a measure of the adherence) and the variation in quantitative weekly measures of client's problem (e.g. PQ score).

Sceptic Case
A sceptic position requires a good-faith effort to find nontherapeutic processes that could account for an observed or reported client change. Elliott (2002) identified eight alternative explanations that the sceptic position may consider: four non-change explanations and four non-therapy explanations.
The four non-change explanations assume that change is really not present, and should consider: 1. Trivial or negative change which verifies the absence of a clear statement of change within qualitative outcome data (e.g. CI), and the absence of clinical significance and/or reliable change in quantitative outcome measures (e.g. PHQ9); 2. Statistical artefacts that analyse whether change is due to statistical error, such as measurement error, regression to the mean or experiment-wise error; 3. Relational artefacts that analyse whether change reflects attempts to please the therapist or the researcher; 4. Expectancy artefacts, analysing whether change reflects stereotyped expectations of therapy.
The four non-therapy explanations assume that the change is present, but is not due to the therapy, and should consider: 5. Self-correction which analyses whether change is due to self-help and/or self-limiting easing of a temporary problem or a return to baseline functioning; 6. Extra-therapy events that verify influences on change such as those due to a new relationship, work, or financial conditions; 7. Psychobiological causes which verify whether change is due to factors such as medication, herbal remedies, or recovery from medical illness; 8. Reactive effects of research, analysing the effect of change due to participating in research, such as generosity or goodwill towards the therapist.
The formulation of affirmative and sceptic interpretations of the case consists of a dialectical process, in which affirmative rebuttals to the sceptic position are constructed, along with sceptic rebuttals of the affirmative position.
Finally, each position is summarised in a narrative that offers a customised model of the change process that has been inferred, including therapeutic elements and an account of the chain of events from cause (therapy) to effect (outcome), including mediator and moderator variables.

Adjudication Procedure
Each single judge received the rich case record (session transcriptions, therapist and supervisor adherence forms and session notes, quantitative and qualitative data and also a transcript of the Change Interview) as well as the affirmative and sceptic cases and rebuttals by email, together with instructions. The judges were asked to examine the evidence and provide their verdict. They were required to establish: • If the case were a clearly good outcome case, a mixed outcome case, or a poor outcome case; • If the client had changed; • To what extent these changes had been due to the therapy; • Which aspects of the affirmative and sceptic arguments had informed their positions.
Furthermore, the judges had to observe which mediator factors in the therapy they considered to have been helpful and which characteristics about the client did they think had contributed to the changes as moderator factor(s).

Results
In earlier published HSCED's the rich case records, along with hermeneutic analysis and judges' opinions were often provided as online appendices (Benelli et al, 2015). Since all the material is in Italian language, we adopted here the solution of providing a summary of the main points, as proposed in MacLeod, Elliott and Rodger (2012). The complete material (session transcriptions, Change Interview, affirmative and sceptic briefs and rebuttal, judge opinions and comments) is available from the first author on request.

