Predictors of treatment dropout in child psychoanalytical psychotherapy

Abstracts

INTRODUÇÃO: As altas taxas de abandono em psicoterapia e a lacuna na literatura sobre abandono de tratamento na psicoterapia de crianças justificam a realização de pesquisas com esse foco. A literatura aponta que algumas variáveis sociodemográficas e clínicas poderiam predizer o desfecho da psicoterapia. O presente estudo objetivou verificar preditores de abandono de tratamento na psicoterapia psicanalítica de crianças na amostra pesquisada. MÉTODO: Trata-se de um estudo documental, retrospectivo, com os prontuários de duas instituições de atendimento psicológico a crianças em Porto Alegre. RESULTADOS: Foram pesquisados prontuários de 2.106 crianças. Dessas, 200 tiveram alta e 793 abandonaram seus atendimentos. Os grupos foram comparados, e os resultados indicam que meninos apresentam mais risco de abandonar o tratamento; já crianças encaminhadas por neurologistas ou por psicólogos apresentam menos risco de abandono. Após o sexto mês de atendimento, o risco de abandono decai consideravelmente. DISCUSSÃO: Os resultados são discutidos à luz da literatura sobre gênero, abandono de tratamento e crianças em psicoterapia. Algumas hipóteses são levantadas para as associações e não-associações encontradas neste estudo. CONCLUSÕES: Conhecer preditores de abandono em psicoterapia possibilita aos terapeutas identificar precocemente pacientes pertencentes ao grupo de risco para abandono, oportunizando-lhes trabalhar preventivamente e mais diretamente aspectos de resistência e transferência negativa desses pacientes e seus familiares, principalmente nos primeiros 6 meses de tratamento. Criar técnicas de intervenção precoce com os pais de tais crianças e realizar tratamentos transdisciplinares também são saídas possíveis para evitar o abandono.

Abandono; psicoterapia psicanalítica; crianças

ORIGINAL ARTICLE

Predictors of treatment dropout in child psychoanalytical psychotherapy

Marina Bento Gastaud^I; Maria Lúcia Tiellet Nunes^II

^ISpecialist in Psychoanalytical Theory and Child and Adolescent Psychotherapy, Contemporâneo. MSc in Clinical Psychology, PUCRS. PhD student in Medical Sciences: Psychiatry, UFRGS. Porto Alegre, RS, Brazil.

^IIPhD in Psychology, Treatment and Prevention, Freie Univesitat Berlin. Professor, PUCRS, Porto Alegre, RS, Brazil.

Brazilian Federal Agency for the Improvement of Higher Education (Coordenação de Aperfeiçoamento de Pessoal de Nível Superior, CAPES).

Correspondence

ABSTRACT

INTRODUCTION: The high rates of dropout in psychotherapy and the shortage of studies on treatment dropout in child psychotherapy are strong reasons for further research on this topic. The literature suggests that some sociodemographic and clinical variables could predict psychotherapy outcome. The objective of the present study was to identify predictors of treatment dropout in child psychotherapy.

METHODS: This is a retrospective study involving the analysis of medical records of two institutions that provide psychological care to children in Porto Alegre, RS, Brazil.

RESULTS: We analyzed the medical records of 2,106 children. Of these, 200 children were discharged from treatment and 793 dropped out. The two groups were compared, and the results suggest that boys have higher risk of dropping out, while children referred by neurologists or psychologists are at a lower risk of dropping out. After six months of treatment, the dropout risk is considerably reduced.

DISCUSSION: The results are discussed based on the literature on gender, treatment dropout and child psychotherapy. Some hypotheses were proposed to explain the associations and lack of associations found in the present study.

CONCLUSIONS: Knowing predictors of dropout from psychotherapy makes it possible for therapists to identify patients belonging to the risk group in the beginning of treatment. Therefore, they can try to directly prevent aspects related to resistance and negative transference of these patients and their relatives, mainly during the first 6 months of treatment. Implementing early intervention techniques with the parents and delivering transdisciplinary therapies are possible solutions to avoid dropout.

Keywords: Dropout, analytical psychotherapy, children.

Introduction

Psychotherapy dropout — involving children, adolescents or adults — has been often investigated and discussed within the clinical and scientific environment, since several studies have demonstrated dropout rates ranging from 25 to 60% of cases,^1-4 depending on the study and service characteristics.

Among the studies on treatment dropout, most of them involve adult patients. Therefore, there is a gap in the Brazilian and international literature regarding dropout in child psychotherapy. With regard to studies on psychotherapy dropout, only 1 to 2% investigated children or adolescents.⁵ The fact that studies involve mainly adult samples suggests the need to investigate children, since factors related to treatment dropout in adults may not be the same for children. For instance, some differences between patient and therapist in terms of sociodemographic characteristics (such as male patients being treated by female therapist or older patient being treated by very young therapist) are important dropout predictors when considering adult psychotherapy, but this is not true for child psychotherapy according to Midgley & Navridi.⁴ Therefore, there is a lack of data identifying the factors that may interfere with dropout in child psychotherapy. It seems important to understand the factors that contribute to treatment dropout, both to establish criteria for the identification of cases at risk and to interfere to promote deeper engagement of the family in the treatment.⁵

Hence, there is need of conducting investigation on the patient's characteristics that lead him/her to decide terminating treatment, since often even the patients do not know exactly which factors are involved in their decision. In most cases, patients are not able to report any reasons for dropping out from treatment or, when they do report a reason, they mention financial difficulties.⁶

