How Many Families in Child Welfare Services Are Affected by Parental Substance Use Disorders? A Common Question that Remains Unanswered

Warning: The NCBI web site requires JavaScript to function. more...

Child Welfare. Author manuscript; available in PMC 2016 Jun 6.

Published in final edited form as:

Child Welfare. 2015; 94(4): 19–51.

PMCID: PMC4894838

NIHMSID: NIHMS790387

PMID: 26827475

How Many Families in Child Welfare Services Are Affected by Parental Substance Use Disorders? A Common Question that Remains Unanswered

Kristen Seay

Author information Copyright and License information Disclaimer

Abstract

Associated with extensive negative outcomes for children, parental substance use disorders are a major concern within the child welfare system. Obtaining actual prevalence rate data has been difficult, however, and there are no recent published reports on this issue. Using a systematic search, this paper examines: (1) Prevalence estimates of parental substance use disorders in the child welfare population; (2) the types of child welfare involvement for reported prevalence estimates; and (3) how prevalence information is being collected. Prevalence rates were found to have a wide range, from 3.9% to 79%, with regional prevalence estimates being higher than national estimates. Prevalence rates of parental substance use disorders varied by type of child welfare involvement of the family and method of data collection. This study points out the need for improvements in prevalence estimates in the United States and national data collection procedures to ensure that child welfare and substance abuse treatment systems are adequately responding to children and families with substance use disorders.

Parental substance use disorders are a major concern within the child welfare system (Hill, Tessner, & McDermott, 2011; Seay & Kohl, 2013; Seay & Kohl, 2015; Staton-Tindall, Sprang, Clark, Walker, & Craig, 2013). Yet, estimates of the prevalence of parental substance use disorders in the child welfare system have been cited to range from 5% to 90% (Jaudes, Ekwo, & Van Voorhis, 1995; Jones, 2004). Such a wide range indicates a lack of precision and consensus on this topic. Although prevalence is often cited as a range in the literature, little description is provided for the basis of these estimates. Researchers have frequently discussed the need for future research to examine how many clients involved with child welfare have substance use disorders (Jones, 2004; Semidei, Radel, & Nolan, 2001; Young, Boles, & Otero, 2007). To report on the current state of the field and to recommend policy changes to better track parental substance use disorders, this article examines and systematically compares the ten empirical studies that provide estimates of the prevalence rates. Within the article, child welfare indicates a continuum of agency involvement including investigations, in-home services, foster care, and adoption. Estimates are compared based on (a) whether they are from regional or national samples; (b) by the type of child welfare involvement (i.e., all child welfare reports, families receiving in-home services, and families with a child in foster care); and (c) by the source of data used to measure substance use.

Issues in the Measurement of Substance Use Disorders

The Statewide Automated Child Welfare Information System (SACWIS) does not require states to collect information that could be used to determine the percentage of families in the child welfare system that are affected by parental substance use, and the Adoption and Foster Care Analysis and Reporting System (AFCARS) treats these data as voluntary for states to submit in their reporting to Children's Bureau (Young et al., 2007).

Explanations for the wide variation in prevalence estimates have been proposed by a number of researchers. The type of child welfare involvement of the sample is one key distinction that may be contributing to inconsistencies in rates (Besinger, Garland, Litrownik, & Landsverk, 1999; Jones, 2004; Semidei et al., 2001; U.S. Department of Health and Human Services [USDHHS], 1999; Young et al., 2007). The prevalence rates of substance use disorders among families involved with child welfare has been proposed to increase across the continuum of services, from cases in which children are not removed (often referred to as in-home services), to more intensive services in which children are placed in protective custody in foster or kinship placements, to those for whom parental rights are terminated and children are adopted or in legal guardianships. The data collection method has also been proposed to impact prevalence estimates (Jones, 2004; USDHHS, 1999; Young et al., 2007). Researchers have used a wide array of data collection methods to assess for prevalence because no standard method or definitions exist. Case record reviews of case files may include examinations of risk assessment measures, case notes, court records, and other completed forms and documents. While in some studies workers are interviewed about their perceptions of parental substance use, other studies may use strict diagnostic guidelines for alcohol dependence and drug dependence. Additional explanations include inconsistencies across studies in definitions for substance use disorders (Besinger et al., 1999; Jones, 2004; Semidei et al., 2001; USDHHS, 1999; Young et al., 2007); whether or not both parents are assessed (Besinger et al., 1999; Jones, 2004); and in the onset, recency, or duration of the substance use disorder (Jones, 2004).