Quantitative Outcome Data
Deborah's initial depressive score (PHQ-9, 12) indicated a moderate level of depression. The anxiety score (GAD-7, 14) indicated a moderate level of anxiety. The severity score of personal problems (PQ, 4.9) indicated that the client perceived her problems as bothering her somewhere between moderately and considerably. YSR scores show that the sub-scale anxious/depressed symptoms was above the clinical cut off, whereas the sub-scale thought problems was borderline and the other sub-scales were in the normal range. Internalizing problems scale was in the borderline range and externalizing problems scale and total problems scale were in the normal range. Deborah reported in her YSR symptoms such as tics, nervous movements, confusion and difficulties in avoiding negative thoughts, especially severe self-criticism ("I'm wrong").
At session 8, (mid-therapy), depression passed into the subclinical mild range (6), with RCSC. Severity of personal problems decreased to little bothering (3.9). In the YSR, the Anxious/Depressed Thought Problems subscales decreased to the normal range whereas the Somatic and Social problems sub-scales increased to the borderline range. All the other sub-scales except that of attention problems increased from values around zero to close to the borderline cut-off. The Internalizing, Externalizing and Total scales were all within the borderline range. Deborah reported in her YSR an increase in alcohol and tobacco consumption and reductions in the use of inappropriate behaviours with adults and peers. At the same time, she described herself as very shy, suspicious, and unable to build intimate relationships with others and she reported episodes of self-harming behaviours. Between the somatic symptoms she reported dizziness, being tired without reason and vomiting.
By the end of the therapy, depression scores passed into the healthy range (3) maintaining RCSC, and personal problems decreased to very little bothering (2.6). Deborah's YSR score revealed a decrease in all the subscales except for the Anxious/Depressed one which was in the borderline range and for the Rule-Breaking Behaviour that increased to the clinical range. Similarly, while the Internalizing and Total scales slightly decreased to the bottom of the borderline range, the Externalizing scale increased to the clinical range. Deborah reported that she preferred to spend time with peers older than herself, that she had stolen things from home and from outside home, that she had used marijuana, and craved for new experiences.
At the 1-month follow-up, depression scores remained in the healthy range (3) with RCSC. Anxiety passed to subclinical mild range (8), with both clinical significance and reliable improvement (RCSC) compared to assessment. Personal problems returned to moderately bothering (4.3) without reliable deterioration.
At the 3-month follow-up, depression remained in the healthy range (3), anxiety remained in the mild range (5), and personal problems passed to be considered as not bothering at all (1.3), with RCSC.
Finally, at the 6-month follow-up depression returned to the mild range maintaining RCSC (5), personal problems returned to very little bothering, maintaining RCSC (2.3). YSR scores indicated a global increase in all the symptoms except for the Social problems sub-scale. In particular, the Rule-Breaking Behaviour and Anxious/Depressed sub-scales and all global scales were in the clinical range. A great number of items obtained the maximum score of 2 with the co-presence of anxious behaviours such as being nervous and worried, excessive crying and eating, and deviant behaviours such as the use of drugs, disregarding social and parents' rules and bullying behaviours. Table 2 shows the 10 problems that the client identified in her PQ at the beginning of the therapy and their duration. Problems are related to: symptoms (1, think; 3, hurt), specific performance or activity (4, tell; 6 stop, 9 be understood), relationships (7 attacked), mood (5, disappointing, 8, uncontainable) and self-esteem (2, fattening; 10 unaccepted). Two problems were rated as bothering her maximum possible, two as very considerably, two were rated considerably bothering, two as moderately. Two problems were rated under the clinical cut off, and hence without clinical significance. Three problems lasted from 3-5 years, two from 1-2 years, four from 6-11 months and one from 1-5 months, representing an almost stable and longstanding baseline. The longer lasting problems were related to relationships, mood and self-esteem, where symptoms were present for less than two years.
At the end of the therapy 5 out of the 8 problems above the clinical cut off showed a RCSC, and 2 showed a reliable improvement (RCI). At the first follow-up almost all items reliably deteriorated. At the 3-month follow-up, all the problems above the clinical cut off showed a RCSC, and four maintained RCSC at the 6-month followup. Her symptoms (items 1 and 3) showed the larger and more stable change.
Figures 1 to 4 below allow time series' visual inspection s of the weekly scores of primary (PHQ9 and GAD-7) and secondary (PQ and YSR) outcome measures. Concerning the YSR we report in Figure 3 the sub-scales that evidenced behavioural symptoms within the borderline or clinical range during the therapy Figure 1: Deborah's week ly depressive (PHQ-9) sc ore Note. 0A, 0B, 0C and 0D = assessment sessions. PHQ-9 = Patient Health Questionnaire 9-item for depression . FU = follow-up. Note. The first available score was in session 1. 0A, 0B and 0C = assessment sessions. FU = follow-up. PQ = Personal Questionnaire (Elliott, Shapiro, & Mack, 1999).

Figure 3: Deborah's Yout h Self Report (YSR) sub-sc ales T-scores
Note. FU = follow-up. YSR = Youth-Self Report (Achenbach, 1991). The grey area indicates the borderline range. Above the grey area scores are considered in the clinical range, below in the normal range.

