Factor Structure for an Assessment of Barriers to Substance-Related and Mental Health Services Following Release from Incarceration

Many individuals reentering the community following a period of incarceration need mental health and substance-related treatment services; receiving these services often makes the difference between successful and unsuccessful reentry. Many barriers prevent individuals’ receipt of services during reentry but few barrier assessment tools exist.  This paper reports efforts to refine measurement of barriers to mental health and substance-related treatment services during reentry. The current study adapted the Allen Barriers to Treatment Inventory, which originally addressed barriers to the substance-related treatment of incarcerated women.

In this study, 309 incarcerated women and men were interviewed during the weeks prior to their release from jail, prison, or community-based correctional facility (CBCF). Of this total, 230 respondents anticipated a need for either substance-related, mental health, or both types of services following release.  Respondents rated 30 potential barriers on a 5-point scale (1 = not a barrier at all to 5 = a major barrier).  To avoid over-representation of individuals in these analyses, 135 individuals who reported both types of need were randomly assigned to have only their substance-related or their mental health responses included in Exploratory Factor Analysis (EFA). EFA is a statistical technique used for data reduction and in scale development. A preliminary Kaiser-Meyer-Olkin (KMO) measure equaling 0.813 was obtained; KMO values close to 1.0 indicate that EFA may be useful. Furthermore, Bartlett’s Test of Sphericity, significant at the p<.001 level, suggested that correlations among the 30 variables were strong enough to warrant EFA.

EFA was conducted using a principal axis factor extraction and oblique rotation, forcing the variables into six possible solutions ranging between four and nine factors. The data were best represented by the seven factor solution, in which 21 of the 30 items were retained and explained 56% of the total variance. All eigen values remained greater than 1. The Intra-Class Correlation (ICC) values for the seven scales were all in the .628 to .849 range. It should be noted that three factors had only two items each, which can contribute to factor instability. Based on review of the contents for each factor, the seven scales were labeled as follows: Characteristics of Treatment (4 items), Not Knowing about Treatment Programs (2 items), Competing Responsibilities and Children (4), Shame about the Problem or Treatment (3 items), No Resources to Pay for Treatment (2 items), Transportation Difficulties (2 items), and Not Ready to Make Changes (4 items). This factor structure overlaps considerably but not perfectly with one previously reported in the literature involving women leaving jail.

The scales derived through EFA yield a picture of seven categories of barriers these women and men anticipate and could experience during reentry.  Assessing barriers allows social workers to match different intervention strategies to individuals’ specific need and barrier patterns; some focused on “micro” individual change and case advocacy efforts, others on more “macro” levels involving community resources and cause advocacy. The scale structure could be useful for improving referral adherence and further human services research in both the mental health and substance use arenas.