The following research questions will be answered in this study:

What are the prevalence rates being reported for parental substance use disorders in the child welfare population?
What are the types of child welfare involvement for the reported estimates?
What sources of data are used to measure substance use disorder prevalence?

Methods

This review located studies reporting prevalence rates of substance use disorders among samples of parents in the child welfare population. Prevalence is the total number or percentage of individuals in a given population that have a particular condition. Prevalence is distinct from incidence (i.e., the number of new cases with a particular condition in a given population), and is the focus of this review. In the study, the term parent is used to describe primary caregivers who may have been biological/step/foster/adoptive parents, custodial grandparents, or legal guardians providing long-term care of a child.

A detailed literature search was conducted to identify journal articles or government documents discussing data collection of prevalence estimates of substance use disorders among parents involved with child welfare in the United States. In the initial search for published journal articles, a total of six databases were searched: Academic Search Premier, CINAHL Plus, MEDLINE, PsycINFO, Social Work Abstracts, and SocINDEX. The terms used in this search were substance abuse, prevalence, and child welfare, as well as synonyms of these terms. The search was not limited by publication year due to the small number of studies meeting inclusion criteria. This initial search resulted in 239 references published between 1983 and 2004. Titles, abstracts, and the text of these articles were reviewed to determine if they reported prevalence rates of parental substance use in the child welfare system. In articles in which prevalence rates were cited from another article, citations were used to find the original article or government report listed in the 239 articles.

Government reports are publically available documents that were created for submission to committees of the U.S. House of Representatives or U.S. Senate. Citations listing information received through personal communication or studies lacking basic details on how the data were collected were excluded from the study. A small number of highly cited articles that reported prevalence rates based on anecdotal information (e.g., Judge Doe stated that over half of his cases involve substance abuse) were excluded from the study. Three studies that provided prevalence estimates were excluded from this analysis due to extremely specific samples (e.g., all children were survivors of parent-perpetrated sexual abuse) not widely generalizable to the child welfare population. This search resulted in the inclusion of nine studies (see Table 1 on page 36) measuring the prevalence rates of substance use disorders among parents in the child welfare population.

