Assessment of psychoanalytic psychotherapy outcomes

Abstracts

INTRODUÇÃO: Este artigo relata uma investigação de coorte retrospectiva que avaliou a efetividade da psicoterapia psicanalítica com pacientes adultos em um serviço de atendimento comunitário na cidade de Porto Alegre, Brasil. MÉTODO: A amostra foi constituída por 34 pacientes divididos em dois grupos: grupo 1, 17 pacientes que realizaram tratamento por até 11 meses (média de 6,4 meses), e grupo 2, 17 pacientes com 1 ano ou mais de psicoterapia (média de 24,7 meses), contatados após o término de seu tratamento psicoterápico, em média 20,9 meses no grupo 1 e 29,9 meses no grupo 2. Os instrumentos utilizados foram: entrevista semi-estruturada, questionário de efetividade e escala de avaliação global do funcionamento. Experts independentes aplicaram a escala de avaliação global do funcionamento na entrevista inicial de tratamento (realizada pelo psicoterapeuta e encontrada no arquivo da instituição) e de seguimento. RESULTADOS: Os pacientes melhoraram significativamente seu funcionamento global, comparando a avaliação inicial à de seguimento da psicoterapia (p < 0,001), independentemente do grupo de tratamento. Pacientes e experts avaliaram o tratamento psicoterápico de forma satisfatória. Entretanto, a opinião dos experts não apresentou correlação com a opinião dos pacientes investigados. CONCLUSÕES: A psicoterapia psicanalítica foi efetiva na amostra estudada. A duração da psicoterapia, analisada isoladamente, não foi um fator decisivo para o resultado do tratamento, isto é, outros fatores também devem ser considerados na avaliação de resultados (relação terapêutica, qualidade das relações objetais, potencial auto-reflexivo, reação às interpretações). Constatou-se que a maioria dos pacientes fica satisfeita com resultados em um tempo menor do que aquele que seus psicoterapeutas considerariam ideal.

Psicoterapia; terapia psicanalítica; avaliação de resultados; avaliação de resultado de intervenções terapêuticas; pesquisa

ORIGINAL ARTICLE

Assessment of psychoanalytic psychotherapy outcomes

Simone Isabel Jung^I; Maria Lúcia Tiellet Nunes^II; Cláudio Laks Eizirik^III

^IPsychologist. MSc. in Medical Sciences - Psychiatry, Universidade Federal do Rio Grande do Sul (UFRGS), Porto Alegre, RS, Brazil

^IIPsychologist. PhD in Psychology, Free University of Berlin, Berlin, Germany. Professor, Faculdade de Psicologia, Pontifícia Universidade Católica do Rio Grande do Sul (PUCRS), Porto Alegre, RS, Brazil. Coordinator, Graduate Program in Psychology, PUCRS, Porto Alegre, RS, Brazil

^IIIPsychiatrist. PhD in Medicine, UFRGS. Associate professor, Department of Psychiatry and Legal Medicine, UFRGS. Training analyst, Sociedade Psicanalítica de Porto Alegre, Porto Alegre, RS, Brazil. President of the International Psychoanalytic Association (IPA)

^{Correspondence}

ABSTRACT

INTRODUCTION: This article reports a retrospective cohort investigation, which assessed the effectiveness of psychoanalytic psychotherapy in adult patients at a community health center in the city of Porto Alegre, Brazil.

METHODS: The sample was composed of 34 patients, who were divided into two groups: group 1, 17 patients who underwent treatment for up to 11 months (mean 6.4 months) and group 2, 17 patients who underwent psychotherapy for 1 year or longer (mean 24.7 months). The patients were contacted after their psychotherapeutic treatment was over, in average 20.9 months (group 1) and 29.9 months (group 2). Instruments used were semi-structured interview, effectiveness questionnaire and global assessment of functioning scale. Independent experts applied the global assessment of functioning scale in the initial interview (carried out by the psychotherapist and found in the institution's files), as well as in the follow-up interview.

RESULTS: The patients significantly improved their global functioning, when initial assessment was compared to psychotherapy follow-up (p < 0.001), regardless of treatment group. Patients and experts considered psychotherapy satisfactory. However, the experts' opinion was not related to the patients' opinion.

CONCLUSIONS: Psychoanalytic psychotherapy was effective in the studied sample. Length of psychotherapy, when assessed alone, was not a decisive factor for treatment outcome, i.e., other factors should also be taken into account when assessing outcomes, such as therapeutic relationship, quality of object relations, self-awareness potential and reaction to interpretations. Most patients were found to be satisfied with outcomes in a shorter time than their psychotherapists would consider optimal.

Keywords: Psychotherapy, psychoanalytic therapy, outcome assessment (health care), outcome assessment of therapeutic interventions, research.

Introduction

Despite the current state of psychoanalytic knowledge and relevance of previously performed studies, opinions concerning effectiveness of psychoanalytic psychotherapy (PP) are still divergent, both in the media in general^1-3 and among professionals in the so-called "psy" area.^4-7 In this setting, systematic investigation of outcomes in PP is still needed, especially as a response to criticisms made against duration and quality of psychotherapeutic outcomes.^5,7

Assessing the effectiveness of psychotherapies started with Freud himself,⁸ through the study of a single case in his Studies on hysteria.⁹ Based on that study, a significant number of research reports on PP outcomes has been published, especially in the 2000's.^10-19 Wallerstein,²⁰ for example, already mentioned four generations of psychoanalytic researches on outcomes. However, in a search in LILACS, PubMed (from January 1990 to April 2007) and PsycINFO (1989 to April 2007) databases, there were only four outcome research studies in adults (non-psychotic and without drug abuse) in long-term outpatient PP, conducted or being conducted in Latin America;^21-24 one of these was performed in Brazil.²² Therefore, outcome assessments in psychotherapy in Brazil are a fertile field for psychoanalytic investigation.

The study presented in this article is a retrospective cohort investigation approved by the Ethics Committee of Universidade Federal do Rio Grande do Sul and is part of the master's thesis by the first author.

