Substance abuse during pregnancy is not only a medical problem for the infant and mother, but also results in major social consequences for other family members and the community.^1,2 In the past decade, the medical literature has focused on the potential impact of maternal substance abuse on the fetus and subsequent child development.³ The sequelae of intrauterine drug exposure include impaired fetal growth,¹⁷ prematurity,²⁴ neurologic deficits,²⁵ behavioral changes,²⁸ developmental delays,³⁴ sudden infant death syndrome (SIDS),³⁵ and increased risk for child abuse.⁴¹ Because of the perception that drug-using mothers are unable to care for their infants adequately, or that drug-affected environments are unsafe for infants, hospitals have taken measures to identify infants exposed to drugs during pregnancy, and child protective authorities in many jurisdictions have acted to separate the infants from their mothers at birth. The impact of these interventions has not been studied.

Public policies have been implemented in response to the increased use of drugs and the reported risks for the infant. In some locales, public attitudes have encouraged criminal prosecution of the mother; other jurisdictions emphasize a more therapeutic approach. It is not uncommon for a positive drug test to be interpreted as prima facie evidence of child endangerment, resulting in allegations of child neglect against the mother and interruption of maternal custody.⁴⁴ Indeed, positive newborn urine toxicology screening results have been used as evidence to prosecute mothers criminally as drug users or even as drug traffickers.⁴⁵ A more therapeutic model holds that a diagnosis of child neglect is intended to lead to the provision of rehabilitative services to further the best interests of the child.^45,47 The American Academy of Pediatrics considers unethical the practice of performing drug screening for the primary purpose of detecting illegal use.⁵⁰

The medical literature has not ignored the controversial legal ramifications of criminal prosecution of new mothers.^45,49 Studies conducted to examine which infants are most likely to be placed in foster care show that the race and age of the mother are significant factors.¹ However, to our knowledge, no study has been conducted to assess the long-term consequences for the infants or their mothers resulting from newborn drug screening results being reported to judicial authorities. Many questions need answering: Are adequate resources to assist the families provided? Are effective rehabilitative services available and are they used by the mothers? Do the infants return to their mothers or languish in foster care for long periods of time? For how long do the courts retain legal custody and control of the infant? What are the factors that lead to termination of court involvement? In short, does the reporting of these children really benefit either the infant or the mother? Answering these questions through large studies will be difficult because of complex interactions among the medical, legal, and social systems. Moreover, because of the lifestyles accompanying illicit activity and substance abuse, the study population is diverse and difficult to follow prospectively.

The purpose of this study was to examine what happened to mothers and infants in one county in Northern California after the infants' positive toxicology screening results were reported to county authorities. This report presents descriptive information intended to begin the process of answering the questions presented above. It is based on a retrospective and concurrent review of medical and social service agency records, following a group of drug-exposed newborns for 5 years.

METHODS

Policies and Drug Testing

Physical custody and legal control of the infant are two distinct and somewhat independent parameters. Although physical custody could be awarded to a mother, her infant might remain a dependent of the court, ensuring that county agencies could monitor the infant's health and safety and the mother's activity and rehabilitative efforts. Provision of preventive and rehabilitative services was a policy requirement of the county; the current emphasis of child welfare services is on keeping the family intact. Cases could be closed and court involvement terminated at any of the review hearings, but a final determination of custody and placement was required before 18 months of age. The length of time the infants were dependents of the court was determined by the judicial review of each mother's compliance with court orders and by reports from social service workers regarding her progress and stability and the safety of the infant's environment.

Subjects

The people using the teaching service of the well-baby nursery represented a diverse population: 75% were Hispanic, 15% were non-Hispanic Caucasians, 8% were African-American, and 2% were Asian or South Pacific Islanders. Most were enrolled in or eligible for Medicaid.

Statistical Analysis

RESULTS

Cocaine was the principal drug identified, being present in 47 infants. Opiates were identified in 5, amphetamines in 2, and other drugs in 4. Five infants had more than one drug identified.

During the period subjects were entered into this study, four other infants were also discharged from the well-baby nursery to foster homes. They were not included in this study because their needs for social services were not related to substance abuse.

Subject Characteristics

Table 1. Characteristics of the Mothers and Infants [View Table]

All of the infants had normal physical examinations, with about one in three described as jittery. Five (9%) were slightly premature, but mature and well enough to remain in the level 1 nursery; the others were full-term and of normal weight. No behavioral or physical characteristics specific to substance abuse were recorded.

