Long-Term Economic Benefit of Treatment Foster Care Oregon (TFCO) for Adolescent Females Referred to Congregate Care for Delinquency

Abstract

Treatment Foster Care Oregon (TFCO) is an alternative to congregate care, for youth involved in the juvenile justice and/or child welfare systems. Though demonstrated as clinically-and cost-effective across multiple rigorous trials, the long-term cost benefit of TFCO has not been considered. This study follows n = 166 females from adolescence to young adulthood, who were involved in both systems and referred for out-of-home-care. Records of arrest, court, incarceration (juvenile, jail, and prison), monitoring (parole and probation) and child-welfare services were included in a long-term cost-benefit analysis. Outcomes highlight ongoing benefit of the TFCO intervention, nearly 10 years post-intervention.

Female adolescents represent approximately 30% of the juvenile justice population (Ehrmann, Hyland, & Puzzanchera, 2019). For youth involved in the juvenile justice system (JJS) with repeated offenses and severe problem behaviors, out-of-home care, such as a group home or residential treatment, often is recommended. Although only 20% of young women who are involved in the JJS are referred for out of home placement, those who are, present with a range of high mental health and substance abuse treatment needs. Youth involved in both the JJS and child welfare system (CWS) are more likely to recidivate, and be placed out of the home than those involved in a single system; young women who are dually involved demonstrate similar patterns of delinquency as boys who are involved in the JJS alone (Lee & Villagrana, 2015). Of note, of youth involved in the JJS, young women are represented disproportionately as dually-involved (American Youth Policy Forum, 2019).

Youth who are referred to both the JJS and CWS are recognized as being at higher risk, with greater service needs, than youth involved in either system alone (Feyerherm & Johnson, 2012). Youth who are dually involved are 2–3 times more likely to use public services in adulthood that their peers (American Youth Policy Forum, 2019). Moreover, providers report spending more on females involved in the JJS than males (Office of Program Policy Analysis & Government Accountability, 2007). Thus, comprehensive, cost-effective treatments are essential to meet the needs of youth who are dually involved across systems, especially for young women who pose the greatest risk for ongoing system involvement and increasing costs to public systems.

One Solution: Treatment Foster Care Oregon (TFCO)

Treatment Foster Care Oregon (TFCO; Chamberlain, 1998), is an evidence-based practice (EBP) for the treatment of youth involved in the JJS and/or CWS who are referred for out-of-home or congregate care. Backed by multiple randomized controlled trial, TFCO has demonstrated effectiveness in reducing a range of symptoms including youth recidivism, delinquency, unplanned pregnancies, internalizing symptoms, deviant peer relations, and psychotic symptoms (Chamberlain, Leve, & DeGarmo, 2007; Leve & Chamberlain, 2005; Leve & Chamberlain, 2007; Kerr, Leve, & Chamberlain, 2009; Harold, Kerr, Leve et al., 2013). Consistent positive outcomes have led TFCO to be selected as a Blueprint Program for Healthy Development (https://www.blueprintsprograms.org/programs/treatment-foster-care-oregon/), and highlighted in two US Surgeon’s General reports (US Department of Health and Human Services [USDHHS], 2000a, b). TFCO was selected by the Coalition for Evidence-Based Policy as meeting “top tier” evidence of effectiveness (http://www.toptierevidence.org), and is rated as a “well-supported” program by the California Clearinghouse for Evidence-Based Practice (https://www.cebc4cw.org/program/treatment-foster-care-oregon-adolescents/).

