Abstract
Objective
to assess factors impacting child welfare involvement and child abuse and neglect outcomes among prenatally substance exposed infants.
Study design:
Retrospective review of case registry data regarding substance exposed infants tracked statewide in Delaware from 2014–2018. Differences in maternal, infant, and substance exposure factors by level of child welfare involvement (screened-in vs screened-out case status) and CAN outcomes were examined. Screened-in status was defined as case acceptance for investigation, family assessment, or treatment referral. Using logistic regression, associations between factors and screened-in status and between factors and CAN outcomes were assessed. Cases involving CAN were analyzed qualitatively.
Results:
Among 1222 SEI, 70% were screened-in by child welfare for ongoing involvement; 28 (2.3%) of SEI were identified as CAN victims sustaining serious physical or fatal injury before age 1. Most SEI remained with caregivers; few entered foster care. Poly-substance exposure and maternal mental health condition were factors associated with screened-in status. Neither substance exposure type nor maternal mental health condition reliably predicted future CAN.
Conclusions:
SEI had a significant risk for CAN. Although maternal and substance exposure factors were associated with screened-in case status, they unreliably predicted future risk of CAN.
Caregiver substance use negatively impacts families and increases the likelihood that infants and children experience child abuse or neglect (CAN) (1-10). More than 25% of substantiated CAN cases nationally involve caregiver substance concerns (11). Substance use may impair the ability to appropriately parent (2,12,13), impact child permanency (3,4,14), and result in removal and placement into foster care for nearly twice as long as children from families without substance use (15). Families affected by substance use often face significant co-occurring adversities, including psychological trauma, mental health conditions, economic insecurity and domestic violence that impact child safety and pose challenges to reunification and substance use disorder recovery (1,13,16-19).
Recent national trends in substance use reveal women of childbearing age are disproportionately affected, and growing numbers of infants are born prenatally substance exposed. Maternal substance use has been identified as a key factor predictive of increased reports to child welfare agencies (4,6,7,13,20) and infants specifically prenatally substance-exposed may be at higher risk for child maltreatment than non-substance exposed peers (4-7,21-24). Although this suggests prenatal exposure confers unique vulnerability to CAN that requires enhanced child welfare involvement and protection, ensuring safety for substance exposed infants (SEI) poses complex challenges. Likelihood and level of involvement with child welfare is highly variable (4) across jurisdictions despite federal mandates through the Child Abuse Prevention and Treatment Act (CAPTA) that state child welfare agencies comprehensively address needs of SEI and their families (25).. Protocols for maternal/infant toxicology testing and drug screening, treatment programming, child welfare hotline notification, and criminal prosecution (1,4,14,26,27) lack uniformity.
State-specific data from Delaware (DE) indicate that SEI were disproportionately represented among CAN victims in 2014, including abuse fatalities. As a result, development and implementation of a case registry system to track SEI-related safety outcomes specifically was undertaken through an innovative partnership between the statewide child welfare system and Office of the Investigation Coordinator, an oversight body for child safety investigations. The primary aim of the case registry system was to inform child welfare policy around SEI at the state level by analyzing data trends and identifying risk and protective factors associated with CAN outcomes. Using this state case registry system, we aimed to describe the maternal, infant, and substance-exposure factors associated with level and type of involvement by the state child welfare agency and describe maternal, infant, and substance-exposure factors associated with CAN outcomes—specifically, serious physical and fatal injury—among SEI previously reported to the child welfare agency at birth.
Methods
For this study, we retrospectively reviewed hotline notifications received by DE’s statewide child welfare agency and case-specific data entered into the statewide case registry regarding SEI born between 11/12/2014 and 3/27/2018. The case registry included SEI birth notifications and any subsequent hotline reports for CAN, serious physical injury, or fatalities involving SEI within the first 12 months after birth/registry entry. Infants were tracked specifically throughout the first year of life because rates of physical abuse are highest in this age group.
Use of an evidence-based, Structured Decision Making (SDM) (28) algorithm by child welfare services in DE had previously been implemented in 2013 to provide clearly defined, consistent criteria for screening hotline reports, determining priority response for screened-in reports, identifying immediate harm, and estimating the risk of future CAN at the time of initial child welfare notification. Levels of child welfare involvement were categorized as “screened-in” if child welfare accepted the case for investigation, family assessment, and/or treatment after conducting a risk assessment or “screened-out” if nature of the concern failed to meet accepted threshold definitional criteria in use at the time of the study (28) (Figure 1 and Figure 2; available at www.jpeds.com). Cases were screened in if the following criteria were met: the infant or mother had positive toxicology screening at birth, the infant was symptomatic/substance affected, the mother had a positive toxicology within 60 days prior to birth, or the mother disclosed using substances within 60 days prior to SEI birth (Figure 1). Cases were screened out if the identified concerns were linked to a duplicate report or active child welfare case, if insufficient information existed to locate the family, or if the SEI was prenatally exposed to a prescription medication used as prescribed and there were no other risk factors present (Figure 2). Notably, structured decision-making algorithms in use by child welfare hotline personnel at the time of this study did not risk-stratify hotline reports by substance exposure type (ie, reports involving cocaine were not considered higher risk than marijuana exposures), and identified parental risk factors were not weighted cumulatively (although presence of multiple risk factors could change the timing/priority of response).
Figure 1.
Text. Sample safety assessment tool (Sections 1 and 2) in use by Delaware Child Protective Services at the time of study completion
Figure 2.
Text. Sample safety assessment tool (Sections 3 and 4) in use by Delaware Child Protective Services at the time of study completion
Substance exposure was determined prior to the hotline report either through results of maternal verbal drug screening or toxicology performance, infant toxicology performance, or identification of infant symptoms consistent with prenatal substance exposure at/immediately after birth.