Table 1

Prevalence Data Collection Articles

Article	Data From; Location	Prevalence Estimate	Sample size	Sample Characteristics	Type of Child Welfare Involvement	Substance Abuse Definition	Degree to which substance abuse is a factor in the case decisions	Who is the parent?	How information was collected
Gibbons et al., unpublished	1999–2001 National NSCAW I	9.6%	N = 4073 Only one child per household	Children: 17% 0–2 21% 3–5 37% 6–10 24% 11–14 47% White& Non-Hispanic 27% Black & Non-Hispanic 19% Hispanic 7% Other 50% male Parents: 64% <35 yr 28% 35–44 7% 45–54 2% 55+ 90% female 10% male	Cases reported to child welfare Reported to child welfare and child remained in the home	Combining two questions on the risk assessment, child welfare caseworker stated yes that caregiver had an active alcohol and/or drug abuse problem at the time of the investigation	Child welfare caseworker reported he/she perceived caregiver active alcohol and/or drug abuse—Likely impacted case decisions	Primary or secondary caregiver who was living in the home of the child at the time of the report	National probability study; child welfare caseworkers interviewed
	1999–2001 National NSCAW I	3.9%	N = 4073	Same as above	Same as above	By self-report, met criteria for alcohol or drug dependence based on Composite International Diagnostic Interview Short Form (CIDI-SF; Kessler, Andrews, Mroczek, Ustun, & Wittchen, 1998)	Unknown if alcohol and/or drug use were even recognized in the case	Same as above	National probability sample, Primary caregiver completed CIDI-SF about their alcohol and drug use using audio computer-assisted self-interview technology
Sedlak et al., 2010	2005–2006 National NIS-4	11.4%	N = 29,488 children	Children investigated by child welfare, reported but screened out, or not reported to child welfare but recognized as maltreated by community professionals Not reported separately for children meeting Harm Standard	Combined sample of maltreated children, some reported to child welfare Children (0–17) investigated by child welfare, children reported but screened out, and children not reported to child welfare who were recognized as maltreated by community professionals. All children meet Harm Standard	ALCOHOL USE For each case meeting the Harm Standard, data was collected from child welfare agencies and community professionals on “the extent to which sentinels or CPS investigators considered [the perpetrator's alcohol use] to be a factor in the maltreatment”	The caregiver's alcohol use was perceived to be a factor in the child maltreatment	Perpetrating caregiver with the closest biological relationship to the child	Forms completed by child welfare investigators and community professionals
	2005–2006 National NIS-4	10.8%	Same as above	Same as above	Same as above	DRUG USE For each case meeting the Harm Standard, data was collected from child welfare agencies and community sentinels on “the extent to which sentinels or CPS investigators considered [the perpetrator's drug use] to be a factor in the maltreatment”	The caregiver's drug use was perceived to be a factor in the child maltreatment	Same as above	Same as above
USDHHS, 1997	1994 National National Study of Protective, Prevention, and Reunification Services Delivered to Children and Their Families	26%	N = 499,700 primary caretaker	46% Caucasian 41% African American 11% Hispanic 2% Other 48% children had a disability Primary caretaker: 75% bio/step/adoptive mom 5% bio dad 4% grandparent Mean age 37 yrs	Cases reported to child welfare Families and children who had an open case in the child welfare system on 3/1/94	During the families' current service episode, the caseworker reported that at least one of the “Presenting problem of the primary caretaker” was “Substance abuse”	Presenting problem among families with an open child welfare case Parents could have many presenting problems	Primary caretaker—person typically responsible for the day-to-day care of the child If more than one person jointly cared for the child, the mother was always considered the primary caregiver if she was one of them	Phone interviews with case workers of a random national sample of children and their families with an open child welfare case on 3/1/94
Besinger et al., 1999	1990–1991 San Diego County, California	79%	N = 639 Only one child per family	Children: 44.3% male 55.7% female 44% Caucasian 33% African American 19% Hispanic 4% Asian or Pacific Islander 1% American Indian Mean age 5 years (SD=4.88 years) Caregiver characteristics not reported.	Foster Care Sample of children (0 to 16 years) placed in out-of-home care due to maltreatment and who remained in care at least 5 months	At least one of the following: 1. Review of court reports, face sheets, and psychological evaluations with data abstraction tool indicated evidence of substance abuse and addiction. 2. Child welfare coded the reason for removal as related to drug or alcohol use 3.DSM III-R diagnosis of abuse or dependence of alcohol or drugs in file	Varied from contributed to the removal to not enough information to know if contributed	Mother, father, an adult who lived in the home at the time of the child's removal by child welfare, or another adult who assumed caregiver responsibilities for the child	Review of child welfare case records for children who entered out-of-home care between May 1990 and October 1991 for a period of 18 months following the removal of the child. Files may contain police reports, court reports/documents, medical and psychiatric evaluations, and caseworker reports
	1990–1991 San Diego County, California	16%	Same as above	Same as above	Same as above	Must have a DSM III-R diagnosis of abuse or dependence of alcohol or drugs in file	Not enough information to know if contributed	Same as above	Same as above
	1990–1991 San Diego County, California	33%	Same as above	Same as above	Same as above	Child welfare must have coded the reason for removal as related to drug or alcohol use	Drug or alcohol use contributed to the child's removal	Same as above	Same as above
Famularo et al., 1989	Not provided Case load of one urban juvenile and family court Boston	67.4% 147 of the 218 caregivers were involved in substance abuse based on RDC criteria; Prevalence in 136 cases was not reported	N = 218 caregivers More than one caregiver could be from the same case—218 caregivers from a sample of 136 separate cases	Demographics not reported Of 147 individuals: 55.1% Alcohol abuse 26.5% Polydrug abuse 18.4% Other	Foster Care “children maltreated enough to warrant custodial transfer of the child from the parents to the state” who were referred for clinical evaluation	Involved in substance abuse based on RDC criteria (Spitzer, Endicott, & Robins, 1978). The RDC defines alcoholism as “at least three specific manifestations of alcoholism in an illness of at least one month's duration” and drug use disorder as addiction, dependence, or abuse of “drugs other than alcohol, tobacco, and ordinary caffeine-containing beverages” (p. 778)	Unspecified but 137 of the 147 caregivers involved in substance abuse were court ordered to treatment for substance abuse indicating it was likely a factor in the foster care case	Person living with the child at the time of the court proceedings	Review of court records by the authors