Initial Disposition

Table 2. Duration of Services Provided to Infants and Separation From Relatives (Days) [View Table]

Duration of Foster Care and Subsequent Separation From Parent

Of the 14 (26%) who remained in nonrelative foster care for >1 month, 3 were united eventually with relatives (after 2 to 10 months of foster care) and 5 were returned to a parent (2 in the second and third months, 2 as a result of the 6-month hearing, and 1 after 18 months). Thirteen infants in this study (25%) were never reunited with their parents; only 6 (11%) were never reunited with relatives.

No infant was placed for adoption in the first 6 months of life. Yet no infant was in foster care >18 months without being in the process of being adopted. Of the 26 infants who were returned to their parents in the first month of life, 7 (27%) eventually reentered the foster care system because of continued drug use by the mother. At 12 months of age, 10 infants (19%) remained in foster care. All of the others had been returned to relatives or parents, or they had been placed in long-term guardian relationships or adopted. Most infants who were successfully reunited with their mothers did so in the early postpartum period. The median time spent in foster care for infants who returned to their parents was only 6 days.

Duration of Dependent-of-the-Court Status

Justification for Closing the Cases and Terminating Court Dependency

Eventual Disposition

Table 3. Eventual Custody Arrangements for the Infants [View Table]

Table 3 also lists the place of residence of the person with legal custody of the child at the time the case was closed. A total of 16 (30%), including those adopted by nonrelatives, were residing in a two-parent home. Fourteen (26%) lived with one biological parent; 20 (38%), including some who were with their mother, resided in the home of relatives.

Services Rendered and Compliance With Orders

Table 4. Requirements and Court-Ordered Services for the Mothers [View Table]

Subsequent Maternal Problems

Predicting Eventual Disposition

Table 5. Maternal Characteristics Grouped by Duration of Infant's Separation From Mother [View Table]

In contrast, mothers who never regained custody of their infants were more likely to be African-American, to have had a history of previous involvement with CPS, or to have had a child removed from the home previously. They were also more likely to have failed drug rehabilitation.

Logistic regressions were used to explore these data further, using the maternal characteristics that appeared significant by ² tests. Although the model as a whole did not exhibit statistical significance, two characteristics that were related strongly to infants never returning to a parent were the mother's previous involvement with CPS and the mother having previously lost custody of a child. The odds ratio for a mother with previous CPS involvement never having her infant returned to her was 3.4.

Maternal failure at previous rehabilitative attempts and the absence of a father capable of helping were the two variables related strongly to not reuniting in the first week after hospital discharge, but they did not distinguish between late reunification and nonreunification. Five of the 16 mothers who had failed rehabilitation previously eventually did succeed this time and regained full custody of the infant involved in this study. An additional characteristic, the amount of prenatal care received, was related linearly to the timing of the infant's return to the mother, with more prenatal care being associated with early return of the infant.

Impact of Policy on Living Situation of Infants and Mothers

Twenty-four (45%) of the mothers completed a drug rehabilitation program and produced negative urine drug screening results for a prolonged period, as a consequence of court orders. Twenty-four mothers (45%) were unable to comply with the orders to complete a rehabilitation program, some because they never got involved in a program, others because they tried and failed, and a few because they moved. Fourteen of the 22 mothers ordered to undergo psychiatric treatment or counseling complied. Twenty-one (40%) of the mothers eventually established housing considered by the courts to be stable. In 26 cases (49%), services offered were used fully by the mothers. In 6 cases (11%), the mothers partially used services available, and in 9 additional cases (17%), only the children received benefits of services.

DISCUSSION

However, the sample size of this study was small, and the subjects were drawn from patients who were in the nursery reserved for "well babies." The study involved a subset of patients in that nursery who do not have a private pediatrician on the hospital's medical staff. During the period of subject enrollment, no drug screenings were performed on the infants cared for by private pediatricians in this nursery, suggesting that no other drug-exposure referrals to CPS from the well-baby nursery occurred. The subjects were identified by selective screening based on identified risk factors. If universal screening had been used or if other screening methods had been used (eg, meconium or hair), it is possible that the sample size would have been larger or different. Such universal screening is not endorsed by the American Academy of Pediatrics.⁵⁵