The general objective was to verify effectiveness of PP in adult patients who were given treatment at a community care service. Correlations between treatment duration and outcome and between the opinion of patients and PP specialists were analyzed.

It is worth stressing that in discussions on psychotherapy, the terms psychoanalytic and psychodynamic have been increasingly used as synonyms.²⁵ Therefore, in this study, the terms PP and psychodynamic psychotherapy will be used as equivalent, representing the same modality of psychotherapy originated from psychoanalysis.

Methods

Patients

Inclusion criteria were 18 years or older when psychotherapy was started; minimum frequency of 18 treatment sessions; treatment discontinuation due to discharge (objectives achieved); interruption (without achieving all objectives) and dropout (not going to sessions) from December 1999 to July 2005 and at least 6 months before being included in the study.

Patients with professional activities or students of psychology and medicine were excluded, according to an institutional regulation, as well as patients who returned for psychotherapeutic treatment. Such exclusions were due to ethical issues and to avoid a possible confounding factor in outcome assessment.

The research was planned to include patients with at least 28 sessions, based on dose/effect studies by Howard et al.²⁶ and Kopta et al.,²⁷ but the number of sessions was reduced with the aim of increasing the sample. Through a screening of the number of sessions used in studies of short-term psychodynamic treatment conducted since 2000^28-39 and that indicate a wide spectrum of one to 63 sessions, it was possible to reach the minimum number of 18 sessions, and five patients with this frequency were included.

Follow-up time of 1 year was initially planned, following the tendency found in other investigations.¹⁹ However, to make this research feasible in the available time, cases with lower follow-up time were included. Five patients were included with follow-up time lower than 1 year.

Gabbard²⁵ proposes, as long-term treatments, those performed with more than 24 sessions or more than 6 months. Leichsenring⁴⁰ refers that, in Germany, treatments with 25 to 50 sessions are not considered long. Therefore, efforts made to define what a long-term treatment is might be arbitrary.^25,40 Long-term treatments were considered those that lasted 1 year or more, according to Crits-Christoph & Barber,⁴¹ and short-term treatments as those that lasted for at least 18 sessions or for a maximum period of 11 months.

Instruments

Effectiveness questionnaire

The effectiveness questionnaire (EQ) is a reduced version of the questionnaire developed by the Consumer Report Study.⁴² The translation and adaptation of the EQ into Portuguese was carried out by Freedman et al.²² EQ is comprised of 28 questions that investigate the reason for seeking treatment, emotional and physical status at treatment onset and at the current moment, and the patient's perception as to satisfaction and benefits of psychotherapeutic treatment.

Only the effectiveness score (ES) was assessed, which is obtained in three parts, each originated from separate questions of the EQ: specific progress (100 points), patient's satisfaction with the treatment of his problems by the therapist (100 points) and overall progress (100 points). Maximum therapeutic effectiveness corresponded to 300 points; 150 meant no effectiveness; and values under 150 represented negative effectiveness, in the sense of increased psychic complaints throughout the treatment.

Semi-structured interview

The semi-structured interview was recorded in audio and later transcribed. It was used to encourage verbalizations as to the patient's current status. Interventions to remove any doubts were performed whenever necessary.

Global assessment of functioning scale

The global assessment of functioning scale (GAF) is described in axis V of the Diagnostic and Statistic Manual of Mental Disorders.⁴³ It is a widely used scale to screen clinical progress of patients in overall terms, by attributing a single measurement, which may range between 0 and 100. GAF was applied to the interviews at treatment onset (they were found dialogued at the institution's file) and to the follow-up interviews (final). Each interview was assessed by at least two specialists, being careful to avoid specialists from applying GAF, initial and follow-up interview of the same patient. The specialists were blinded to assess GAF regarding the patients' treatment duration, but the same procedure was not possible regarding the interview being initial or follow-up.

Procedures

The patients were contacted by mail, followed by a telephone call. At the telephone call, the first patient's doubts about the research were clarified and an interview was scheduled. In that interview, the patients were informed about the research procedure, their doubts were clarified, the consent term was signed and the EQ was filled. The semi-structured interview was also conducted during that initial contact. Those stages were performed by the first author.

The first treatment interview was found, reported by the patient's psychotherapist, at the institution's files.

After typing the first patient's interview and transcribing the follow-up interview, independent specialists applied GAF in both interviews.

Institution

The patients included in this study are individuals who underwent PP at the care service of Integrated Services on Psychoanalytic Psychotherapy (ESIPP). ESIPP is a teaching institution based in Porto Alegre, Brazil, recognized by the Brazilian Federal Council of Psychology and by the Brazilian Association of Teaching in Psychology, whose main goal is the formation of PP specialists.

Treatment

The psychotherapy offered by ESIPP is directed to the insight, performed face to face, once to three times a week and without using any manuals. Throughout the psychotherapeutic treatment, psychotherapists have weekly meetings with institution supervisors (experienced professionals in PP) for collective and individual supervision of cases being treated. They also hold, every semester, one to three seminars of psychoanalytic theory and technique once a week.

Statistical analysis

Sample size (17 patients in each group) was calculated by comparing means (software WINPEPI, Compare module, version 1.45), to detect a clinically relevant difference (size effect = 1 or more, i.e., 1 or more standard deviations - SD) between mean changes in GAF (initial and follow-up) in both treatment groups, considering α = 0.05 and power of 80%.

To compare characteristics of treatment groups between themselves and non-participants, the following were used: chi-square or Fisher's exact test for qualitative variables, Student's t test for quantitative variables with symmetrical distribution and Mann-Whitney test for quantitative variables with asymmetrical distribution.

Student's t test for paired samples was calculated to verify the difference between scores attributed to GAF (initial and follow-up) by the specialists, considering both treatment groups. As an attempt to reach a more homogeneous score, as the test result did not show any significant difference between GAF scores performed by both specialists, a third rater was introduced to score GAF when the difference between both previous scores was lower than ± 1 SD. This occurred in five cases of initial interview and in seven cases of follow-up interview in group 1; and in four cases of initial interview and in three cases of follow-up interview in group 2. Final score was performed using the mean between scores by all three raters.