Despite its demonstrated clinical effectiveness, TFCO is a relatively expensive model. TFCO intervention costs $10,766 (2016 dollars) more than treatment as usual (TAU) annually (Washington State Institute for Public Policy [WSIPP], 2017) per youth. However, youth receiving TFCO typically complete treatment in less than a year, compared to ongoing lengths of stay for those in TAU. Further, costs for TAU do not include therapeutic costs, but rather, the costs of congregate care alone. In contrast, costs reported for TFCO are inclusive of total program costs, including treatment. In TFCO, youth are placed individually in specialized foster homes, with foster parents who are well-trained and supported by a coordinated clinical team. Close consultation, training, and support of the foster parents form the cornerstone of the TFCO model. Programs are led by a supervisor with small caseloads (ten families). The supervisor leads a team of individual and family therapists, a foster parent recruiter and caller, and a skills coach. The program maintains daily contact with foster parents through the daily caller who collects data regarding youth functioning and in-home behavioral treatment plan delivery. The basic components of TFCO include (a) daily (M–F) telephone contact (PDR; Chamberlain & Reid, 1987); (b) weekly foster parent group meetings focused on supervision, training in parenting practices, and support; (c) an individualized behavior management program implemented daily in the home by the foster parent; (d) individual therapy for the youth; (e) individual skills training/coaching for the youth; (f) family therapy (for biological/adoptive/relative family of the youth) focused on parent management strategies; (g) close monitoring of school attendance, performance, and homework completion; (h) case management to coordinate the TFCO, family, peer, and school settings; (i) 24-h on-call staff availability to foster and biological caregivers; and (j) psychiatric consultation as needed. Thus, TFCO provides an intensive alternative to congregate or residential care, within the community setting.

Research has found a benefit to cost ratio of $1.85 for TFCO compared to usual care (WSIPP, 2017). Although positive, this largely underestimates the potential benefit of TFCO long-term. Previous outcomes have been limited to analysis of data collected in the months immediately following treatment completion. Leveraging a long-term follow-up of a sample of young women randomized to either TFCO or treatment as usual (TAU) during adolescence, the current analysis describes the cost-benefit of TFCO compared to TAU across two cohorts, up to ten years post- randomization. Now young adults, analyses consider outcomes in relation to participants’ own children and ongoing public service system involvement.

Method

A detailed description of the original randomized controlled trial study procedures can be found elsewhere (e.g., Leve, Kerr, & Harold, 2013; Rhoades et al., 2014). The current study is limited in scope to cost-benefit procedures.

Participants

Of the original 166 young women participating in the randomized controlled trial during adolescence, 152 young adult women were successfully recruited and assessed for long-term follow-up (91.6%). Two former participants were deceased at the time of the follow-up study, and the remaining were unable to be located. Originally, the young women were referred by juvenile court judges and had been mandated to community-based, out-of-home care because of problems with chronic delinquency and serious family adversity. The young women from the original trial were approximately 15 years old at baseline (M = 15.31, SD = 1.17) and 25 years old at the last young adult follow-up for the current study (M = 24.87, SD = 3.12). (Leve et. al., 2015).

The ethnicity of the sample was consistent with local demographics: Caucasian (68.1%), African-American (1.8%), Hispanic (11.4%), Native American (0.6%), Asian (0.6%), and mixed ethnic heritage (16.9%). Less than 1% reported other or unknown ethnicity. At baseline, 63% of the young women lived with single-parent families, and 54% of the young women lived in families earning less than $10,000. (Leve et. al., 2015). All participants were consented for each study following Institutional Review Board approved procedures.

Cohorts.

Young women were recruited across two cohorts. Cohort 1 (TFCO =37 and TAU=44) began recruitment in 1997 and Cohort 2 (TFCO=44 and TAU=41) completed in 2006. The follow-up study commenced approximately 10 years postbaseline for Cohort 1 (M = 9.81, SD = 1.73) and 5 years postbaseline for Cohort 2 (M = 4.69, SD = 1.16), and included an in-person assessment and a series of telephone interviews. Thus, Cohort 1 had a longer length of post treatment duration than those in Cohort 2. Every participant in Cohort 1 had at least 3,250 days (8.90 years) of follow-up data and those in Cohort 2 had at least 1,638 days (4.49 years). There were no group differences on the rates or types of pre-baseline offenses for either cohort (Leve, Chamberlain, & Reid, 2005).

Treatment Conditions

Treatment Foster Care Oregon (TFCO).