Variables abstracted from the registry included infant date of birth, sex, medical condition (defined as prematurity versus other), CAN or fatality outcome, age at injury, and type of substance exposure. Abstracted variables regarding type of child welfare involvement included visiting home nursing referrals, foster care placements, and safety plan implementation. Safety plan was defined as an agreement by child welfare and the SEI caregiver that a supervising party would be required to monitor their caregiving activities for a specified period due to safety concerns. Substance exposure type was categorized as marijuana, non-medication assisted treatment opiates (ie, excluding methadone and suboxone/subutex), methadone, suboxone/subutex, cocaine, benzodiazepines, amphetamines, phencyclidine, barbiturates, and alcohol. Infants may have been exposed to one (single substance) or more than one (poly-substance) exposure type. Maternal variables included race, number of prior SEI births, mental health condition, personal childhood child-welfare history, and prior child-welfare case substantiation of CAN.
Statistical Analyses
The primary outcome of interest was screened-in status; the secondary outcome was CAN, defined as either serious physical injury or fatality. Data were analyzed using statistical software R, version 3.4.4. Descriptive statistics were computed for all maternal and infant factors using frequencies and proportions for categorical variables and medians and interquartile range (IQR) for continuous variables. Differences in maternal, infant, and substance-exposure factors by screened-in versus screened-out status (ie, level of child welfare involvement) and CAN were examined using chi-square or Fisher exact test, as appropriate, for categorical variables and Mann-Whitney U test for continuous variables. Univariable and multivariable logistic regression models were used to examine the associations between these factors and screened-in versus screened-out status and between these factors and CAN. Cases involving CAN also were analyzed qualitatively for trends.
Ethical Considerations
The hospital institutional review board approved the study protocol.
Results
The case registry included 1222 unique infants born between 11/12/2014 and 3/27/2018 who were subjects of SEI hotline notifications to child welfare (Table I). Of these, 70.0% of infants were screened in for investigation and/or treatment. 61.6% were exposed to a single substance. Visiting nurse referrals were arranged for 65.7% of infants who were screened in for investigation and/or treatment, and safety plans were instituted for 33.7% screened in.
Table 1.
Maternal and Infant Characteristics
| Infant Characteristics | |
|---|---|
| Sex | n (%) |
| Female | 582 (47.6) |
| Male | 640 (52.4) |
| Unknown/Missing | 0 |
| Medical Condition | |
| Yes | 79 (6.5) |
| No | 873 (71.4) |
| Unknown/Missing | 270 (22.1) |
| Medical Condition Type | |
| Prematurity | 53 (4.3) |
| Others | 28 (2.3) |
| Unknown/Missing | 1141 (93.4) |
| Reports Screened In For Investigation/Treatment by Child Welfare | 855 (70.0) |
| Reports Screened Out by Child Welfare | 367 (30.0) |
| Infants with Single Substance Exposure | 753 (61.6) |
| Infants with Poly-Substance Exposure | 469 (38.4) |
| Serious Physical Injuries or Fatalities | |
| Yes | 28 (2.3) |
| No | 1192 (97.5) |
| Unknown/Missing | 2 (0.2) |
| Infant Placement in Out of Home/Child Welfare Custody | |
| Yes | 210 (17.2) |
| No | 1010 (82.6) |
| Unknown/Missing | 2 (0.2) |
| Infant Visiting Nurse Referrals | |
| Yes | 803 (65.7) |
| No | 314 (25.7) |
| Unknown/Missing | 105 (8.6) |
| Infant Safety Supervision Plan | |
| Yes | 412 (33.7) |
| No | 741 (60.6) |
| Unknown/Missing | 69 (5.7) |
| Maternal Characteristics | |
| Race | n (%) |
| Caucasian | 704 (57.6) |
| African American | 461 (37.7) |
| Hispanic | 39 (3.2) |
| Other | 5 (0.4) |
| Unknown/Missing | 13 (1.1) |
| Mental Health Condition | |
| Yes | 302 (24.7) |
| No | 644 (52.7) |
| Unknown/Missing | 276 (22.4) |
| Maternal Childhood Child Welfare Involvement | |
| Yes | 350 (28.6) |
| No | 583 (47.8) |
| Unknown/Missing | 289 (23.6) |
| Maternal Prior Prenatally Substance Exposed Infant Delivery | |
| Yes | 296 (24.2) |
| No | 920 (75.3) |
| Unknown/Missing | 6 (0.5) |
| Maternal Prenatally Substance Exposed Infant Delivery | |
| 1 Prior Infant Delivery | 232 (19.0) |
| 2 Prior Infant Deliveries | 48 (3.9) |
| 3 | 9 (0.7) |
| 4 | 4 (0.3) |
| 5 | 1 (0.1) |
| 6 | 1 (0.1) |
| 7 | 1 (0.1) |
| No prior SEI deliveries | 920 (75.3) |
| Unknown/Missing | 6 (0.5) |
| Maternal Prior Substantiation for Abuse/Neglect | |
| Yes | 59 (4.8) |
| No | 544 (44.5) |
| Unknown/Missing | 619 (50.7) |
For 56.1% of infants, maternal race was Caucasian and 36.6%, African American. For 24.7% of infants, mothers had mental health diagnoses, and for 28.6%, mothers had personal prior child welfare involvement in DE during their childhood. Greater than 75% of mothers had not delivered a previous SEI, although 232 (19%) had delivered one prior SEI, 48 (3.9%) had delivered two, and 16 (1.3%) had delivered three or more. Although only 4.8% of infants had mothers with prior substantiations for CAN, one-half of infants had unknown/missing maternal information around prior or previous abuse substantiations.
The four most common categories of substance exposure were marijuana, non-medication assisted treatment opiates, methadone, and cocaine (Table 2). Most infants exposed to marijuana (68.9%) were not exposed to additional substances.
Table 2.
