Assessment of a penitentiary relapse prevention program

Ten structured meetings of cognitive-behavioral therapy for relapse prevention were applied with prison inmates. The project had the participation of 28 subjects, divided into a work group and control group (15 and 13 subjects respectively), who were evaluated before and after the intervention. There was no significant difference in reoffending. Despite this, the program reduced the fear of negative evaluation (especially among non-reoffenders) and the Social Stress and Escape Scale score. After one year, reoffenders had lower scores in the Social Stress and Escape Scale and a tendency to have higher scores in the Dysfunctional Attitude Scale. Uniterms: Cognitive therapy; Crime prevention; Prisons; Rehabilitation.

also necessary to "provide criminal individuals with the conditions to recover and return to society" (Salla, 1999, p.24).
In 2005, Wilson conducted a meta-analysis on structured cognitive-behavioral programs delivered to groups of offenders.The studies usually included a control group that received no treatment, a noncognitive-behavioral intervention, or a minimal treatment intervention clearly hypothesized to be less effective.The research had to verify whether the subjects relapsed or not after the intervention.Results showed that cognitive-behavioral programs were effective in reducing criminal behavior.
Based on the principle that maladaptive thinking influences criminal behavior, an intervention was created that focused on some thinking distortion characteristics, which were: control, cognitive immaturity, and egocentrism.Control corresponds to the "need for power and control over oneself, other people and the environment" (Mandracchia et al., 2007(Mandracchia et al., , p.1038)).Cognitive immaturity implies the lack of maturity and difficulty to generalize.Finally, egocentrism is when the subject interprets "the actions of the others, constructing a view of importance of oneself in relation to the environment" (Mandracchia, Mogan, Garos & Garland, 2007, p.1038).The interesting part of these interventions is that these thinking distortions also represent non-criminal maladaptive thinking.Some criminal rehabilitation studies also dealt with the Risk Principle.The Risk Principle establishes that the need for supervision and treatment of a criminal should be proportional to his or her risk level, which corresponds to their probability of recidivating (Lowenkamp, Latessa & Holsinger, 2006;Lowenkamp, Smith & Bechtel, 2007).These studies concluded that, intensive cognitive behavioral programs associated with other resources are for high risk offenders; rehabilitation programs for low risk offenders tend to increase reoffending.The problem with these studies is that none of the authors defined what kinds of interventions are appropriate for low risk offenders.
A cognitive-behavioral intervention program was carried out with medium and high risk offenders in the British Channel Island of Jersey (Raynor & Miles, 2007).The model used was the risk and needs assessment, which also focuses on prevention and treatment.The authors discussed whether the success of interventions was due to the small size of the service, which facilitated staff communication, or to the interventions themselves.
Many studies have shown the importance of evaluating the inmate's risk level prior to an intervention, however, some of them did not establish how this could be done, and, furthermore, each study used a different method to perform the risk assessment.This makes it impossible to establish whether high risk offenders are a homogeneous group or not.

Criminal relapse versus prison relapse
It is difficult to evaluate the rate of criminal relapse in Brazil due to the unreliability of data, which is not treated scientifically.The concept of penitentiary relapse used in this study is that was used by Adorno and Bordini (1989) -when a subject has already served his or her term and is imprisoned again.

Penitentiary relapse prevention program
A study on the role of psychologists, conducted within the State Prison System, showed a lack of specific resources that are essential for psychologists to perform their jobs and systematize their work within the context of the prison (Saffi, Bemvenutto, Martins, Caires & Rigonatti, 2002).This study prompted the creation of the Penitentiary Relapse Prevention Program, based on a proposal by Marlatt and Gordon (1993) for chemical and alcohol addiction and adapted to the needs of the prison population.This program uses the cognitivebehavioral therapy technique.This theory establishes that the way one interprets and thinks about a particular situation is what determines one's feelings and behavior.In other words, the behavior of individuals is guided by the way they structure their worlds, that is, their cognitions.The goal of a cognitive-behavioral therapist is to produce cognitive changes -i.e., changes in the patient's thinking and beliefs -so that a long-lasting emotional and behavioral change can take place.Therefore, the goal of therapeutic work is to access and change what is present but not necessarily apparent.
The Penitentiary Relapse Prevention Program focuses on the fact that people may not have caused the problems they have encountered; however, once they occur, these problems are their problems and they have to cope with them and their consequences.Taking all of the above considerations into account, this study was devised to raise awareness of the process that leads a detainee to perform antisocial actions.