Young women randomized to TFCO were individually placed in highly trained and supervised homes with state-certified foster parents. Experienced program supervisors supervised all clinical staff, coordinated all aspects of each youth’s placement, and maintained daily contact with foster parents to provide ongoing consultation, support, and crisis intervention services. The intervention was individualized and included all basic TFCO components. Cohort 2 young women received additional intervention components targeting HIV-risk behaviors (i.e., information on dating, sexual behavior norms, and HIV-risk behaviors) and substance use behaviors (i.e., motivational interviewing, peer refusal skills, random urinalyses, and contingency management).

Group care (TAU).

As described elsewhere (Leve et al., 2013), young women randomized to TAU were placed in 1 of 35 community-based group care programs located throughout the state. Each site served 1–12 study participants (M = 2.18, SD = 2.95). The programs had 2–83 youth in residence (M = 13) and 1–85 staff members (Mdn = 9); facilities served young women only (68%) or both sexes (32%) but housed males and females in separate units. Sites required on-grounds schooling (41%), sent only some young women to off-grounds school (38%), or sent all young women to off-grounds school (21%). The program philosophies were primarily behavioral (67%) or multi-perspective (33%).

Measures

Official court records.

Official court records were obtained from every known area of residency for the participants (23 US States); the records listed the charges associated with an incident date, level of severity for each charge, whether there was arrest associated, and court proceedings of pre-hearing, hearing or trial.

Official incarceration records.

Official Incarceration lengths of stay for jail, prison, probation, and post-prison supervision were obtained from the Department of Corrections (DOC). Using the court records (described above), participants were identified who were convicted and sentenced to jail, prison, probation, or post-prison supervision. These records were compared to the records from the DOC. The DOC records adjusted the court sentenced length to the actual length of stay per outcome. DOC records contained all court sentenced prison records in addition to prison terms of stay from a probation or parole violation. Court sentence duration for jail, probation, and parole was very close to the DOC records that were present. Therefore, DOC records replaced court sentences when present and were included in the absence of any court sentencing record. Costs related to criminal charges and the associated proceedings were estimated based on public reports, and when needed, previous reports from the state (e.g., Carey, Waller, & Marchand, 2006; Carey and Waller, 2011).

Official CWS records.

Official records were obtained from the Department of Human Services (DHS), Children, Adults and Families Division at the conclusion of the final young adult assessment. These official CWS records included any participant with one or more substantiated maltreatment incidents against them for at least one of their children.

Self-report.

As part of a longer assessment battery described elsewhere, participants were interviewed at an average of 8.5-, 9-, 9.5, 10, 10.5, 11-years postbaseline. Contributing to the current study, participants reported on demographics, including number of children.

Results

Outcomes are presented for the combined cohorts across the longest period of time for which all participants have data (i.e., 1,638 days/4.49 years post-baseline), and separately for Cohort 1, whose members had the opportunity for data to be collected from a longer period (3,250 days/8.9 years post-baseline).

Court Involvement

Table 1 illustrates charges identified in convicted court case records. Across both cohorts, 60% of women in the TAU condition experienced a conviction compared to 42% of women in the TFCO condition. Of those, a similar number of women across conditions were convicted for theft (n = 28 and 29, respectively, for TAU and TFCO). Though similar in nature to theft, a significantly greater number of convictions were found for women in the TAU condition than the TFCO condition for assault (TAU n = 22; TFCO n = 7) and robbery (TAU n = 6; TFCO n = 0). Assault and robbery are both associated with high levels of victimization and incarceration costs. Of note, for the first cohort, 98% of participants in the TAU condition had been convicted in a court by the end of the nearly 9-year post-baseline data collection period, compared to 68% of women receiving TFCO.

Table 1.