Variance analysis for repeated measures with two factors (ANOVA) was used to compare initial and follow-up GAF scores, that is, onset and post-treatment, according to the investigated group. Magnitude of mean difference between both groups was verified using the effect size calculated by Cohen's methodology.⁴⁴ To verify the difference of ES means between groups, Student's t test was used for independent samples.

Pearson's correlation was calculated between final GAF and ES (partial and total scores). Spearman's correlation was used to correlate therapist's experience time and outcome measured by final GAF.

The outcomes were analyzed by the software Statistical Package for the Social Sciences (SPSS, version 14.0), with significance level of 0.05.

Results

Sample characteristics

A total of 82 patients who met the inclusion criteria were found at the institution's file. Of these, 23 were not located, nine did not accept participating in this research, nine were excluded because they were undergoing some type of treatment and seven quit participating before the interview.

The sample was comprised of 34 patients (minimal number found in the sample calculation), who were divided into two groups: group 1 - 17 patients who underwent treatment over a period of up to 11 months; and group 2 - 17 patients who underwent treatment for 1 year or more. Choice for this cut-off point was an attempt to compare short- and long-term treatments.

The patients were contacted in average 20.9 months (SD = 10.7) and 29.9 months (SD = 16.7) after the psychotherapeutic treatment was over in groups 1 and 2, respectively.

Table 1 shows the sociodemographic characteristics of the patients participating in the research in both treatment groups.

Clique here to enlarge

Figure 1 shows the reasons for seeking psychotherapy in percentage in both treatment groups. It is worth stressing that the patient could choose in the questionnaire (EQ) more than one reason for seeking treatment, i.e., there was multiple choice.

To proceed to a comparison, the sociodemographic characteristics of non-participating individuals were screened, i.e., 48 patients (23 were not found, nine refused to participate, nine were excluded because they were undergoing some type of treatment and seven quit). These will be presented only in comparison with the groups of participants in the comparability item.

Psychotherapists

The psychotherapists who conducted the treatments graduated in medicine or psychology and were taking a specialization course in psychotherapy at that institution. In the group of patients receiving treatment for up to 11 months, psychotherapy was conducted by 15 psychotherapists, one physician and 14 psychologists. Of these, 14 were female and one was male, with mean age of 34.6 years (SD = 11.0) and mean experience time in psychotherapy of 2.6 years (SD = 2.1). Two psychotherapists conducted two treatments each in this group of patients.

The group of patients who underwent treatment for 1 year or more was treated by 13 psychotherapists, one physician and 12 psychologists. Of these, 12 were female and one was male, with mean age of 39.3 years (SD = 8.7) and mean experience time in psychotherapy of 4.6 years (SD = 3.4). In this group, four psychotherapists conducted two treatments each with the patients in the sample. It is worth stressing that, as in other studies, in this research two psychotherapists conducted treatments in both groups.

Raters

The 13 independent raters who participated in the study (two male and 11 female) majored in psychology or medicine; six were psychologists and six were psychiatrists. Of these, two are psychoanalysts and 11 have a graduation at master (6) or doctorate (5) level. Mean age of raters was 41.2 years (SD = 9.9), and mean experience time in PP was 15.4 years (SD = 9.2).

Comparability between groups

Comparison between treatment groups is crucial for further analysis of psychotherapy effectiveness. There were no significant differences between patients in groups 1 and 2 as to gender, town of residence, age, schooling level, marital status, treatment frequency, use of medication, psychiatric hospitalization, previous treatment, therapist's gender and age, time of contact after termination and global functioning at treatment onset. Time of therapist's experience who treated group 2 patients, despite being higher (approximately 2 years), was not statistically significant (Mann-Whitney test = 92.000, p = 0.066). There was significant difference between groups (chi-square test = 7.404, p = 0.007) as to reason for terminating treatment: in group 2, there were more discharges, and in group 1, more interruptions.

Comparing patients in groups 1 and 2 as to reasons for seeking treatment, there is significant difference only concerning depression (depression, depressive mood and mourning), which was higher in group 2 (Fisher's exact test, p = 0.039).

With regard to the comparison of non-participants who underwent up to 11 months of treatment and group 1 patients, there was no significant difference in compared characteristics - age, marital status, number of sessions, treatment duration, reason for discharge, town of residence, schooling level, psychotherapist's age, gender and time of experience, except for patients' gender variables - there are more men among the non-participants (chi-square test = 5.929, p = 0.015) - and time of contact after termination, which was higher in non-participants (Student's t test for independent samples = 5.987, p < 0.001). A higher number of men was among the non-participants who were not found.

Non-participants who underwent treatment for 1 year or more, when compared with group 2 patients, had a higher number of sessions (Student's t test for independent samples = - 2.093, p = 0.046), were contacted at a longer time after discharge (Student's t test for independent samples = - 5.044, p < 0.001) and most were single (chi-square test = 7.243, p = 0.027). Such differences are statistically significant. Most singles belong to the group of non-participants who were not found to participate in the study. For other variables, age, gender, treatment duration, reason for discharge, town of residence, schooling level, psychotherapist's age, gender and time of experience, there were no significant differences.

Treatment effect

Table 2 shows the treatment effect based on assessment of initial and final (follow-up) GAF in both treatment groups.

Clique here to enlarge

In this investigation, it was not possible to have a control group; however, the data were compared with the index of expected mean change (effect size) in control group, developed by Leichsenring & Rabung,⁴⁵ based on 26 randomized controlled studies of psychoanalysis and psychodynamic psychotherapy (mean = 0.12, SD = 0.19). This type of comparison was also used by Leichsenring et al.,¹³ in 2005, when they published the preliminary results of the Göttingen study of effectiveness of psychoanalytic and psychodynamic therapy, i.e., the data from patients who underwent psychoanalysis. The effect size found in this study, when compared with the change index - effect size of control group⁴⁵ - was 8.38 in the total sample, 6.08 in group 1 and 8.01 in group 2, that is, a significant difference both in total as in groups separately (p < 0.001).