Figure 4: Deborah's Yout h Self Report (YSR) scales T-scor es
Note. FU = follow-up. YSR = Youth-Self Report (Achenbach, 1991). The grey area indicates the borderline range. Above the grey area scores are considered in the clinical range, below in the normal range.  Qualitative Data Deborah compiled the HAT form at the end of every session (Table 3), reporting positive/helpful events and one hindering event. The hindering event was reported in session 6 and rated 3 (moderately hindering): "My family lack of Parents", suggesting a misunderstanding of the meaning of in-therapy hindering events. All positive events were rated from 7 (moderately helpful) to 9 (extremely helpful). She reported helpful aspects on: symptoms (1 guilty, 2 hurt, 16 addiction); relationships (6 I'm the child, 9 boundaries, 10 support, 11 change); mood/emotion (2 angry); self-esteem/inner experience (3 deserve, 5 feeling less, 7 accept, 8 worry about me, 12 special, 13 boundaries, 14 comfortable, 15 protect myself).
Deborah participated in a Change Interview 1-month after the conclusion of the therapy. In this interview she identified her main and significant changes (see Table 4). Deborah described her therapy as "a place where I could talk to a friend… it was a liberation, like emptying everything… It has been nice" (Client line 12), "because I succeeded in talking about things I normally don't speak of" (C6). "When I got back home I felt lighter" (C112), "it made me reach some goals that without a psychologist I would have never reached" (C5). When Deborah started the therapy, she used to hurt herself because she "needed to feel external pain, in order to stop feeling the pain inside" (C33), whereas now she is "not doing this anymore. I try to focus on other things, like school, I go out with my friends" (C33). At the beginning of the therapy, she used to think she was disappointing her parents, whereas now she does not think about it anymore (C33). Deborah in her CI did report two negative, obstructive or unpleasant aspects of therapy: firstly, that she found filling in the outcome measures as "boring" (C121) and secondly that she was "thinking too much" as a consequence of therapy (C52).
Four changes reported by Deborah are related to her ability to think and solve problems (items 1, 3, 6 and 11). Three changes are related to her initial symptoms: hurt herself (2), feeling guilt (4) and drug use (5). Other change refers to self-protection (7), self-esteem (9) and differentiation from parents (10). All changes are rated very or extremely important for her. Three changes are rated unlikely, five somewhat unlikely, one neither and two somewhat likely without therapy. Two changes were surprising, four somewhat surprising, two neither, one somewhat expected and two expected. According to Deborah, all these improvements happened because she "worked hard and found a very attentive therapist, who listened, understood and remembered things I told her, and also because back home you can think about what you said, what happened during the session" (C100).
Deborah also reported that her mother used to check her body each week, to ensure that she was not cutting herself, "but now I understand, and she understands that I'm different, that something has changed, and that I have changed, she is not checking my body any longer" (C59). The client also stated that therapy had been useful in helping her to change her way of thinking (C107).

HSCED Analysis Affirmative Case
The affirmative team identified four lines of evidence supporting the claim that Deborah changed and that the therapy had a causal role in this change.
Change in stable problems Quantitative data (Table 1) shows that there is an early improvement in depressive symptomatology measured with PHQ-9; visual inspection shows a clinically significant improvement since session 6, reliable since session 12, that is maintained until the end of the treatment and throughout the follow-up intervals. Anxious symptomatology, measured with GAD-7 in the phase 0 and then again at the first and second follow-up, shows a reliable and clinically significant improvement. YSR total problem scale was in the non-clinical range at the beginning of the treatment and showed a global, not reliable and not clinical deterioration during the treatment, that became reliable and clinically significant at the 6-month follow-up. The anxiety/depressive subscale was in the clinical range at the beginning of the treatment, and showed an early, clinically significant and reliable improvement in session 8, and in following assessment showed a deterioration, however not reliable in respect the pre-treatment score. In the PQ (Table 2), Deborah identified 10 main problems at the beginning of the therapy that she was trying to solve. All problems reached reliable and clinically significant improvement by the second follow-up (except item 9 and 10 that were in the subclinical range at pre-treatment assessment) and four maintained the reliable and clinically significant improvement at the six-month follow-up. All the stable problems, lasting from 1 to 5 years, showed a clinical and/or reliable change at the 6-month follow-up. Overall, there is support for claiming of a global reliable change.
Qualitative data supports these changes in stable problems. "Some aspects of my life changed drastically, before, when I woke up in the morning I thought 'Here we are, another day to deal with', whereas now I wake up and I feel normal" (C10). A very helpful aspect of therapy for Deborah was "feeling free from judgements" (C101). She also affirmed that there were no negative aspects, obstacles or unhelpful aspects to her therapy.
Association between outcome and process (outcome to process mapping) The HAT completed at the end of each session provides us with regular and immediate reports of what Deborah found helpful in each session. All reported events (but one) are considered moderately to extremely useful and are coherent with the diagnosis, the treatment plan and the interventions reported in the therapist's notes. Changes in depression and anxiety symptoms (Table 1), and in particular, feelings of guilt and worry, and personal problems (  Table 4, item 2) appear tied to the growing awareness and comprehension that she tended to cut herself, starve herself, self-induce vomiting, and made use of alcohol and drugs in order to punish herself for thinking that she had disappointed her parents (HAT, Table 3, session 2). Changes related to thinking and problem solving (CI ,  Table 4, item 1, 3, 6, 11) appear tied to intervention of the problem-solving protocol (HAT , Table 3, session 1, 3,8,9,15).

