Criminal Justice and Sexually Predators

Sexually violent predators and civil commitment: is selection evidence based?

Shayna A. Wrightena*, Marlene B. Al-Barwanib, Robert R. Moranc, Geoffrey R. McKeed and R. Gregg Dwyere

aDepartment of Biology, Francis Marion University, Florence, SC, USA; bCollege of Social Work, University of South Carolina, Columbia, SC, USA; cDepartment of Biostatistics, University of South Carolina, Columbia, SC, USA; dForensic Psychological Services, University of South Carolina, Columbia, SC, USA; eCommunity and Public Safety Psychiatry Division, Department of Psychiatry and Behavioral Sciences, Medical University of South Carolina, Charleston, SC, USA

(Received 3 December 2014; accepted 31 March 2015)

Sexual offenses represent an alarming proportion of crimes committed yearly. To address these concerns, several states, including South Carolina (SC), have enacted laws requiring sexually violent predators (SVPs) to be civilly committed to treatment. To date, no published study has examined sexual offenders recommended for treatment in SC. This study used a spe- cially designed statewide database (SC-SVP research database) to deter- mine which offender and offense characteristics were associated with increased likelihood of being recommended for civil commitment. Factors correlated with being more likely to be recommended included: being of a younger age at time of evaluation, prior sex convictions, having related and unrelated victims, a higher number of victims, frequent substance use, and a history of suicide attempts. Prior sex convictions, having both related and non-related victims, and a higher total number of victims align with characteristics associated with sexual recidivism. Frequent substance abuse and a history of suicide attempts do not mirror previous findings regarding sexual recidivism. These findings present new information regarding the civil commitment process of offenders being committed to the SC-SVP treatment program, characterize types of offenders committed to SC-SVP treatment program, and provide a foundation for using a computerized database in conducting sex offender research.

Keywords: sexually violent predators; sexual offenses; sexual offenders; sexual offender treatment programs; recidivism; civil commitment

Introduction

Sexual offenses, including rape, child molestation, and other sexually violent acts, represent a critical national concern. In 1997, the United States Supreme

*Corresponding author. Email: [email protected]

© 2015 Taylor & Francis

The Journal of Forensic Psychiatry & Psychology, 2015

Vol. 26, No. 5, 652–666, http://dx.doi.org/10.1080/14789949.2015.1040439

Court ruled in favor of the state of Kansas in Kansas vs. Hendricks declaring involuntary civil commitment of sexually violent predators (SVPs) under Kansas’ Sexually Violent Predator Act (SVPA) to be constitutional (Kansas v. Hendricks, 1997). Subsequently, in 1998 the South Carolina Sexually Violent Predator Act (SC-SVPA), which allows offenders determined to be SVPs to be involuntarily committed to an in-state treatment program designed for SVPs, was passed (Sexually Violent Predator Act, 1998).

The SC-SVPA established a multi-step screening process to assess sex offender eligibility for involuntary confinement to SVP treatment programs. South Carolina’s process for determining civil commitment recommendations includes seven steps. First, the sexual predator multidisciplinary review team (SPMDT), consisting of five qualified persons, reviews relevant records to assess if an offender meets criteria to be classified as a SVP (step 1: multidisci- plinary review team). If an offender is determined to be a SVP, his/her SPMDT case summary is forwarded to the prosecutor’s review committee (PRC), where it is reviewed along with offender records, and additional information submit- ted by the circuit solicitor to determine whether or not probable cause exists to believe the offender is a SVP (step 2: prosecutor’s review committee). If the PRC determines probable cause, the Attorney General files a petition for a probable cause hearing by the court with jurisdiction where the offense was committed (step 3: Attorney General).

If the court determines that the probable cause allegation is supported, the offender is transported to a secure facility, if not already held in one, and given the opportunity to contest the court’s probable cause determination at a proba- ble cause hearing (step 4: probable cause determination). During the probable cause hearing, the court reviews arguments and additional evidence to assess whether probable cause exists to believe the offender is a SVP (step 5: proba- ble cause hearing). Offenders receiving a court determination of probable cause are evaluated at the SC Department of Mental Health (SCDMH) by an expert examiner to determine SVP status (step 6: mental health evaluation). Next, a judge or jury hears the testimony of the evaluator(s) and other witnesses to determine whether or not the offender should be civilly committed to the SVP treatment program (step 7: SVPTP, trial by jury or judge). Only offenders receiving a unanimous jury decision or a judicial ruling for commitment are committed. The case against the offender may be dropped, with no further bearing on the offender, at any step throughout the process if it is determined at that step that the offender does not meet criteria for a SVP. Current treat- ment programs are long-term placements making it imperative that offenders committed to these programs truly classify as SVPs.

