FFT Research

Summary of research outcomes by study type: Efficacy, Effectiveness, Dissemination / Implementation, Basic, and Process

Citation

Bias*

Population / Referral

Demographics

Design

Outcomes

Basic Research

Alexander (1973)

Dev

Delinquent

n = 20;

55% (11) male; Mostly White

Non-clinical ; Observational

Change Mechanisms

-Delinquent families engaged in more defensive communications

-non-Delinquent families engaged in more supportive communications

-Delinquent families reciprocated defensive behaviors and did not reciprocate supportive behaviors

-non-Delinquent families reciprocated supportive behaviors and did not reciprocate defensive behaviors

Alexander et al. (1989); Study 1

Dev

Delinquent

n = 32;

50% (16) male; Primarily White

Non-clinical ; Observational; Cooperative or Competitive set

Change Mechanisms

-Parents of delinquent youth respond with more defensive communications than families of non-delinquent youth in competitive set

-Parents of delinquent youth engage in significantly less defensive communications in cooperative vs. competitive set

Alexander et al. (1989); Study 2

Dev

Delinquent

n = 49;

Primarily White

Non-clinical ; Manipulation of attribution set

Change Mechanisms

-Parents made as many positive dispositional attributions in the satisfied set as negative attributions in the dissatisfied set

-Parents used more internal than external attributions for both the child's successful behaviors and problem behaviors

Alexander et al. (1989); Study 3

Dev

Delinquent

n = 61; Primarily White

Observational; Compared impact of relabeling vs. 3 interventions

Change Mechanisms

-Neither positive information or relabeling were able to reduce blaming attributions

Barton, Alexander, & Turner (1988)

Dev

Delinquent

n = 32;

56% (18) males Primarily White

Age: 14-17

Observational; Non-clinical

Change Mechanisms

-Delinquent families expressed significantly lower rates of negative communication in a cooperative set than in a competitive set

-Delinquent families expressed significantly lower rates of adaptive communication than non-delinquent families in the cooperative set

Mas, Alexander, & Turner (1991)

Dev

Substance use or delinquent; Mental Health

n = 49;

61% (30) male; Primarily White

Age: 13-18

Observational ;

Interactions in two priming conditions

Change Mechanisms

-Low conflict family members made fewer blaming attributions about other family members dissatisfying vs. satisfying behaviors

-High conflict family members made equivalent amounts of blaming attributions for both satisfying and unsatisfying behaviors

-High conflict family members behaved more defensively than low conflict family members

Morris, Alexander & Turner. (1991)

Dev

Non-clinical; Undergraduate students

n = 120;

Primarily White

Anecdotal;Observational ;

Randomized

Change Mechanisms

-Experimental reattributions (similar to relabel/reframe) significantly reduced the intensity of blaming attributions

Process Research

Alexander, Barton, Schiavo, & Parsons (1976)

Dev

Delinquent

Youth/family

n = 21;

48% (10) male;

Majority White

Therapists :

n = 21;

67% (14) male

Clinical ; Experimental; Observational

Change Mechanisms

-Therapist relational skills as opposed to structuring skills were associated with good clinical outcomes

-Family member supportive to defensive behaviorsat the end of treatment (not earlier in treatment) were associated with clinical outcomes

Flicker, Waldron, & Turner, Brody, & Hops (2008a)

Dev

Substance Use

n = 86;

64% (72) male;

50% Hispanic;

50% White;

Age: M = 15.7 (13-19)

Random assignment :

a. FFT

b. integrated FFT+CBT

Assessment Period

-4 months post-randomization

Clinical Outcomes

-Significant pre-post reductions in substance use for all youth in FFT and FFT+CBT.

Change Mechanisms

-Hispanic youth with Hispanic therapists showed greater decreases in substance use compared to Hispanic youth with Anglo therapists. Ethnic match was not related to treatment outcome for Anglo youth

Flicker, Turner, Waldron, Brody, & Ozechowski (2008b)

Dev

Substance Use

see Flicker 2008a

Clinical interaction ; Observational

Change Mechanisms

-Hispanic families who dropped out of treatment had greater unbalanced alliances than Hispanic families that completed treatment

-No differences were observed between dropouts and completers among White families

Mas, Alexander, & Barton (1985)

Dev

Delinquent

n = 49;

61% (30) male;

100% White;

Age: 13-18

Clinical interaction ; Observational

Change Mechanisms

-Adolescents spoke less with female therapists than male therapists

Newberry, Alexander, & Turner (1991)