Famularo et al., 1992	1985–1988 Case load of one urban juvenile and family court Boston	67% 127 case records of the 190 records involved 1 or more parent classified as engaging in substance abuse	N = 190 case records	Average caregiver age of 26 for total sample; No other demographics reported	Foster Care—“cases in which the state took legal custody of the children following a finding of significant child maltreatment”	One or more of the following: 1. Substantiated allegations by two or more separate professionals (social services or mental health) of alcohol and/or drug misuse Self-report of substance abuse of sufficient severity to meet RDC; excludes recreational or occasional drug us (Spitzer et al., 1978)	Court records for cases where the state took legal custody of the children; No information about the extent to which substance use impacted the decision	A parent or step-parent living in the home	Review of court records by the authors Sample of cases only included those in which the perpetrator of the maltreatment was one of the child's parents but not both of the child's parents
Jones, 2004	1995 San Diego County, CA	68% of children had a biological mother who abused alcohol or drugs	N = 443 Only one child per family	Biological Mother: 49.0% white 26.1% Hispanic 18.3% African American 31.9% married and living with spouse 26.3% employed 47.3% on public assistance	In-home services Substantiated child welfare case of abuse and initially received in-home services	At least one of the following: 1. Mother received services or referral for drugs or alcohol abuse 2. Child tested positive for drug at birth 3. Mother self-reported a problem with drugs or alcohol 4. Helping professional (e.g., psychologist) reported to the social worker that mother abused drugs or alcohol 5. “someone in a position to know, such as a parent or spouse” reported person had a substance abuse problem	Not enough information 39.1% of the total sample of families received services for alcohol and/or drug abuse during the service period indicating that a smaller percentage of families had a case with substance abuse treatment included than the total indicated need	Biological mother	Review of child welfare case records for families who initially received in-home services between 1/1/95 and 6/30/95
Murphy et al., 1991	1985–1986 Boston 1 juvenile court	43%	N = 206	67% single parent families 22% 2-parent non-married 11% 2-parent married AFDC present in 56% 53% rejected court ordered services No race, age reported	Classified as in-home services because court-involved case where foster care is unspecified “serious child abuse or neglect brought before a metropolitan juvenile court on care and protection petitions”	At least one caretaker had “documented problem with either alcohol or drugs” Alcohol or substance abuse had to be documented in a written report by a psychiatrist, psychologist, or a court-ordered screening for substance abuse	Not enough information to know how much it factored in but it was on the court investigators' risk checklist and then it was documented by a professional—Likely impacted the case	Original caretaker from which the petition was involved including biological parents, grandparents, foster parents, mother's boyfriend	Consecutive sample of cases at the Boston Juvenile Court. Cases in which the primary complaint was sexual abuse or where the child was older than 12 were excluded Court investigators completed 92 item risk checklist which documented “observable, behavioral characteristics of the parent, child, or type of mistreatment”. Master's students then reviewed the risk checklist for substance abuse variables
	1985–1986 Boston 1 juvenile court	50%	Same as above	Same as above	Same as above	“Alleged instances of substance abuse” Alcohol or substance abuse was alleged but NOT documented in a written report by a psychiatrist, psychologist, or a court-ordered screening for substance abuse	Not enough information to know how much it factored in but it was on the court investigators' risk checklist and then it was NOT documented by a professional—May or may not have impacted the case	Same as above	Same as above
USGAO, 1994	1986 Examination of statewide foster care databases for California and New York and review of random samples of case files for LA county, New York City, and Philadelphia County	51.8% abused substances in 1986	N = 376 children 0 to 36 months in foster care	Not reported	Foster care	“abusing drugs or alcohol” at the time child was removed from the home Not further defined	Difficulties faced around the time the child was removed from the home; Unknown how much it contributed to removal	At least one parent; “Parent” not operationalized	Examination of statewide foster care databases for California and New York and review of random samples of case files for LA county, New York City, and Philadelphia County
	1991 Examination of statewide foster care databases for California and New York and review of random samples of case files for LA county, New York City, and Philadelphia County	78.2% abused substances in 1991	N = 383 children 0 to 36 months in foster care	Same as above	Same as above	Same as above	Same as above	Same as above	Same as above
USGAO, 1998	1997 California	65.2%	N = 227 foster care cases	Not reported	Foster care	Parent who was required to undergo drug or alcohol treatment as part of the case plan for family reunification	Likely impacted case as treatment was required before reunification	At least one parent	Data collected in two ways: 1. Questionnaire mailed to foster care caseworker on a random sample of foster care. 2. Administrative
	data reviewed 1997	Illinois	74.3%	N = 292	foster care cases	Same as above	Same as above Same as above Likely impacted case as treatment	was required	before reunifica-
tion Same as above Same as	above AFCARS, 2013 data, unpublished With the exception of Illinois, all U.S. states including	District of Columbia, and Puerto Rico	31% national average		N = 637314 children in foster care Children removed from the	home and placed in foster care Combining two questions from the AFCARS data: As a condition associated with a child's removal from home and contact with the foster care system, the principal caretaker's compulsive use of alcohol/	drugs that is not of a temporary nature Alcohol or drug use was the reason the child was removed	from the home	“principal caretaker” Each state is asked to report this information to the federal government through the federal reporting system