To properly identify offenders who are SVPs, various assessments for risk of sexual recidivism have been created and employed. These assessments include: Historical-Clinical-Risk-20, the Sexual Violence Risk-20, the Violence Risk Appraisal Guide, Psychopathy Checklist-Revised (although this assessment was originally designed to assess the presence of psychopathy),

The Journal of Forensic Psychiatry & Psychology 653

and sexual arousal profiles such as penile plethysmography (Rogers & Jackson, 2005). These tools are designed to aid decision-makers in determining which sex offenders are most likely to reoffend and are therefore most in need of being placed in treatment programs (Rogers & Jackson, 2005). Currently, in SC, the use of these aforementioned assessments during the civil commitment process is left up to the examiner(s) at the various stages of the commitment procedure. Although there is evidence supporting low recidivism rates of sex- ual offenses by both adult (Hanson & Bussière, 1998; Hanson, Harris, Helmus, & Thornton, 2014; Rettenberger, Briken, Turner, & Eher, 2004) and juvenile (Hagan, Anderson, Caldwell, & Kemper, 2010) sexual offenders, several studies have identified characteristics of sex offenders that correlate with a higher rate of sexual offense recidivism (Carpentier & Proulx, 2011; Harris & Hanson, 2004; Nicholaichuk, Oliver, Gu, & Wong, 2013; Rettenberger et al., 2004). Therefore, the SPMDT should be more inclined to recommend for commitment individuals who possess characteristics resembling those that have been associated with higher rates of sex offense recidivism.

It has been found that sex offenders under the age of 50 years at time of release are more likely to recidivate than offenders over the age of 50 at the time of release (Harris & Hanson, 2004; Nicholaichuk et al., 2013). Addition- ally, Harris and Hanson reported that adult offenders with a prior sexual offense have recidivism rates twice as high as first-time sex offenders (Harris & Hanson, 2004); a finding that is true of adolescent sex offenders as well (Carpentier & Proulx, 2011; Riser, Pegram, & Farley, 2013; Rettenberger et al., 2004). Rettenberger and colleagues found that in juvenile sex offenders, there is recidivism of 5% for first-time offenders and 13% for repeat offenders when examining recidivism within a five-year period after prison release (Rettenberger et al., 2004). Other factors that have predicted sexual offense recidivism are having victimized a stranger as opposed to a known victim and generally having higher levels of delinquency (Carpentier & Proulx, 2011).

Since recidivism rates are commonly low among sexual offenders, it is imperative that individuals who are committed to treatment programs, which are typically long-term placements, are those who pose the highest actual threat to the general population as well as those who will gain the most benefit from the program. Although several researchers have investigated the role of jury characteristics in civil commitment cases, very few studies have focused on examining the actual offenders who are committed to treatment programs. Gaining a better understanding of which sex offenders are more likely to be recommended for commitment will aid in better designing the programs to treat the specific characteristics and needs of those who are committed to treat- ment programs. Additionally, having a standardized approach for comparing who is being recommended to treatment programs compared to those who most fit the profile of individuals likely to recidivate may aid in determining the necessity of implementing required use of particular assessments or aid in

654 S.A. Wrighten et al.

creating an uniform procedure of identifying individuals recommended to treat- ment programs, since current recommendations are semi-subjective.

The current study utilized data extracted from SPMDT case summaries to compare sex offender features to determine which offenders are most likely to be recommended for civil commitment. It is hypothesized that if offenders recommended for commitment by the SPMDT align with reported characteris- tics of sex offenders likely to recidivate, then SC sex offenders most likely to be recommended for treatment under the SC-SVPA would be under the age of 50, have more than one reported sex offense, have a younger age at first offense, and will have shown no preference for relatedness to victim.

Methods

The SC-SVP research database was created to compile profiles of all sex offenders screened by the SPMDT for civil commitment in SC under the SC-SVPA. The SPMDT conducts an extensive review of perpetrator charac- teristics, criminal history, and victim profiles to determine perpetrator eligibility for long-term confinement within a secure facility for treatment of sexually aggressive behavior. Data extracted from a sex offender’s SPMDT case sum- mary are converted into an electronic sex offender profile and entered into the SC-SVP research database. The database currently contains profiles for 800 sex offenders convicted within state jurisdiction, but grows with each new sex offender case presented to the SPMDT.

