Comparative Effectiveness Trail to Reduce Child Maltreatment, Improve Client Outcomes and Examine Client Burden

In child welfare services, structured behavioral parenting programs have been documented to reduce important child-welfare outcomes, including child maltreatment recidivism.1-3 In this study, we attempt to learn which factors impacted implementation of an evidence-based practice (EBP) in a diversity of child-welfare serving sites and systems. The primary aim of this study was to identify and assess barriers and facilitators of implementation of a structured behavioral parenting program (SC). We utilized a qualitative research strategy that included semi-structured interviews and focus groups with several levels of staff responsible for implementing the model: program administrators, supervisors, and frontline staff (providers). Our second aim was to understand parent and provider reactions to SafeCare (SC) services and Supportive Case Management (SCM), especially parents' perceptions related to trajectory of burden, engagement, satisfaction, and perceived impact across intervention receipt. We employed mixed methods (both quantitative and qualitative data collection) to inform this question. Specifically, we (1) conducted qualitative interviews with families at two time points during the course of service, (2) collected session-by-session ratings from families on service reaction (perceived burden satisfaction, perceived effectiveness) and providers on family engagement, and (3) collected organizational environment surveys from providers at two time points. The final aim of this study wass to examine the short-term impact of SC versus SCM on client-centered outcomes. Quantitative surveys collected in the family's home at the beginning and end of services measured parenting variables, parent mental health and well-being, and child behavioral, social, and emotional well-being.

Study Overview

• Status: Completed
• Conditions: Child Maltreatment
• Intervention / Treatment: Behavioral: SafeCare; Behavioral: Supportive Case Management

Detailed Description

The study design was a cluster randomized trial, with randomization occurring at the team level within each study site. We randomized providers within each site to be trained in SafeCare (n = 96) or to continue to deliver SCM (n = 96) which allowed us to control for site differences. From a statistical power perspective, it would have been preferable to randomize clients to interventions, but the fact that teams at community-based organizations typically served a defined geographic area would have meant that home visitors within each team would have had to deliver two distinct interventions raising the likelihood of cross-contamination which, in our experience, would have not been well managed by the study. Providers that were randomized were invited to participate in the study by completing a survey at baseline and 1-year follow-up which included measures of demographics, work experience, organizational factors (culture, climate, leadership) and individual attitudes and beliefs that may affect implementation and service quality. We recruited caregivers into the research study that were receiving services from randomized providers. Measurement of client-level outcomes were collected at baseline and 6-month follow up, as well as during service provision. Both quantitative and qualitative data were collected to capture both breadth and depth of family outcomes and experiences of the interventions.74 By focusing on measures of well-being, this study holds potential to expand the way in which the program effectiveness is conceptualized. This could be particularly important for dissemination of parenting programs to at-risk parents, who may be interested in different outcomes (e.g., improving their child's behavior) than child welfare systems, and findings may be useful for making interventions more appealing to consumers. To measure barriers and facilitators to implementing an evidence-based practice (EBP), we used a qualitative research strategy that included semi-structured interviews and focus groups with several levels of staff, responsible for implementing the model: program administrators, supervisors, and frontline staff (providers). This strategy would yield insight into an array of stakeholder perspectives concerning implementation. In contrast to most implementation studies, ours is the first to conduct a complementary set of interviews and focus groups with staff who have not been trained in SC, the SCM providers.

• Study Type: Interventional
• Enrollment (Actual): 289
• Phase: Not Applicable

Contacts and Locations

Study Locations

• United States
  • Iowa
    • Council Bluffs, Iowa, United States, 51301
      • Southwest Iowa Family Access Center
    • Des Moines, Iowa, United States, 50314
      • Children & Families of Iowa
    • Des Moines, Iowa, United States, 50324
      • Mid Iowa Family Therapy Center
    • Iowa City, Iowa, United States, 50644
      • Four Oaks
    • Waterloo, Iowa, United States, 50701
      • Families First

Participation Criteria

Eligibility Criteria

• Ages Eligible for Study: 16 years to 63 years (Adult, Older Adult)
• Accepts Healthy Volunteers: No
• Genders Eligible for Study: All

Description

Inclusion Criteria:

• Home Visitors: Home Visitors who are providing Supportive Case Management or being trained in SafeCare.
• Parents: English or Spanish speaking parents with a child between the ages of 0-5 years who are receiving services from a participating site.

