Who receives substance abuse treatment in the ‘real world’ of the prison? A register-based study of Finnish inmates

Abstract

Assessing and targeting substance abuse-related treatment needs according to evidence-based practice has become the norm in most prison administrations. Not everyone with a need will, however, receive support in practical settings. Drawing on Finnish prison registers, we show that of all prisoners released in 2011, 60% were assessed as having a need for substance abuse-related support. Of these, 22% received an intervention in prison. Two multivariate models were used to examine the factors related to selection into interventions. The main factors associated with receiving an intervention with evidence-based programmes were a longer sentence, Finnish nationality, younger age and treatment motivation. The predictors for any substance abuse interventions (including non-evidence-based) were the same, with the exception that motivation was no longer a significant predictor, and female gender remained significant. The selective use of motivation as an entrance criterion for interventions can be a means of systemic adaptation to a combination of a high prevalence of substance abuse problems and a lower prevalence of treatment motivation in the prison population. We discuss how our results might reflect a Scandinavian way of doing evidence-based, prison-based drug treatment—one that stresses rehabilitation as a goal in itself in addition to reducing recidivism. Most prisoners, however, do not receive any intervention at all.

Keywords::

• substance abuse
• prison
• risk assessment
• treatment needs
• selection to treatment
• RNR model

Acknowledgements

This work was supported by the Finnish Foundation for Alcohol Studies.

Notes

¹ The data were collected by the second author, who has a thorough knowledge of the limitations and possibilities of the database.

² In this study, we presume that a person ‘has been assessed’ in some way if a sentence plan or a risk and needs assessment was made. It was assumed that a sentence plan was made, if a date indicating a completed sentence plan was registered or if there was one or more needs/goals registered for the prisoner. The risk and needs assessment was assumed to be made if one was registered.

³ The questions concerning substance abuse resemble the OASys Offender Assessment System (Home Office, 2002).

⁴ The reliability between motivation measured in the sentence plan and motivation measured in the risk assessment was low (0.372 for the substance abuse motivation variables and 0.316 for overall motivation). The reason should be that prison staff will record motivation only in one place—the risk and needs analysis if one is made and otherwise in the sentence plan (Of the persons for whom motivation was mentioned in the sentence plan, only 51.9% had a risk and needs analysis made). Thus it was important to include motivation recorded in both the sentence plan and the risk and needs assessment in the analysis in order to get a comprehensive picture of (measured) motivation among prisoners.

⁵ Finnish open prisons are units for prisoner who are expected to ‘make it’ with less supervision; the prisoner has more freedom to govern over his or her time concerning for an example outdoor recreation and leisure-time activities (Pratt, 2008); however, the prisoner is expected to follow timetables and prison rules (Mäkipää, 2013) or to take part in group activities, work or study. In closed prisons, activities are more carefully scheduled (outdoor visits are restricted) and security is at a higher level (Mäkipää, 2013). However, also these facilities are characterized by Pratt (2008, p. 121) as being of ‘relative material comfort’ in comparison to international counterparts (Cf. Pratt, 2008).

⁶ Some prisoners had spent several years imprisoned, and the locations of some of the substance-free wards had changed during their sentence. This was inquired about and taken into account. One prison also operated a TC on a half-year basis (at partly variable times), which was also taken into account. Only stays that lasted for more than one day were taken into account.

⁷ All models were tested for multicollinearity by OLS regression using two separate regression models with a bivariate response variable (received intervention/did not receive intervention) and the same covariates. No multicollinearity was detected. Also relevant interactions between age, gender, nationality (cf. Guerrero, Marsh, Cao, Shin, & Andrews, 2014) and the other covariates were tested for but not detected. The models also controlled, in Model 2, for the type of goal in the sentence plan (substance abuse goals or not) and, in Model 3, for principal crime and recidivism, none of which were significant predictors when controlling for motivation and are, thus, not shown in the tables.

⁸ An assessment with a sentence plan or sentence plan or a risk assessment was conducted for 97.5% of all prisoners.

⁹ At the bivariate level, persons who stayed in a prison psychiatric hospital and in an outside unit also received evidence-based interventions significantly more often (For any substance abuse intervention, there was no difference.). The differences between those who had spent time in (any) medical care versus those who had not, were altogether very small and did not persist in controlling for covariates, so these variables were dropped from the multivariate models. One explanation is that the health care/treatment was already given at the healthcare/treatment units. The registers do not indicate which health care visits/stays were due to substance abuse and which due to other reasons.

¹⁰ The percentages for attending programmes and staying in contracts wards are not shown in tables.

¹¹ The model also initially contained motivation to change behaviour, recidivism and principal crime but they were statistically insignificant and did not contribute to the model, so they were removed.

¹² Table available from the first author.