Whether the results of this study are capable of being generalized to other communities or larger populations is unclear. This study was performed in a county offering progressive services. Although dominated by suburban living, San Mateo County has pockets of poverty, crowding, and drug use more characteristic of inner cities. For example, the finding that only one of the infants was involved in reported child abuse is in marked contrast with other reports.^43,56 Whether such differences are related to the subjects or the policies was not addressed by this study. Nevertheless, the beneficial changes in the life and environment observed in half of the mothers and most of the infants can serve as an endorsement of the policies of reporting these patients, involving services, and maintaining legal custody as a tool for monitoring and coercing progress and reform. This study did not address directly the question of whether the assessments leading to the "initial" disposition could occur before the mother or infant is discharged from the hospital and thus obviate the 1 week of foster care experienced by so many of the study subjects. Another issue worthy of mention is the use of court outcome as a predictor of staying off of drugs in the long term. Because the data were collected from social service files, no information about the use of drugs after the cases were closed was available unless the mother reconnected with social service agencies.

The mothers of this study did not have the same demographic characteristics as those of the nursery as a whole. As mentioned in the text, these women were predominantly African-American, often with complicated obstetric histories. This finding is consistent with a study of the epidemiology of drug use in pregnancy throughout the entire state of California,⁵⁷ suggesting that this group of subjects may represent larger populations.

The finding that certain maternal characteristics predicted failure of the mother to regain custody of the infant suggests that an individualized approach to the placement of children from the hospital is practical. However, the sample size in this study was small, and the description of certain maternal characteristics as being predictive resulted from exploratory analysis of the data, not true hypothesis testing. Although 53 subjects were included in the logistic model, this was not a random sample of a large population, and this analysis served best simply to describe the data. Firm conclusions about which maternal characteristics are most useful in determining policy must await additional investigation of larger sample groups. The number of mothers who eventually underwent rehabilitation and regained custody of their children even after months of noncompliance and continued drug use suggests that early decisions leading to permanent out-of-home placement might be improper. Even a few mothers carrying all the predictors for failure succeeded in breaking out of their dysfunctional lifestyles.

The one infant who eventually experienced abuse was an atypical case throughout many months of court dependency. The mother of this infant was the only mother in this study who had her child returned to her despite failing to comply with drug rehabilitation and urine testing. Although she was monitored by the courts for 18 months, it was felt she had developed a stable relationship with her child, and the caseworker argued for the return of her child despite her lack of compliance with court orders. If this case serves as an example, it suggests that the close monitoring and strict demand for compliance with court orders that the other mothers experienced were appropriate and beneficial to the children.

Other findings in this study show the impact of court-ordered long-term monitoring of the mothers' progress and compliance. Seventeen of the mothers who underwent random testing of their urine were found to have a continuing problem with drugs. Only one of these mothers regained full custody of her child. Clearly, a monitoring system was in place and was truly being used in the disposition process. However, some court orders, eg, keeping all medical appointments, appeared to be not as critical in the decision to return a child to the mother. In addition, some orders appeared to be used more as a test of mother's compliance rather than truly having a rehabilitative function. For example, all the mothers in this study were ordered to participate in a class to learn cardiopulmonary resuscitation (CPR). This order was said to be based on literature suggesting that cocaine-exposed infants might be at increased risk for SIDS in the first few months of life.³⁵ However, if a mother had not received CPR training by the time of the 6- and 12-month review hearing, long after the theoretic risk of SIDS, she would still be ordered to undertake CPR training. This suggests a need for more communication between the court and health care providers in establishing guidelines for intervention and therapy. In addition, the impact of placement decisions on the child's development and behavior was never monitored or formally addressed. Issues such as discontinuity of care and the impact of kinship care versus foster care may have a significant impact on the children, but data specific to these issues were not available.

One finding that does not appear in group data but was apparent to those involved in reviewing each chart is that in a few cases, the threat of losing her child through adoption and the finality of the decisions of the 18-month hearing were critical in motivating the mother to finally enter and complete drug rehabilitation.

This study did not address the question of what happens to infants exposed to drugs in utero but not detected or reported by the nursery. A number of women had the same characteristics as these mothers but were never reported to CPS because the infant's urine drug test result was negative. A future study should address questions about these infants. How many eventually came to the attention of CPS because of other issues? How many suffered abuse or neglect in the meantime? Could this group serve as a control population in assessing the effectiveness of the reporting and rehabilitation process? Until such a study is performed, we are left with only the descriptive study presented here. However, this study does begin to answer the questions presented initially and serves to reduce the skepticism that had been voiced by practitioners who felt that nothing productive was being accomplished or that a more punitive approach would be the only effective route to helping the infants.