To incorporate unpublished but relevant scholarly work, the author conducted a separate search with the search terms described above using online search engines. This resulted in one additional article for the sample (Gibbons, Barth, & Martin, unpublished) that was drawn from an academic website listing the article as under review. Additionally, unpublished data from the Adoption and Foster Care Analysis and Reporting System (AFCARS) was included in the analysis. This resulted in a total of eleven studies. The search was replicated by a trained graduate assistant. No additional articles were found in the replicated search. Using SAS 9.4, mean and standard deviations were obtained for the total studies and for smaller groups of studies within the sample. These means and standard deviations do not provide any weighting based on the sample size of the eleven studies or the number of estimates per study.

Results

Description of Studies

Of the eleven studies reporting primary data collection of prevalence rates for parental substance use disorders in the child welfare system, three studies provide national estimates (Gibbons et al., unpublished; Sedlak et al., 2010; USDHHS, 1997) and eight provide state or regional estimates (AFCARS, unpublished; Besinger et al., 1999; Famularo, Kinscherff, Bunshaft, Spivak, & Fenton, 1989; Famularo, Kinscherff, & Fenton, 1992; Jones, 2004; Murphy et al., 1991; United States General Accounting Office [USGAO], 1994, 1998). Prevalence rates across the eleven studies in the sample ranged from 3.9% to 79% (Figure 1; M= 35.04, SD = 18.06). In Figure 1, national prevalence estimates are located to the left of the solid black line and regional prevalence estimates are to the right. National prevalence estimates (n = 5) were lower than regional estimates, ranging from 3.9% to 26% (M = 12.34, SD = 8.19). State and regional prevalence estimates (n = 63) ranged from 5.2% to 79% (M = 36.84, SD = 17.42).