Different Convicted Charges Present in Court Cases

Sample	N a	Burglary	Arson	Assault	Theft	Fraud	Forgery	Robbery	Total
			Combined Cohorts (Baseline to 1638 days)
TAUb	51	9	0	22	28	0	4	6	69
TFCOc	34	4	1	7	29	2	6	0	49
			Cohort 1 (Baseline to 3250 days)
TAUd	43	6	0	20	19	0	7	5	57
TFCOe	28	4	0	9	21	3	9	0	46

Note. TAU = Treatment as usual (Group Care); TFCO = Treatment Foster Care Oregon.

^aNumber of participants represented as having at least one of the charges in the subsequent columns.

^bn = 85.

^cn = 81.

^dn = 44.

^en = 41.

As shown in Table 2, a greater number of women randomized to TAU (39%) were arrested than those randomized to TFCO (26%), however, of those arrested, the average number of arrests were similar. Of women sentenced to jail, those in the TAU condition had over 3 times the length of stay than those in the TFCO condition. Note that within the durations, multiple stays were aggregated, with a maximum of 3 stays per individual. Of women who were sent to jail, the average number of stays for those in TAU were higher than those in TFCO (p = .04), suggesting that the difference in “Jail Duration” was due to women in TAU being sentenced to jail 1.5 times more than women in TFCO. Although slightly fewer women in TFCO (6%) were sentenced to prison than women in TAU (8%), the number of days that women in TFCO were sentenced was greater than women in TAU across both the combined cohorts and Cohort 1. On average, participants served approximately 70% of their sentenced prison term.

Table 2.

Description of Arrests, Supervision, and Incarceration within the Sample

Sample	Mean Arrests	Mean Arrests for those Arrested (n)	Jail Durationa (n)	Prison Durationa (n)	Probation Durationa (n)	Parole Durationa (n)
		Combined Cohorts (Baseline to 1638 days)
TAUb	0.9	2.4 (33)	105* (15)	668 (7)	1,013 (22)	584* (5)
TFCOc	0.5	2.1 (21)	30* (14)	987 (5)	1,233 (14)	287* (4)
		Cohort 1 (Baseline to 3250 days)
TAUd	2.4	3.9 (27)	172 (19)	655 (8)	2,214 (19)	871 (8)
TFCOe	2.2	4 (20)	64 (16)	1,036 (5)	1,785 (17)	1,041 (6)

Note. TAU = Treatment as usual (Group Care); TFCO = Treatment Foster Care Oregon.

^aMean excludes those participants with zero corresponding days.

^bn = 85.

^cn =81.

^dn = 44.

^en = 41.

^*p < .10.

CWS Involvement

Child welfare involvement was assessed for women who had their own child(ren) since baseline. Within the combined cohort, nearly half (49%) of all women had a child during the study period. Of those, slightly more were in the TAU condition (n = 43; M = 1.37, SD = 0.58) than the TFCO condition (n = 38; M = 1.26, SD = 0.55) with both conditions averaging between one and two children. Within Cohort 1, with more time passing since baseline, the average number of children increased within both groups (TAU M = 2.19, SD = 0.78; TFCO M = 2.26, SD = 0.94). As shown in Table 3, a striking difference was found between women in the TAU and TFCO conditions regarding child welfare involvement, with women in the TAU condition experiencing significantly more child welfare involvement regardless of cohort. Within the combined cohort, women randomized to the TAU condition had a 5 times greater likelihood of having child welfare involvement than those in the TFCO condition.

Table 3.

Description of the Child Welfare Services Present within the Sample

	Combined Cohorts (Baseline to 1638 days)		Cohort 1 (Baseline to 3250 days)
	Percentage with Children	Percentage with DHS Involvementa	Percentage with Children	Percentage with DHS Involvementb
TAU	43 of 85 (51%)	11 of 43 (26%)	32 of 43 (74%)	11 of 32 (34%)
TFCO	38 of 81 (47%)	2 of 38 (5%)	27 of 37 (73%)	6 of 27 (22%)

Note. TAU = Treatment as usual (Group Care); TFCO = Treatment Foster Care Oregon. One TAU CWS case was dropped as involvement was due to the death of the participant.

^ax²(1)=6.181.

^bx²(1)=1.054.