Method
The study was approved by the Research Ethics Committee of the Universidade de São Paulo, protocol 919/03 on 16/12/2003, and all the participants signed the Terms of Free Prior Informed Consent (TFPIC) before their inclusion in the sample.
The Penitentiary Relapse Prevention Program was composed of 10 structured meetings that applied various techniques: Socratic questioning, brainstorming, problem-solving etc.In the first meeting, the program coordinator and the participants were introduced to each otherand the goals of the program were presented.In the second, third and fourth meetings, the participants worked on automatic thoughts and distorted automatic thoughts.Beliefs were dealt with in the fifth meeting.Participants worked on risk situations in the sixth and seventh meetings and worked with the behavioral chain in the eighth.The topic of the ninth meeting was "overcoming undesirable thoughts".Closure and assessment of the program was carried out in the final meeting.
Twenty-eight inmates (13 from the control group and 15 from the work group), participated from the beginning to the end of the study.The maximum sentence was less than 15 years and they had already served enough time to be eligible for an alternative sanction hearing (from medium security to halfway houses) and/or parole.
The program did not include inmates convicted for felonies defined by Article nº 5 of Law nº 8072 of July 25, 1990 (traffic of illegal drugs and narcotics, rape, indecent assault, felony homicide, terrorism, and torture) or detainees who worked outside the prison facility.

Instruments
A guided interview was created to outline the profile of individuals who participated in the program, as well as the research topics that conceptualize the cognitive approach.Additionally, some questionnaires or scales where applied both before and after the program, such as: the Stress and Social Escape Scale -to study discomfort in social situations and avoidance of and/or desire to avoid such situations (Echeburúa, 1997); the Rosenberg Self-Esteem Scale -self-esteem is the favorable opinion a subject has about him or herself (Romano, Negreiros & Martins, 2007); the Fear of Negative Evaluation Scale -which assesses the fear individuals have of being judged by others; a Questionnaire on Automatic Thoughts -to study the frequency of negative automatic thoughts associated with depression (Joseph, 1994), and the Dysfunctional Attitude Scale -to verify cognitive vulnerability for depression (Orsini, Tavares & Troccoli, 2006;Remor, 1997).Study participants were asked to sign the TFPIC, i.e., a document that explains the aims of the project and who it is directed toward, in a few simple words.The purpose of this document was to ensure the confidentiality of the participants' identities and inform them of the responsibilities of the researcher.

Procedures
The Penitentiary Relapse Prevention Program was presented to the Department of Social Reintegration of the São Paulo State Prison Administration Division.After approval by the department, the project was implemented in two prison facilities.
Inmates that met the sample characteristics and were interested in collaborating with the study were invited to participate.They then received a brief explanation on the aims of the study and the TFPIC.
All the participants first signed the TFPIC and were then divided into two groups: a work group and control group.Once the groups were separated, psychologists who did not know which group each participant belonged to applied the scales and intellectual level tests.The next step was a guided interview conducted by two psychologists -one supervising the work group and the other the control group.The psychologist who interviewed the work group carried out the Prevention Program with two weekly meetings per group of no more than 10 participants.At the end of the program, the same psychologists conducted new interviews, still without knowing to which group each participant belonged.Members of the control group were only submitted to the first and last phases of the study (assessment before and after the program) together with members of work group.Thus, the members of the control group were submitted twice to the interviews and the application of the scales, however, did not participated in the Penitentiary Relapse Prevention Program.All study participants (control group and work group) were monitored on a monthly basis for twelve months after the dissolution of the groups.
Re-offenders were considered those who had escaped from prison and were recaptured with or without another crime on their record.Non-re-offenders were those who had been judicially released (halfway house, parole, or bail), those who had escaped and had not been recaptured, and those who were serving terms in medium security prisons.
The data was analyzed using the Statistical Package for the Social Sciences (SPSS) 14.0 program.The first step was to carry out a statistical test to verify that the sample distribution was normal, which proved to be the case.Subsequently, the Student's t-test was used for the analysis of the means, the Fisher's exact test for the categorical data, and the Mann-Whitney test to compare the two groups.The stipulated significance level was 0.05, or 5%.