Data from the effectiveness questionnaire

Mean ES found in EQ was 179.71 (SD = 40.06) in the total sample; 180.30 (SD = 40.87) in group 1; and 179.12 (SD = 40.48) in group 2. There was no significant difference between ES scores in both treatment groups (Student's t test) for independent samples = 0.084, p = 0.933).

The first ES score, shown in Table 3, measured how much the psychotherapist helped the patient with the specific problems that led him to seek treatment.

Clique here to enlarge

In the second ES score, assessed measure was patient's satisfaction with the treatment of his problems by the therapist (Table 4).

Clique here to enlarge

The third ES score compared the patient's emotional status in the beginning of the psychotherapeutic process with the current status in the follow-up interview, i.e., the patient's overall progress. The patients chose one of the following scores: 1 - very poor; 2 - fairly poor; 3 - so-so; 4 - quite good; 5 - very good. Of the 17 patients in group 1 who characterized their emotional status as very poor, as fairly poor or so-so in the beginning of psychotherapy, 70.59% (12) advanced one to three scores at the end of the treatment, 23.53% (4) remained in the same status and 5.88% (1) worsened. Of the 17 patients in group 2 who started psychotherapy feeling very poor, as fairly poor or so-so, 64.71% (11) characterized their status at the end of the treatment better than in the beginning, advancing one to three scores, 29.41% (5) remained in the same status and 5.88% (1) worsened.

Correlations

The correlation between final GAF score (specialists' opinion) and ES in the EQ (patients' opinion) in the total sample of patients was r = 0.129 (p = 0.466), i.e., a weak and non-significant correlation. The correlation between final GAF and patient's current emotional status was r = 0.333 (p = 0.054), i.e., a regular correlation and in the limit of significance.

In group 1, the correlation between final GAF and EQ was r = 0.342 (p = 0.179), i.e., it represents a non-significant correlation. The correlation between final GAF and EQ score, which measured patient's satisfaction, was r = 0.0358 (p = 0.159), representing a regular correlation, but with no statistical power.

In group 2, the correlation between final GAF score and EQ was not significant (r =.- 0.121, p = 0.645). Pearson's correlation between final GAF and patient's current emotional status was r = 0.413 (p = 0.099): a regular and statistically non-significant correlation.

Spearman's correlation was then calculated between psychotherapist's time of experience and treatment outcome obtained by final GAF, with no significance (r = 0.007, p = 0.969).

Discussion

In this retrospective naturalistic study, PP effectiveness was evaluated in adult patients in both treatment groups: group 1, patients who underwent treatment over a period of up to 11 months; and group 2, patients who underwent treatment for 1 year or more.

When both groups were compared in demographic terms, there were no significant differences, except as to reason for seeking treatment (informed by patients in the EQ) due to depression (depression, depressive mood and grief) and to reason for treatment termination.

A significantly higher number of patients in group 2 sought treatment due to depression (depression, depressive mood and grief); however, group 1 had a large number of patients presenting the same complaint. The study corroborates data from other reports, i.e., individuals who seek psychotherapy are chiefly those who have symptoms of mood changes,^46-48 as well as the areas in which individuals manifest more suffering when they seek psychotherapy are: marriage/partner, family/children and conflicting with work.^46,47,49

In group 2, there were more discharges, and in group 1, more interruptions: an expected difference, since group 2 is comprised of patients who underwent treatment for a longer period of time - which allows a higher probability of being discharged - and because interruptions tend to occur in the beginning of treatments.

In the specialized literature, variety of definitions of what is considered dropout and interruption is a confounding factor.^50-52 Independent of those definitions, the percentage of patients who quit psychotherapy (dropout or communicated interruption) in different ages, diagnostic groups and treatment modalities range between 30 and 60%.⁵³ In studies conducted in Brazilian psychotherapeutic services, dropout rates ranged between 35 and 68.7% of the population.^54-56 Such high index draws attention, but the data from our sample show even higher rates: in group 1, there were 16 interruptions (94.12%), and in group 2, 13 interruptions (76.47%).

The process of psychotherapy termination involves experience of important emotions, which may lead to a higher or lower difficulty in conducting the treatment to its conclusion. Patient and psychotherapist will have to face feelings associated with separation, grief and acceptance of limits of patients and psychotherapeutic couple. Therefore, variations in the process of treatment termination are very complex, and termination is often characterized by interruptions with patients returning to treatment later.⁵⁷ It is known that less than 20% of patients in community mental health centers undergo treatments whose termination is established in a mutual agreement.⁵⁰ Our sample confirms those statements: besides being mostly comprised of patients who interrupted treatments, a significant number had received some type of previous treatment.

Another issue to be discussed regarding the high number of interruptions is the fact that the psychotherapists were beginners. Psychotherapists' inexperience may lead to idealization and highly perfectionist expectation, aiming at a structural change or at a transference resolution that is not possible considering the patients' reality; it may also favor countertransference errors, and psychotherapists no longer support termination due to their own personal ambitions, which are most of the times higher than the patient's own ambitions.⁵⁷ This means that the psychotherapist's interpretation of psychotherapy termination as treatment interruption can often be influenced by those factors, leading to a distortion of what really occurred and overestimating interruptions.

According to a study by Sandell et al.,¹⁰ more experienced psychotherapists tend to achieve better results in their cases, and have a positive correlation between amount of time a person has been working and outcome of their treatments. There was no significant correlation between psychotherapist's time of experience and treatment outcome in our sample. It is believed that, in the case of psychotherapists with little experience in psychotherapy, difference in time of experience between them is very small and possibly does not have any significance in clinical performance.

A sample predominantly comprised of female patients corroborates data from other studies regarding prevalence of women who seek psychotherapy.^{12,16,19,58,59} Psychotherapist's gender, who were also mostly female, shows the tendency of a higher number of female psychotherapists among PP professionals in Brazil. Presence of male psychotherapists, which could be interpreted as a confounding factor, was controlled, since it occurred in an equal number in both treatment groups (one in each group).