Event-shift sequences (process to outcome mapping)
The greater effect on depressive symptoms appeared to be tied to interventions in first sessions, on changing the self-critical internal dialogue associated with her feelings of guilt and worthlessness. Interventions on changing this internal dialogue and self-esteem are mirrored in HAT forms (session 1, 2, 4, 5) and reflected in subsequent changes in PHQ-9 scores on item 2 feeling down, depressed and item 6 feeling bad about yourself. Since the beginning of therapy, the therapist focused on Deborah's self-injury, connecting it with unexpressed anger (Table 3, HAT session 2), that led to an improvement in her PQ scores (Table 2, item 3), which was maintained at the six-month follow-up. In session 3, the therapist worked on decontaminating script beliefs about Deborah's way of relating to others. Deborah realised she had adopted her mother's model of relating, responding passively and powerlessly to her boyfriend, in the same way her mother did with her own husband. Deborah recognised that she was used to avoiding problems and not thinking about solution, instead of actively problem-solving. This was reflected in following changes in PQ items 1, 8, 9. From session 6, the therapist worked on restoring the roles of daughter and parents. This is reflected on HAT 6 ("I'm not the Parent, I'm the Child") and led to a change in role-assumption and her relationship with her parents that is reflected in HAT 11 ("parents are behaving like parents", "My relationship with them has changed").
Sceptic Case 1.The apparent changes are negative (i.e. involved deterioration) or irrelevant (i.e. involve unimportant or trivial variables).
Three of the quantitative measures used (PHQ-9, GAD-7 and PQ) are not validated for adolescence, and should therefore not be used in this case study. The only measure validated for adolescents is the YSR, which at the 6-month follow-up showed a reliable deterioration in all cumulative scales (internalizing, externalizing, total problems). Anxiety/depression subscale indicated a nonreliable improvement at the end of the therapy, and a deterioration worse than the pre-therapy score at the 6month follow-up. For such reasons, we reject the claim of a global reliable change. Indeed, the deterioration observed in all cumulative scales at session 8, at the end of the therapy and at the follow-up, and the deterioration observed on the rule-breaking sub-scale, suggest a negative impact of the therapy. Also, in qualitative data we note evidence of inconsistent change: in the CI Deborah tells about having stopped using drugs, yet in the 6-month follow-up she indicated in the YSR, at item 105 I use drugs for non-medical purposes (this item did not include alcohol or tobacco) she scored 2, very true/often true.
2. The apparent changes are due to statistical artefacts or random errors, including measurement error, experiment-wise error from using multiple change measures, or regression to the mean. Even considering the PHQ-9 as a valid measure for depression in adolescence, the visual inspection of the three-point baseline shows an unstable pattern, making it difficult to calculate a reliable change. The PQ scores have not been collected in the pre-treatment assessment, so the case lacked a stable baseline to examine subsequent scores.
3. The apparent changes reflect relational artefacts such as global hello-goodbye effects on the part of a client expressing his or her liking for the therapist, wanting to make the therapist feel good, or trying to justify his or her ending therapy.
In her CI, Deborah reported almost only positive comments about the therapy and the therapist, and in her HAT forms she reported only one hindering event (that is also unclear since it appeared to refer to extra-therapy events). She described the therapist as if she were "her best friend" (C6). For this reason, it is possible that Deborah's tendency to 'compliance' might have affected her outcome measures. Furthermore, in almost all sessions, she showed compliance to the therapist's interventions.
4. The apparent changes are due to cultural or personal expectancy artefacts; that is, expectations or scripts for change in therapy.
Deborah searched independently and spontaneously for therapy, and three out of eleven changes reported in her CI were expected or somewhat expected, and two neither expected nor unexpected. This suggests that the change can be partially tied to self-persuasion and personal expectancy of a resolution of her problems.
5. There is credible improvement, but it involves a temporary initial state of distress or dysfunction reverting to normal baseline via corrective or self-limiting processes unrelated to therapy. Even accepting data from PHQ-9 and GAD-7, changes and widely fluctuating scores are normal in adolescence and all observed changes can be attributed to normal fluctuations associated with adolescence. Furthermore, Deborah discovered that she was suffering from a metabolic disorder the previous year. In fact, two problems out of ten on her PQ concern her illness ( Table  2, items 2 and 10). Changes in depression and anxiety symptoms may represent a normal reaction to her illness, followed by a return to the previous condition.
6. There is credible improvement, but it is due to extratherapy life events, such as changes in relationships or work.
The improvements in Deborah's quantitative data may have been due to extra-therapeutic events, such as: her father's long absence for work, might have led to a reduction in conflicts (session 6, C4; and 7, C5); the end of the relationship with her devaluing boyfriend (session 5, C 23); meeting a new boy she liked (session 7, C5); and finally, participation in a support group for adolescents with the same illness (session 11, C3).
7. There is credible improvement, but it is due to psychobiological processes, such as psychopharmacological mediations, herbal remedies, or recovery of hormonal balance following biological insult. Depression may be tied to Deborah's metabolic illness and change in depression may be tied to hormonal balance following pharmacological treatment. The use of marijuana might also have had some effect on her mood, making it difficult to differentiate between effect of the therapy and effect of the biological processes.
8. There is credible improvement, but it is due to the reactive effects of being in research. Participating in the research, talking about her problems and being recorded made Deborah feel embarrassed (CI, C4), thinking "what if I swear, I must be very careful" (CI, C5), which might have affected quantitative and qualitative data. She also found outcome measures very boring (CI, C122), suggesting a possible inaccurate compilation of the forms.
Affirmative Rebuttal 1) A search for existing literature indicated that there are several studies which support the use of PHQ-9 and GAD-7 with adolescents. Studies indicate that disorder sensitivity and specificity in adolescence are similar to those of adult population, suggesting only a slightly higher clinical cut off. Even if there is not a validated version for Italian adolescents, there is no reason to suppose a different result might occur in light of the other validations. Thus, we affirm the presence of a global reliable change. The global deterioration observed in YSR may reflect the progressive trust of the client towards the therapist and consequent self-disclosure in reporting her problems accurately. The items of the YSR are defined as concrete behaviours, and in the pretherapy assessment the client may have had some hesitation in describing herself, in fact she signed the item "I use drugs for non-medical purpose (don't include alcohol or tobacco)" with 0 (never) however in following sessions she stated that she used marijuana with friends.
2) Fluctuation in the PHQ-9 scores in the pre-treatment phase are inferior to the reliable change index, thus are not reliable and may reflect the error measure of the test. The PQ is a client-generated measure that is built with the client during the pre-treatment phase, so the first score is usually available only at the first session. The stable baseline is supported by the duration form that assesses how long the personal problems were bothering the client.
3) The creation of a friendly relationship is considered a necessity to engage adolescents in therapy. In her CI, the client appears able to describe unpleasant aspects of the therapy, such as boredom in filling the forms (CI, C122) and freely spoke about some goals that had not been reached. For example, regarding the relationship with her father, she admitted she hoped to change it, but during the course of therapy she realised that it was not possible: "this did not change, of course it didn't, I somewhat expected it… but this is who he is, he did not take part in these sessions, so my head changed, not his, and I accept it" (C55), suggesting that the relationship was friendly but not compliant. 4) Most of the change reported by the client in her CI are rated as unexpected or somewhat unexpected, suggesting that it is not present as a personal expectancy artefact.