SC-SVP research database design

The SC-SVP research database provides detail for a broad range of sexually violent individuals who are often excluded from empirical study. According to the SC-SVPA, an individual is subject to a SVP civil commitment evaluation if he or she has:

(a) pled guilty to, pled nolo contendere to, or been convicted of a sexually violent offense; (b) been adjudicated delinquent as a result of the commission of a sexually violent offense; (c) been charged but determined to be incompetent to stand trial for a sexually violent offense; (d) been found not guilty by reason of insanity of a sexually violent offense; or (e) been found guilty but mentally ill of a sexually violent offense. (Sexually Violent Predator Act, 1998)

All conviction types subject to this statute render an offender eligible for inclusion in the SC-SVP research database. Therefore, the SC-SVP research database is demographically inclusive and contains profiles of sex offenders from all genders, age groups, conviction statuses, and levels of mental health competency.

The Journal of Forensic Psychiatry & Psychology 655

Data available on the SPMDT case summaries are culled from South Carolina Department of Corrections records, National Crime Information Centre records, South Carolina Department of Juvenile Justice records, and the SCDMH records for offenders adjudicated guilty but mentally ill or not guilty by reason of insanity. With the exception of protected health information (PHI), the majority of these data are publicly available and were originally col- lected for non-research purposes. No PHI was available to or utilized by the research team for this study. Analysis of these archived SPMDT case sum- maries and development of the SC-SVPA research database was reviewed and ruled exempt from informed consent requirements by the Institutional Review Board.

Participants

SPMDT case summaries of 800 sex offenders were analyzed. The sample consisted of 8 women and 702 men. Offenders younger than 18 years of age at time of conviction for index offense and offenders assigned to DJJ were classi- fied as juveniles. All others were classified as adults. Sex offenders were classified as recommended/committed or released based on SPMDT recommendations to refer for civil commitment review or release from further consideration at either the initial SPMDT stage of review or at the second review stage.

Measures

Perpetrator characteristics

SPMDT case summaries provided extensive details about perpetrator demo- graphics, criminal history, educational background, psychiatric profile, and completed stages of review. For this study, perpetrator demographics of interest included age. Educational background included highest grade completed, IQ score with reading and math subscores, and participation in vocational training (yes/no). Criminal history was defined as the presence (yes/no) and number of prior sex convictions and prior felony convictions. Other criminal history vari- ables included age at first offense, age at first sex offense, number of victims, and victim–perpetrator relationship (related/unrelated/both). Mental health variables of interest include frequency of substance use (none, occasional and frequent), history of suicide attempts (yes/no), and presence of visual or auditory hallucinations (yes/no).

SPMDT statistical analyses

All data were analyzed using SAS version 9.2). Chi-square tests were used to assess whether statistically significant differences in nominal variables between

656 S.A. Wrighten et al.

the juvenile and adult groups existed. Student’s t-tests were conducted to deter- mine differences in the continuous variables between released and committed groups. Any controlling variable found to be statistically significant between the juvenile and adult group was included in the regression models used for analysis. Multiple logistic regression models were used to assess relative risks and odds ratios for dichotomous outcome variables while controlling for any statistically significant variables. All statistical tests were conducted using a significance level of α = .05. All models were tested for their specific model assumptions such as linearity of relationships, independence of errors, homoscedasticity of errors, and normality of errors.

Results

Demographics and criminal history

As can be seen in Table 1, this study included analysis on a total of 800 sex offenders. There were 8 females and 45 males who were either dismissed from the civil commitment process at the SPMDT level or recommended for com- mitment, but dropped from further consideration at the second stage. All those that moved forward (n = 327) in the sample were male. The overall sample age range was 12–72 years (mean 42.8 and SD 11.64). Completed education ranged from 1 to 23 years (mean 10.4 and SD 2.3) and the IQ range was from 43 to 164 (mean 93.3 and SD 15.5). To follow are the findings based on which final determination stages recommended offenders reached compared to those not recommended for commitment by the SPMDT (Tables 2 and 3).