Exclusion Criteria:

• Parents under the age of 18, who have no children under the age of 6 years, or do not speak English or Spanish will be excluded from this study.

Study Plan

How is the study designed?

Design Details

• Primary Purpose: Treatment
• Allocation: Randomized
• Interventional Model: Parallel Assignment
• Masking: None (Open Label)

Number of Arms

2

Arms and Interventions

Participant Group / Arm	Intervention / Treatment
Experimental: SafeCare; SafeCare, an evidence-based home visiting program	Behavioral: SafeCare; An evidence-based home visiting program
Active Comparator: Supportive Case Management; Child welfare services as usual	Behavioral: Supportive Case Management; Child welfare services as usual; Other Names: Services as Usual

What is the study measuring?

Primary Outcome Measures

Outcome Measure	Measure Description	Time Frame
Parenting Young Children Scale-Supporting Positive Behavior Sub-scale	The Supporting Positive Behavior sub-scale of the Parenting Young Children Scale is made up of 7 items that assess supporting positive behavior (Example question: "Notice and praise your child's good behavior"). This subscale score is generated by calculating a mean of seven items on a 7-point scale (1=Not at all - 7=Almost Always). Theoretical range of means: 1-7; Actual range of means: 1-7. Higher scores represent higher degree of positive parenting skills. Because the data were heavily skewed toward the positive end of the scale, the measure was dichotomized into high vs. low based on an approximate median split. Specifically, participants scoring 6 or below were rated as low (n=125) and participants rating higher than 6 were rated as high (n=159).	Families were assessed at two time points: 1) after being invited into the study during a visit with their provider (Baseline) and, 2) approximately 6-months later (follow-up).
Parenting Young Children Scale-Proactive Parenting Sub-scale	The Proactive Parenting sub-scale of the Parenting Young Children Scale is made up of 7 items that assess proactive parenting (Example question: "Avoid struggles with your child by giving clear choices"). This subscale score is generated by calculating a mean of seven items on a 7-point scale (1=Not at all - 7=Almost Always). Theoretical range of means: 1-7; Actual range of means: 1-7. Higher scores represent higher degree of positive parenting skills. Because the data were heavily skewed toward the positive end of the scale, the measure was dichotomized into high vs. low based on an approximate median split. Specifically, participants scoring 6 or below were rated as low (n=148) and participants rating higher than 6 were rated as high (n=136).	Families were assessed at two time points: 1) after being invited into the study during a visit with their provider (Baseline) and, 2) approximately 6-months later (follow-up).
Parenting Young Children Scale-Setting Limits Sub-scale	The Setting limits sub-scale of the Parenting Young Children Scale is made up of 7 items that assess limit setting (Example question: "Stick to your rules and not change your mind"). This subscale score is generated by calculating a mean of seven items on a 7-point scale (1=Not at all - 7=Almost Always). Theoretical range of means: 1-7; Actual range of means: 1-7. Higher scores represent higher degree of positive parenting skills. Because the data were heavily skewed toward the positive end of the scale, the measure was dichotomized into high vs. low based on an approximate median split. Specifically, participants scoring 6.3 or below were rated as low (n=141) and participants rating higher than 6.3 were rated as high (n=143).	Families were assessed at two time points: 1) after being invited into the study during a visit with their provider (Baseline) and, 2) approximately 6-months later (Follow-up).
Parenting Stress Inventory - Short Form	Parenting Stress Inventory - short form is a 36-item scale designed to measure stressors in parenthood including parental distress, dysfunctional interactions, and stressors related to having a difficult child. (Example question: "Sometimes I feel like my child doesn't like me and doesn't want to be close to me"). A total score generated by summing all 36 items on a 5-point scale (1=Strongly Agree - 5=Strongly Disagree). Theoretical total range: 36-180; Actual total range: 38-146. Lower scores represent more stress/dysfunction. Because the data were heavily skewed toward the positive end of the scale, the measure was dichotomized into high vs. low based on an approximate median split. Specifically, participants scoring 71 or below were rated as low (n=136) and participants rating higher than 71 were rated as high (n=148).	Families were assessed at two time points: 1) after being invited into the study during a visit with their provider (Baseline) and, 2) approximately 6-months later (follow-up).
Protective Factors Survey-Parent Knowledge Sub-scale	The Parent knowledge sub-scale of the Protective Factors is made up of 5 items that assess parent knowledge. (Example question: "There are many times when I don't know what to do as a parent"). This subscale score is generated by calculating a mean of 5 items on a 7-point scale (1=Never - 7=Always). Theoretical range of means: 1-7; Actual range of means: 3.4-7.0. Higher scores indicate higher parent knowledge. Because the data were heavily skewed toward the positive end of the scale, the measure was dichotomized into high vs. low based on an approximate median split. Specifically, participants scoring below 7 were rated as low (n=149) and participants rating 7 and higher were rated as high (n=133).	Families were assessed at two time points: 1) after being invited into the study during a visit with their provider (Baseline) and, 2) approximately 6-months later (follow-up).