5) When
Deborah presented for therapy, her condition was worsening, to the point she felt the need for help. Despite the typical fluctuating patterns of adolescents, the client identified in her PQ and CI different changes in problems that were not resolved in previous years by the simple passage of time or natural course of the disorder, which contradicts a 'reverse to normal baseline' hypothesis.
6) The observed extra-therapeutic events are normal events in the life of an adolescent and in the change interview the client did not refer to them as a cause of her change. The meeting with the new boy is not followed with a stable relationship, and happens after the wider improvement in depressive symptomatology observed in session 6. Furthermore, the client described a wide network of friendships at the beginning of the therapy, and it appears improbable that the new friendship group would substantially affect her depression.
7) The use of marijuana and pharmacotherapy began largely before the worsening that led the client to the therapy, thus suggesting the absence of a causal role in observed improvement.
8) The overall transcriptions of the sessions show that Deborah expressed without censoring herself, furthermore in her CI Deborah reported feeling initially embarrassed, "then I got the hang of it, I got used to it, I don't even realise it" (C5), thus supporting the claim that the client was not influenced by the participation in research. As for boredom, quantitative measures, Deborah reported also "they were very boring, but then you see how you have changed… when you conclude therapy you say 'Crap! I reached this goal, I made it!' and I think that without these evaluations you cannot realise it" (CI, C122), suggesting an adequate compliance in filling in tests and forms.