Offender characteristics

For those subjects identified by the SPMDT as needing commitment, offender age at first sex offense related to likelihood of referral for commitment by a factor of 1.08 for every year of age decreased (p = .05). That same relation and factor applied when comparing those not recommended to those commit- ted at the level of court decision (p = .0001). When adding those recom- mended for commitment by the SPMDT, but not moving past the PRC stage of the process, every year of age decreased at first sex offense resulted in 1.1 times greater chance of being committed (p = .0001). For every year of decreased age, offenders were 1.07 times more likely to be recommended for commitment compared to those not recommended once they reached the mental health evaluation stage (p = .03).

Comparing those recommended for commitment to those who progressed only to the mental health evaluation stage, being the former was 1.08 times more likely to be committed for every year increase in age at time of evalua- tion (p = .02). When compared to those not recommended for commitment by the SPMDT to those who were recommended and eventually went to trial, the

The Journal of Forensic Psychiatry & Psychology 657

Table 1. Characteristics of sex offenders released or recommended for civil commitment as a SVP.

Released (1’s and 2’s) Recommended (6’s,

7’s, and 8’s) p-Value

N Mean (%) Std. N

Mean (%) Std.

Demographics Male 465 – – 327 – .0181* Female 8 – – 0 – Age 472 31.85 13.78 327 42.85 11.64 .0001* Education 365 10.41 2.29 286 10.73 2.83 .1175 IQ 438 93.29 15.56 307 91.56 17.02 .1502

Offender characteristics Age at first sex offense 469 25.90 12.19 326 24.11 11.55 .0001* Age at time of release 472 31.85 13.78 327 42.85 11.64 .0181* Prior sex offense(s) 462 .29 .60 324 .96 1.03 .0001*

Number of prior sex offenses 0 prior sex offenses 356 77.1 – 128 39.5 – – 1 prior sex offense 83 18.0 – 118 36.4 – – 2 prior sex offenses 17 3.7 – 49 15.1 – – 3 prior sex offenses 6 1.3 – 23 7.1 – – 4 prior sex offenses 0 .0 – 4 1.2 – – 5 prior sex offenses 0 .0 – 1 .3 – – 6 prior sex offenses 0 .0 – 1 .3 – –

Victim characteristics Relationship to victim(s) .0001* Related 139 30.7 – 55 17.9 – – Unrelated 287 63.4 – 191 62.0 – – Both 27 37 – 62 20.1 – – Number of victims 466 1.64 2.00 322 3.39 3.46 .0001*

Educational level and vocational training Reading scores 396 8.26 3.55 306 7.74 3.90 .0635 Math scores 388 6.89 2.55 304 6.88 2.67 .9582 Received vocational training

53 37.3 – 139 51.9 – .0050*

Psychiatric profile Alcohol/drug use .0120* None 270 58.4 – 211 65.3 – – Occasional 53 11.5 – 18 5.6 – – Frequent 139 30.1 – 94 29.1 – –

Suicide attempt(s) 43 9.2 – 64 19.8 – .0001*

*p < .05.

658 S.A. Wrighten et al.

likelihood of going to court increased 1.05 times for every year older they were at time of evaluation (p = .0028). The rate increased to 1.15 when com- paring those not recommended to those committed (p = .0001). When adding those recommended for commitment by the SPMDT, but not moving past the PRC stage of the process every one year increase in age resulted in 1.17 times greater chance of being committed (p = .0001).

The presence of prior sex convictions also correlated with the likelihood of reaching various stages of the commitment process. For every prior sex convic- tion, the rate of reaching the stage of a required mental health evaluation, rather than not being recommended at all or being dropped from the process at the second step increased at a rate of 1.98 times (p = .0001). Reaching the next stage of a court decision had the same relationship, but at the lower rate of 1.57 (p = .04). When comparing those recommended for commitment versus not at the mental health evaluation stage, the latter were 2.2 times more likely for every prior sex conviction (p = .01). For those who reached the court stage, being committed was 1.65 times more likely for every prior sex offense con- viction (p = .03). Reaching any of the stages from mental health evaluation through court to actual commitment was 1.82 times more likely not being recommended by the SPMDT or being recommended but dismissed at the next stage for every prior sex offense (p = .0001).

