Bridging Disciplinary Boundaries (January 11 - 14, 2007)

Recently meta-analyses have taken center stage in the process of determining what works in social welfare, education, health, and criminology. Home visiting is a service delivery method whose popularity has waxed and waned with the times. It has been embraced by nursing, social work, child development, and public health as a method to provide services to families in their homes. Home visiting programs may have a number of goals including: prevention of child abuse and neglect, improved health outcomes, improved parenting skills, and improved child development (cognitive and/or behavioral). Past reviews of evidence on home visiting suggested the evidence was mixed but promising. This paper synthesizes the impact of randomized controlled trials of home visiting services provided by non-nursing professionals to families of newborns published between 1990 and 2005. The focus is on outcomes associated with child well-being such as parenting, provision of a safe and stimulating home, cognition and behavior. Abstracts and titles of 148 publications identified through database searches, existing reviews and website searches were reviewed. Ultimately, 17 independent studies provided data in a format from which effect sizes (either standardized differences in means or odds ratios) could be calculated. The 17 studies provided 177 comparisons which could be synthesized in 61 different models when grouped based on: construct of interest, timing of measurement, measure and the use of adjusted and/or unadjusted estimates of impact. Thirty-nine of the 61 syntheses were statistically significant at the p ≤ .10 level. Only 6 of the effect sizes surpassed Cohen's rule of thumb for a small effect (≤ .20). Four studies, all of which included the prevention of child abuse and neglect as a goal, included a measure of the risk of child abuse in the family. A standardized difference in means of -.21 (p = .00) was found when the synthesis used the furthest time point reported for each of the four studies. This finding suggests programs which have the goal of decreasing the risk of a family for child abuse and neglect can achieve that goal using non-nursing home visitors. The impact remains significant if a shorter time period is used and if data adjusted for different characteristics is used. Home visiting by non-nursing providers was also found to have the desired impact on child cognition, the environment in which the infants were being raised (measured by the HOME Inventory), social support for the parents, parenting knowledge and attitudes towards parenting. However, it did not have the desired impact on child behavior or parental mental health. Further analyses examining the specific characteristics of the home visiting programs help identify particular components or practices that may be vital to the success of these programs. The review can inform policy and practice about the possibility of home visiting services to families with infants by non-nursing professionals to affect aspects of the well-being of children which may affect their transition to, and success in, school and adulthood.