Tuesday, March 14, 2017

Beware of Easy Answers: The perils of single studies

The ATSA listserv was recently immersed in a discussion about mixing various clients in treatment groups, including those who are higher risk with lower risk. The discussion arrived at a familiar place: the idea that treatment can actually elevate risk among those who are lower risk. There has been research suggesting that high-intensity interventions can sometimes increase risk among lower risk people in the criminal justice system (Smith, Goggin, & Gendreau, 2002), but the idea that treatment can make people worse is one that we should examine very seriously. Ours is a field in which we continue to ask questions about whether treatment works in reducing future sexual violence, and if so, how, with whom, with which methods, and with what kinds of therapists.

The discussion focused an influential paper by Brian Lovins, Christopher Lowenkamp, and Edward Latessa in 2009, which found that “low-risk sex offenders who successfully completed treatment were 27% more likely to be reincarcerated than sex offenders who did not receive halfway house services” (p. 353).

Although that finding appears at first to say that intensive sex-offender treatment of low-risk sex offenders caused an increase in sexual recidivism, we think it would be a mistake to draw that conclusion, both from the paper and in general application to practice.  Although this finding has apparently been put forward as an indication of treatment somehow increasing recidivism or sexual recidivism, it’s not clear that it’s treatment that accounts for the results.

Page 348 of the Lovins paper includes:

Recidivism was coded as incarceration for any new offense, return to incarceration for a technical violation, and any new arrests for a misdemeanor or felony offense. For the purpose of this study, incarceration for any new offense and return to incarceration for a technical violation were collapsed into a single measure of return to incarceration for any reason.

As we read this study, comparisons were made between people who were released from prison (a) directly into the community or (b) into halfway houses.  It seems important to note that:

1.     The two groups differed in terms of their living situation and level of supervision, not just intensity of treatment, and
2.     The recidivism variable was re-incarceration, not detected SEXUAL recidivism per se.

One has to wonder what other factors might have influenced release decisions that were not accounted for in the study? After all, the groups were not randomly assigned to their conditions and there is therefore no reason to think that they were equivalent.

Our reading of this study is that it might reflect different re-incarceration rates for returning citizens who are faced with different levels of community supervision (halfway house or not) rather than different intensity levels of sex-offender treatment. For example, on pages 347-348, the authors state:

The second sampling frame comprised parolees who were released directly to the community on discharge from the institution. Out of the original 3,273 offenders in the comparison group, 238 had a sexual crime as their instant offense. These comparison offenders may have been mandated to outpatient treatment in the community at release, but they did not receive the more intensive residential sex offender treatment.

This highlights the fact that neither of the comparison groups consisted specifically of untreated low-risk sex offenders. Further, page 348 of the Lovins et al. study states that:

Because there was no consistent measure of risk across programs and parole, the modified SFS (Note: this stands for Salient Factor Score; it is not a validated, stand-alone measure of recidivism risk for people who have sexually abused) was used to determine level of risk. The modified SFS includes the following risk factors: prior arrest, prior commitment, age at current offense, employed at arrest, history of community control violations, and history of drug use. The values for each variable are weighted and the total ranges from 0 to 10. Risk categories were developed based on the raw values. The modified SFS consists of four categories: low, low/moderate, moderate, and high.

As we understand this statement, not only is the study not reporting on differences in detected sexual recidivism, it is also the case that the “low risk offenders” are given that classification on the basis of something other than a commonly accepted assessment of their risk to sexually reoffend.

Taken together, there are good reasons not to consider this study as an indication that providing intensive sex-offender treatment to “low-risk” sex offenders causes an increase in (sexual) recidivism. To sum up our concerns about taking the Lovins et al. (2009) study as compelling evidence that low-risk people should not receive treatment:

·     Recidivism in this study refers to ANY re-incarceration, not specifically to a new sex crime, or even to a new crime at all.
·    None of the groups specifically consisted of people who did not receive sex-offender treatment. More specifically, it is a comparison of people who were or were not required to live in a halfway house after release from confinement.
·    Classification of “low risk” had very little or no relationship to what we would think of as “low risk” for sexual re-offense. Measures specifically developed and tested for this purpose were not used.
·    The number of people in the comparison group was 14, and in the “low-risk” “successful completion” group, even lower.  The actual differences in re-incarceration are likely around 3 of 14 for the comparison group, and perhaps 3 of 11 for the “successful completion” group.

We very definitely intend no disrespect to the authors, Brian Lovins, Christopher Lowenkamp, and Edward Latessa; this is an informative study. Quite the opposite: we want more studies to examine this question. For the moment, however, we are serious in our discouragement of professionals viewing this study as evidence that providing treatment to some released sex offenders causes an increase in (sexual) re-offending.

The authors of this post are people who have read the extant research and overseen programs, and at times have come to many differing conclusions as to the future of our field. This is precisely why we need better research. In our view, a primary take-away is that professionals in our field should continue to humbly, even quietly, try to go about the work of doing no harm and making every effort to prevent and reduce the harm of sexual abuse.


Gregory DeClue, Ph.D., ABPP (forensic), and David S. Prescott, LICSW

No comments:

Post a Comment