Victim characteristics

Subjects recommended for commitment by the SPMDT exhibited a distinctive pattern of choices in victim–offender relationship and number of victims. Subjects who had both related and unrelated victims were 10.81 times more likely to be recommended for commitment than those with only unrelated vic- tims (p = .0001). Compared to subjects with only unrelated victims, those with only related victims were 4.04 times more likely to be recommended for com- mitment by the SPMDT (p = .0013). As the number of victims per subject increased, so did the likelihood of being recommended for commitment at a rate of 1.74 (p = .0025). When comparing those recommended by the SPMDT to those who reached the stage of a mental health evaluation, the latter were 3.97 times more likely to reach that level for every additional victim (p = .0001). The rate decreased to 2.69 when comparing those not recom- mended by the SPMDT to those who reached the level of a court decision (p = .0001). Reaching the level of requiring a mental health evaluation in com- parison to those not recommended for commitment by the SPMDT and those recommended but not moving past the initial assessment at the attorney gen- eral’s office, the former were 1.50 times more likely to be recommended for every victim (p = .0002). Comparing those same groups resulted in a finding of a factor of 2.01 for every additional victim among those committed (p = .0001). Those who reached the mental health evaluation stage, the court stage and actual commitment were 1.9 times more likely to reach these stages

The Journal of Forensic Psychiatry & Psychology 659

for every additional victim than those subjects who were either not recommended by the SPMDT for further consideration or who received such a recommendation but were dismissed from the process at the very next stage of consideration (p = .0001). For every additional victim, the subjects were 3.48 times more likely to be committed at the court stage when compared to those not recommended for further processing by the SPMDT (p = .0001).

Education level and vocational training

As reading ability scores increased, so did the likelihood of being recom- mended all the way to the stage of a court by a factor of .85 compared to not being recommended for commitment by the SPMDT (p = .02). Combined reading and math score decreases correlated with a greater likelihood of reach- ing the stage of a mental health evaluation, court proceeding, or all the way to commitment when compared to those not recommended for commitment by the SPMDT or recommended but dropped from processing at the immediate next step at rates of .92 and 1.09, respectively (p = .01; p = .05).

The presence of vocational training corresponded to a 2.85 time greater chance of being recommended for commitment than not at the mental health evaluation stage (p = .02).

Psychiatric profile

Frequent substance users were .33 times more likely to be referred to court than those not recommended for commitment by the SPMDT (p = .04). Sub- jects who had attempted suicide were 2.69 times more likely to be committed than those deferred for commitment consideration by the SPMDT (p = .0051). Having attempted suicide made persons 2.77 times more likely to move for- ward in the process when comparing those who were recommended for com- mitment by the SPMDT, but were dismissed from the process at the next step with those who made it to the evaluation stage regardless of final commitment (p = .0027). Comparing offenders with suicide attempts who were not commit- ted to those offenders actually committed to treatment, the subjects who had attempted suicide were 3.59 times more likely to be committed (p = .0001).

Being processed to the stage of a mental health evaluation to determine recommendation for commitment was 9.52 times more likely for those offenders who had experienced auditory hallucinations (p = .0074).

Discussion

This study identified a profile of who is being committed to the SVP treatment program in SC. This identification will enable the use of more standardized criteria for commitment determination and treatment in SC, as well as bring forth considerations for programs elsewhere to tailor treatment to the needs of

660 S.A. Wrighten et al.

the individuals being committed. Very few studies nationally and no studies in SC have focused on examining the characteristics of the offenders who are committed to SVP programs across as many variables and with as large a sam- ple size as this effort. The SC-SVP research database is the only one in the state of SC and one among few in the nation to contain a wealth of informa- tion on sex offenders, mental health histories of offenders, the nature of their crimes, and their victims with specific detail on victim characteristic. This study is unique in the literature to date in its depth and breadth of data points and number of subjects from a data-set of 800 subjects and growing monthly.