Results
The two groups were similar regarding age, total sentence, and time already served.There was no significant difference in the penitentiary relapse rate between inmates that participated in the program and those who did not participate (control group).
Regarding the scales, when comparing the results from the inmates who finished the program, it was noticed that after the program there was a decrease in the Stress and Social Escape Scale (p=0.03), and in the Fear of a Negative Evaluation Scale (p=0.02)(Table 1).
When looking exclusively at the work group members, they had already served more time (p=0.09).Control group recidivist inmates were more likely to show less fear of a negative evaluation prior to the initiation of the program (p=0.07) and had lower Stress and Social Escape Scale scores (p=0.01).
All the inmates were analyzed again one year after the conclusion of the program and re-offenders were more likely to have a lower score in the Stress and Social Escape Scale (p=0.05)(Table 2) and in the Questionnaire on Automatic Thoughts (p=0.07)(Table 2), as well as a tendency to increase the score in the Dysfunctional Attitude Scale (p=0.07)(Table 3).
For the non-recidivists there was a noticeable trend towards a reduction in negative evaluation after the program (p=0.01) (Table 4).When exclusively considering the work group non-recidivists, there was a tendency towards a lower score in the Fear of a Negative Evaluation Scale (p=0.06)(Table 5).

Discussion
Based on analysis of the collected data, it can be said that the Penitentiary Relapse Prevention Program reduces fear of negative evaluations.The reduction in the Fear of Negative Evaluation Scale scores can be associated with improved self-esteem, as the participants felt safer and more confident.Valliant and Antonowics (1991) mention an increase in self-esteem and a decrease in anxiety in inmates who undergo Cognitive Therapy in weekly two-hour sessions.
Another point that appeared was that participants in the control group had a decreased score in the Stress and Social Escape Scale, and showed improvement despite not having participated in the Penitentiary Relapse Prevention Program.This might be because prison is a very hostile environment and the study gave participants the opportunity to be heard (in the interviews, before and after the program) and to tell their stories to people who were interested in them.This attitude alone can be considered as an intervention that relieves symptoms.This can be related to Beck's Theory (Beck, Whight, Newman & Liese 1993), which states that one of the essential points of a therapeutic approach is empathy, that is, therapists must look at a patient's world with the patient's eyes.This is attained when one is able to understand the other person's history and beliefs.
There was no significant difference in the penitentiary relapse rate between inmates that participated in the program and those who did not participate (control group).Some authors found no significant difference in their prevention program studies.Caldwell, Skeem, Salekin and Van Rybroek (2006) examined the impact of an intensive treatment program on male adolescent recidivism.Their conclusion showed that there were no differences for general and nonviolent recidivism, because offenses are more influenced by life circumstances than psychological factors.On the other hand, the program was more effective when it focused on high-risk offenders.Lowenkamp et al. (2007), when covering studies with the same aspects, quoted a study by Bonta (2000) that conclude "offender treatment was effective in reducing recidivism for higher risk offenders" (p.327).
A study conducted by Golden, Gatchel and Cahill (2006), which evaluated the effectiveness of a 22 session cognitive-behavior program with adult offenders classified as medium and high risk, in a National Correctional Institute, showed that, even though re-offending differences were not statistically significant, inmates who finished the program were more likely to present a reduction in the number of offenses.
A meta-analyses study confirmed that the analysis of correctional programs showed increases in recidivism rates in lower risk offenders.Another point mentioned was that the program's effectiveness increased when cognitive behavior therapy was associated with other services, and also when the program was more intensive (Lowenkamp et al., 2006).Petersila (2004) concluded, after summarizing studies on what is effective and what is ineffective in prison re-entry programs, that the programs should be intensive (at least 3 months), need to focus on higher risk individuals, and should use cognitive treatment techniques.The author highlights a problem: the academic literature results do not often reflect the experience of the staff practice.