Categories of Substance Exposure
| N | % | |
|---|---|---|
| Marijuana | 627 | |
| Marijuana only | 432 | 68.89 |
| Marijuana plus 1 other drug | 119 | 18.99 |
| Marijuana plus 2 other drugs | 76 | 12.12 |
| Opiates (Excluding Medication Assisted Treatment) | 443 | |
| Opiate only | 128 | 28.89 |
| Opiate plus 1 other drug | 193 | 43.57 |
| Opiate plus 2 other drugs | 122 | 27.54 |
| Opiate Medication Assisted Treatment: Methadone | 312 | |
| Methadone only | 86 | 27.56 |
| Methadone plus 1 other drug | 126 | 40.38 |
| Methadone plus 2 other drugs | 100 | 32.06 |
| Cocaine | 206 | |
| Cocaine only | 40 | 19.42 |
| Cocaine plus 1 other drug | 166 | 80.58 |
| Cocaine plus 2 other drugs | 0 | 0 |
| Benzodiazepine | 78 | |
| Benzodiazepine only | 13 | 16.67 |
| Benzodiazepine plus 1 other drug | 28 | 35.90 |
| Benzodiazepine plus 2 other drugs | 37 | 47.43 |
| Opiate Medication Assisted Treatment: Subutex/Suboxone | 70 | |
| Subutex/suboxone only | 15 | 21.42 |
| Subutex/suboxone plus 1 other drug | 36 | 51.43 |
| Subutex/suboxone plus 2 other drugs | 19 | 27.15 |
| Amphetamines | 55 | |
| Amphetamines only | 9 | 16.36 |
| Amphetamines plus 1 other drug | 25 | 45.45 |
| Amphetamines plus 2 other drugs | 21 | 38.19 |
| Phencyclidine (PCP) | 24 | |
| Phencyclidine only | 9 | 37.5 |
| Phencyclidine plus 1 other drug | 5 | 20.83 |
| Phencyclidine plus 2 other drugs | 10 | 41.67 |
| Barbiturate | 11 | |
| Barbiturate only | 3 | 27.27 |
| Barbiturate plus 1 other drug | 2 | 18.18 |
| Barbiturate plus 2 other drugs | 6 | 54.55 |
| Alcohol | 11 | |
| Alcohol only | 0 | 0 |
| Alcohol plus 1 other drug | 3 | 27.27 |
| Alcohol plus 2 other drugs | 8 | 72.73 |
Characteristics Associated with Case Disposition and CAN
Distribution of maternal and infant factors by screened-in versus screened-out status (level of child welfare involvement) and CAN outcomes is included in Table 3. In univariate analysis, poly-substance exposure, prescription drug misuse, maternal mental health conditions, and prior maternal substantiation for CAN were all associated with screened-in status; race, prior SEI birth, and maternal personal childhood child-welfare history were not (Table 4). In multivariable analysis, only poly-substance abuse and maternal mental health conditions were associated with screened-in status (Table 5). None of the examined variables were statistically associated with CAN in either univariate or multivariate analysis.
Table 3:
Distribution of Maternal and Infant Factors by Screen-In, Screen-Out, No Fatality or Abuse, and Fatality or Abuse
| Variables | Level | Screen-Out | Screen-In | P | No Fatality or Abuse | Fatality or Abuse | P |
|---|---|---|---|---|---|---|---|
| Total n | 367 | 855 | 1194 | 28 | |||
| Infant Characteristics | |||||||
| Sex (%) | Female | 182 (49.7) | 399 (46.7) | 0.358 | 570 (47.9) | 11 (39.3) | 0.48 |
| Male | 184 (50.3) | 456 (53.3) | 621 (52.1) | 17 (60.7) | |||
| Medical Condition/Prematurity (%) | No | 273 (93.2) | 600 (91.0) | 0.332 | 847 (91.9) | 25 (89.3) | 0.495 |
| Yes | 20 (6.8) | 59 (9.0) | 75 (8.1) | 3 (10.7) | |||
| Infants with Poly-substance Exposure (%) | No | 300 (81.7) | 453 (53.0) | <0.001 | 732 (61.4) | 19 (67.9) | 0.619 |
| Yes | 67 (18.3) | 402 (47.0) | 460 (38.6) | 9 (32.1) | |||
| Infant Safety Supervision Plan (%) | No | 359 (98.6) | 383 (48.5) | <0.001 | 723 (64.3) | 17 (63.0) | 0.99 |
| Yes | 5 (14) | 407 (51.5) | 402 (35.7) | 10 (37.0) | |||
| Infant Visiting Nurse Referrals (%) | No | 101 (28.3) | 214 (28.1) | 1 | 309 (28.2) | 6 (28.6) | 0.99 |
| Yes | 256 (71.7) | 547 (71.9) | 786 (71.8) | 15 (71.4) | |||
| Infant Placement in Out of Home/ Child Welfare Custody (%) | No | 345 (94.0) | 665 (78.0) | <0.001 | 993 (83.4) | 16 (57.1) | 0.001 |
| Yes | 22 (6.0) | 188 (22.0) | 197 (16.6) | 12 (42.9) | |||
| Maternal Characteristics | |||||||
| Mother Race (%) | Caucasian | 196 (53.4) | 508 (59.4) | 0.086 | 686 (57.6) | 18 (64.3) | 0.883 |
| African American | 148 (40.3) | 313 (36.6) | 449 (37.7) | 10 (35.7) | |||
| Hispanic | 14 (3.8) | 25 (2.9) | 39 (3.3) | 0 (0.0) | |||
| Other | 9 (2.5) | 9 (1.1) | 18 (1.5) | 0 (0.0) | |||
| Maternal Prior Prenatally Substance Exposed Infant Delivery (%) | No | 292 (79.6) | 628 (74.0) | 0.044 | 898 (75.7) | 20 (71.4) | 0.764 |
| Yes | 75 (20.4) | 221 (26.0) | 288 (24.3) | 8 (28.6) | |||
| Mother Mental Health Condition (%) | No | 312 (85.0) | 608 (71.1) | <0.001 | 901 (75.6) | 19 (67.9) | 0.473 |
| Yes | 55 (15.0) | 247 (28.9) | 291 (24.4) | 9 (32.1) | |||
| Maternal Childhood Child Welfare Involvement (%) | No | 268 (73.0) | 604 (70.6) | 0.438 | 854 (71.6) | 17 (60.7) | 0.292 |
| Yes | 99 (27.0) | 251 (29.4) | 338 (28.4) | 11 (39.3) | |||
| Maternal Prior Substantiation for Abuse/Neglect (%) | No | 359 (97.8) | 804 (94.0) | 0.007 | 1136 (95.3) | 26 (92.9) | 0.388 |
| Yes | 8 (2.2) | 51 (6.0) | 56 (4.7) | 2 (7.1) |
Notes: P generated by Fisher exact test
Table 4.