In adult patients, many studies have found clinical characteristics associated with dropout, such as: therapeutic alliance; change of therapist; number of sessions missed during treatment;⁷ psychopathological characteristics, namely "psychoticism" and "sentimentalism";⁸ interpersonal motivation and sensitivity;⁹ psychosomatic symptoms; dissatisfying sexuality; alcohol and pharmaceutical abuse; and exposure in infancy to violence and sexual abuse.¹⁰ Also in studies focused on adults, sociodemographic data have also demonstrated associations with dropout, among them: to live outside the referral area, lack of bus-passes, spontaneous demand to the service, male patient,¹¹ patient's age and patient with children.⁷

With regard to adolescents, Pelkonen et al.¹² investigated patients between 12 and 18 years old who dropped out from treatment. Their study included both clinical and sociodemographic data. The authors found association between early dropout and low parental socioeconomic status, not having mood disorder, not having psychotropic medication, and having substance abuse.

Differently from what occurs with studies involving adult psychotherapy, mental health studies on children are rare, and most data obtained from this population are related to school-aged children. However, the magnitude of the problems affecting children's mental health is a cause of concern. Based on a recent study,¹³ Nunes et al.¹⁴ estimated that 10% the population of Porto Alegre between 0 and 9 years old needed specialized mental health care.

Although there are few studies, the literature shows that there are sociodemographic and clinical predictors of dropout from child psychotherapy. The most relevant factors in cases of child psychotherapy according to researchers include: socioeconomic status, source of referral, delayed delivery of care, geographic distance from care facility, previous history of treatment, parental stress, parental expectations, type of child psychopathology, child's gender, child and parents' age, ethnic origins, language spoken at home, unemployment, mother's educational level, parental and family configuration, income, family size and duration of child's complaint.^3-5 Findings regarding the impact of these factors are discordant, and evidence suggests that there is not an isolated factor that can be necessary or sufficient to predict treatment dropout in children. In spite of that, Luk et al.³ suggest that dropout seems to be associated with family configuration (young mothers and single-parent families), socioeconomic status (socioeconomically disadvantaged families) and reason for seeking treatment (children with severe antisocial behavior). Kazdin,⁵ on the other hand, suggested there is a positive and direct association between dropout and socioeconomically disadvantaged family (lower educational level, lower income, financial support from government and/or living in very small houses and dangerous neighborhoods), family configuration (younger mothers, single-parent families, minority families), parents' psychopathology (maternal history of antisocial behavior during childhood), children's symptoms (great number of aggressive, antisocial symptoms and behaviors, regardless of diagnosis) and child's academic/educational functioning (lower than normal intelligence level, history of school failure). Midgley & Navridi⁴ demonstrated that treatment duration is associated with dropout: children who remain under treatment for longer than 2 years are not likely to dropout. These authors also found association between dropout and gender; with female children having a higher probability of completing their treatments without dropping out.

These studies show that unfavorable economic conditions of patients and their families is a factor associated with psychotherapy dropout. Garcia & Weisz¹⁵ studied the reasons for treatment dropout in child and adolescent psychotherapy and demonstrated that therapeutic relationship problems and money issues were the only factors distinguishing those patients who dropped out from those who completed their treatments. Due to the high financial cost of psychotherapeutic treatment, care delivered in institutions seems to be a good alternative when the patient needs to receive low-cost care. Nevertheless, there is discrepancy among the types of services provided by institutions and the types of patients who seek these institutions, since teaching outpatient clinics (main institutions providing psychotherapy to the general population) were organized according to the needs and possibilities of the professionals who work there and not necessarily take into consideration the demands of the population being treated.^16,17 Such discrepancy might be another variable associated with high rates of treatment discontinuation, since studies have proved that dropout is related to the lack of connection between therapist and patient.¹⁸ Therefore, there is need of studies that define the profile of those patients treated at teaching outpatient clinics that is more closely associated with treatment dropout with the purpose of training the professionals who work at these institutions to deal with these possible characteristics.

Child therapies are greatly influenced by parents or guardians: they decide about the time of the psychotherapy session, they take the patient and pay for the treatment. For instance, Garcia & Weisz¹⁵ conducted a survey involving parents of children and adolescents aimed at finding out the reasons for deciding on the discontinuation of their children's treatment. The authors concluded that therapeutic relationship problems was the main reason reported, followed by money issues. Luk et al.³ studied the parents of children with behavioral problems who dropped out from treatments delivered in a mental health service. Dropout showed an association with parental dissatisfaction with the treatment service and perception that the treatment was less organized. The same authors also associated dropout with mothers who were younger and less educated.

However, it seems that some of the child's characteristics are also involved in psychotherapy dropout, as evidenced by Di Gallo et al.¹⁹ The authors studied children and adolescents treated at a psychiatry outpatient clinic in Basel, Swiss. Treatment dropout had a higher incidence among children and adolescents who had been referred by emergence services and general hospitals, and there was lower incidence when the patients themselves sought treatment.

In addition, a study²⁰ conducted with parents of children being treated with psychotherapy aimed at investigating if providing informational material to parents would increase attendance rates. Parents watched videos and read brochures about the importance of their involvement in the process, how play is used in therapy, confidentiality and the importance of persisting with treatment until goals are met. The authors found that the preparation procedures did not increase the attendance rates. Such finding provides evidence, among other possibilities, that factors related to the child also need to be considered so that he/she keeps attending the psychotherapy sessions.