It is pertinent to stress that differences in scores of patients' initial GAF in comparison groups were nor significant, i.e., the patients started their treatments with a similar level of global functioning. Such information is relevant, considering that patients' initial diagnosis was not performed and that there is only an idea of the clinical constellation in our sample through the screening of reasons for seeking treatment.

With regard to comparison between participants and non-participants, significant differences tend to partly justify non-participation of patients, such as lower number of sessions and contact after longer period of time of termination. The higher percentage of male non-participants who underwent up to 11 months of treatment was significant when compared with group 1, as well as the number of single patients among non-participants who underwent treatment for 1 year or more, when compared with group 2. Such information is interesting, since it mostly occurred in non-participants who were not found. It is possible to raise the hypothesis that being male and single may lead to a more intense search for new opportunities, which would partially justify the fact of not being found.

With regard to sample representativeness, a questioning that is always possible is whether the patients who were found and participated in the study are those who achieved their treatment objectives the most, and whether the patients who were not found or refused to participate are those to whom treatment failed.

In case the patients found in the institution's file are considered as a total number, i.e., 82 patients, the loss rate would be high (58.54%), but still in the same percentage of the outcome research conducted by Freedman et al.⁵⁸ When the patients who were not found (23) and those excluded because they were undergoing another treatment (9) are removed from that calculation, the rate drops to 32% (nine who refused to participate and seven who quit during the study) of the sample of 50 patients. That rate is above the losses found in the studies by Leuzinger-Bohleber et al.¹² and Grande et al.;¹⁹ however, it is below the losses presented in other outcome studies.^10,15,16

The statistically significant difference (p < 0.001) found in the total sample (initial GAF vs. final GAF) demonstrates that the patients' level of global functioning improved, independent of treatment group. However, there was no evidence of interaction between treatment duration and outcome (p = 0.581), i.e., in our sample treatment duration alone had no influence on improvement in global functioning, since the patients in both groups improved in the same proportion. Such fact differs from previous investigations, whose results confirm that treatment duration or time have a direct relationship with outcome.^{10,22,58,60,61} Increasing the number of participants in this study is necessary, since a larger sample could detect differences in treatment outcome in both groups.

Thus, some questions can be formulated: could the patient present intense improvement in global functioning in the beginning of the treatment? Would GAF be an instrument sensitive enough to detect structural and transference changes that usually occur in long-term treatments? Symptomatic improvement in the beginning of treatments and the possible lack of sensitivity in GAF for deeper changes could justify, to a certain extent, similar outcomes in both groups, i.e., this can be studied in future research.

Another issue to be considered, despite being controversial, is what Malan⁶² called "therapeutic Eros," that is, the enthusiasm that invades beginning psychotherapists. The author described a contagious predisposition by psychotherapists that lead to positive repercussions on therapy outcome, despite having later proposed the rejection of such hypothesis.⁶³ However, this seems to be a clinical fact waiting for a better understanding or confirmation by research studies.

In general, more experienced psychotherapists are expected to promote better treatment outcomes, such as confirmed by Sandell et al.¹⁰ Therefore, in more experienced psychotherapists there would probably be a higher probability of having a significant difference between treatment outcome of patients in groups 1 and 2.

It is relevant to consider that, in this sample, treatment duration, when analyzed separately is not necessarily a decisive factor for psychotherapeutic outcome; certainly other factors should be considered, such as therapeutic relationship, quality of object relations, self-awareness potential and reaction to interpretations.

Treatment effect size (initial GAF vs. final GAF) can be considered large (Cohen⁴⁴) in the total sample and in groups separately. Those rates, when compared with mean change index (effect size), developed by Leichsenring & Rabung,⁴⁵ were significantly higher, independent of treatment group (rates were similar to those found by Leichsenring et al.¹³). Those data, besides corroborating the results found in our study, allow us to consider that improvement found in patients have a higher probability of being attributed to psychotherapeutic treatment than to any other variable.

The high number of interruptions, previously pointed out, is in contrast with the patients' opinion about their treatments, issued by filling in the EQ. Most patients in both groups felt completely satisfied, very satisfied or reasonably satisfied with the treatment given by the psychotherapist to their problems; they considered that psychotherapy/psychotherapist made things much better or better in relation to reason for seeking treatment and characterized their status at the end of the treatment as better than in the beginning of the psychotherapy.

Specifically regarding treatment satisfaction, adding the percentages of completely satisfied, very satisfied and reasonably satisfied patients, group 1 had a satisfaction rate of 88.24% (11.76% of dissatisfaction), and group 2 had 94.12% of satisfaction (5.88% of dissatisfaction). Those data reinforce the idea that most patients are satisfied with outcomes at a shorter time than what would be considered ideal by their psychotherapists and often terminate the treatment without the consent of the psychotherapist, who wishes more improvement to his patient; that is, for the patient it was a case of discharge, but the psychotherapist considered it as an interruption. In other words, the number of discharges is underestimated by psychotherapists. These findings can be corroborated by Leuzinger-Bohleber et al.'s study,¹² in which analysts were much more critical regarding treatment outcome than patients.

EQ mean effectiveness score of 179.71 (SD = 40.056) in the total sample is relatively lower than what was found in other studies.^60,64 It should be taken into account that the samples in those studies are not homogeneous, and that such differences can express the settings in which the investigations were performed. As to comparison of EQ in treatment groups, there was no significant difference in ES score, i.e., the patients have similar opinions about the treatment, independent of treatment group.

There was no significant correlation between final GAF score and ES in the EQ in the total sample of patients and in treatment groups. This means that the specialists' opinion, issued by final GAF score, has no correlation with the patients' opinion, evaluated through ES. The data show a correlation tendency (a regular correlation and in the limit of significance) between final GAF and patient's current emotional status.

It is possible to assume that, despite both specialists and patients generally considering that the psychotherapy offered was effective, there was discrepancy as to the specific opinion for each case. In addition, the subjective dimension of each patient in their self-evaluation should be considered, since it leads to possible inclusion of their fantasies, expectations and transferences as an apparent sign of better or worse outcomes. Because there were two different measures (GAF and EQ), it was not possible to verify who the most rigorous were (raters or patients) when evaluating treatments.