Offender characteristics

The presence of prior sex convictions correlated with increased likelihood of reaching the various stages of the commitment process. For every prior sex conviction, the rate of reaching the stage of a required mental health evaluation rather than not being recommended at all or being dropped from the process at the second stage of the process increased significantly. These findings are con- sistent with the fact that individuals with a previous sex offense history are more likely to recidivate than offenders without a history of sex offenses (Hanson & Bussière, 1998; Hanson, Morton, & Harris, 2003; Rettenberger et al., 2004). Additionally, in our study, subjects identified as needing commit- ment were more likely to be referred if they were of a younger age at first sex offense. This is an interesting finding with little in the recent literature for comparison as most age-related studies in the last several years have focused on age at release and the effect, if any, of aging on sexual offense recidivism. A frequently referenced meta-analysis of sexual recidivism factors did include started ‘offending sexually at an early age’ as having a small to moderate correlation with sexual recidivism (Hanson & Bussière, 1998). Harris con- ducted a study of 751 men with sex offenses referred for civil commitment (Harris, 2012). She looked at age of onset for offending and compared the sex offenders to nonsexual, but violent offenders, and property offending men. Her findings revealed that those with an onset of property crimes during middle adolescence had longer criminal histories and more offenses (Harris, 2012). Rice and Harris suggest that although not clearly definitive or independent of other age-related factors, onset of criminal, aggressive, and antisocial behavior at a young age could be indicators of enduring characteristics, and could be linked with later risk for recidivating (Rice & Harris, 2014). The authors also noted that adolescence-limited offending would not fit such a trajectory and that use of simply chronological age is unlikely to provide significant data for risk of sexual reoffending (Rice & Harris, 2014). Given findings such as in Dunsieth et al.’s study of 113 men convicted of sex offenses in which those with a paraphilia had younger ages of offending onset, the etiology of offend- ing might prove to be the critical factor (Dunsieth et al., 2004).

The Journal of Forensic Psychiatry & Psychology 661

Victim characteristics

Subjects who had both related and unrelated victims were more likely to be recommended for commitment than those with only unrelated victims. Hanson and Bussière found having an extra familiar victim as producing a small to

Table 2. Regression analyses for offender and victim characteristics at various stages of the civil commitment process.

Variable Comparisons Coefficient variable Measure p-Value

Age of first offense

8 vs. 1 −1.08 Likelihood to be committed .05

8 vs. 7 −1.08 .0001 8 vs. 2 −1.1 .0001 8 vs. 6 −1.07 .03

Age at time of evaluation

8 vs. 6 1.08 .02

8 vs. 2 1.17 .0001 7 vs. 1 1.05 Likelihood to reach court .003 8 vs. 1 1.15 .0001

Prior sex convictions

6 vs. 1/2 1.98 Likelihood to reach mental evaluation

.0001

7 vs. 1/2 1.57 Likelihood to reach court .04 8 vs. 6 2.22 Likelihood to be committed .01 8 vs. 7 1.65 .03 8 vs. 6/7 −1.82 .0001

Number of victims

1.74 .003

6 vs. 1 3.97 Likelihood to reach mental evaluation

.0001

7 vs. 1 2.69 Likelihood to reach court .0001 6 vs. 1 1.50 Likelihood to move to the next

stage .0002

6/7/8 vs. 1/2 1.9 Likelihood to reach the next stage

.0001

7 vs. 1 3.48 Likelihood to reach court .0001

Suicide attempts

2 vs. 1 2.77 Likelihood to move to the next stage

.003

8 vs. not 3.59 Likelihood to be committed .0001 Auditory

hallucinations 8 vs. 6 9.52 .0007

Notes: The seven stages include: review by the multidisciplinary review team (1), PRC (2), Attorney General review (3), probable cause determination (4), probable cause hearing (5), mental health evaluation (6), and trial by jury or judge (7). Following a trial an offender may be released or committed (8). Significance set at p ≤ .05.

662 S.A. Wrighten et al.

moderate correlation with recidivism (Hanson & Bussière, 1998). Levenson’s study of 229 sex offenders recommended for civil commitment in Florida revealed that 97% of those selected had extra familiar victims (Levenson, 2004). In Becker et al.’s study of 120 civilly committed men in Arizona, just 3% of offenders had only family member victims (Becker, Stinson, Tromp, & Messer, 2003). Compared to subjects with only unrelated victims, those with only related victims were more likely to be recommended for commitment by the SPMDT. These findings are curious in that it has been reported that one of the high risk factors for sex offense recidivism is offending against unrelated victims (Carpentier & Proulx, 2011; Hanson & Bussière, 1998).

We found that as the number of victims per subject increased so did the likelihood of being recommended for commitment. This is consistent with Levenson’s findings that for every additional victim, the subjects were more likely to be committed at the court stage when compared to those not recom- mended for further processing (Levenson, 2004, Levenson & Morin, 2006). However, a confounding variable might be the presence of a paraphilia given the potential for more victims such as found in Dunsieth et al.’s study (Dunsieth et al., 2004).

Table 3. Regression analyses for offender and victim characteristics with various factors.