Univariable Logistic Regression to Determine the Association with Outcomes
| Screen-In by CPS | Child Abuse, Neglect, or Fatality Output |
|||
|---|---|---|---|---|
| OR (95% CI) | P | OR (95% CI) | P | |
| Maternal Race/Ethnicity | ||||
| Caucasian | Referent | Referent | ||
| AA | 0.82 (0.64, 1.06) | 0.13 | 0.76 (0.37, 1.59) | 0.472 |
| Hispanic | 0.7 (0.36, 1.37) | 0.296 | ||
| Others/missing | 0.39 (0.15, 1.00) | 0.05 | ||
| Single Substance Exposure | ||||
| No | Referent | Referent | ||
| Yes | 0.26 (0.19, 0.35) | <0.001 | 1.1 (0.54, 2.26) | 0.795 |
| Poly Substance Exposure | ||||
| No | Referent | Referent | ||
| Yes | 3.89 (2.9, 5.22) | <0.00 | 0.91 (0.44, 1.87) | 0.795 |
| Prescription Drug Use | ||||
| Appropriate Use | Referent | Referent | ||
| Misuse | 1.5 (1.17, 1.93) | 0.002 | 2.03 (0.87, 4.71) | 0.1 |
| Prior Substance Exposed Infant Delivery | ||||
| No | Referent | Referent | ||
| Yes | 1.32 (0.99, 1.77) | 0.063 | 1.56 (0.75, 3.25) | 0.237 |
| Maternal Mental Health Issue | ||||
| No | Referent | Referent | ||
| Yes | 2.28 (1.66, 3.14) | <0.001 | 1.35 (0.63, 2.86) | 0.44 |
| Maternal Childhood CPS History | ||||
| No | Referent | Referent | ||
| Yes | 1.12 (0.85, 1.46) | 0.431 | 1.86 (0.92, 3.76) | 0.083 |
| Maternal Prior Substantiation for Abuse/Neglect | ||||
| No | Referent | Referent | ||
| Yes | 2.88 (1.35, 6.12) | 0.006 | 1.31 (0.31, 5.61) | 0.717 |
Note: OR* = odds ratio of screen-in, OR# = odds ratio of fatality (yes). In multivariable model, the poly substance cannot be included as this variable is the complementary of single substance. Missing OR indicates that there was insufficient data to estimate OR for that group.
Table 5.
Multivariable Logistic Regression to Determine the Association with Outcomes
| Variable | Screen In by Child Welfare |
Child Abuse, Neglect or Fatality Outcome |
||
|---|---|---|---|---|
| OR (95% CI) | P | OR (95%) | P | |
| Maternal Race | ||||
| Caucasian | Referent | Referent | ||
| African American | 1.17 (0.89, 1.55) | 0.257 | 0.65 (0.3, 1.4) | 0.27 |
| Hispanic | 0.97 (0.48, 1.97) | 0.931 | ||
| Other/missing | 0.51 (0.19, 1.37) | 0.18 | ||
| Single Substance Exposure | ||||
| No | Referent | Referent | ||
| Yes | 0.26 (0.19, 0.36) | <0.001 | 1.43 (0.67, 3.06) | 0.361 |
| Prescription Drug Use | ||||
| Appropriate Use | Referent | Referent | ||
| Misuse | 1.12 (0.85, 1.49) | 0.417 | 1.77 (0.72, 4.34) | 0.21 |
| Maternal Prior Substance Exposed Infant Delivery | ||||
| No | Referent | Referent | ||
| Yes | 1.02 (0.74, 1.4) | 0.908 | 1.45 (0.68, 3.11) | 0.337 |
| Maternal Mental Health Condition | ||||
| No | Referent | Referent | ||
| Yes | 1.96 (1.38, 2.78) | <0.001 | 1.03 (0.47, 2.28) | 0.939 |
| Maternal Childhood CPS History | ||||
| No | Referent | Referent | ||
| Yes | 0.89 (0.66, 1.20) | 0.44 | 1.62 (0.78, 3.40) | 0.198 |
| Maternal Prior Substantiation for Abuse/Neglect | ||||
| No | Referent | Referent | ||
| Yes | 1.82 (0.83, 4.00) | 0.138 | 0.99 (0.22, 4.4) | 0.985 |
Note: OR* = odds ratio of screen-in, OR# = odds ratio of fatality (yes). In multivariable model, the poly substance cannot be included as this variable is the complementary of single substance. Missing OR indicates that there was insufficient data to estimate OR for that group.
Characteristics of Infant CAN Victims
In qualitative analysis, 28 SEI were victims of serious physical or fatal CAN injury during the study period (Table 6). Most infant victims of serious physical or fatal CAN injuries (66%) were exposed to a single substance. Most victimized infants were screened in (86%), but most did not have a safety plan instituted (61%). Mental health conditions were diagnosed among 32% of mothers of victimized infants. Eight of the 28 infants (28%) were exposed to marijuana; 10 of 28 (36%) were exposed to methadone, and five of 28 (18%) were exposed to cocaine. Eight of 28 (28%) sustained head trauma, 7 of 28 (25%) sustained fractures, and 2 of 28 (7%) sustained abusive cutaneous injury. Four of 28 (14%) suffered opiate toxicity, including 3 infants younger than 2 months and 1 older infant (age 9 months, whose developmental status may support exploratory ingestion). Six infants (21.43%) died related to bed-sharing or hazards in the sleep environment. Almost half (46%) of the 28 infants experienced CAN within the first seven weeks of life, and most (85%) were victimized by age 4 months.