Among the study limitations are small sample, retrospective data collection, use of undergraduate psychotherapists, use of few measurement instruments, absence of control group and diagnosis of assessed patients.

Using a larger sample, there is a possibility of identifying differences in treatment outcome of patients in group 1 and group 2, i.e., it is possible to find an interaction between treatment duration and outcome, as in other studies, considering that, with the number of participants in this sample, the outcome statistical power is small (8-15%). However, with that number of patients (17 in each group), it would have been possible to detect a clinically significant difference (1 or more SD) in means of GAF changes (initial and follow-up) in both treatment groups (power of 80% and α = 0.05).

Reconstruction of the past, after outcome occurrence, may induce a memory bias in the patient's perception toward the psychotherapeutic treatment. However, evaluation that the treatment was generally effective, in the opinion of both specialists and patients, suggests that the memory bias did not significantly distort outcome appraisal. The advantage of a retrospective study is not influencing psychotherapy process and outcome, which is significant in our country due to a certain resistance to research by psychotherapists.

The treatment conducted by inexperienced psychotherapists could be a confounding factor; however, since improvement outcomes in our sample were significant, this fact contributes to confirm treatment effectiveness, even by beginning professionals. What may not have been clear is the comparison made between treatment outcome and duration, since, as previously mentioned, more experienced psychotherapists could show significant differences in outcome between both treatment groups.

Use of multiple measurements seems to be the most adequate to evaluate outcomes in psychotherapy.^65,66 The instruments used in this study were limited to semi-structured interview, self-administered questionnaire and clinical scale (GAF) applied by specialists. GAF and interviews are widely used instruments in research studies, but EQ has its limits, such as non-validation into Portuguese (despite a translation adapted by Freedman et al.²² already being used in Brazil) and the fact that it is not well consolidated as a research instrument.

Follow-up interviews made by the first author of this study may mean a researcher bias, but it was balanced by the interview evaluation performed by independent specialists. It was not possible to blind the interviews as to their application time (initial or follow-up interviews), which could have generated some influence in the result of evaluations. However, specialists were prevented from evaluating initial and follow-up interview of the same patient.

Absence of a control group is mainly a consequence of ethical and logistic difficulties. Outcomes, however, are strengthened by the comparison performed with the expected mean change index in a control group developed by Leichsenring & Rabung,⁴⁵ in which a significantly higher change was detected in the group of psychotherapeutic treatment.

Lack of an initial diagnosis of the sample is another limitation. In this study, there was a certain compensation of diagnosis absence, by screening reasons for seeking treatment and the fact that the patients' global functioning level was not different in the beginning of the treatment. It is relevant to state that, according to Jones,⁶⁷ there is no direct relationship between formal diagnosis and process of change in psychotherapy, and it is rarely useful in anticipating treatment events.

Conclusions

The results of the present study indicate that the patients undergoing PP in our sample improved their global functioning and have a favorable opinion about the benefits of a psychotherapeutic treatment. However, specialists' and patients' opinion were not positively correlated, despite both demonstrating significant improvement of patients. Treatment duration alone was not a determiner of PP outcome. This is certainly an area of great relevance for further research in PP.

Investigating is one of the challenges of mental health professionals, independent of reach, limits and difficulties in researching treatment effectiveness. As De La Parra et al.⁶⁸ stated, "investigating represents an ethical imperative: is it really efficacious what we are proposing when we indicate psychotherapy?"

Acknowledgments

We would like to thank Dr. Falk Leichsenring and Dr. Rabung for sending us the article Change norms: a complementary approach to the issue of controls groups in psychotherapy outcome research (unpublished at that time) to be used in our study.

References

1. Vitória G. A busca da felicidade. Rev Isto É. 1995;1356:100-7.

2. Cavalcante R. A psicanálise no divã. Rev Superinteressante. 2002;181:42-50.

3. Lima JG. Freud é um vencedor: entrevista com Renato Mezan. Rev Veja. 2006;39(17):13-7.

4. Solms M. Freud está de volta. Sci Am Bras. 2004;3(25):51-6.

5. Hobson AL. Freud está de volta? Como num pesadelo. Sci Am Bras. 2004;3(25):57.

6. Eizirik CL. Psychoanalysis as a work in progress. Int J Psychoanal. 2006;87:645-50.

7. Kandel ER. A biologia e o futuro da psicanálise: um novo referencial intelectual para a psiquiatria revisitado. Rev Psiquiatr RS. 2003;25:139-65.

8. Nunes MLT, Lhullier AC. Histórico da pesquisa empírica em psicoterapia. Rev Bras. Psicoter. 2003;5(1):97-112.

9. Freud S. Estudos sobre histeria. Rio de Janeiro: Imago; 1989. Vol. VII. p. 12-115.

10. Sandell R, Blomberg J, Lazar A, Carlsson J, Broberg J, Schubert J. Varieties of long-term outcome among patients in psychoanalysis and long-term psychotherapy. A review of findings in the Stockholm Outcome of Psychoanalysis and Psychotherapy Project. Int J Psychoanal. 2000;81(Pt 5):921-42.

11. Brockmann J, Schluter T, Eckert J. [Therapy goals, change of goals and goal attainment in the process of psychoanalytically oriented and behavior long-term therapy - a comparative study from the private practices of insurance-registered psychotherapists]. Psychother Psychosom Med Psychol. 2003;53(3-4):163-70.

12. Leuzinger-Bohleber M, Stuhr U, Ruger B, Beutel M. How to study the quality of psychoanalytic treatments and their long-term effects on patients well-being: a representative, multi-perspective follow-up study. Int J Psychoanal. 2003;84(Pt 2):263-90.

13. Leichsenring F, Biskup J, Kreische R, Staats H. The Gottingen study of psychoanalytic therapy: first results. Int J Psychoanal. 2005;86(Pt 2):433-55.

14. Huber D, Klug, G e Von Rad, M. Die Münchner Prozess: Outcome Studie. Ein Vergleich Zwischen Psychoanalysen und psychodynamischen Psichotherapien unter besonderer Berücksichtigung therapiespezifischer Ergebnisse. In: Stuhr U, Leuzinger-Bohleber M, Beutel M, eds. Psychoanalytische Langzeittherapien. Stuttgart: Kohlhmmer; 2001.

