Presentation is loading. Please wait.

Presentation is loading. Please wait.

Engagement and Treatment Completion in a Correctional Sample

Similar presentations


Presentation on theme: "Engagement and Treatment Completion in a Correctional Sample"— Presentation transcript:

1 Engagement and Treatment Completion in a Correctional Sample
Stephanie A. Van Horn, M.A. & Robert Morgan, Ph.D. Texas Tech University June 22, 2016 Engagement and Treatment Completion in a Correctional Sample

2 What’s Working in CLCO? Psychoeducation Treatment Engagement
Increased management of symptoms (Lukens & McFarlane, 2006) Reduced symptom severity and institutional misconduct (Liau, Shively, Horn, Landau, Barriga, & Gibbs, 2004; Pomeroy, Kian, & Abel, 1999) Treatment Engagement Varying Definitions Attendance, homework, participation Significant predictor of non-completion (Drieschner and Verschuur, 2010) To get closer look at what’s working, we wanted to focus on two key components: psychoeducation and treatment engagement Large focus on psychoeducation throughout the program. Research shows psychoeducation important piece of symptom management and behavioral change, in both clinical and correctional populations Treatment engagement is defined in various ways across the literature. Attendance, homework compliance, homework quality, participation in sessions, etc. Dry-shner and Vur-shurr found that attendance and active participation was predicitve of treatment completion in an outpatient sample of justice involved individuals

3 Research Questions Are participants gaining knowledge in each of the modules? Is knowledge acquisition predictive of treatment completion? To what extent are attendance and homework completion related to program completion? Is engagement early in the program predictive of completion?

4 Sample Three correctional facilities (N = 177)
Two separate residential facilities for dually- diagnosed offenders (n = 39; n = 130) One psychiatric correctional facility (n = 8) Sample size for analyses differed based on research question Sample size differed based on research question due to differences in implementation of the program. One residential facility used open group format so order effects could not be tested. In addition, only individuals who completed both pre and post module quizzes were included in analysis of knowledge acquisition.

5 Demographics M age = 34.38 years (SD = 10.36) 68.9% Male 31.1% Female
Race/Ethnicity 63.8% White 34.5% African-American 1.7% Other 12.4% Hispanic 36.2% Unreported Education 62.3% Less than HS 35.3% High school 20.8% Post- secondary Marital Status 64.4% Single 15.8% Married 19.2% Divorced 0.6% Widowed

6 Measures Knowledge Acquisition Attendance and Homework
Content quizzes administered before and after completion of Modules 2-9 Raw pre/post change scores Attendance and Homework Binary variables (0 = no; 1 = yes) Percent total for program completion Raw count for early prediction

7 Knowledge Acquisition
Pre Post n M (SD) t dav Awareness 117 11.12 (4.94) 14.76 (2.38) 9.49* .94 Thinking 142 4.50 (3.09) 9.08 (4.19) 11.81* 1.24 Medication 129 7.29 (1.83) 8.19 (1.84) 4.93* .49 Coping 179 6.98 (1.80) 8.83 (2.60) 10.34* .83 Emotion 122 7.12 (1.90) 9.89 (3.91) 7.06* .90 Associates 119 4.87 (1.61) 6.29 (1.63) 7.61* .84 Skills 66 4.58 (3.36) 6.55 (3.13) 3.96* .62 Substance 27 5.04 (3.04) 9.48 (4.41) 4.65* 1.17 Dependent t tests were conducted using pre and post test scores. Post scores significantly higher than pre scores in all modules with large effect sizes in most of the modules

8 Knowledge Acquisition & Program Completion
Knowledge acquisition across program (N = 169) M = 9.09 SD = 8.97 Statistically significant but minimal effect β = .04, p = .04 OR = 1.04, 95% CI [1.00, 1.08] 53% likelihood for +1 SD 51% likelihood for -1 SD Mean and SD somewhat difficult to interpret as each module had differing numbers of questions. Logistic regression conducted with total KA as predictor and program completion as outcome variable. Used estimates to calculate probability of program completion at mean, 1 sd above and 1 sd below. Effect of KA on program completion was negligible, as you can see by small difference in likelihood of completion.

9 Engagement and Program Completion
Percent of Total Attendance M = 71.2% SD =26.2% β = 1.29, p = .03 OR = 3.65; 95% CI [1.15, 12.11] 79.3% likelihood for +1 SD 51.3% likelihood at mean 22.4% likelihood for -1 SD To control for fact that individuals who completed the program had the opportunity to attend more sessions and complete more homework than those who didn’t, attendance was measured by calculating the percentage of sessions attended and homework completed of each person’s total possible attendance/homework. Using same method, calculated likelihood of completion at mean, and +/- 1 SD. Attendance was significantly related to program completion. Those who attended only 50% of sessions available to them were only 22% likely to complete the program. Those with average attendance, 71% in this sample, were just as likely to complete as not complete.

10 Engagement and Program Completion
Percent of Total Homework M = 64.1% SD = 26.8% β = 1.06, p = .07 OR = 2.89; 95% CI [.94, 9.19] 75.2% likelihood for +1 SD 51.3% likelihood at mean 26.7% likelihood for -1 SD Again, using same methods of data analysis, we found that homework completion was not significantly related to program completion. However, the effect sizes suggest there may be some clinical meaning here. Individuals who completed less than half the homework assigned to them were almost three times less likely to complete the program as those who completed almost all their homework assignments.

11 Identifying At-Risk Participants
Attendance in Module 3 related to program completion n = 47 M = 4.49 SD = 1.18 β = .84, p = .05 OR = 2.32; 95% CI [1.01, 5.36] 84% likelihood for 5 sessions 66% likelihood at 4 sessions 41% likelihood for 3 sessions Wanted to examine if it was possible to predict non-completion within the first few weeks of the program, using attendance and homework in each of the first 3 Modules as individual predictors. Only significant predictor was attendance in Module 3 (Medication at that time). Module was six sessions in length, those who attended only half those sessions only had 41% likelihood of completing the whole program, compared to those who attended all the sessions, who were over 80% likely to complete the program.

12 Summary Participants learning program content
Knowledge has negligible effect on completion Attendance and homework completion increase likelihood of program completion More investigation needed to identify those at-risk for non-completion in the early phases of program

13 Limitations & Future Directions
Quiz reliability Small sample size for early prediction Primarily inpatient Update and standardize quizzes Test quiz reliability Randomized Clinical Trial

14 Thank You


Download ppt "Engagement and Treatment Completion in a Correctional Sample"

Similar presentations


Ads by Google