Bridging the gap: building and sustaining recovery capital in the transition from prison to recovery residences

Figures & data

Table 1. Assessment numbers (total for each REC-CAP assessment) from baseline assessment for HARP and SRR groups.

	REC-CAP 2	REC-CAP 3	REC-CAP 4	REC-CAP 5	REC-CAP 6	REC-CAP 7	REC-CAP 8
SRR	406	255	250	130	78	32	19
HARP	20	26	18	21	12	6	3

Table 2. Comparison between HARP and SRR groups at baseline (summary means for binary and numerical measures).

	SRR (n = 1,852)	HARP (n = 126)
Binary prognostics	%	%
Barriers (Accommodation)	28.0	21.4
Barriers (Substance Use)	60.7	54.8
Barriers (Risk Taking)	16.9	18.3
Barriers (Criminal Justice Involvement)	62.3	86.5
Barriers (Employment, Training and Education)	33.6	42.1
Needs (Mental Health)	44.0	59.5
Needs (Drug Treatment)	35.6	52.4
Needs (Alcohol Treatment)	15.6	15.1
Needs (Housing)	32.5	53.2
Needs (Employment, Training and Education)	41.8	59.5
Needs (Health Services including primary care)	39.9	55.6
Needs (Family)	17.2	35.7
Needs (Other)	6.9	19.0
Ethnicity (White)	34.0	25.4
Sex (Males)	65.6	61.9
Numeric Prognostics	Mean	Mean
Psychological Health (0–20)	13.5	11.8
Physical Health (0–20)	14.3	12.8
Accommodation Satisfaction (0–20)	14.6	10.8
Quality of Life (0–20)	12.7	10.4
Quality of Support Network (0–20)	14.2	12.0
Age (range 19–76)	37.8	35.5
RGPS (0–14)	7.1	9.5
REC-CAP Composite Index (0–50)	37.4	38.1

Table 3. Average treatment effect estimator of disengagement and retention length from recovery residences: HARP compared to SRR.

Outcome metric	Potential outcome means for SRR	Potential outcome means for HARP	Average Treatment Effect (ATE)	p-Value	95% CI Lower	95% CI Upper
Disengagement in percentage points	35.1	18.8	−0.163	0.004	−0.275	−0.051
Retention length in number of days	203	228	25	0.466	−42	92

Figure 1. Generalized Estimated Equation (GEE) model with stabilized weights for changes in REC-CAP from baseline: HARP compared to SRR.

Figure 2. Generalized Estimated Equation (GEE) model with stabilized weights for changes in REC-CAP from baseline within the HARP program, comparing the Therapeutic Community and Community components.

Figure 3. GEE model with stabilized weights for changes in RGPS from baseline: HARP compared to SRR.