15. Bond M, Perry JC. Long-term changes in defense styles with psychodynamic psychotherapy for depressive, anxiety, and personality disorders. Am J Psychiatry. 2004;161(9):1665-71.

16. Wilczek A, Barber JP, Gustavsson JP, Asberg M, Weinryb RM. Change after long-term psychoanalytic psychotherapy. J Am Psychoanal Assoc. 2004;52(4):1163-84.

17. Gerber AJ, Fonagy P, Target M, Bateman A, Higgit A. Structural and symptomatic change in psychoanalysis and psychodynamic psychotherapy of young adults: a qualitative study of treatment process and outcome, 2004. Disponível em: http://www.ipa.org.uk/research/pdf/gerber.pdf. Acessado jul 2006.

18. Bond M, Perry JC. Psychotropic medication use, personality disorder and improvement in long-term dynamic psychotherapy. J Nerv Ment Dis. 2006;194(1):21-6.

19. Grande T, Dilg R, Jakobsen T, Keller W, Krawietz B, Langer M, et al. Differential effects of two forms of psychoanalytic therapy: results of the Heidelberg-Berlin study. Psychother Res. 2006;16(4):470-85.

20. Wallerstein RS. Outcome research. In: Cooper A, Person ES, eds. The American psychiatric publishing textbook of psychoanalysis. Washington (DC) : Am Psychiatric; 2005. p. 301-15.

21. Bernardi RE. Investigación clínica e empírica sistemática em psicoanálisis. In: Lhullier AC, org. Novos modelos de investigação em psicoterapia. Pelotas: Educat; 1998. p. 23-42.

22. Freedman N, Hoffenberg J, Borus N, Eizirik C, Knijinik D. Tiempo en tratamiento: observaciones de Mid-Manhattan y Porto Alegre, Brasil. Invest Psicoan Psicoter. 2001:87-101.

23. Rascón SR, Corona PC, Latirgue T, Rios JM, Garza DL. A successful trial utilizing the Leuzinger-Bohleber methodology for evaluation of psychoanalytic treatment: preliminary report. Int J Psychoanal. 2005;86(Pt 5):1425-40.

24. Moreno CML, Schaleyeff C, Acosta S, Vernengo P, Roussos A, Lerner BD. Evaluation of psychic change through the application of empirical and clinical techniques for a 2-year treatment: a single case study. Psychother Res. 2005;15(3):199-209.

25. Gabbard, GO. Major modalities: psychoanalytic/psychodynamic. In: Gabbard GO, Beck JS, Holmes J. Oxford textbook of psychotherapy. EUA: Oxford University; 2005. cap 1. p. 3-13.

26. Howard KI, Kopta SM, Krause MS, Orlinsky DE. The dose-effect relationship in psychotherapy. Am Psychol. 1986;41(2):159-64.

27. Kopta SM, Howard KI, Lowry JL, Beutler LE. Patterns of symptomatic recovery in psychotherapy. J Consult Clin Psychol. 1994;62(5):1009-16.

28. Milrod B, Busch F, Leon AC, Shapiro T, Aronson A, Roiphe J, et al. Open trial of psychodynamic psychotherapy for panic disorder: a pilot study. Am J Psychiatry. 2000;157(11):1878-80.

29. Dare C, Eisler I, Russell G, Treasure J, Dodge L. Psychological therapies for adults with anorexia nervosa: randomized controlled trial of out-patient treatments. Br J Psychiatry. 2001;178:216-21.

30. Junkert-Tress B, Schnierda U, Hartkamp N, Schmitz N, Tress W. Effects of short-term dynamic psychotherapy for neurotic, somatoform, and personality disorders: a prospective 1-year follow-up study. Psychother Res. 2001;11(2):187-200.

31. Abbass A. Intensive Short-term Dynamic Psychotherapy in a private psychiatric office: clinical and cost effectiveness. Am J Psychother. 2002;56(2):225-32.

32. Burnand Y, Andreoli A, Kolatte E, Venturini A, Rosset N. Psychodynamic psychotherapy and clomipramine in the treatment of major depression. Psychiatr Serv. 2002;53(5):585-90.

33. Cooper PJ, Murray L, Wilson A, Romaniuk H. Controlled trial of the short- and long-term effect of psychological treatment of post-partum depression. I. Impact on maternal mood. Br J Psychiatry. 2003;182:412-9.

34. Hilsenroth MJ, Ackerman SJ, Blagys MD, Baity MR, Mooney MA. Short-term psychodynamic psychotherapy for depression: an examination of statistical, clinically significant, and technique-specific change. J Nerv Ment Dis. 2003;191(6):349-57.

35. Svartberg M, Stiles TC, Seltzer MH. Randomized, controlled trial of the effectiveness of short-term dynamic psychotherapy and cognitive therapy for cluster C personality disorders. Am J Psychiatry. 2004;161(5):810-7.

36. Muran JC, Safran JD, Samstag LW, Winston A. Alliance-focused treatment for personality disorders. Psychother Theor Res Pract Train. 2005;42(4):532-45.

37. Maina G, Forner F, Bogetto F. Randomized controlled trial comparing brief dynamic and supportive therapy with waiting list condition in minor depressive disorders. Psychother Psychosom. 2005;74(1):43-50.

38. Hersoug AG, Bogwald KP, Hoglend P. Changes of Defensive Functioning: Does Interpretation Contribute to Change? Clin Psychol Psychother. 2005;12(4):288-96.

39. Stiles WB, Barkham M, Twigg E, Mellor-Clark J, Cooper M. Effectiveness of cognitive-behavioural, person-centred and psychodynamic therapies as practised in UK National Health Service settings. Psychol Med. 2006;36:(4):555-66.

40. Leichsenring F. Are psychodynamic and psychoanalytic therapies effective? A review of empirical data. Int J Psychoanal. 2005:86(Pt 3):841-68.