Protective Factors Survey-Family Functioning Sub-scale	The Family Functioning sub-scale of the Protective Factors is made up of 5 items that assess family functioning. (Example question: "My family pulls together when things are stressful"). This subscale score is generated by calculating a mean of 5 items on a 7-point scale (1=Never - 7=Always). Theoretical range of means: 1-7; Actual range of means: 1-7. Higher scores indicate higher family functioning. Because the data were heavily skewed toward the positive end of the scale, the measure was dichotomized into high vs. low based on an approximate median split. Specifically, participants scoring below 6 were rated as low (n=146) and participants rating 6 and higher were rated as high (n=133).	Families were assessed at two time points: 1) after being invited into the study during a visit with their provider (Baseline) and, 2) approximately 6-months later (follow-up).
Brief Symptom Inventory-Significant Case Percentage	The Brief Symptom Inventory is a 53-item scale designed to measure a range of emotional health states including depression, anxiety, somatization, and others. (Example question: "How much were you distressed by nervousness or shakiness inside"). The 'significant case' definition from the BSI was used and includes those with elevated scores (higher than 2) on any of the subscales. The percentage of participants that were considered a 'significant case' is reports. The 'case' definition from the BSI, which includes elevation on any of the subscale. The Brief Symptom Inventory is a 53-item scale designed to measure a range of emotional health states including depression, anxiety, somatization, and others. A total score or the, Global Severity Index, is generated by calculating a mean of all 53 items; lower scores indicate lower levels of distress.	Families were assessed at two time points: 1) after being invited into the study during a visit with their provider (Baseline) and, 2) approximately 6-months later (follow-up).
BSI-Global Severity Index	The Brief Symptom Inventory is a 53-item scale designed to measure a range of emotional health states including depression, anxiety, somatization, and others. (Example question: "How much were you distressed by nervousness or shakiness inside"). The Global Severity Index calculated a mean of all of the BSI subscales which includes 53 items on a 5-point scale (0=Not at all-4=Extremely). Theoretical range of means: 0-4; Actual range of means: 0.0-3.5). Higher scores indicate higher existence of symptoms. Because the data were heavily skewed toward the positive end of the scale, the measure was dichotomized into high vs. low based on an approximate median split. Specifically, females scoring below .77 and males scoring below .57 were rated as low (n=197) and females rating .78 and higher and males rating .58 and higher were rated as high (n=87).	Families were assessed at two time points: 1) after being invited into the study during a visit with their provider (Baseline) and, 2) approximately 6-months later (Follow-up).
Devereaux Early Child Assessment-Initiative Sub-scale	The Initiative sub-scale of the Devereaux Early Child Assessment (DECA) is made up of 11-18 items (depending on child's age) that assesses the child's initiative behavior. (Example question: "Did the child do things for himself"). This subscale score generates a t-score of standardized norms from a sum of the 11-18 items on a 5-point scale (0=Never - 4=Very frequently). Theoretical range of means: 28-72; Actual range of means: 28-72. Higher ratings represent a higher degree of child initiation. Because the data were heavily skewed toward the positive end of the scale, the measure was dichotomized into high vs. low based on an approximate median split. Specifically, participants scoring below 56 were rated as low (n=110) and participants rating 56 and higher were rated as high (n=118).	Families were assessed at two time points: 1) after being invited into the study during a visit with their provider (Baseline) and, 2) approximately 6-months later (Follow-up).
Devereaux Early Child Assessment-Attachment Sub-scale	The Attachment sub-scale of the Devereaux Early Child Assessment (DECA) is made up of 8-18 items (depending on child's age) that assesses the child's attachment behavior. (Example question: "Did the toddler accept comfort from a familiar adult"). This subscale score generates a t-score of standardized norms from a sum of the 8-18 items on a 5-point scale (0=Never - 4=Very frequently). Theoretical range of means: 28-72; Actual range of means: 28-72. Higher ratings represent a higher degree of child attachment. Because the data were heavily skewed toward the positive end of the scale, the measure was dichotomized into high vs. low based on an approximate median split. Specifically, participants scoring below 52 were rated as low (n=101) and participants rating 52 and higher were rated as high (n=138).	Families were assessed at two time points: 1) after being invited into the study during a visit with their provider (Baseline) and, 2) approximately 6-months later (Follow-up).