Sceptic Rebuttal
Even accepting the use of outcome measures not validated for adolescence, there still remain the difficulties in the use of questionnaires for adolescents. There are several indications that Deborah found some difficulties, inconsistency and confusion in completing the questionnaires. For example, during the CI, when reviewing the PQ, Deborah noticed that she made a mistake in rating the item 10 in the first session: she stated "'very little'… it was supposed to be the opposite, I don't know what happened, it was supposed to be 'maximum possible', before I didn't think, whereas now it's 'not at all'" (CI, C148). She also misunderstood how to rate how likely the change would have been without therapy, scoring the attribution by inverting the scale: somewhat unlikely without therapy… "I believe therapy helped me a lot, so 4 (somewhat likely)" (C68), and repeating this procedure to be sure she understood how to rate it correctly she asked "without therapy, what did I say?", the interviewer replied "4", "yes, somewhat unlikely, that's correct" (C76-77). Furthermore, in the CI Deborah referred "I have always thought that I was a disappointment for my parents" (C33), whereas, in her PQ's duration form she wrote it lasted only from 6-11 months. In her HAT, she described as a hindering intherapy event that in her family there was a lack of clarity regarding who would take parental roles. Overall, this suggests that scores on quantitative data could reflect her difficulties in understanding the measure and in quantifying improvements in different problems.

Affirmative Conclusion
Deborah's depression, anxiety, personal and behavioural problems were related to difficulties in sustaining a self-nurturing internal dialogue, harsh self-criticism and difficulty in solving problems. The therapist created a warm relationship where the client felt free to be open and experienced strong support for her compromised self-esteem. The focus on awareness of her internal dialogue, differentiation between internal dialogues from Critical and Nurturing Parent, the ability to think about herself and her future, the systematic engagement in problem solving activity and the systematic provision of permission contrasting her injunctions, led to a stable change in depressive and anxiety symptoms, especially guilt and worthlessness. This also resulted in a behavioural change in self-injury and marijuana use, and a change in interpersonal relationships, where she now expresses her anger and puts boundaries in place with others. This change was reflected also in the adoption of new roles and rules within her family, with the client accepting her role of daughter rather than parent.

Sceptic conclusion
Deborah's symptoms arose after discovering her metabolic illness that made her gain 10 kilos, just while entering adolescence, a phase of life generally difficult and abundant in different problems. Some quantitative measures are not validated for adolescents and the tests also present several errors in their compilation. Several extra-therapeutic events may have had a prominent role in the reversal of symptomatology. The observed change could be due to a spontaneous remission.