41. Crits-Christoph P, Barber JP. Long-term psychotherapy. In: Ingram RE, Snyder CR. Handbook of psychological change: psychotherapy processes an practices for the 21st century. New York: Wiley; 2000. p. 455-73.

42. Seligman MEP. The effectiveness of psychotherapy. The Consumer Reports study. Am Psychol. 1995;50:965-74.

43. American Psychiatric Association. Manual diagnóstico e estatístico de transtornos mentais. 4th ed. Porto Alegre: Artmed; 1995.

44. Cohen J. Statistical power analysis for the behavioral sciences. 2dn ed. Hilllsdale (NJ): Erlbaum; 1988.

45. Leichsenring F, Rabung S. Change norms: a complementary approach to the issue of controls groups in psychotherapy outcome research. Psychother Res. 2006;16:594-605.

46. Urtiaga ME, Almeida G, Vianna MED, Santos MV, Botelho S. Fatores preditivos de abandono em psicoterapias: um estudo na clínica Sérgio Abuchaim. J Bras Psiquiatr. 1997;46:279-83.

47. Pinheiro SD. Vínculo e abandono em psicoterapia psicanalítica [dissertação]. Porto Alegre: Pontifícia Universidade Católica do Rio Grande do Sul; 2002.

48. Kessler RC, Berglund P, Demler O, Jin R, Koretz D, Merikangas KR, et al. The epidemiology of major depressive disorder: results from the National Comorbidity Survey Replication (NCS-R). JAMA. 2003;289(23):3095-105.

49. Montado G. Psicoterapia focal psicoanalítica: investigación de proceso y resultados. In: Bernardi R, Defey D, Hebert EJ, Fiorini H, Fonagy P, Gril S, et al. Psicoanalisis focos y aperturas. Montevideo: Agora/Psicolibros; 2001. p. 248-61.

50. Beck NC, Lamberti J, Gamache M, Lake EA, Fraps CL, McReynolds WT, et al. Situational factors and behavioral self-predictions in the identification of clients at high risk to drop out psychotherapy. J Clin Psychol. 1987;43(5):511-20.

51. Garfield SL. Research on client variables in psychotherapy: research pertaining to continuation in psychotherapy. In: Bergin A, Garfield S. Handbook of psychotherapy and behavior change. New York: Wiley; 1994. p. 190-228.

52. Lhullier AC. Abandono de tratamento em psicoterapias realizadas em uma clinica-escola. [tese]. Porto Alegre: Pontifícia Universidade Católica do Rio Grande do Sul; 2002.

53. Baekeland F, Lundwall L. Dropping out of treatment: a critical review. Psychol Bull. 1975;82(5):738-83.

54. Lopez MA. Considerações sobre o atendimento fornecido por clínicas-escola de psicologia. Arq Bras Psicol. 1983;35(2):123-35.

55. Carvalho RMLL, Térziz A. Caracterização da população atendida na clínica-escola do Instituto de Psicologia: PUCCAMP. Est Psicol (Campinas). 1988;5(1):112-25.

56. Lhullier AC, Nunes MLT, Antochevis AF, Porto AM, Figueiredo D. Mudança de terapeuta e abandono da psicoterapia em uma clínica-escola. Aletheia. 2000;11:7-11.

57. Gabbard GO. Elaboração e término. In: Gabbard GO. Psicoterapia psicodinâmica de longo prazo: texto básico. Porto Alegre: ArtMed; 2005. p. 161-78.

58. Friedman RC, Garrison WB, Bucci W, Gorman BS. Factors affecting change in private psychotherapy patients of senior psychoanalysts: an effectiveness study. J Am Acad Psychoanal Dyn Psychiatry. 2005;33(4):583-610.

59. Vargas F. Abandono de psicoterapia em instituição de formação de psicoterapeutas [dissertação]. Porto Alegre: Pontifícia Universidade Católica do Rio Grande do Sul; 2002.

60. Freedman N, Hoffenberg JD, Vorus N, Frosch A. The effectiveness of psychoanalytic psychotherapy: the role of treatment duration, frequency of sessions and the therapeutic relationship. J Am Psychoanal Assoc. 1999;47(3):741-72.

61. Kordy H, Von Rad M, Senf W. Empirical hypotheses on the Psychotherapeutic treatment of psychosomatic patients and short and long-term time unlimited psychotherapy. Psychother Psychosom. 1989;52(1-3):155-63.

62. Malan D. Toward the validation of dynamic psychotherapy. Nem York: Plenun Medical; 1976.

63. Malan D. Resultados sobre o segundo estudo em psicoterapia breve. In: Malan D. As fronteiras da psicoterapia breve: um exemplo de convergência entre pesquisa e prática Médica. Porto Alegre: Artmed; 1981. p. 57-65.

64. Hartman S, Zepf S. Saarland study of psychotherapy effectiveness and patient satisfaction. In: Fonagy P, ed. An open door review of outcome studies in psychoanalysis. London: International Psychoanalytical Association; 2002. p. 147-9.

65. Edmonstone Y, Freeman C. Research in psychotherapy. In: Freeman C, Tyler P, eds. Research methods in psychiatry: a beginner's guide. EUA: American Psychiatric; 1992. p. 209-32.

66. Aveline M, Strauss B, Stiles WB. Psychotherapy research. In: Gabbard GO, Beck JS, Holmes. Oxford textbook of psychotherapy. EUA: Oxford; 2005. p. 449-62.

67. Jones EE. Therapeutic action: a guide to psychoanalytic psychotherapy. New Jersey: Aronson; 2000.

68. de la Parra G, von Bergen A, del Rio M. Primeros hallazgos de la aplicación un instrumento que mide resultados psicoterapéuticos en una muestra de pacientes y de población general. Rev Chil Neuro-Psiquiatr. 2002;40(3):201-9.

Received May 27, 2007

Accepted July 5, 2007

Simone Isabel Jung was given a scholarship by CAPES from March 2005 to February 2006.

Publication Dates

• Publication in this collection
  19 Dec 2007
• Date of issue
  Aug 2007

History

• Accepted
  05 July 2007
• Received
  27 May 2007