Therefore, the objective of the present study is to investigate the relationships between sociodemographic and clinical characteristics of children treated at psychoanalytical psychotherapy outpatient clinics and psychotherapy dropout. Based on results that contribute to the explanation of this phenomenon, strategic preventive measures might be taken to reduce the high dropout rates in child psychotherapy.

Objectives

- To analyze the relationship between sociodemographic variables (age, gender, family configuration, educational level and city of residence) and dropout in child psychoanalytical psychotherapy.

- To analyze the relationship between clinical variables (reason for seeking treatment, source of referral and treatment duration) and dropout in child psychoanalytical psychotherapy.

- To identify which sociodemographic and clinical variables can predict dropout in child psychoanalytical psychotherapy.

Method

This is a quantitative, retrospective, descriptive study based on medical reports on the treatment of children undergoing psychoanalytical psychotherapy.

The study was carried out at two institutions: Contemporâneo (Institute of Psychoanalysis and Transdisciplinarity) and CEAPIA (Center of Childhood and Adolescence Studies, Care and Research). Contemporâneo and CEAPIA are institutions that provide graduate education in psychoanalysis to specialists in child psychoanalytical psychotherapy. These institutions have outpatient clinics that offer therapeutic care to low- and middle-income population charging lower fees than the usual fees charged at private offices. The treatments offered at these institutions do not have a predefined termination (open termination), and the weekly frequency of sessions is defined by patient and therapist in the beginning of the treatment. Therapists' fees are established by therapist and patient based on the minimum value set by the institutions. We adopted the principle of the Brazilian Child and Adolescent Statute (Article 2nd, 1990) that states that a child is "the person younger than 12 years old".²¹ All children who sought psychotherapy at Contemporâneo and CEAPIA between 1979 and 2007 were included in the analysis. Medical records related to the second time the same patient sought care at a different moment were excluded.

Database was developed using SPSS (Statistical Package for the Social Sciences) 13, based on a form containing sociodemographic and clinical variables retrieved from the medical records. The variables of interest were defined according to the criteria listed next.

Treatment termination

Defined according to information entered on the medical records. When therapists did not objectively expressed a clear endpoint, information was interpreted based on data entered on the medical records regarding the case evolution according to the following definitions:²²

Nonadherence

Treatment is discontinued during psychotherapy evaluation phase, that is, before the goals set for the treatment are clear for both participants or when there is not indication of treatment. We assume that the evaluation phase lasts for 1 month.

Dropout

Psychotherapy is discontinued before the goals established in the contract are met, regardless of the reasons that led the patient or the therapist to terminate the treatment and regardless if this was a uni- or bilateral decision. The treatment should have lasted for at least 1 month for the patient to be considered a dropout.

Discharge

Psychotherapy is terminated when the goals set in the contract are met.

Gender, age, family configuration, educational level, child's city of residence, and source of referral

Defined according to the data entered on the patient's record, considering information provided by the child's parents or guardians.

Treatment duration

According to information provided by the child's therapist at the end of treatment, as included on medical records.

Reason for seeking treatment

Defined according to the description provided: a) by parents/guardians during screening or when filling out the initial contact form at the institutions; b) by the professional who carried out the patient's screening while filling out the evaluation record when the patient attended the institution for the first time; and c) by the therapist in charge of the case while filling out the patient's psychological evaluation form. The three first complaints reported in each one of these moments were collected and entered on the database. Information regarding the reason for seeking treatment included on the medical record were stored in an unstructured manner. For classification of this variable through the clinical analysis of each patient, we used the internalizing, externalizing, neutral and social behavior scales of the Child Behavior Checklist (CBCL), Syndrome Scale 6-18. The scales used include:

- Anxious/depressed (crying, fears, not feeling loved, etc);

- Withdrawn/depressed (shy, sad, prefers being alone, etc);

- Somatic complaints (dizziness, fatigue, nausea, headache, etc);

- Relationship problems (does not get along with people, dependent, people tease him/her, etc);

- Thought problems (hears voices, has visions, weird behaviors, etc);

- Attention problems (cannot concentrate, restless, has reveries, etc);

Oppositional defiant behavior (vandalism, thefts, lies, etc);

- Aggressive behavior (fights, screaming, arguments, etc);

- Learning problems (failure, difficulty in specific subjects, etc).

This analysis was performed by a group of referees that classified each patient according to the possible categories proposed by the CBCL based on the clinical evaluation of the complaints entered on the database.

In the beginning of the treatments, both at Contemporâneo and CEAPIA, the children's guardians signed a consent form allowing their data to be used with scientific and teaching purposes. Patients who for any reason refused to sign the consent form were treated with psychotherapy at both institutions, but were not included in the study. Data entered on the database did not include the patients' names.

Data analysis was carried out using the statistical software SPSS 13. For sample characterization, all variables of interest were analyzed in terms of survey (frequency and percentages). We used the chi-square test to evaluate the relations between sociodemographic and clinical variables and dropout. All variables were categorized, that is, they were discrete variables. Poisson regression was used to calculate the adjusted relative risk of dropout in order to identify independent (predictive) variables related to dropout. Cumulative probability of treatment dropout was estimated using the Kaplan-Meier curve. Results with a p value < 0.05 were considered significant.

Results

We analyzed the medical records of 2,106 children. Tables 1 and 2 show the characteristics of the sample. Some data were not entered on the medical records. Therefore, the number of cases without certain information is quite high for some variables.