Table 6.
Qualitative Case Review – Substance Exposed Infants with Child Abuse and Neglect Outcomes (n = 28)
| Single/Poly Substance Exposure |
Substances Identified | Screen In or Out |
Safety Plan Instituted |
Maternal Mental Health Condition |
Maternal Childhood CPS History |
Maternal Prior Substantiation for Abuse/Neglect |
Child Abuse/Neglect Type |
Age at Injury |
|---|---|---|---|---|---|---|---|---|
| Single Substance | Marijuana | In | Yes | No | No | No | homicide (drowning) | 3.5 months |
| Poly Substance | Opiates, benzos | In | No | No | Yes | Yes | homicide (blunt force trauma) | 7 months |
| Single Substance | Benzos | In | No | Yes | No | No | SUID (cause/manner undetermined) | 5 weeks |
| Poly Substance | Opiates, cocaine | In | No | Yes | No | No | SUID (cause/manner undetermined) | 2.5 months |
| Single Substance | Methadone | In | No | Yes | Yes | No | Bed-sharing/unsafe sleep | 4 weeks |
| Single Substance | Opiate | Out | No | No | No | Yes | Bed-sharing/unsafe sleep | 2 months |
| Poly Substance | Cocaine, Marijuana | In | Yes | No | Yes | No | Bed-sharing/unsafe sleep | 4 months |
| Poly Substance | Methadone, cocaine, benzos | In | No | No | No | No | Bed-sharing/unsafe sleep | 2 weeks |
| Poly Substance | Methadone, Opiates, Benzos | Out | Yes | Yes | No | No | Bed-sharing/unsafe sleep | 3 weeks |
| Single Substance | Methadone | In | No | No | No | No | Bed-sharing/unsafe sleep | 3 weeks |
| Single Substance | Methadone | In | No | Yes | No | No | head trauma | 7 weeks |
| Single Substance | Methadone | In | No | No | Yes | No | head trauma | 2 months |
| Single Substance | Marijuana | In | No | Yes | No | No | head trauma | 7 months |
| Poly Substance | Opiates, cocaine, methadone | In | Yes | No | No | No | head trauma | 5.5 months |
| Single Substance | Methadone | Out | No | No | Yes | No | head trauma | 4 months |
| Single Substance | Opiates | In | No | No | No | No | head trauma, fractures | 2.5 months |
| Single Substance | Marijuana | In | Yes | Yes | Yes | No | head trauma, fractures | 5 weeks |
| Poly Substance | Methadone, opiates, cocaine, benzos, marijuana | In | Yes | Yes | No | No | head trauma, fractures | 3 months |
| Single Substance | Marijuana | In | No | No | No | No | fractures | 7 weeks |
| Single Substance | Marijuana | In | Yes | No | Yes | No | fractures | 3 weeks |
| Single Substance | Marijuana | Out | No | No | Yes | No | fractures | 3.5 months |
| Poly Substance | Opiates, subutex | In | Yes | No | Yes | No | fractures | 4 weeks |
| Single Substance | Methadone | In | Yes | Yes | No | No | bruising, bite | 7 weeks |
| Poly Substance | Methadone, suboxone | In | Yes | No | No | No | burns | 2 months |
| Single Substance | Opiates | In | No | No | Yes | No | opiate toxicity* | 2 weeks |
| Single Substance | Opiates | In | No | No | Yes | No | opiate toxicity* | 9 months |
| Single Substance | Opiates | In | Yes | No | No | No | opiate toxicity* | 2 months |
| Single | Opiates | In | No | No | No | No | opiate toxicity* | 1 week |
opiate toxicity implies ingestion, poisoning, or other administration
Discussion
How best to ensure the safety of children in families affected by substance exposure and level of support and involvement by child welfare agencies has become a critical, but controversial public health issue and national policy priority. Statistics for CAN in DE alone from 2014–2018 indicate approximately one-third of serious or fatal CAN involved SEI; in 2017 SEI notifications were received for 4.1% of all births statewide, but SEI were associated with 30% (4 of 13) and 27% (8 of 30) of CAN-associated deaths and serious injuries respectively that year in the state (29,30). This suggests SEI are at uniquely heightened risk for poor safety outcomes and that enhanced involvement, scrutiny, and support by child welfare agencies may potentially benefit and protect this vulnerable population. Of the 28 SEI in our study who experienced serious or fatal CAN injury, 85% were injured by age 4 months, suggesting very young infants were at highest risk.
Maternal substance use has been identified as a key factor predictive of increased reports to child welfare agencies (4,6,7,13,20), and previous studies suggest prenatal exposure confers unique vulnerability to CAN (4-7,21-24). Our study supports this hypothesis, and concludes that identification of this risk factor warrants universal, supportive, child welfare-based protective intervention to reduce risk of infant harm. Specifically, study results suggest that maternal and substance-exposure factors influence child-welfare involvement (screened-in status), but are not universally predictive of future risk of harm to SEI during the first year of life. Risk to infant safety is likely multifactorial, and evaluation of maternal and substance exposure factors alone may provide an inappropriate basis for decision-making around level and type of child-welfare interventions for SEI. Infants exposed to any substance are at risk for CAN; as evidenced by this study, no single substance-exposure type was considered protective against future CAN, and CAN outcomes occurred among infants with heterogeneous, variable substance-exposure types.
At the time of this study, SEI cases reported to the state child-welfare agency had only two levels of involvement: screened in for investigation/treatment or screened out. The majority of cases reported to the hotline were screened in for ongoing services and monitoring, which may explain why most SEI did not experience CAN or fatal outcomes. Notably, in contrast to data indicating caregiver substance use impacts permanency (3) and recent data indicating disproportionate placement of infants into foster care due to caregiver substance use concerns (31), the majority of SEI in this study (82.6%) were not placed into kinship care or foster care. Safety plan implementation and visiting nurse referrals were likely appropriate and protective interventions for the majority of infants, suggesting that this level and type of involvement by child welfare agencies with families affected by substance use were effective strategies to mitigate risk of CAN.