Cost Breakdown and Benefit-Cost Ratio

As shown in Table 4, statistically significant cost differences were found between young women randomized to TAU versus those randomized to TFCO during adolescence. In Cohort 1, 15% (n = 8 TAU; n = 5 TFCO) of the 85 participants had either no cost incurred or a minor court pre-hearing (< $300) during the timeframe of baseline through Cohort 1. In the larger combined cohort, with a shorter period of time, 26% (n = 17 TAU; n = 26 TFCO) of the 166 young women had either no cost incurred or a minor court pre-hearing (< $300) during the study period. Combined, the cohorts’ cost difference was $33,875 -- a statistically significant difference; Cohort 1, with a reduced sample size, showed a cost difference of $48,965 -- trending toward significance.

Table 4.

Total Costs for Two Different Timeframes

Sample	CWS	Court & Lawyer	Victimization	Incarceration	Arrest	Total Condition Costs
		Combined Cohorts (Baseline to 1638 days)
TAUa	$9,548	$1,025	$29,180	$24,321	$644	$64,718
TFCOb	$1,472	$706	$9,413	$18,876	$376	$30,843
Difference	$8,076**	$319	$19,767**	$5,445	$268	$33,875**
		Cohort 1 (Baseline to 3250 days)
TAUc	$25,642	$2,232	$50,104	$43,909	$1,613	$123,499
TFCOd	$15,577	$1,895	$25,094	$30,515	$1,454	$74,534
Difference	$10,065	$337	$25,010*	$13,394	$159	$48,965*

Note. All dollars are expressed in the base year chosen for this analysis (2016). CWS = Child Welfare Services; TAU = Treatment as usual (Group Care); TFCO = Treatment Foster Care Oregon.

^an = 85.

^bn = 81.

^cn = 44.

^dn = 41.

^*p < .10.

^**p < .01.

Benefit-cost ratio.

Taking these factors into consideration, and examining dollars spent on women referred to out-of-home care in adolescence, the long-term benefit to cost ratio for the combined cohort was $3.15, indicating that for every dollar spent in excess on TFCO, the return was $3.15. These benefits show increase over time -- the benefit to cost ratio for Cohort 1 was $4.55. This is an advance above the 2017 WSIPP report indicating a TFCO benefit to cost ratio of $1.85, based on outcomes from the rigorously evaluated trials with relatively small periods of post-treatment timeframes (WSIPP, 2017). Thus, though TFCO is an expensive model, findings suggest that these expenses – again that are inclusive not just of placement but of treatment – continue to show a return on investment nearly 10 years later.

Discussion

This paper examined the benefit to cost ratio of TFCO, an evidence-based practice for the treatment of youth referred to out-of-home care referred by the juvenile justice and/or child welfare systems. Billions of dollars are allocated annually for the sole purpose to protect and care for youth with costs increasing over recent years (Child Welfare League of America, n.d.), some of whom are on a trajectory to be highly costly to the adult social service systems. Spending extra dollars during adolescence on services for youth who are system-involved can yield vastly reduced future cost to society, benefiting not only society, but also the individualized needs of youth and their future families. Youth, and particularly young women, who are dually involved in the JJ and CW systems are at extremely high-risk for ongoing problems into adulthood, and as seen from this study, their children might face similar challenges.

TFCO is an intervention targeted at the discrete moment that youth are identified as being appropriate for a group home or residential setting, with a high potential of being released into adulthood having limited positive life skills, prosocial settings, or goals. Alternatively, TFCO aims to increase structure, healthy functioning, and the ability to abide by social norms within the regular school and community setting. Although the TFCO program costs $10,766 (2016 dollars) more per youth as compared to TAU, annually, that expense has a return of nearly 400% when long-term criminal and CWS costs are considered. Thus, using the conservative estimate of 4 years post-baseline for the combined cohorts, for every dollar spent in excess on the TFCO intervention, $3.15 is returned in future adverted criminal and child welfare costs.