Dev

Delinquent

N = 34; Majority White

Clinical interaction ; Observational

Change Mechanisms

-Mothers and fathers responded more supportively to female therapists' supportive statements than to male therapists' supportive statements

-Fathers responded more supportively to structuring statements than mothers, irrespective of therapist gender

-Female therapists were more likely than male therapists to respond to family members' supportive statements with structuring statements

Robbins, Alexander, Newell, & Turner (1996)

Dev

Delinquent

n = 35;

57% (20) male; Primarily White

Clinical interaction ; Observational

Change Mechanisms

-Therapist reframing more likely than alternative intervention strategies to increase family member positive statements

Robbins, Alexander, & Turner (2000)

Dev

Delinquent

n = 37;

70% (26) male; Primarily White

Age: M=15 (12-17)

Clinical interaction ; Observational

Change Mechanisms

-Following defensive family member communications, therapist reframing interventions were more effective than alternative interventions in reducing subsequent defensive behaviors

-Family member non-defensive reactions to a defensive communication were associated with lower levels of subsequent defensive behaviors than therapist reflection and elicit/structure interventions (but not reframing)

Robbins, Turner, Alexander, & Perez (2003)

Dev

Delinquent

n = 34;

59% (20) male; Primarily White

Age = 12-18

Clinical interaction ; Observational

Change Mechanisms

-Dropouts had significantly greater unbalanced alliances (parent minus adolescent) with therapists than completers

-Overall level of alliance did not predict outcome

-Analysis of role showed significantly higher unbalanced alliances in father-adolescent dyads in dropout vs. completer cases. No differences were observed for mother-adolescent dyads.

Sholevar, Baron, Aussetts, & Spiga (2010)

Ind

Delinquent

n = 187;

66% (125) male;

76% African Am;

Age M=14.3 (11-17)

Quasi-experimental ;

-Within group analyses of youth who were re-arrested following FFT

Change Mechanisms

-Youth who completed 6 or fewer sessions had a shorter time to arrest than youth who complete 7 or more sessions (406 vs. 510 days, respectively)

-Substance use, association with deviant peers, and poor session attendance (high numbers of cancellations/no shows) were associated with a lower number of days to first arrest

Efficacy Research

Alexander, 1971

Dev

Delinquent

n = 40;

Primarily White

Random assignment

a. FFT only,

b. individual therapy only (IT),

c. FFT+IT

d. minimal probation supervision

Assessment Period

-Post-treatment

Change Mechanisms

-Family therapy plus individual therapy produced significantly greater improvements in communication style than other conditions

Alexander & Parsons (1973)

Dev

Delinquent; Juvenile Courts

n = 86

44% (38) male; Primarily White

Age: 13-16

Random assignment (a-d):

a. FFT

b. client-centered family groups,

c. psychodynamic family therapy,

d. no treatment control,
e. post hoc selected controls, n=46

f. county wide(n =2800) recidivism rates 1971 = 51%

Assessment Period

-18 month post-randomization

Clinical Outcomes

-FFT recidivism was 26%, compared to 50% for no treatment control, 47% for client-centered family groups, and 73% for psychodynamic family therapy

Change Mechanisms

-FFT produced significant improvements in family interactions compared to all other comparison conditions

Friedman (1989); Stanton & Shadish (1997)

Ind

Substance Use

n = 135;

90 % (121) male;

89% White;

Age M=17.2 (14-21);

Random assignment

a. FFT

b. parenting group intervention

Assessment Period

-15+ months post-randomization

Clinical Outcomes

-Significant pre-post reductions in substance use at all follow-up points, with greater reductions in FFT, compared to parenting intervention

Change Mechanisms

-FFT produced greater involvement of parents, lower family dropout rate, improved psychiatric and family functioning in both conditions

Hansson, Cederblad, & Hook (2000)

Ind

Delinquent(Lund, Sweden)

n = 89;

87% (77) male; Primarily White;

Age M=15 (10-18)

-Random assignment

a. FFT

b. treatment as usual (TAU)

Assessment Period

-24 month post-randomization

Clinical Outcomes

-FFT more effective and TAU in reducing recidivism

-FFT associated with greater reductions in youth and parent reports of youth externalizing and internalizing symptoms

Hops, Ozechowski, Waldron, Davis, Turner, Brody, & Barrera (2011)

Dev

Substance Use; HIV Risk

n = 225;

83% 187 male; 51% Hispanic;

49 % White;

Age: 13-19

Random assignment :

a. individual CBT (IT)

b. integrated FFT+CBT

Assessment Period

-19 months post-randomization

Clinical Outcomes

-Significant pre- to post-treatment reductions in HIV-risk behaviors for high-risk youth in both treatment conditions, with greater reductions in IT than FFT+CBT and greater reductions for high risk Whites, compared to Hispanics