Secondary Outcome Measures

Outcome Measure	Measure Description	Time Frame
Mother-Child Neglect Scale (MCNS)	The Mother-Child Neglect Scale (MCNS) is a 22-item scale designed to assess caregiving behaviors in four domains: physical, cognitive, supervision, and emotional needs. (Example question: "I make sure my child sees a doctor when he/she needs one"). A total score generated by calculating a mean of the 22 items rated on a 4-point scale (1=Strongly Agree - 4=Strongly Disagree). Theoretical range of means: 1-4; Actual range of means: 1.2-3.9. Lower scores indicate less neglectful behaviors. Because the data were heavily skewed toward the positive end of the scale, the measure was dichotomized into high vs. low based on an approximate median split. Specifically, participants scoring 3.65 and below were rated as low (n=139) and participants rating higher than 3.65 were rated as high (n=145).	Families were assessed at two time points: 1) after being invited into the study during a visit with their provider (Baseline) and, 2) approximately 6-months later (Follow-up).
Confusion, Hubbub, and Order Scale (CHAOS)	The CHAOS scale (Confusion, Hubbub, and Order) is a 15-item scale used to measure structure and chaos in the home environment. (Example question: "There is very little commotion in our home"). A total score generated by calculating a mean of all 15 items on a 4-point scale (1=Very much like your own home - 4=Not at all like your own home). Theoretical range of means: 1-4; Actual range of means: 1.0-2.7. Lower scores indicate less chaos. Because the data were heavily skewed toward the positive end of the scale, the measure was dichotomized into high vs. low based on an approximate median split. Specifically, participants scoring 1.5 or below were rated as low (n=151) and participants rating higher than 1.5 were rated as high (n=133).	Families were assessed at two time points: 1) after being invited into the study during a visit with their provider (Baseline) and, 2) approximately 6-months later (Follow-up).
Family Resources Scale - Revised	The Family Resources Scale - Revised is a 40-item scale that assesses the adequacy of family needs in four domains: basic needs, money, time for self, and time for family. (Example question: "How well is the following need being met: House or apartment"). A total count of resources needed out of 40 assessed. Resources were considered needed if the participant indicated the resource was 'Not at all' met, 'A little' met, or 'Sometimes' met (5-point scale: 1=Not at all - 5=Almost always). Theoretical range of means: 0-40; Actual range of means: 0-40. Higher ratings indicating a higher number of resources needed. Because the data were heavily skewed toward the positive end of the scale, the measure was dichotomized into high vs. low based on an approximate median split. Specifically, participants with 9 or fewer needs not met were rated as low (n=152) and participants with 10 and higher needs not met were rated as high (n=132).	Families were assessed at two time points: 1) after being invited into the study during a visit with their provider (Baseline) and, 2) approximately 6-months later (Follow-up).

Collaborators and Investigators

• Sponsor: Daniel Whitaker
• Collaborators: Patient-Centered Outcomes Research Institute
• Investigators: Principal Investigator: Daniel J Whitaker, PhD, Georgia State University

Publications and helpful links

General Publications

• Whitaker DJ, Self-Brown S, Hayat MJ, Osborne MC, Weeks EA, Reidy DE, Lyons M. Effect of the SafeCare(c) intervention on parenting outcomes among parents in child welfare systems: A cluster randomized trial. Prev Med. 2020 Sep;138:106167. doi: 10.1016/j.ypmed.2020.106167. Epub 2020 Jun 20.

Study record dates

Study Major Dates

• Study Start (Actual): September 1, 2015
• Primary Completion (Actual): May 1, 2018
• Study Completion (Actual): May 1, 2018

Study Registration Dates

• First Submitted: August 27, 2015
• First Submitted That Met QC Criteria: September 11, 2015
• First Posted (Estimate): September 15, 2015

Study Record Updates

• Last Update Posted (Actual): January 13, 2020
• Last Update Submitted That Met QC Criteria: December 20, 2019
• Last Verified: December 1, 2019

More Information

Terms related to this study

• Other Study ID Numbers: 90CU0062