Adjudication
Each judge examined the rich case record and hermeneutic analysis and independently prepared their opinions and ratings of the case (Table 5). The judges overall conclusions are that this was a clearly good outcome case, that the client changed substantially, and that the changes are between substantially and completely due to the therapy.
Opinions about the treatment outcome (good, mixed, poor) Judge A (VC). This is a clearly good outcome (60% of certainty) with aspect of a mixed outcome (40% of certainty) Quantitative data show a reliable and clinically significant change on measures of depression (PHQ9), anxiety (GAD7) and Personal problems (PQ), maintained at the follow-ups, allowing a claim for a global reliable change. Qualitative data supports claims of change in self-injury, marijuana use, self-esteem and problem-solving ability, all of which are changes that the client attributes to the therapy. Aspects of mixed outcome are mainly connected to the YSR scores, which at the 6-month follow-up indicated a global increase in most problem behaviour dimensions.
Judge B (SM). This is a clearly good outcome (80% certainty) or a mixed outcome (20%). Quantitative and qualitative data described in affirmative brief and rebuttal appear to provide clear support for a claim that this was a good outcome case. Aspects of mixed outcome are related to the sceptic briefs and rebuttal, that suggest a more conservative position.
Judge C (AP). This case is classifiable as good outcome case (80% certainty) to mixed case (20% certainty). Quantitative data suggests a reliable and global change, partially maintained at the six-month follow-up. Behavioural change in self-injury, use of marijuana, and problem-solving ability appear to be the most important achievements for this client, supported by session transcripts and Change Interview. The client became more differentiated from her father by ending her controlling behaviours towards him and understanding that they are two different persons. At the same time, she began to ask her parents to fulfil their role. These kinds of change are not only tied to a symptomatic remission, but appear tied to a deeper change in thinking and selfesteem.
Opinions about the degree of change Judge A. The client's change is considerably (60%, with 80% of certainty), considering the global reliable change and the importance of change in self-injury behaviours. The change in quantitative measures appears stable during the follow-ups. Moreover, qualitative data (Change Interview) shows a consistent change in Deborah's self-esteem, (disappointing her parents vs no more disappointing, not protecting herself vs protecting herself), indicating a change in deep dimensions of personality and not merely symptomatic changes.
Judge B. The client changed substantially (80% with an 80% certainty). There is strong evidence that Deborah has more self-esteem, she stopped hurting herself, she likes herself more, she understands her emotions and feelings, she does not lose control anymore, she protects herself and she is able to think about the consequences of her actions. This shows a significant change. Furthermore, in her CI Deborah reports not feeling that she is a disappointment for her parents anymore, being able to think about negative things and understanding that making use of drugs was not a solution for feeling better and drawing her parents' attention.
Judge C. The client showed a substantial change (80% with 80% of certainty). The client's quantitative data (clinical significance, reliable change index and stability during follow-up) showed a clear improvement, especially in her depressive symptomatology. Furthermore, qualitative measurements represent a full description of Deborah's improvement, traceable in the Change Interview and in relational episodes described in her last sessions. These aspects support a widely significant change in her self-representation: she seems to feel more certain of herself, to have more self-esteem, to feel more comfortable with her body, less worried about taking care of her parents and more focused on taking care of herself, and she ceased self-injuring behaviours and taking drugs.
Opinions about the causal role of the therapy in bringing the change Judge A. The observed change is substantially (80% with 80% of certainty) due to the therapy. Qualitative data in the HAT form (summarised in Table 3) of the client is extremely helpful to understand what the client felt important in the course of therapy, such as increasing self-esteem and stop hurting herself in many different ways. Qualitative data in the Change Interview (summarised in Table 4) reported a list of changes attributed to the therapy, for example the client explicitly recognised that without the therapy she would still be addicted. There were clear connections between interventions, HAT forms and changes drawn in the affirmative case which appear credible.
Judge B. The change is substantially (80% with 80% of certainty) due to the therapy. Since the beginning of therapy, the therapist focused on Deborah's tendency to think she was not doing enough to be a good child and on her difficulty in appreciating and loving herself. All sessions involved deep work on her self-esteem, also reported in Deborah's HAT forms (Table 3: session 1 , 4, 5, 7, 8, 9, 12 and14). Furthermore, in the CI she reported that eight out of eleven big changes were due to therapy (Table 4).
Judge C. The change appears substantially due to the therapy (80% with 80% of certainty). One of the most important changes is that the client gave up taking care of her parents, which had previously created a frustrating situation where she was the parent of her own parents. This frustrating situation appeared to be tied to depressive symptoms and negative self-representation.
The therapist focused on helping the client understand she is not responsible for her parents, and can ask them to protect her as a young girl of her age requires. Furthermore, the therapist helped the client to focus on her needs and to express them. These aspects can be found in her HAT forms (Table 3) and CI, where she described how she began to feel like she is a child rather than a parent (HAT 6). This produced a change in the relationship with her parents, a decrease in her frustration, and as a result a decrease in her self-injuring behaviours and her use of drugs.