Klein, Alexander, & Parsons (1977)

Dev

Younger siblings of delinquent youth (see Alexander & Parsons, 1973)

n = 99 referred 86 families followed; 44% (38) male; Primarily White

Age: 13-16

Random assignment :

a. FFT

b. client-centered family groups,

c. psychodynamic family therapy,

d. no treatment

Assessment Period

-30-40 months post-treatment

Clinical Outcomes

-Siblings of youth receiving FFT showed lower arrest rates than siblings from alternative treatment conditions 2 ½ to 3 ½ years post-treatment

Parsons & Alexander (1973); Alexander & Barton, (1976, 1980)

Dev

Delinquent

n = 40;

45% (18) male; Primarily White;

Age M=15.1

Random assignment:

a. FFT

b. client centered family therapy,

c. no treatment

Assessment Period

-Post-treatment

Change Mechanisms

-FFT families displayed significant improvements in family interactions

-No improvements in controls

Rohde, Waldron, Turner, Brody, & Jorgensen (2014)

Dev

Substance use; Depression

n = 170;

78% (133) male;

22% Hispanic; 61% White;

17% Other

Age M=16.4 (13-18)

Random Assignment; Sequenced Interventions :

a. FFT followed by Coping with Depression (FFT/CWD)

b.CWD followed by FFT (CWD/FFT)

c. coordinated FFT and CWD (CT)

Assessment Period

-18 months post-randomization

Clinical Outcomes :

-FFT/CWD yielded better substance use outcomes than CT

-For participants with baseline Major Depression, CWD/FFT had lower substance use outcomes than FFT/CWD and CT

-Depressive symptoms decreased significantly for youth in all three treatment conditions, with no differences between treatments

Slesnick & Prestopnik (2009)

Ind

Alcohol abusing, runaway youth

n = 119;

45% (54) male; 29% White;

44% Hispanic;

11% Native Am;

5% African Am;

11% Other;

Age: M=15.1 (12-17)

Random assignment :

a. home-based ecological family therapy

b. office-based FFT

c. services as usual

Assessment Period

-15 months post-randomization

Clinical Outcomes

-Significant pre- to post-treatment reductions inalcohol and drug use for all three conditions

Waldron, Slesnick, Brody, Turner, & Peterson (2001); Waldron, Ozechowski, Turner, & Brody, 2011; French, Zavala, McCollister, Waldron, Turner, & Ozechowski, (2008)

Dev

Substance Use

n = 120;

80% (96) male;

38% White;

47% Hispanic;

8% Native Am;

7% other;

Age M=15.6 (13-17)

Random assignment :

a. FFT

b. individual CBT (ICBT)

c. group therapy (GT)

d. integrated FFT+CBT (IBFT)

Assessment Period

-19 months post-randomization

Clinical Outcomes

-FFT, GT, and IBFT all showed significant reductions in substance use

-FFT and IBFT superior to ICBT

Change Mechanisms

-Improvements in family functioning associated with substance use reductions in the FFT conditions, but not GT, supporting family improvement as a mechanism of change in FFT

Cost analyses

-FFT and IBFT were more cost effective than IT or GT at post-treatment -GT was more cost effective than the other treatment conditions at follow-up

Effectiveness Research

Barnoski (2004)

Outcome Evaluation WSIPP (Follow-up to Barnoski, 2002; Sexton & Turner, 2010)

Ind

Delinquent; Community-based sample

n =700

Age M = 15.35 (13-17)

Random assignment

a. FFT

b. probation services as usual

Assessment Period

-18 months post-randomization

Clinical Outcomes

-No overall differences between conditions in adjudicated felony recidivismrates

Change Mechanisms

-Adjudicated recidivism felony recidivism rates were lower for competent FFT therapists than usual probation services, and non-competent FFT therapists

Lantz (1982)

Ind

Delinquent

n = 46

Random assignment :

a. FFT

b. alternative treatment

Assessment Points

-Post-treatment

Clinical Outcomes

- FFT had lower rates of recidivism than alternative treatment

-Lower rates of outplacements were observed in FFT than alternative tx

Lewis, Piercy, Sprenkle, & Trepper (1990)

Ind

Substance Use

n = 84;

81% (68) male;

Age M=16 (12-22)

Random assignment

a. Purdue Brief Family Therapy (based on FFT)

b. Family Drug Education

Assessment Period

-Post-treatment

Clinical Outcomes

-Reductions in substance use only for family therapy condition involving an adaptation of FFT but not the Family Drug Education condition