Of the 2,106 children included in the total sample, 681 of them (32.3% of cases) lack information about the type of treatment termination. Table 3 shows the distribution of the sample in terms of treatment endpoint.

Since the objective of the present study was to analyze the relation between sociodemographic and clinical variables and treatment dropout, its final sample was comprised of children belonging to the categories "dropout" (n = 793) and "discharge" (n = 200). Both groups (discharge and dropout) were compared in order to identify which characteristics were related to dropout. The final sample consisted of 993 children. Results regarding the relation between dropout and sociodemographic and clinical variables are presented next.

Dropout and age

In this sample, variables "age" and "dropout" were not associated (chi-square = 3.476/degrees of freedom or DF = 3/p = 0.324).

Dropout and gender

Our results showed an association between treatment dropout and gender (chi-square = 5.911/DF = 1/p = 0.015). Male children had higher risk of dropping out from psychotherapy than girls.

Dropout and family configuration

In order to calculate the association between dropout and family configuration, it was necessary to allocate the several possibilities of caregivers into categories. Since "family" is a term that can be described, but it has a difficult conceptual definition because it encompasses different structures of human grouping throughout time,²³ we decided to group caregivers according to three different possibilities. As the literature demonstrates that there is a relation between treatment dropout and absence of a father figure, we questioned: which category do children who live with stepfathers belong to? Similarly, some studies have suggested that dropout rates are related to single-parent families; therefore, we proposed the following question: can the family consisting of stepfathers and stepmothers be considered a single-parent family?

The first possibility was to consider that children who live with their mother and stepfather belong to the category "living only with the mother" and the children who live with their father and stepmother belong to the category "living only with the father." According to this option of categorization, there was no association between dropout and family configuration (chi-square = 3.427/DF = 3/p = 0.330). The second possibility was to consider families consisting of stepfathers and stepmothers as nuclear families, which have the father and mother figures. According to this possibility, the variables "dropout" and "family configuration" are not associated (chi-square = 1.982/DF = 3/p = 0.576). The third possibility was to exclude from the sample those children who live with stepfathers and stepmothers because it is not possible to identify the peculiar characteristic of each relationship, which makes it impossible to decide between the options "living only with the mother/father" and "living with both parents." Such measure was aimed at avoiding the bias of including in the calculation a confounding variable. Calculating the relation between dropout and family configuration, defined according to the third possibility of categorization (that is, excluding families with stepfathers and stepmothers), we also found that there was no association between these variables (chi-square = 2.576/DF = 3/p = 0.462).

Thus, our results suggest that, in the present sample, there was no relation between treatment dropout and family configuration.

Dropout and child's educational level

The variables "dropout" and "educational level" are not associated (chi-square = 6.687/DF = 5/p = 0.245).

Dropout and city of residence

Our results suggest that there is not a relation between treatment dropout and the city where the child lives (chi-square = 1.727/DF = 2/p = 0.422).

Dropout and reason for seeking treatment

The relation between dropout and reason for seeking treatment was calculated in two different ways.

First, the calculation was performed using the nine specific categories suggested by the CBCL. According to this method, there was not association between dropout and reason for seeking treatment (chi-square = 4.087/DF = 8/p = 0.849).

Next, the nine options of complaints were recategorized according to the broad categories suggested by the CBCL, that is: a) social competence = learning problems; b) internalizing behavior problems = anxiety/depression, withdrawn/depression and somatic complaints; c) externalizing behavior problems = oppositional defiant behavior and aggressive behavior; and d) neutral scale = relationship problems, thought problems and attention problems. Again, there was not association between dropout and reason for seeking treatment (chi-square = 0.922/DF = 3/p = 0.820) using the recategorized complaints.

Thus, dropout and reason for seeking treatment were not associated in this sample.

Dropout and source of referral

There is also two ways to categorize source of referral.

The first option was to perform the calculations using the initial categories. According to this method, there was association between dropout and source of referral (chi-square = 23.672/DF = 7/p = 0.001). Children referred by neurologists and psychologists had lower dropout rates than the other children.

The second option was to perform the calculation defining new categories for source of referral. All children referred to psychotherapy by pediatricians, psychiatrists, neurologists, other medical professionals, speech therapists and education specialists were grouped into the same category, which was called "Combined Treatments." We assumed that these professionals refer children to psychotherapy so that they can receive combined treatments regarding their problems, and we aimed at checking whether the delivery of combined treatments was associated with psychotherapy dropout. The result showed that there is a threshold significance in this association (chi-square = 7.302/DF = 3/p = 0.063), suggesting that children undergoing combined treatments might have a lower dropout rate depending on the sample.

Therefore, there is a relation between dropout and source of referral for the sample studied.

Dropout and treatment duration

Treatment dropout was associated with treatment duration for the sample studied (chi-square = 131.924/DF = 4/p < 0.001). Between the first and the sixth month of treatment there is higher risk of treatment dropout. After the seventh month of treatment, the child has less chances of dropping out than up to the sixth month. There is a decrease in the percentage of dropout as the treatment continues, as shown in Table 4.

Treatment dropout is, therefore, related to child's gender, source of referral and treatment duration. In order to isolate the possible interference that a variable can have on another variable, we calculated the adjusted relative risk of dropout. Our results suggest that boys have 10% higher risk of dropping out from treatment than girls (adjusted relative risk = 1.10/95%CI 1.03-1.19/p = 0.009). If compared to referrals by neurologists, the adjusted relative risk of a patient dropping out from treatment when referred by another source was significant:

- Family referral:

Adjusted relative risk = 1.26 (95%CI 1.05-1.52/p = 0.013).

