NCJ Number
204455
Journal
Child Abuse Review Volume: 12 Issue: 6 Dated: November-December 2003 Pages: 416-439
Date Published
November 2003
Length
24 pages
Annotation
After identifying instruments designed to predict future child maltreatment, this article evaluates their predictive accuracy.
Abstract
In the search of a number of electronic databases to identify relevant published studies, the researchers selected only those studies that had used a prospective methodology and tested at least one standardized instrument designed to identify families during the antenatal or postnatal period that were at increased risk of child maltreatment. Eight studies were selected; however, only six of them followed up on the entire sample; of these, only two did not provide an intervention for some or all of the high-risk group. The instruments are based on a range of socioeconomic, demographic, and clinical criteria. Six instruments require the professional to make subjective judgments. Seven of the instruments have summative scales and cutoff points above which respondents are deemed to be "high risk." Overall, the results indicate that three of the instruments have a high level of accuracy regarding both sensitivity and specificity for at least one outcome being measured; however, only two of the instruments tested using the entire sample obtained acceptable levels of accuracy for both specificity and positive predictive value, i.e., above 90 percent and 25 percent, respectively (Muir et al., 1989 and Hunter et al., 1978). Even with the most accurate instrument, however, more than half of the families who are diagnosed as being at risk do not actually go on to abuse or maltreat their children. This suggests that currently available instruments should only be used to focus nonpunitive interventions, possibly within the context of research studies. An alternative to the use of screening instruments to increase the cost-effectiveness of interventions would be the provision of universal parenting support that begins during the antenatal period, in which professionals already working with families would identify parents in need of higher levels of intervention. Such an approach would avoid contributing to the fallacy that it is possible to predict accurately parents who will maltreat their children. This approach would also avoid submitting parents to the stigmatization of being labeled "at risk." 25 references