Regas & Sprenkle (1982)

Ind

ADHD (referrals to Child Welfare)

n = 55

-Random assignment

a. FFT

b. group therapy

c. no tx control

Assessment Period

-Post-treatment

Clinical Outcomes

-FFT and group therapy produced significant improvements in ADHD behaviors at home and at school

Change Mechanisms

-Only FFT also led to significantly more positive perceptions of the family

Sexton & Turner (2010)

Ind

Delinquent; Community-based sample

n =917;

79% (724) male;

78% White;

10% African Am;

5% Asian Am;

3% Native Am;

Age M= 15.35 (13-17)

Random assignment

a. FFT

b. probation services as usual

Assessment Period

-12 months post-randomization

Clinical Outcomes

-Overall, no differences were found between FFT and services as usual in adjudicated recidivism

Change Mechanisms

-When therapists were adherent to the model, FFT showed significantly greater reductions in felonies, and violent crimes, with a marginally significant reduction in misdemeanors , compared to services as usual

Dissemination/Implementation Research

Baglivio, Jackowski, Greenwald, & Wolff (2014)

Ind

Delinquent

n = 2203;

72 % (1599) male;

53% White;

47% non-White;

Age: M=15.57 (10 to 19)

-Matched (Propensity) Assignment

a. FFT

b. MST

Assessment Period

-12 months post-discharge

Clinical Outcomes

-Female youth referred to MST had higher offense rates during service than females referred to FFT

-Low risk youth referred to MST having a higher rate of new arrests and/or violations of probation while receiving the therapy than FFT-referred low risk youth

-The recidivism rate differences pre-matching for moderate-high to high risk to reoffend youth remained significant (p < .05), but not at the Bonferroni-corrected level (p < .025), with those who received MST having a higher recidivism rate than those receiving FFT

-Post matching, a new significant difference emerged with the "all youth referred" sample (the full sample) FFT youth having significantly lower number of offenses during service (at the non-corrected p < .05 level).

Change Mechanisms

-FFT has a significantly higher completion rate than MST

-FFT had a significantly lower length of service than MST

Barton, Alexander, Waldron, Turner, & Warburton (1985); Study 1

Dev

Delinquent

n = 27;

Primarily White

Non-random assignment

a. FFT

b. district juvenile justice base rates

Assessment Period

-13 months post-treatment

Clinical Outcomes

-FFT had lower recidivism rates than thepopulation base rate

Change Mechanisms

-Significant reductions in family defensiveness in FFT

Barton et al. (1985) Study 2

Dev

Child Welfare

n = 325;

Primarily White

Non-random assignment :

a. FFT

b. community-based social workers

Assessment Period

-Post-treatment

Clinical Outcomes

-Reduction in foster care placement referrals FFT (11%) versus non-FFT (49%)

Change Mechanisms

-Reduction in units of service per family to less than half (14.7-6.2)

Barton et al. (1985) Study 3

Dev

Delinquent

n = 74;

Primarily White

Non-random assignment :

a.FFT

b. Alternative treatment

Assessment Period

-15 months post-treatment

Clinical Outcomes

-Lower recidivism rate observed in FFT compared to alternative tx

-Those in the FFT group who did reoffend did so at a lower rate/frequency than those in the regular services group

Celinska, Furrer, & Cheng (2013)

Ind

Delinquent / Child Welfare

n = 72;

69% (50) males;

36% African Am;

36% Hispanic;

19% White;

9% Other;

Age: M=15.3 (11-17)

Matched Assignment

a. FFT

b. Matched control

Assessment Period

-Post-treatment assessment

Clinical Outcomes

-Only FFT youth showed improved functioning in life domains, such as living arrangements, school behavior/achievement/attendance, legal concerns and vocational concerns

-Only FFT youth showed significant reduction in emotional and behaviors needs as well as risk behavior

Change Mechanisms

-Older youth responded better than younger youth

-Hispanic youth responded better on life domains and child risk behaviors

-White youth responded better on child strengths

-African American youth responded better on child emotional and behavioral needs

Darnell & Schuler (2015)

Ind

Delinquent

N = 8713;

78% (6796) male;

29% AA;

61% Hispanic;

8% White;

Age M= 17 (11-18)

Quasi-experimental

Propensity Matched ;

a. Standard Probation (SP)

b. FFT plus SP

c. Functional Family Probation Services (FFP®)

d. FFT plus FFP®

Assessment Period

-36 months post discharge

Clinical Outcomes

-Youth receiving FFT (both FFT+SP and FFT+FFP®) compared to SP alone had significantly lower likelihood for outplacements in first two months following treatment…but this advantage disappeared in later months.