Cases involving maternal poly-substance exposure and mental health issues were highly associated with screened-in level of child welfare involvement. Poly-substance use complicates treatment and recovery for the mother (2,13), and previous studies have well-established that caregivers with substance use are likely to have co-occurring mental health issues threatening child permanency and warranting more substantial supportive service delivery, policy development and intervention by child welfare (13,32-36). The fact that these factors were associated with increased child welfare involvement suggests they were recognized appropriately as substantial impediments to family stability. Supportive child welfare-based services were prioritized to address co-occurring adversities in these families.
Study results support that substance exposure type, including single or poly-substance use, however, does not inform future CAN risk. Eight of 28 cases were associated with marijuana exposure (which is legal in many jurisdictions and culturally perceived as low risk with minimal negative health and safety outcomes), whereas only five cases involved cocaine. Serious CAN and CAN fatalities similarly occurred in infants exposed to methadone; 10 of 28 cases were associated with methadone, and 6 cases involved single-substance exposure to methadone only. Methadone is commonly prescribed as medication-assisted treatment for opioid dependence among mothers stably in recovery from substance-use disorders, but it is perceived to be high risk for abuse and diversion (2,3). These results support the hypothesis that use of any substance can put children at risk for CAN (3), as substance use generally has been associated with risk-taking behaviors in the user (2) that negatively impact child safety. Attempts to risk-stratify substances to confer a safety profile have hinged primarily upon scientific understanding of physiologic dependence, risks of addiction potential, or behavioral effects in the user (37,38). Though medically and psychiatrically relevant, such criterion may be inappropriate in characterizing the safety of a substance when creating child-welfare policy to prevent maltreatment outcomes as it fails to consider the broader impact of any substance use on caregiver risk-taking behaviors.
Data on substance-specific influence to CAN risk are limited; although, in one study, methamphetamine use evidenced greatest risk for CAN (39). Marijuana-specific risk posed to SEI remains unclear, but notably cases existed in our state registry with serious physical or fatal injury outcomes wherein sole substance of exposure was marijuana. Algorithmic decision-making at the hotline level based primarily on substance-exposure type, without additional involvement or monitoring by child welfare and without additional caregiver, family and environmental psychosocial risk factor screening, may presume safety and miscalculate future risk of harm to the child. Had protocols been in place in DE automating the screening out of cases based on substance type, such as marijuana, it is likely infants at CAN risk would have been missed.
Prior studies indicate families affected by substance exposure have received differing child-welfare agency resources based on perceived risk of the substance exposure involved. Agencies have historically provided more resources and supportive services to individuals who use illicit or non-prescribed substances other than marijuana and alcohol (40), extending enhanced protections to these infants traditionally perceived by child-welfare agencies to be at higher risk for CAN. Child welfare agencies have also historically utilized type of parental substance use to guide child placement and removal recommendations (41); however, this could underestimate risk to children whose caregivers are perceived arbitrarily as using “lower-risk” substances. No single substance type was more protective against CAN versus another in our study cohort, contraindicating this pattern of historical behavior by child welfare and supporting instead that substance-exposure type alone should not be a criterion impacting the degree and level of child-welfare involvement and provision of supportive services. Instead, study results suggest a universal, supportive services child welfare-based approach, irrespective of substance exposure type, may be most effective to mitigate future risk of harm.
Likelihood and level of involvement with SEI by child-welfare agencies generally varies by jurisdiction despite recent amendments to federal CAPTA and Comprehensive Addiction and Recovery Act 2016 legislation requiring states implement policies, known as Plans of Safe Care (POSC), to address needs of SEI and their families. Plans of Safe Care include an in-depth inventory of maternal substance use that does not discriminate based on identified substance-exposure type. A statutory change in language removed the term “illegal” from the types of substances warranting POSC development, broadening the potential exposure types amenable to supportive service planning. A comprehensive assessment of SEI and family needs identifies which services and supports should be included in individualized program development. Agency notifications regarding SEI are increasing exponentially, and states are struggling to comply with CAPTA requirements and align systems and stakeholders with differing perspectives through community partnerships. The data in this study were collected prior to implementation of POSC supportive services for SEI and their families in DE, where POSC was disseminated statewide late 2018.
As one of the first states in the nation to comply with current federal mandates, DE is leading research efforts to understand the impact of this universal, family-centered programming on rates of CAN among SEI. Our case registry currently tracks the impact of POSC programming on CAN outcomes; we are particularly interested in identifying which types of POSC supportive programming may be associated with reductions in CAN outcomes. Data from this current study indicate that over 80% of SEI with CAN were actually screened in, indicating that child welfare was appropriately discerning infants with safety risks; although, type of involvement was ultimately inadequately protective against CAN. Tracking specific supportive service programmatic delivery to SEI through our state case registry is underway and may aid future prevention efforts against CAN among SEI. Notably, POSC implementation statewide has also resulted in practice change by DE’s child welfare agency. Although positive maternal/infant toxicology studies generally result in screened-in hotline reports in almost every situation, currently those reports involving marijuana only or mothers adherent to medication assisted treatment where no other risk factors are present are contracted to providers external from the statewide child welfare agency. How these changes have impacted CAN outcomes in DE is also under study.
A universal maternal drug screening protocol existed in Delaware at the time of this study, but universal screening did not necessarily indicate universal toxicology performance, nor universal notification of child-welfare agencies around SEI birth if clinically suspected or confirmed by screening or toxicology. The decision to report SEI birth to child welfare was variable across the state, and our study population only assessed those infants reported to child welfare. Therefore it is possible this underestimates the true proportion of SEI in the state or that included SEI differ from non-reported infants by undetermined maternal, psychosocial, or substance exposure characteristics that skew data interpretation. Factors influencing health care provider decision to report SEI to child welfare was not specifically assessed, but study results do not support an association with maternal race and likelihood of screened-in status. Previous literature suggests minority infants with prenatal substance exposure are no more likely than white infants to be reported to child welfare (27), but CAN outcomes for SEI based on race and substance exposure remain understudied.
