Manuscript Details

Source

Barth, R. P. (1991). An experimental evaluation of in-home child abuse prevention services. Child Abuse and Neglect: The International Journal, 15(4), 363-75.
Rating
Moderate
Author Affiliation

Richard Barth, the study author, developed CPEP and was involved in program implementation.

Funding Sources

Department of Health and Human Services, Office of Human Development Services, Administration on Children, Youth, and Families, Grants 90-CA-0988 and 90-PJ-000101. Bio-Medical Research Support Grant, Grant 2-507-RR07006. National Institute of Health, Division of Research, State of California Office of Child Abuse Prevention, Grant CB 33015-AI.

Study Design

Design Attrition Baseline equivalence Confounding factors Valid, reliable measures?
Randomized controlled trial Not applicable Established on race and socioeconomic status None
Notes

Mothers in the CPEP group who refused services or received fewer than five home visits were reassigned to the control group. Thus, we treat the study as a quasi-experimental design for purposes of the review, and an evaluation of attrition does not apply. The moderate rating applies to outcomes that were not assessable at baseline (health habits during pregnancy, child health, child development, and reductions in child maltreatment); those outcomes that could be measured at baseline (parent well-being and parent support) but were not controlled in the analysis receive a low rating.

Study Participants

Expectant mothers deemed at-risk for child abuse were referred to the study team by service providers in public health, education, and social services. Eligible mothers who expressed interest in the program were randomly assigned to either CPEP or a no-treatment control group. Random assignment occurred before the formal initial assessment interview by the CPEP team. Mothers in the CPEP group who refused services or received fewer than five home visits were reassigned to the control group. The analysis sample for the study included 97 mothers in the CPEP group and 94 mothers in the control group. The ethnic composition of the combined groups was 45 percent Caucasian, 17 percent African American, 31 percent Hispanic, and 7 percent of other or multiple races. Four in 10 mothers in the sample were on Aid to Families with Dependent Children, and 70 percent had incomes of less than $10,000. On average, mothers were 5.7 months pregnant when they began the program; 56 percent had at least one additional child at enrollment.

Setting

The study was conducted in Contra Costa County, California.

Home Visiting Services

CPEP clients received on average 11 home visits over the course of the six-month study, with a range of 5 to 20 visits. Paraprofessionals, known as parenting consultants, administered the intervention. Home visits were centered around task lists created by mothers in conjunction with the parenting consultant. Tasks might be completed by the mother alone, the parenting consultant alone, or together. Examples of tasks include visiting prenatal care or obtaining transportation.

Comparison Conditions

Control group mothers received referrals to appropriate social and health services identified by the two-hour assessment interview.

Findings that rate moderate or high in this manuscript

Maternal health
Outcome measure Timing of follow-up Rating Effect size Stastical significance Sample size Sample description
Eat bad 6 months Moderate -0.06 Not statistically significant, p ≥ 0.05 191 mothers Reassigned sample
Eat right 6 months Moderate 0.04 Not statistically significant, p ≥ 0.05 191 mothers Reassigned sample
Prenatal care 6 months Moderate 0.05 Not statistically significant, p ≥ 0.05 191 mothers Reassigned sample
Pregnancy problems 6 months Moderate 0.12 Not statistically significant, p ≥ 0.05 191 mothers Reassigned sample
Hospital stay 6 months Moderate -0.33 Statistically significant, p = 0.02 191 mothers Reassigned sample
Discomfort (birth outcome) 6 months Moderate -0.14 Not statistically significant, p ≥ 0.05 191 mothers Reassigned sample
Worries (birth outcome) 6 months Moderate 0.12 Not statistically significant, p ≥ 0.05 191 mothers Reassigned sample
Effect rating key
Favorable finding / Statistically significant
UnFavorable finding / Statistically significant
Ambiguous finding / Statistically significant
No effect / Not statistically significant
Child health
Outcome measure Timing of follow-up Rating Effect size Stastical significance Sample size Sample description
Birthweight (grams) 6 months Moderate 0.21 Not statistically significant, p ≥ 0.05 191 mothers Reassigned sample
Health (child welfare) 6 months Moderate 0.06 Not statistically significant, p ≥ 0.05 191 mothers Reassigned sample
Emergency (child welfare) 6 months Moderate 0.00 Not statistically significant, p ≥ 0.05 191 mothers Reassigned sample
Baby care (child welfare) 6 months Moderate 0.09 Not statistically significant, p ≥ 0.05 191 mothers Reassigned sample
Effect rating key
Favorable finding / Statistically significant
UnFavorable finding / Statistically significant
Ambiguous finding / Statistically significant
No effect / Not statistically significant
Child development and school readiness
Outcome measure Timing of follow-up Rating Effect size Stastical significance Sample size Sample description
Activity (ITQ subscale) 6 months Moderate -0.23 Not statistically significant, p ≥ 0.05 191 mothers Reassigned sample
Mood (ITQ subscale) 6 months Moderate -0.39 Statistically significant, p = 0.01 191 mothers Reassigned sample
Distractibility (ITQ subscale) 6 months Moderate -0.07 Not statistically significant, p ≥ 0.05 191 mothers Reassigned sample
Effect rating key
Favorable finding / Statistically significant
UnFavorable finding / Statistically significant
Ambiguous finding / Statistically significant
No effect / Not statistically significant
Reductions in child maltreatment
Outcome measure Timing of follow-up Rating Effect size Stastical significance Sample size Sample description
Need care (child welfare) 6 months Moderate 0.07 Not statistically significant, p ≥ 0.05 191 mothers Reassigned sample
Effect rating key
Favorable finding / Statistically significant
UnFavorable finding / Statistically significant
Ambiguous finding / Statistically significant
No effect / Not statistically significant

This study included participants with the following characteristics at enrollment:

Race/Ethnicity

The race and ethnicity categories may sum to more than 100 percent if Hispanic ethnicity was reported separately or respondents could select two or more race or ethnicity categories.

Black or African American
17.00%
Hispanic or Latino
31.00%
White
45.00%
Some other race
7.00%

Maternal Education

Data not available

Other Characteristics

Enrollment in means-tested programs
40.00%