-Youth receiving FFP® alone (as compared to SP alone) had lower likelihood for outplacements in first two months (but not significant)

-Ultimately, at the end of the 36-month outcome observation period, there were no significant differences in outplacements between any of the three intervention groups

-12 month survival analysis illustrated that youth in the FFT group remained less likely to have an outplacement than comparison youth

Gordon, Arbuthnot, Gustofson, & McGreen (1988); Gustofson, Gordon, & Arbuthnot (1985); Gordon, Graves, & Arbuthnot, 1995; Gordon (1995) -Study 1

Ind

Delinquent Rural; Low SES

n = 54;

70% (38) male;

100% White;

Age: M15.4

Matched assignment

a. FFT

b. probation services as usual

Assessment Period

-30-60 month post-treatment assessment of adult convictions

Clinical Outcomes

-FFT group had lower recidivism rates compared to regular services group at 30- and 60-month follow-up

Cost analyses

-Cost-benefit analysis on these groups indicated that FFT had significantly lower direct costs than treatment as usual

Gordon & Arbuthnot (1990); Gordon, (1995) -Study 2

Ind

Delinquent

n = 49

Age 17-18

Non-random assignment

a. FFT

b. statistical control (empirically derived risk of recidivating)

Assessment Period

-18 months post-baseline treatment

Clinical Outcomes

-FFT had lower new convictions after treatment and institutional commitments than statistical control group, and

Gordon (1995) -Study 3

Ind

Delinquent; Re-entry

n = 52;

Age 16-17

Matched assignment :

a. FFT, n=27

b. probation services as usual

Assessment Period

-16 months post-baseline

Clinical Outcomes

-FFT showed a significantly lower recidivism rate, compared to the services as usual group.

Hansson, Johansson, Drott-Englén, & Benderix (2004)

Ind

Community-based sample; Mental Health; Child Welfare; Lund, Sweden

n = 62;

90% (54) male; Primarily White;

Age: M=15 (13-18)

Matched Assignment

a. FFT

b. social work services as usual

Assessment Period

-18 months post-baseline

Clinical Outcomes

-FFT had lower recidivism rates than services as usual

-FFT group associated with greater reductions in youth and parent reports of externalizing and internalizing symptoms

Change Mechanisms

-Improved family functioning, and reduced maternal depression, somatization, and anxiety in FFT group

Stout & Holleran (2013)

Ind

Child Welfare

Sample includes all outplacements in New Jersey from 2005 to 2011

Time Series Analysis

-Comparison of FFT, MST, and other services

-Dependent variable includes any outplacement for youth between 2005 and 2011

Clinical Outcomes

-MST and FFT were estimated to yield an approximate reduction of 31 outplacements a month or an annual reduction of 372 outplacements

Cost Savings Estimates

-Projected annual savings of $1.33 million for FFT and $2.16 million for MST

-Since 2005, estimated total savings of $17.33 for FFT and $18.16 for MST

White, Frick, Lawing, & Bauer (2013)

Ind

Delinquent; Callous-Un-emotional (CU); Community Mental Health

n = 134;

71.6% (96) male;

59% African Am;

35.1% White; 4.5% Hispanic;

Age: M =15.34 (11-17)

Non-randomized study ; All youth received FFT

Assessment Period -12 month post-treatment

Change Mechanisms

-CU traits associated with poorer adjustment (behavioral, emotional, social) prior to treatment

-CU traits correlated with poorer levels of adjustment post-treatment, less perceived change over treatment by youth and parents and increased likelihood of violent offending during treatment

-CU traits NOT associated with significantly lower rates of participation

-CU traits NOT associated with higher rates of treatment dropout

--CU traits associated with greater improvements in adjustment over course of FFT

-Association between CU traits and risk for violent charges decreased over time



*Refers to primary authors on the study. Dev (developer) denotes that the study was conducted by an investigator(s) that included the Developer of the model (Alexander) or by an investigator(s) that were trained by the Developer at the University of Utah. Ind (independent) denotes that the study was conducted by an investigator(s) that did not include the Developer (Alexander) or other investigators that were trained by the Developer at the University of Utah. Most of the studies in the independent category can be considered replications of prior research or an extension of FFT into new clinical populations or settings.

**Demographic data on total sample, gender, race/ethnicity, and age included when reported. White is used to denote non-Hispanic White; Hispanic is used to denote Hispanic/Latino (very few youth/family members self-identified as Black Hispanic); African Am is used for African American; Native Am is used for Native American, American Indian, and First Nations.