- School referral:

Adjusted relative risk = 1.22 (95%CI 1.03-1.45/p = 0.019).

- "Other" referral:

Adjusted relative risk = 1.24 (95%CI 1.13-1.59/p = 0.001).

Such results suggest that the patient referred by the family has 26% higher risk of dropping out from treatment than the patient referred by a neurologist. School referrals have 22% higher risk of dropping out than those referred by a neurologist. Since they were too few to enter the statistical analysis as an isolated item, patients referred by guardianship councils, social worker, education specialist and by the sources mentioned in the footnote of Table 2 were grouped into the category "others". Such category was associated with dropout, since patients referred to psychotherapy by these sources had 24% higher risk of dropping out than patients referred by a neurologist, which can be understood as a spurious result.

If compared to referrals by psychologists, the adjusted relative risk of a patient dropping out from treatment when referred by another source was significant:

- Family referral:

Adjusted relative risk = 1.21 (95%CI 1.02-1.42/p = 0.026).

- School referral:

Adjusted relative risk = 1.17 (95%CI 1.01-1.35/p = 0.039).

- "Other" referral:

Adjusted relative risk = 1.28 (95%CI 1.10-1.49/p = 0.001).

Such data demonstrate that the patient referred by the family has 21% higher risk of dropping out from treatment than the patient referred by a psychologist. School referrals have 17% higher risk of dropping out than those referred by psychologists. Patients referred to psychotherapy by "other" sources have 28% higher risk of dropping out than patients referred by psychologists, another spurious result.

The Kaplan-Meier curve demonstrates that, once the psychotherapeutic treatment has began (that is, not taking into consideration the first month of evaluation), the patient's probability of dropping out from psychotherapy in the first month of treatment is 13%. During the first 6 months of treatment, the child's probability of dropping out from psychotherapy is 54%. This means that, during the first 6 months, the patient has a higher probability of dropping out from treatment than being retained. The previously mentioned results showed that boys and children who were not referred by neurologists or psychologists contributed to the increase in this probability.

Discussion

Regarding the characteristics of the sample, similar data have been found in other studies on Brazilian children undergoing psychotherapy,^24-30 which have suggested higher prevalence of boys among the children referred to psychotherapy in psychology outpatient clinics. This population comprised school-aged children between 6 and 12 years old referred by their schools. The literature^31,32 demonstrates that boys have a higher incidence of externalizing behavior problems (such as aggressive and antisocial behavior) than girls, which is more harmful to school and family relationships than internalizing problems (such as anxiety, depression and withdrawn), and this is a possible explanation for boys being referred to psychotherapy more often than girls. Externalizing problems, in addition to being more visible, cause trouble in a more direct manner to relatives, which explains a greater demand of psychotherapeutic treatment for children presenting such problems. The age group more often referred to psychotherapy consists of school-aged children, and this explains the fact that the main source of referrals to child psychotherapy is school. It might be possible that school is better equipped to detect child problems than children's parents or guardians. And this is the reason why some symptoms are only noticed when the child starts attending school.³³ Considering that the school plays an important role both in socialization and knowledge acquisition, it is actually expected that problems such as aggressive behavior, attention problems, learning problems and relationship problems — the most prevalent complaints in this sample — become evident when the child starts attending school and are more easily detected by teachers than by parents, doctors or other caregivers. Nevertheless, we assume that school-related problems might be hiding other previous difficulties, which, upon school entrance, are noticed at school due to the child's difficulty in making friends, complying with the rules and adapting to the school environment, for instance, thus explaining the higher prevalence of school referrals.²⁸

The sample distribution regarding family configuration represents the change in the family structure that can be observed currently. Since the 20th century, the emergence of the "contemporaneous" family has been identified: marriage and family have been influenced by social changes; the patriarchal model has started to be questioned; there has been a higher tendency to a redefinition of male and female roles; women have been participating in activities that were previously restricted to men. Within this context, the increase in the number of divorces, the decrease in the number of formal marriages and the reduction in the number of children appear as significant aspects.³⁴ The increase in the number of divorces and remarriages and the resulting rupture of the matrimonial bond of the parental couple suggest that the new family structures can be defined based on the variables cohabitation and kinship.³⁵ The current plurality of family compositions was observed in our sample, since the supposed "traditional family" — consisting of father, mother and children — accounts for 57.6% of the sample, whereas 40.8% are scattered among the other possibilities of contemporary grouping.

Taking into consideration that both institutions are located in the city of Porto Alegre, the highest prevalence of children living in Porto Alegre and the metropolitan area is coherent and reasonable. The fact that 2.5% of the children receiving treatment in Porto Alegre are from other cities in the state of Rio Grande do Sul raises the hypothesis that there is an imbalance between the demand of care and the availability of psychological care for low- and middle-income children from cities located outside the metropolitan area, which constrains the families to cover long distances to reach the state capital and get treatment for their children. The percentage of children referred to the institutions by private psychologists (10.2%) might be a sign that many families that seek treatment at private offices cannot afford the fees charged by private care, being referred to the institutions.