Mediator Factors
Judge A. A good therapeutic alliance and empathic listening were essential in order to reinforce self-esteem. During the therapy self-injuring behaviours were connected to depressive feelings, beneath which there were: anger toward her parents, need for boundaries from parents and holding of emotions, and feelings of having disappointed her parents. Changes in selfcriticism and problem solving strategies appear to be at the base of the change process.
Judge B. A good therapeutic alliance has been essential for exploring Deborah's worries, emotions and feelings in order to work on her depressive symptoms and to have more self-esteem. Also internal dialogues which generated feelings of guilt and low self-esteem have been successfully analysed.
Judge C. The most significant mediator factor seems to be the ability to differentiate between the role of child and parent, to feel her own needs, recognise they are allowed and experience permission to express them. Furthermore, the therapist succeeded early in the therapy in establishing a good therapeutic alliance. The therapist acted as a model of affective, protective and Nurturing Parent, allowing the client to have a new thein in establishing a good therapeutic alliance. The positive experience and change her internal dialogue which increased her self-esteem. Tabl e 5: A djudic ation results.

Moderator Factors
Judge A. Being very clever and smart, and having a wide network of relationship with peers, have been important resources for the client in order to make the best use of this therapy, which she sought spontaneously.
Judge B. The client was very motivated since the beginning; that and her cleverness were both surely powerful resources that helped Deborah make the best use of the therapy.
Judge C. The client showed a positive attitude toward the therapist and therapy and was very motivated to work on herself during therapy. She was also extremely intelligent and introspective.

Discussion
This case aimed to investigate the effectiveness of a manualised TA treatment for depression in adulthood in an adolescent with moderate MDD in comorbidity with GAD, self-injury and substance use. Primary outcomes were depressive and anxiety symptomatology, that showed a reliable and clinically significant change, maintained throughout follow-ups. Secondary outcomes were personal problem and behavioural problems, that showed a mixed outcome. The therapist conducted the treatment in a good to excellent adherence to the manual. Hermeneutic analysis evidenced quantitative and qualitative aspects of the change, pointing also the problems in assessing change in adolescence.
The judges concluded that this is a clearly good outcome case, with a considerably to substantial degree of change, and which was substantially to completely due to the therapy. These conclusions provide further support to the effectiveness of the manualised TA treatment for depression in adults, and the case provides the first evidence that the manualised treatment was effective in treatment with a female adolescent with comorbid depression and anxiety.
This provides a new line of evidence for the effectiveness of the manualised TA therapy, which has now demonstrated effectiveness with adults and adolescents, with both 'straightforward' depression as well as comorbid anxiety, in two European cultures. The case also provides some initial indications that TA therapy may be effective for self-injuring behaviours and substance use problems amongst adolescents and further research is warranted to investigate this further.
Creating an early therapeutic alliance, supporting selfesteem, changing self-critical internal dialogues, developing an internal Nurturing Parent, providing appropriate permission tailored to the specific needs of the client and developing problem-solving ability all appeared to be mediators of change in this case, which were moderated by the intelligence of the client, her introspective capacity and the relational networks. These mediator and moderator factors all suggest areas for further research and can be used to examine and aggregate findings from all HSCED investigations of manualised TA therapy conducted so far.

Limitations
The first author has a strong allegiance to TA, is a teacher of the members of the hermeneutic groups and a colleague of the three judges. The author was also funded for this research by TA institutions (see Funding below). Despite the reflective attitude adopted in this work, these factors may have influenced in subtle ways both the hermeneutic analysis and the judges' evaluations.

Conclusion
This case study provides evidence that the specified manualised TA treatment for depression in adulthood  has been effective in treating a major depressive disorder in an adolescent treated by a female therapist. Despite results from a case study being difficult to generalise, this study adds evidence to the growing body of research supporting the efficacy and effectiveness of TA psychotherapy, and notably supports the effectiveness of the manualised TA psychotherapy for depression as applied to adolescents.

Funding
This study was supported by grants from the European Association for Transactional Analysis, as part of the project 'Transactional Analysis meets Academic Research in order to become an Empirically Supported Treatment: an Italian two-year plan for publishing evidence of Transactional Analysis efficacy and effectiveness into worldwide recognized scientific journals' and from the Center for Dynamic Psychology -Padua, a transactional analysis-oriented School of Specialization in psychotherapy.