Secondly, the small sample size of CAN-victimized infants limited our ability to identify characteristics predictive of abuse. It is possible that results of our qualitative case review are not nationally representative or generalizable and instead reflect nuances specific to SEI from DE.
Thirdly, we cannot estimate the magnitude of the effect of child welfare involvement or safety plans on subsequent CAN victimization in our study. Seventy percent of the infants were screened in for services, and about half of those screened in had safety plans instituted as part of child-welfare involvement. To what degree child welfare interventions and safety plans prevented CAN in these cases is not measureable. Additionally, the effect of specific interventions (safety plan versus home nursing versus other supportive services) requires further study, and impact of results is unknown.
Finally, our study does not provide any estimate of risk of abuse for SEI beyond the first year of life. Prospective data collection may provide information on continued level of risk and factors predictive of future child-welfare involvement.
Further study is needed to better understand the impact of level and type of supportive services by child welfare on CAN outcomes for SEI.
Acknowledgement:
The authors would like to acknowledge John Leventhal, Sam Stubblefield, Matthew Di Guglielmo, and Mia Papas for their recommendations and feedback regarding this study.
Supported by the Centers for Medicare and Medicaid Services Innovation Award (CMS331027-01) and National Institute of Health Institutional Development Award from the National Institute of General Medical Sciences (U54-GM104941). An inter-agency Memorandum of Agreement was established for purposes of this project; signatories included the Delaware Department of Services for Children, Youth and Families, Office of the Investigation Coordinator, and Nemours/Alfred I. duPont Hospital for Children. The authors declare no conflicts of interest.
Abbreviations:
- DE
Delaware
- CAPTA
Child Abuse Prevention and Treatment Act
- POSC
Plans of Safe Care
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
References
- 1.Ghertner R, Baldwin M, Crouse G, Radel L, Waters A. The relationship between substance use indicators and child welfare caseloads. ASPE Research Brief, Office of the Assistant Secretary for Planning and Evaluation, U.S. Department of Health and Human Services. 2018. [Google Scholar]
- 2.Radel L, Baldwin M, Crouse G, Ghertner R, Waters A. Substance use, the opioid epidemic, and the child welfare system: key findings from a mixed methods study. ASPE Research Brief, Office of the Assistant Secretary for Planning and Evaluation, U.S. Department of Health and Human Services. 2018. [Google Scholar]
- 3.Radel L, Baldwin M, Crouse G, Gherter R, Waters A. Medication-assisted treatment for opioid use disorder in the child welfare context: challenges and opportunities. ASPE Research Brief, Office of the Assistant Secretary for Planning and Evaluation, U.S. Department of Health and Human Services. 2018. [Google Scholar]
- 4.Prindle JJ, Hammond I, Putnam-Hornstein E. Prenatal substance exposure diagnosed at birth and infant involvement with child protective services. Child Abuse Negl 2018; 76:75–83. [DOI] [PubMed] [Google Scholar]
- 5.Puls HT, Anderst JD, Bettenhausen JL, Clark N, Krager M, Markham JL, et al. Newborn risk factors for subsequent physical abuse hospitalizations. Pediatrics 2019; 143:e20182108. [DOI] [PubMed] [Google Scholar]
- 6.Smith BD, Test MF. The risk of subsequent maltreatment allegations in families with substance-exposed infants. Child Abuse Negl 2002; 26:97–114. [DOI] [PubMed] [Google Scholar]
- 7.Leventhal JM, Forsyth BW, Qi K, Johnson L, Schroeder D, Votto N. Maltreatment of children born to women who used cocaine during pregnancy: a population-based study. Pediatrics 1997; 100. [DOI] [PubMed] [Google Scholar]
- 8.Howell S, Bailey L, Coffman J. Evaluation of drug-endangered children: The yield of toxicology and skeletal survey screening. Child Abuse Negl 2019; 96:104081. [DOI] [PubMed] [Google Scholar]
- 9.Crowley D, Connell C, Jones D, Donovan M. Considering the Child Welfare System Burden From Opioid Misuse: Research Priorities for Estimating Public Costs. Am J Manag Care 2019; 25:S256–S263. [PMC free article] [PubMed] [Google Scholar]
- 10.Walsh C, MacMillan HL, Jamieson E. The relationship between parental substance abuse and child maltreatment: findings from the Ontario Health Supplement. Child Abuse Negl 2003; 27:1409–1425. [DOI] [PubMed] [Google Scholar]
- 11.US Department of Health and Human Services, Administration for Children and Families, Administration on Children, Youth and Families and Children’s Bureau. Child Maltreatment. 2015. Available at: https://www.acf.hhs.gov/cb/resource/child-maltreatment-2015. Accessed July 30, 2019. [Google Scholar]
- 12.Fuller TL, Wells SJ. Predicting maltreatment recurrence among CPS cases with alcohol and other drug involvement. Child Youth Serv Rev 2003; 25:553–569. [Google Scholar]
- 13.Hammond I, Eastman AL, Leventhal JM, Putnam-Hornstein E. Maternal mental health disorders and reports to child protective services: a birth cohort study. Int J Environ Res Public Health 2017; 14. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Rebbe R, Mienko J. Brown E, Rowhani-Rahbar A. Child protection reports and removals of infants diagnosed with prenatal substance exposure. Child Abuse Negl 2019; 88:28–36. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Brook J, McDonald TP, Gregoire T, Press A, Hindman B. Parental substance abuse and family reunification. J Soc Work Pract Addict 2010; 10:393–412. [Google Scholar]