The dropout rate found in the present study (37.7%) is in agreement with the literature with regard to child psychotherapy.^3,36 However, not taking into consideration the cases still being treated and those with undetermined treatment termination (data missing from medical records), the dropout rate was 62.1%. The study by Midgley & Navridi,⁴ conducted with the medical records of the Anna Freud Centre, in London, found a dropout rate from child treatment of 60%. Even though, it is not possible to define whether there are differences or similarities between the local reality in Porto Alegre and the international reality in terms of dropout rates because of the confounding bias caused by the use of different definitions for "treatment dropout" in the studies.

With regard to the association between treatment dropout and sociodemographic and clinical variables, we expected (based on the literature) to find association with age, family configuration, child's educational level, distance from the child's home and the outpatient clinic and reason for seeking treatment. However, these variables were not associated with dropout in our sample. Several hypotheses may be raised based on the absence of statistically significant association. Many studies we reviewed did not present the characteristics of their samples and only demonstrated the associations or lack of association with the variable of interest (dropout). Therefore, it might be possible that our sample (specifically children treated in Porto Alegre, belonging to low- or middle-income families, seen at outpatient clinics linked to graduate programs in psychoanalytical psychotherapy) is not similar to the samples studied by other researchers.

Furthermore, in order to characterize the variable "reason for seeking treatment," we collected the reasons to begin psychotherapy reported by parents, screeners and therapists; however, we did not carry out an analysis of the diagnosis of the child's problems. We raise the hypothesis that there is conceptual similarity between "reason for seeking treatment" and "diagnosis" in some studies that demonstrated association between dropout and complaint.^3,5 In the present study, only the complaint reported was considered, regardless of the child's diagnosis.

Similarly, conceptual differences may be involved in the absence of statistically significant association between dropout and family configuration for the present sample. The definition of "single-parent family" poses some difficulties, and the solutions depend on the theory adopted in the study. Thus, we suggest that authors provide as much details as possible in their reports about the definitions adopted to establish categories for their variables in order to enable a more reliable comparison between the findings and replication of the study in different samples.

The present study found association between dropout from child psychoanalytical psychotherapy and gender, source of referral and treatment duration.

Boys had slightly higher risk (10%) of dropping out from treatment than girls, which is in agreement with the literature.^4,5 In the studies that found an association between gender and dropout, gender was usually associated with complaint and diagnosis: in general, boys had higher chances of being referred to psychotherapy because they presented externalizing behavior problems and for being diagnosed with antisocial functioning and behavioral disorders. Severe cases like these, according to the literature, are at a higher risk of dropping out from treatment; therefore, allocating boys at the so-called "risk group" for psychotherapy dropout. However, in the present study, there was no association between dropout and reason for seeking treatment, what makes it difficult to determine or suggest explanations for the higher risk of dropout in male children. The fact that we did not find an association between complaint and dropout does not exclude the possibility of an association between gender and complaint, since — as mentioned above — the present investigation considered the complaints reported by parents, screeners and therapists in a free, spontaneous and unstructured manner, not taking into consideration the diagnosis. In addition, the model used to allocate complaints into categories (CBCL) has some limitations, such as, for instance, limited number of categories (9), which makes categories very broad and little specific. Very different complaints, such as phobias and depressions, are grouped into the same category for analysis purposes (anxiety/depression). Such lack of specificity may have caused a confounding bias in our findings.

In the present study, the combined treatment has proved to be a preventive factor against dropout, since it enables transdisciplinary and a more specialized understanding of the patient's problem, possibly increasing the effectiveness of the treatment and meeting parents' expectations. Referrals by neurologists are compliant with this multidisciplinary perspective. In general, these are cases with developmental difficulties that require permanent psychological follow-up for children and parents. When parents are able to notice and agree with the need of treatment, as it is usually the case for children with neurological problems, the risk of dropout is reduced.⁵

Kazdin⁵ points out that the more coercive the source of referral the higher the chance of dropout. Coercive sources are those that "force" patients to be treated (such as school), in contrast with the volunteer search for treatment (such as parents' spontaneous search for care). In the present study, however, parents' spontaneous search for care increases in 26% the risk of dropout when compared to the referral by neurologists, which is discordant with Kazdin, but is in agreement the ideas by Midgley & Navridi.⁴ These authors consider that parents' motivation to take their child to psychotherapy is ambivalent, and that their expectations regarding therapy often are too high, which leads to an increase in dropout rates.

Children referred by psychologists may have undergone psychological evaluation (which determined the need of psychotherapy) or may be referred by private psychologists who send those families that cannot afford private fees to institutions. In any case, referrals by psychologists enable a more accurate indication of psychotherapy, decreasing the risk of dropout. The lack of specific criteria for indication and contraindication of psychotherapeutic treatment is one of the factors responsible for high dropout rates.³⁷

Another possible origin of children referred by psychologists is the referral performed by the therapists who treat the child's parents when they notice their patients' children need treatment. In such cases, during their treatments, parents try to improve their participation in the child's conflict and receive help to better deal with phases of negative therapeutic response generated by their children's psychotherapy, which favors the continuity of the child's treatment.

Treatment duration is the variable most often associated with treatment dropout in the literature. Treatment dropouts are more common during the initial phases of treatment.^3,11,18 Establishing a solid therapeutic alliance between therapist and patient (and his/her relatives) is a protective factor against treatment dropout;⁷ however, a long-term relationship between those involved is necessary so that this alliance can be established. The cutoff point in which the risk of dropout starts being significantly reduced varies among the studies. In the present study, the risk of dropout started to significantly decrease at 6 months of treatment. The study by Urtiaga et al.¹⁸ shows that the risk starts to decrease after the tenth session (although the higher risk is present in the first four sessions). Midgley & Navridi⁴ reported that most children whose treatment ended due to discharge received treatment for longer than 2 years and these authors revealed that each year of treatment significantly reduces the risk of dropout. Luk et al.³ found results similar to those of the present study: their study demonstrated that, at the end of the sixth month of treatment, 48% of the children had dropped out from treatment, 24% had been discharged and 28% were still being treated.

