Skip to main content
. Author manuscript; available in PMC: 2011 Oct 1.
Published in final edited form as: J Child Fam Stud. 2010 Oct 1;19(5):629–645. doi: 10.1007/s10826-009-9350-2

Table 1.

Randomized-controlled trials of interventions to improve engagement and retention: pre-treatment or early treatment strategies

Study Sample/setting Intervention Measures Impact of intervention on outcomes?
Engagement Retention
Watt et al. (2007) 262 families requesting treatment for conduct problems at two family outpatient therapy clinics in Australia
Families scheduling initial appointments were consecutively assigned to groups
Pre-treatment telephone reminders
I = call reminder prior to appt. (n not reported)
C = outpatient treatment as usual (no reminders) (n not reported)
Engagement
% of first 5 appts. attended
Retention
% of enrolled families who later refused treatment
No
I > C for families with higher conduct problems (subgroup analyses)
No
McKay et al. (1996a); McKay and Bannon (2004) 107 low-income, inner-city families requesting child mental health services in urban child outpatient clinic
Families scheduling initial appointments were consecutively assigned to providers with training (I) or without training (C)
Brief first session engagement interview
I = provider received training to conduct engagement interview to clarify roles and therapy process, identify practical barriers, develop collaborative plan (n = 33)
C = outpatient treatment as usual (providers received no training) (n = 74)
Engagement
% returning for 2nd session
# of completed sessions in 18-week period
Yes
I > C (97 vs. 83%) attended 2nd session
I > C (7.1 vs. 5.4 completed sessions)
N/A
McKay et al. (1996b) 109 low-income, inner-city families requesting child mental health services in urban child outpatient clinic
Randomization in two steps: (1) intake requests were evenly divided among three groups; (2) after intake appointment was scheduled, I & I-2 families were randomly assigned to providers with (I) or without (I-2) engagement interview training
Combined brief telephone and first session engagement interview
I = combined condition; intake worker delivered 30-min telephone engagement discussion ? provider addressed engagement in 1st session (n = 35)
I-2 = telephone engagement discussion only (n = 35)
C = outpatient therapy as usual (no engagement intervention) (n = 39)
Engagement
# of completed sessions in 18-week period<vr>% of scheduled sessions completed
Yes
I > I-2 & C (7.3 vs. 5.0 vs. 5.9 completed sessions)
I > I-2 & C(74 vs. 49 vs. 58% scheduled sessions completed)
N/A
Szapocznik et al. (1988) 108 Hispanic families with adolescents with substance use problems requesting family therapy at University community mental health center
No information on type of randomization
Brief telephone and early sessions Strategic Structural-Systems Engagement (SSSE) intervention
I = prior to therapy, clinician implemented levels of the SSSE model (n = 56):
  1. schedule family members for therapy

  2. minimal joining (encouragement)

  3. more joining (re: values, interests, issues)

  4. restructuring (advising on inviting family members; call to engage members)

  5. lower-level direct engagement (joining and gathering information from many family members)

  6. higher-level direct engagement (out-of-office visits to family members to engage personally)

    C = family therapy as usual (only level a listed above to engage families) (n = 52)

Retention
% oompleted treatment
N/A Yes
I > C (77 vs. 25%) (ITT)
I > C (83 vs. 59%) (of enrolled families)
Santisteban et al. (1996) 193 Hispanic families with adolescents with substance use problems requesting family therapy at University community mental health center
No information on type of randomization
SSSE (Szapocznik et al. 1988; see above)
I = providers delivered SSSE prior to family therapy (n = 52)
C = outpatient family therapy as usual (n = 67)
C-2 = outpatient group-based family therapy as usual (n = 74)
Engagement
% adolescent and 1 family member attended intake and 1 session
Retention
% completing at least 8 h of therapy and termination assessment
Yes
I > C & C-2 (81 vs. 57 vs. 63%) (ITT)
No
Coatsworth et al. (2001) 104 Hispanic and African-American families with adolescents with behavior problems identified by parents or teachers referred to a University child mental health services center and community mental health agency
Stratified randomization by ethnicity, after intake assessment
SSSE (Szapocznik et al. 1988, see above)
I = providers delivered SSSE (n = 53)
C = outpatient family therapy at community agency as usual (n = 51)
Engagement
For I families, % adolescent and one family member attended intake and 1 session; for C families, % at least one family member attended 1 session
Retention
% completing clinician’s recommended treatment
Yes
I > C (81 vs. 61%) (ITT)
Yes
I > C (58 vs. 25%) (ITT)
I > C (72 vs. 42%) (of enrolled families)
For conduct problem families, I > C (subgroup analyses)
Dakof et al. (2003) 103 low income, black mothers of infants exposed to drugs in utero referred to outpatient and inpatient drug rehabilitation clinics
Stratified randomization (type of drug treatment, age, HIV status, extent of child welfare involvement)
Brief provider support intervention
I = specialist met with mother to enhance therapeutic relationship, review life history, address barriers for 8 weeks; once enrolled, specialist met with mother for 1st 4 weeks of treatment (n = 51)
C = outpatient and inpatient treatment as usual (n = 52)
Engagement
% in treatment after 4 weeks
Retention
# of days in treatment over 90 day period
Yes
I > C (67 vs. 38%) (ITT)
No
Spoth and Redmond (1994) 387 families from 6 low-income rural school districts interested in a substance use prevention program for 6th–7th grade students at a University prevention research center
Multiple step randomization: (1) 425 families randomly selected from school district lists and assigned to I or C group; (2) 387 were eligible and agreed to participate in study
Offering partial then full vs. full prevention program
I = families invited to participate in brief assessment, then 5- session drug prevention program (n = 137)
C = traditional recruitment (recruited directly into 5 session prevention program) (n = 250)
Engagement
% attending each session
Retention
% completing program (5 sessions)
No No

Note: N/A not assessed/not applicable, I intervention condition, C control condition, ITT intent-to-treat analyses. We note when there is more than one intervention condition with I-2, I-3, or I-4