Study limitations
When the research was initially planned in 2005, a sample of 200 subjects was established as ideal, based on the reliability of 10% and magnitude of 80%, considering a 70% relapse rate.The software used to determine the sample size was Stplan version 4.1, January 1996.The 200 subjects would have been divided into a control group and experimental group using a table of random numbers.
However, circumstances changed after the data collection began in June 2006.An important characteristic of the prison population is their suspicion of strangers; they are suspicious of those who do not belong to the prisoner group or to the technical and administrative group of the prison facility where they are incarcerated.This made data collection more difficult.Many inmates enrolled to participate in the study but dropped out when they were asked to sign the TFPIC, where they had to supply personal information, such as address and phone number.
In May 2006, the city of São Paulo suffered an "attack" by a crime gang known as the First Capital Command (PCC in Portuguese).The State of São Paulo's prison facilities have several crime gangs, and the First Capital Command is the most powerful, with the largest number of members.The First Capital Command is dominant in the two facilities where the study was carried out.In the months following this first "attack, " other "attacks" occurred and were violently countered by the police.Some workers who were part of the prison system were murdered due to actions promoted by the First Capital Command, including one who worked at one of the facilities where the study was being conducted.Another form of protest the inmates practiced during this period was the so-called "white strike", during which there were no uprisings, however, the prisoners refused to participate in any proposed activities.The inmates did not work, go to school or therapy, nor attend legal proceedings.
During the 13-months of data collection, 110 people enrolled to participate, however, 34 did not meet the inclusion criteria.Thirty-three of the remaining 76 refused to sign the TFPIC.This brought the study group down to 43 inmates (20 from the control group and 23 from the work group) who had served terms in medium security prisons, and were serving their second or further term (prison re-offenders).
Due to the problems mentioned above, the final sample was very small, especially regarding the analysis of subgroups.

Final Considerations
Some inmates enrolled in the study and quickly dropped out.As in every psychotherapy processespecially with this type of population -the bond with the professional is very important.The bond is not imposed on the patient but made possible and achieved through the relationship.In this study, the professionals involved were not members of the prison facility staff.They were strangers to the environment, which made adherence to the program very difficult.Those subjects who completed the program, especially those who participated in the work group, created an effective bond with the researchers.This is supported by the fact that two subjects insisted that their names be mentioned in the study, and one of them even asked to be mailed a copy of the study.
Considering the time frame, the analysis of significant changes would have called for a study conducted over a longer period.There is an initial discomfort at the beginning of the psychotherapy process, which is soon overcome.Unfortunately, the participants of this study were not given the time to overcome this discomfort.
Finally, this type of study is more adequate for higher risk individuals, which is why new studies need to focus on these types of people.

Table 1
Comparison of results of the scales at the beginning and end of the program Note: * This result is statistically significant.M: Mean; SD: Standard Deviation.Estudos de Psicologia I Campinas I 30(3) I 375-382 I julho -setembro

Table 2
Comparison of results of the scales before and after, of the subjects who and re-offended 1 year after the end of this

Table 3
Comparison of results of the difference of the results between of scales before and after the intervention Note: * This result is statistically significant.M: Mean; SD: Standard Deviation.

Table 4
Comparison of results of the scales before and after, of the subjects who and no re-offended 1 year after the end of this Note: * This result is statistically significant.M: Mean; SD: Standard Deviation.

Table 5
Comparison of results of the scales before and after, of the subjects who and no re-offended 1 year after the end of this and were of the work group Note: * This result is statistically significant.M: Mean; SD: Standard Deviation.* 0,92 * p F. SAFFI & F. LOTUFO NETO Estudos de Psicologia I Campinas I 30(3) I 375-382 I julho -setembro 2013