One advantage of the risk factor model, such as the one used in the present study, is that information may be used to identify cases of risk for the outcome, in this case, treatment dropout. The usefulness of such information is that it enables the early identification of those cases at risk. Such early monitoring may be very useful in clinical practice. For instance, if a patient does not come to one or two sessions, the therapist's response may vary depending on the risk level. For a case identified as being at risk in the beginning of the treatment, the therapist may decide to interfere more actively in order to avoid an early treatment termination.⁵ The early identification of cases at risk enables the therapist to employ any resource available to retain the patient and to increase the patient's chances to have his/her needs fulfilled.

Conclusions

During the first 6 months of psychotherapy, child psychoanalytical psychotherapists must be attentive to the higher risk of treatment dropout presented by their male patients who were not referred by neurologists and psychologists. Treatment dropout is frustrating both to therapist and patient, since the former feels incompetent and professionally undervalued when he/she loses a patient and the later does not receive the help he/she expected to receive from psychotherapy. However, child psychotherapists need to be aware of the fact that certain patients are at a higher risk of dropping out from treatment (boys referred by sources different from neurologists and psychologists) and must focus all their efforts on aspects related to resistance and negative transference of such patients and their families during the first 6 months of treatment in order to avoid dropout. After the first 6 months of psychotherapy, the risk of treatment dropout is quite reduced.

There is the hypothesis, raised based on the results of the present study, that combined treatments may contribute to the prevention of treatment dropout in child psychoanalytical psychotherapy. Technically speaking, working with professionals of neurology, psychiatry, speech therapy and education might help reduce the risks of treatment dropout, since combined treatment provides a transdisciplinary vision of the patient's clinical status and enables specific technical management of the child's symptoms and suffering. Other possible technical management is the early identification of patients belonging to the dropout risk group and the better preparation of these patients' parents and caregivers to begin treatment, performing more initial interviews, better evaluating the indication and contraindication criteria for psychoanalytical psychotherapy and discussing, since the beginning of the evaluation, aspects related to the resistance and negative transference of the child's parents and relatives with the purpose of reinforcing the therapeutic alliance.

Since this is a study based on medical records, there are inherent limitations to the research method. The results (mainly regarding type of treatment termination and reason for seeking treatment) are based on information entered on the patients' medical records, which were filled out in a subjective manner by each therapist depending on their perception of the case. In addition, although the study involved a large sample of children undergoing psychotherapy, there is lack of information about several variables of interest, which is a typical limitation of retrospective studies based on documentation.

Even though, we hope that the present study may contribute both to the theoretical and technical understanding of child psychoanalytical psychotherapy. We suggest that further similar studies are conducted to broaden the necessary clinical understanding about the complex phenomenon of dropout in child psychoanalytical psychotherapy.