Variable Compared factors

Coefficient variable Measure p-Value

Victim relatedness

Related and unrelated vs. unrelated only

10.81 Likelihood to be committed

.0001

Related only vs. unrelated only

4.04 .01

Education and vocational training

Reading scores .85 Likelihood to reach court

.02

Reading and math scores combined

−.92 Likelihood to reach mental evaluation, court or be committed

.01

Vocational training

2.85 Likelihood to reach court

.02

Substance use Frequent substance use

.33 Likelihood to reach court

.04

Suicide attempt

Suicide attempt

2.69 Likelihood to be committed

.0005

Notes: The seven stages include: review by the multidisciplinary review team (1), PRC (2), Attorney General review (3), probable cause determination (4), probable cause hearing (5), mental health evaluation (6), and trial by jury or judge (7). Following a trial an offender may be released or committed (8). Significance set at p ≤ .05.

The Journal of Forensic Psychiatry & Psychology 663

Education level and vocational training

As reading ability scores increased so did the likelihood of being recom- mended all the way to the stage of a court. However, both reading and math score decreases correlated with a greater likelihood of reaching the stage of a mental health evaluation, court proceeding, or all the way to commitment when compared to those not recommended for commitment by the SPMDT or recommended but dropped from processing at the immediate next step. These findings are counterintuitive to many findings of sex offender characteristics and recidivism rates. Riser, Pegram and Farley found that juvenile sex offend- ers generally have a lower IQ score than other offenders (Riser et al., 2013). Moreover, released offenders had more education in Levenson’s study as well (Levenson, 2004). This is not consistent with the Becker et al. study in which all offenders but 38% had gotten at least a high school diploma or GED (Becker et al., 2003). The Hanson and Bussière meta-analysis did not find that low intelligence correlated with sexual offense recidivism (Hanson & Bussière, 1998). Unfortunately, comparing school success and even scores on reading and math assessments are not necessarily indicative of IQ given potential differences in effort, attention, quality of educational setting, and other non intellectual functioning variables.

Psychiatric profile

Frequent substance users were more likely to be referred to court than those not recommended for commitment by the SPMDT. In Becker’s study, 81% of offenders reported a history of substance use (Becker et al., 2003). In Dunsieth et al.’s study, 85% of offenders were found to have a substance use disorder (Dunsieth et al., 2004). However, the Hanson and Bussière meta-analysis did not find that alcohol abuse correlated with sexual recidivism (Hanson & Bussière, 1998). Our findings showed that subjects who had attempted suicide were more likely to be committed than those deferred for commitment con- sideration by the SPMDT. Comparing offenders with suicide attempts to those actually civilly committed, the subjects who had attempted suicide were more likely to be committed. This was not a common variable in other published studies so it is unknown if this finding is consistent with committed persons in other states, and warrants further study before conclusions can be inferred.

Limitations and summary

This study’s focus on examining the parallels between characteristics of indi- viduals recommended for commitment and features shown to be associated with a higher risk of recidivism has limitations. First, sex offender type was not taken into consideration, although other studies have shown that recidivism rates differ among different types of sex offenders (Glowacz & Born, 2013;

664 S.A. Wrighten et al.

Rettenberger et al., 2004). Additionally, the current study analyzed juvenile and adult offenders together, but recidivism rates may differ between juvenile and adult offenders warranting different assessment criteria for these offender age groups. Lastly, this study was based on retrospective record review so the available data are drawn from what was collected by a variety of persons in some cases and not by a study generated protocol.

Overall, this study provides a foundation for understanding the type of sex offender who is most likely to be recommended for commitment to the SVP treatment program in SC. These findings provide evidence that many of the characteristics possessed by those most likely to be recommended for treatment overlap with factors associated with individuals most likely to recidivate. This work raises questions for sexual offender treatment programs beyond the study site state and for the validity of the criteria used for selection to commitment in such long-term programs. This study provides support for the need of fur- ther investigation into correlations between which individuals are being deferred from treatment, recommended but not committed, and committed and then released, and their respective sex offense recidivism rates. Studies such as this one aid in providing the basis for standardized commitment criteria and evidence based development of treatment program content created to rehabilitate individuals identified as SVPs.

Disclosure statement There are no known conflicts, academic, monetary, personal or otherwise, between any of the authors and the following work.

Funding This work was supported by a South Carolina Department of Mental Health Ensor Foundation Scholars Grant.