Of considerable importance is the finding that women who were randomized to TFCO as adolescents were less likely to experience ongoing child welfare involvement in adulthood. The notion of transgenerational child welfare involvement has long been considered in the literature. For youth with severe presentations such as those in the current study, this risk is even greater. By averting future child welfare involvement as parents themselves, TFCO has the potential to significantly impact the mental and behavioral health of multiple generations.

Contextual Considerations

It is important to consider the context of this evaluation. Analyses are specific to Oregon and Washington in determining the cost difference between TAU and TFCO. In other states, adult system costs might vary and thus, the costs reported here that are being averted might differ. The costs of TFCO and TAU vary relative to state legislature/government allocations of reimbursement rates. For example, in California, TAU costs more than TFCO. In the California context, TFCO is reimbursed at the highest level of intensive treatment foster care, at $5,741 a month per qualifying youth (effective 7.1.2015), whereas youth assessed at similar risk levels who require groups homes that are able to accommodate these high needs, are in homes with a reimbursement rate between $8,935 – $10,130 a month per qualifying youth (effective 6.2.2016). Using this example further, in the California system, there is no reason not to expect the same level of adverted future adult system costs for youth who receive TFCO. Thus, under the California system it is advantageous to send qualifying youth through the TFCO foster care approach prior to utilizing a group home setting, for both benefit of the youth and society.

Before attributing the benefit to cost results of $3.15 established from these findings, it is important to define the context in which the program will exist. Every program and setting continually negotiates reimbursement rate structures on an annual fee schedule. TFCO is best implemented under an intensive foster care rate such as in California; however, many states choose a less labor intensive approach of moving youth into a group home setting. Programs such as TFCO come with additional implementation costs, such as the need to recruit strong qualified foster parents, which are the heart of the program.

Additional Benefits of TFCO

Public policy discussions might move beyond thinking of aggregate solutions to ongoing rates of youth out-of-home experiences. Some youth do not adapt well to traditional foster care and might experience more likelihood of escalating to further public system involvement, if individualized models such as TFCO were not provided. Models like TFCO, that provide comprehensive structure, skills training, regular education, individual and family therapy, within the context of a well-supported foster home, might better meet the needs of this specialized population of youth that present with myriad needs. And, as seen in the current project, the long-term benefits of such treatment help both the youth in need, as well as society as a whole. The challenge is the ongoing administrative burden of managing the varying levels of needs on a case to case basis of every youth involved in the system.

Limitations

Several areas of cost differences that were not included in this analysis were health care and high school graduation. Across the young adult follow-up period, 48% of participants had Medicaid and 10% had private insurance. Other participants reported no insurance coverage, despite Medicaid eligibility, thereby exacerbating the potential cost of medical services. Indeed, Evans and colleagues (2004) identified a problem of those involved in the JJS not receiving Medicaid services despite being eligible. Importantly, although women in both treatment conditions showed an overall reduction in substance use since adolescence, with women who were in TFCO showing greater reductions than those in TAU (Rhoades et al., 2014), a high proportion across both conditions of women, continued to endorse substance dependency within the last 12 months (Tobacco 46%; other drugs 21%). Health Care outcomes for the follow-up were self-reported and it was challenging to disentangle preventative health care costs from avoidable health care costs, thus they were omitted in this study. Future research should consider this important topic. Because all healthcare costs are not created equal, prior to introducing them into a cost benefit model, it is imperative to understand if increased medical care is positive (e.g., receiving care for diabetes that previously was being ignored or mismanaged). Similarly, although high school graduation rates have been included in other analyses of costs (e.g., Levin, Belfield, Muenning, & Rouse, 2007), associated future labor market productivity was beyond the scope of this direct observation analysis.

Conclusions

TFCO is an evidence-based practice for youth referred by the juvenile justice or child welfare systems for problematic, high-risk behavior problems. Though TFCO is a relatively expensive model, outcomes suggest that TFCO is successful in significantly averting costs in other public serving systems, such as criminal justice and transgenerational child welfare involvement. Future research should consider long-term benefits of TFCO as related to the next generation of children.