References

• 1. Botega NJ, Fontanella BJB, Gonçalves EB, Rodrigues RT. Ambulatório de psiquiatria em hospital geral - alguns dados do HC/Unicamp. Rev ABP-APAL. 1988;10(2):57-62.
• 2. Chen A. Noncompliance in community psychiatry: a review of clinical interventions. Hosp Community Psychiatry. 1991;42(3):282-7.
• 3. Luk ES, Staiger PK, Mathai J, Wong L, Birleson P, Adler R. Children with persistent conduct problems who dropout of treatment. Eur Child Adolesc Psychiatry. 2001;10(1):28-36.
• 4. Midgley N, Navridi E. An exploratory study of premature termination in child analysis. J Inf Child Adolesc Psychother. 2006;5:437-59.
• 5. Kazdin AE. Dropping out of child psychotherapy: Issues for research and implications for practice. Clin Child Psychol Psychiatry. 1996;1(1):133-56.
• 6. Vargas F, Nunes ML. Razões expressas para o abandono de tratamento psicoterápico. Aletheia 2003;17(18):155-8.
• 7. Lhullier A, Nunes ML, Horta B. Preditores de abandono de psicoterapia em pacientes de clínica-escola. In: Silvares E, org. Atendimento psicológico em clínicas-escola. Campinas: Alínea; 2006. p. 229-56.
• 8. Fassino S, Amianto F, Abbate Daga G, Leombruni P. Personality and psychopathology correlates of dropout in an outpatient psychiatric service. Panminerva Med. 2007;49(1):7-15.
• 9. Johansson H, Eklund M. Helping alliance and early dropout from psychiatric out-patient care: the influence of patient factors. Soc Psychiatry Psychiatr Epidemiol. 2006;41(2):140-7.
• 10. Nickel M, Nickel C, Leiberich P, Mitterlehner F, Forthuber P, Triit K, et al. [Psychosocial characteristics in persons who often change their psychotherapists]. Wien Med Wochenschr. 2004;154(7-8):163-9.
• 11. Melo APS, Guimarães MDC. Factors associated with psychiatric treatment dropout in a mental health reference center, Belo Horizonte. Rev Bras Psiquiatr. 2005;27(2):113-8.
• 12. Pelkonen M, Marttunen M, Laippala P, Lönnqvist L. Factors associated with early dropout from adolescent psychiatric outpatient treatment. J Am Acad Child Adolesc Psychiatry. 2000;39(3):329-36.
• 13. Fleitlich-Bilyk B, Goodman R. Prevalence of child and adolescent psychiatric disorders in southeast Brazil. J Am Acad Child Adolesc Psychiatry. 2004;43(6):727-34.
• 14. Nunes ML, Silva R, Deakin E, Dian S, Campezatto P. Avaliação psicológica e indicação de psicoterapia psicanalítica para crianças. In: Werlang B, Oliveira M, orgs. Temas em psicologia clínica. São Paulo: Casa do Psicólogo; 2006. p. 29-36.
• 15. Garcia JA, Weisz JR. When youth mental health care stops: therapeutic relationship problems and other reasons for ending youth outpatient treatment. J Consult Clin Psychol. 2002;70(2):439-43.
• 16. Barbosa JI, Silvares EFM. Uma caracterização preliminar das clínicas-escola de Fortaleza. Estud Psicol (Campinas). 1994;11(3):50-6.
• 17. Calejon LMC. Estudos com pacientes de clínica psicológica universitária. Mudanças-Psicoter Estud Psicossoc. 1995;3(3/4):235-54.
• 18. 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.
• 19. Di Gallo A, Amsler F, Bürgin D. [Discontinuation of treatment in child and adolescent ambulatory psychiatric care in Basel: an evaluation within the scope of quality assurance]. Prax Kinderpsychol Kinderpsychiatr. 2002;51(2):92-102.
• 20. Shuman AL, Shapiro JP. The effects of preparing parents for child psychotherapy on accuracy of expectations and treatment attendance. Community Ment Health J. 2002;38(1):3-16.
• 21. Brasil, Estatuto da Criança e do Adolescente. Lei Federal 8.069 de 13 de julho de 1990. Brasília: Diário Oficial da União. 1990. Disponível em: http://www.planalto.gov.br/ccivil_03/LEIS/L8069.htm
• 22. Gastaud M, Nunes ML. Abandono de tratamento na psicoterapia psicanalítica de crianças [dissertação]. Porto Alegre: PUCRS; 2008.
• 23. Osório LC. O que é família, afinal? In: Osório LC. Casais e famílias: uma visão contemporânea. Porto Alegre: Artmed; 2002. p.13-23.
• 24. Campezatto PVM, Nunes MLT. Caracterização da clientela das clínicas-escola de cursos de Psicologia da região metropolitana de Porto Alegre. Psicol Reflex Crit. 2007;20(3):376-88.
• 25. Marturano EM, Magna JM, Murtha PC. Procura de atendimento psicológico para crianças com dificuldades escolares: um perfil da clientela. Psicol Teor Pesqui. 1993;9(1):207-26.
• 26. Marturano EM, Toller GP, Elias LCS. Gênero, adversidade e problemas socioemocionais associados à queixa escolar. Estud Psicol (Campinas). 2005;22(4):371-80.
• 27. Romaro RA, Capitão CG. Caracterização da clientela da clínica-escola de psicologia da Universidade São Francisco. Psicol Teor Prát. 2003;5(1):111-21.
• 28. Santos MA. Caracterização da clientela de uma clínica da prefeitura de São Paulo. Arqui Bras Psicol. 1990;42(2):79-94.
• 29. Silvares EFM. É satisfatório o atendimento psicológico nas clinicas-escola brasileiras? Coletâneas ANPEPP. 1996;1(9):137-47.
• 30. Vanni MG, Maggi A. O que demanda à psicologia na rede pública de saúde em Caxias do Sul? Psico (Porto Alegre). 2005;36(3):299-309.
• 31. Bolsoni-Silva AT, Marturano EM, Pereira VA, Manfrinato JWS. Habilidades sociais e problemas de comportamento de pré-escolares: comparando avaliações de mães e de professoras. Psicol Reflex Crit. 2006;19(3):460-9.
• 32. Graminha SSV, Martins MAO. Procura de atendimento psicológico para crianças: características da problemática relatada pelos pais. Psico (Porto Alegre). 1994;25(2):53-79.
• 33. Gastaud M, Merg M. Diferenças de sexo e idade na psicoterapia de crianças. Rev Soc Psicol RS. In press.
• 34. Gueiros DA. Família e proteção social: questões atuais e limites da solidariedade familiar. Serviço social e sociedade. São Paulo: Cortez; 2002. p.102-20.
• 35. Wagner A. Possibilidades e potencialidades da família. A construção de novos arranjos a partir do recasamento. In: Wagner A. Família em cena: tramas, dramas e transformações. Petrópolis: Vozes; 2002. p. 23-38.
• 36. Kazdin A, Mazurick J. Treatment outcome among children with externalizing disorder who terminate prematurely versus those who complete psychotherapy. J Am Acad Child Adolesc Psychiatry. 1994;33(4):549-57.
• 37. Hauck S, Kruel L, Sordi A, Sbardellotto G, Cervieri A, Moschetti L, et al. Fatores associados a abandono precoce do tratamento em psicoterapia de orientação analítica. Rev Psiquiatria RS. 2007;29(3):265-7.

Publication Dates

• Publication in this collection
  24 Aug 2009
• Date of issue
  2009

History

• Accepted
  28 Nov 2008
• Received
  19 Sept 2008