References Becker, J. V., Stinson, J., Tromp, S., & Messer, G. (2003). Characteristics of individuals

petitioned for civil commitment. International Journal of Offender Therapy and Comparative Criminology, 47, 185–195. doi:10.1177/0306624X03251114

Carpentier, J., & Proulx, J. (2011). Correlates of recidivism among adolescents who have sexually offended. Sex Abuse, 23, 434–455. doi:10.1177/1079063211409950

Dunsieth, N. W., Nelson, E. B., Brusman-Lovins, L. A., Holcomb, J. L., Beckman, D., Welge, J. A., … McElroy, S. L. (2004). Psychiatric and legal features of 113 men convicted of sexual offenses. The Journal of Clinical Psychiatry, 65, 293–300. doi:10.4088/JCP.v65n0302

Glowacz, F., & Born, M. (2013). Do adolescent child abusers, peer abusers, and non-sex offenders have different personality profiles? European Child & Adolescent Psychiatry, 22, 117–125. doi:10.1007/s00787-012-0333-2

The Journal of Forensic Psychiatry & Psychology 665

Hagan, M. P., Anderson, D. L., Caldwell, M. S., & Kemper, T. S. (2010). Five-year accuracy of assessments of high risk for sexual recidivism of adolescents. International Journal of Offender Therapy and Comparative Criminology, 54, 61–70. doi:10.1177/0306624X08325348

Hanson, R. K., & Bussière, M. T. (1998). Predicting relapse: A meta-analysis of sexual offender recidivism studies. Journal of Consulting and Clinical Psychology, 66, 348–362. doi:10.1037/0022-006X.66.2.348

Hanson, R. K., Harris, A. J., Helmus, L., & Thornton, D. (2014). High-risk sex offend- ers may not be high risk forever. Journal of Interpersonal Violence, 29, 2792–2813. doi:10.1177/0886260514526062

Hanson, R. K., Morton, K. E., & Harris, A. J. R. (2003). Sexual offender recidivism risk what we know and what we need to know. Annals of New York Academy of Science, 989, 154–166. doi:10.1111/j.1749-6632.2003

Harris, A. J. R., & Hanson, R. K. (2004). Sex offender recidivism: A simple question [Public safety and emergency preparedness canada]. Retrieved from http://www. publicsafety.gc.ca/cnt/rsrcs/pblctns/sx-ffndr-rcdvsm/index-eng.aspx

Harris, D. A. (2012). Age and type of onset of offending: Results from a sample of male sexual offenders referred for civil commitment. International Journal of Offender Therapy and Comparative Criminology, 97, 1226–1247. doi:10.1177/ 0306624X12448649

Kansas v. Hendricks, United States Supreme Court 521 U.S. 346 (1997). Levenson, J. S. (2004). Sexual predator civil commitment: A comparison of selected

and released offenders. International Journal of Offender Therapy and Comparative Criminology, 48, 638–648. doi:10.1177/0306624X12448649

Levenson, J. S., & Morin, J. W. (2006). Factors predicting selection of sexually violent predators for civil commitment. International Journal of Offender Therapy and Comparative Criminology, 50, 609–629. doi:10.1177/0306624X06287644

Nicholaichuk, T. P., Oliver, M. E., Gu, D., & Wong, S. C. (2013). Age, actuarial risk, and long-term recidivism in a national sample of sex offenders. Sex Abuse, 26, 406–428. doi:10.1177/1079063213492340

Rettenberger, M., Briken, P., Turner, D., & Eher, R. (2004). Sexual offender recidivism among a population-based prison sample. International Journal of Offender Therapy and Comparative Criminology, 48, 1–21. doi:10.1037/lhb0000052

Rice, M. & Harris, G. (2014). What does it mean when age is related to recidivism among sex offenders? Law and Human Behavior, 38, 151–161. doi:10.1037/ lhb0000052

Riser, D. K., Pegram, S. E., & Farley, J. P. (2013). Adolescent and young adult male sex offenders: Understanding the role of recidivism. Journal of Child Sexual Abuse, 22, 9–31. doi:10.1080/10538712.2013.735355

Rogers, R., & Jackson, R. L. (2005). Sexually violent predators: The risky enterprise of risk assessment. Journal of American Academy of Psychiatry Law, 33, 523–528. Retrieved from http://www.jaapl.org/content/33/4/523.long

Sexually Violent Predator Act, 321 U.S.C. § 44-48-30 (1998).

666 S.A. Wrighten et al.

Copyright of Journal of Forensic Psychiatry & Psychology is the property of Routledge and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.