Doris, J. L., Meguid, V., Thomas, M., Blatt, S. & Echenrode, J. (2006). Prenatal cocaine exposure and child welfare outcomes. Child Maltreatment, 11(4), 326-337.
There is an assumption that prenatal cocaine exposure is a risk factor for child maltreatment. Using data from a university-affiliated hospital, this study investigated the relationship between prenatal cocaine exposure and child welfare outcomes with a matched sample. After controlling for prenatal care as well as maternal use of alcohol and tobacco, infants exposed to cocaine had significantly higher health risks (e.g., decrements in birth weight, length, head circumference, and lower 5-min Apgar scores) than the control group. However, after controlling for maternal involvement with child welfare, data collected three-years later suggests that the cocaine-exposed infants were not more likely to experience child maltreatment or foster care placements. According to this study, prenatal cocaine exposure does not increase the risk of future child maltreatment.
This study used the entire sample of the neonates born in a hospital during a 2-year period (n=8,913). Of these, 811 were selected for urine toxicology. The sample is biased such that it includes only children whose mother was suspected of using cocaine. Therefore, the control group had mothers who were suspected of cocaine use but whose child did not test positive. This confounds the results. One hundred tested positive for cocaine use and these were matched with 100 who tested negative. Of the 100 negative controls, 15 admitted to cocaine use during pregnancy and nine tested positive during pregnancy. These children were removed along with their matched controls (leaving n=76 in each group). Maternal involvement with child protection meant that there was at least one substantiated report of child maltreatment within 21 days of the child's birth; therefore controlling for incidents when the case was kept open because of the hospital's concerns about prenatal cocaine exposure. Child abuse was measured the State Central Register and foster care records. Descriptive statistics were used to understand the sample. Logistic regressions were used to determine the association between prenatal cocaine exposure and child welfare outcomes. Cell sizes did become small.