- 16.Canfield M, Radcliffe P, Marlow S, Boreham M, Gilchrist G. Maternal substance use and child protection: a rapid evidence assessment of factors associated with loss of child care. Child Abuse Negl 2017; 70:11–27. [DOI] [PubMed] [Google Scholar]
- 17.Choenni V, Hammink A, van de Mheen D. Association between substance use and the perpetration of family violence in industrialized countries: a systematic review. Trauma Violence Abuse 2017; 18:37–50. [DOI] [PubMed] [Google Scholar]
- 18.O'Shea MT, Collins C, Ailis NR, Daly MD. Domestic Violence During Pregnancy. GP Survey Report, 2016. [Google Scholar]
- 19.Burke KD. Substance-exposed newborns: hospital and child protection responses. Child Youth Serv Rev 2007; 29:1503–19. [Google Scholar]
- 20.Dubowitz H, Kim J, Black MM, Weisbert C, Semiatin J, Magder LS. Identifying children at high risk for a child maltreatment report. Child Abuse Negl 2011; 35:96–104. [DOI] [PubMed] [Google Scholar]
- 21.Chaffin M, Kelleher K, Hollenberg J. Onset of physical abuse and neglect: psychiatric, substance abuse, and social risk factors from prospective community data. Child Abuse Negl 1996; 20:191–203. [DOI] [PubMed] [Google Scholar]
- 22.Jaudes PK, Ekwo E, Van Voorhis J. Association of drug abuse and child abuse. Child Abuse Negl 1995; 19:1065–75. [DOI] [PubMed] [Google Scholar]
- 23.Murphy JM, Jellinek M, Quinn D, Smith G, Poitrast FG, Goshko M. Substance abuse and serious child maltreatment: prevalence, risk and outcome in a court sample. Child Abuse Negl 1991; 15:197–211. [DOI] [PubMed] [Google Scholar]
- 24.Uebel H, Wright IM, Burns L, Hilder L, Bajuk B, Breen C, et al. Reasons for rehospitalization in children who had neonatal abstinence syndrome. Pediatrics 2015; 136:e811–20. [DOI] [PubMed] [Google Scholar]
- 25.Child Welfare Information Gateway. Parental Drug Use as Child Abuse. Washington, DC: U.S. Department of Health and Human Services, Children’s Bureau. Available at: https://www.childwelfare.gov/pubPDFs/drugexposed.pdf#page=2&view=Prenataldrugexposure. Accessed December 10, 2018. [Google Scholar]
- 26.Rebbe R, Mienko J, Brown E, Rowhani-Rabbar A. Hospital variation in child protection reports of substance exposed infants. J Pediatr 2019; 208:141–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Putnam-Hornstein E, Prindle JJ, Leventhal JM. Prenatal substance exposure and reporting of child maltreatment by race and ethnicity. Pediatrics 2016; 138:e20161273. [DOI] [PubMed] [Google Scholar]
- 28.National Council on Crime and Delinquency. Structured Decision Making® (SDM) Model for Child Protection. Available at: https://www.nccdglobal.org/assessment/structured-decision-making-sdm-model. Accessed September 9, 2019.
- 29.CPAC Reports and Publications, Review of CAN Panel Findings. Available at: https://courts.delaware.gov/childadvocate/cpac/cpac_reports.aspx Accessed 12/4/2019
- 30.Delaware Health Statistics Center. 2017. Birth Rate. Available at: https://www.dhss.delaware.gov/dhss/dph/hp/bthsdths_pubdata.html. Accessed 12/4/2019
- 31.The AFCARS Report: Preliminary FY 2017 Estimates as of August 10, 2018. No 25. U.S. Department of Health and Human Services, Administration for Children and Families, Administration on Children, Youth and Families. Children’s Bureau. Available at: https://www.acf.hhs.gov/cb. Accessed July 30, 2019. [Google Scholar]
- 32.Jarlenski M, Barry CL, Gollust S, Graves AJ, Kennedy-Hendricks A, Kozhimannil K. Polysubstance use among US women of reproductive age who use opioids for nonmedical reasons. Am J Public Health 2017; 107:1308–10. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Martins SS, Fenton MC, Keyes KM, Blanco C, Zhu H, Storr CL. Mood and anxiety disorders and their association with non-medical prescription opioid use and prescription opioid use disorder: longitudinal evidence from the National Epidemiologic Study on Alcohol and Related Conditions. Psychol Med 2012; 42:1261–72. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Wu LT, Woody GE, Yang C, Blazer DG. How do prescription opioid users differ from users of heroin or other drugs in psychopathology: results from the National Epidemiological Survey on Alcohol and Related Conditions. J Addict Med 2011; 5:28–35. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Grella CE, Needell B, Shi Y, Hser YI. Do drug treatment services predict reunification outcomes of mothers and their children in child welfare? J Subst Abuse Treat 2009; 36:278–93. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Park JM, Solomon P, Mandell DS. Involvement in the child welfare system among mothers with serious mental illness. Psychiatr Serv 2006; 57:493–7. [DOI] [PubMed] [Google Scholar]
- 37.Ammerman S, Ryan S, Adelman WP; Committee on Substance Abuse; the Committee on Adolescence. The impact of marijuana policies on youth: clinical, research, and legal update. Pediatrics 2015; 135:e769–85. [DOI] [PubMed] [Google Scholar]
- 38.Effective Arguments for advocates of taxing and regulating marijuana. Washington, DC: Marijuana policy project. 2014. Available at: www.mpp.org/assets/pdfs/library/effective-arguments-for-T-R.pdf. Accessed June 10, 2019. [Google Scholar]
- 39.Carlson BE, Williams LR, Shafer MS. Methamphetamine involved parents in the child welfare system: are they more challenging than other substance involved parents? J Public Child Welfare 2012; 6:280–95. [Google Scholar]
- 40.Pelton LH. An examination of reasons for child removal in Clark County Nevada. Child Youth Serv Rev 2008; 30:787–99. [Google Scholar]
- 41.Laslett AM, Room R, Dietze P, Ferris J. Alcohol’s involvement in recurrent child absue and neglect cases. Addiction 2012; 107:1786–93. [DOI] [PubMed] [Google Scholar]