An external file that holds a picture, illustration, etc.
Object name is nihms-790387-f0001.jpg

Figure 1

Prevalence Rates of Parental Substance Use in Child Welfare

Each of the national estimates was obtained from large sample, national probability studies. First, Gibbons and colleagues (unpublished) report data from the National Survey of Child and Adolescent Well-Being I (NSCAW), which assessed the prevalence of caregiver substance use disorders among all families reported to child welfare for maltreatment and a subgroup of these families in which the children remained in the home following the baseline report. Next, Sedlak and colleagues (2010) report estimates based on data from the Fourth National Incidence Study of Child Abuse and Neglect (NIS-4). The NIS-4 sample includes both families reported to child welfare and families that were not reported to child welfare but in which child maltreatment had occurred. The prevalence estimate focuses on a subset of the NIS-4, cases in which there was clear harm to the child (Sedlak et al., 2010). Finally, the USDHHS (1997) National Study of Protective, Prevention, and Reunification Services Delivered to Children and Their Families interviewed the caseworkers of a random sample of children and their families with an open child welfare case on March 1, 1994. Additional details about the sample of each study, type of child welfare involvement, how substance use was defined, and how information was collected is available in Table 1 (see page 24).

The regional studies varied in the number of states for which they provided estimates. Five of the seven regional studies provide data on one state or rates for one county in a state (Jones, 2004; Murphy et al., 1991; Besinger et al., 1999; Famularo et al., 1989; Famularo et al., 1992). The other three studies provide estimates for two states (USGAO, 1998); three states (USGAO, 1994); or all 50 states, the District of Columbia, and Puerto Rico (AFCARS, unpublished). Excluding the AFCARS data, the remaining seven regional studies only collected data in five total states, with four studies reporting data on California (Besinger et al., 1999; Jones, 2004; USGAO, 1994, 1998); three studies reporting data on Massachusetts (Famularo et al., 1989; Famularo et al., 1992; Murphy et al., 1991); and one study each reporting data on Pennsylvania (USGAO, 1994), Illinois (USGAO, 1998), and New York (USGAO, 1994). With the exception of the AFCARS data, which provides the number of children placed in out-of-home care with parental alcohol or drug use as a factor in the child's removal for all states (except Illinois, which does not report the parental alcohol and drug use data), the vast majority of regional data are coming from the northeast or California; less is known about the rest of the country.

To illustrate the available prevalence estimates for each state, AFCARS data and individual state estimates are presented in Figure 2. The extreme variability in AFCARS data is illustrated by the dotted line connecting estimates. Comparing the regional prevalence estimates to the AFCARS data, square data points indicate prevalence estimates for three states: California (Besinger et al., 1999; Jones, 2004; USGAO, 1998), Massachusetts (Famularo et al., 1989; Famularo et al., 1992; Murphy et al., 1991), and Illinois (USGAO, 1998). For the national studies (Gibbons et al., unpublished; Sedlak et al., 2010; USDHHS, 1997) and for USGAO (1994), it was not possible to identify estimates by state.

An external file that holds a picture, illustration, etc.
Object name is nihms-790387-f0002.jpg

Figure 2

Prevalence Estimates by State

Type of Child Welfare Involvement

To examine prevalence rates by the type of child welfare involvement, each of the eleven studies were classified by the type of child welfare services (see Table 1). Two studies reported the prevalence among families referred to child welfare who may or may not have received any services (Gibbons et al., unpublished; Sedlak et al., 2010) (see Figure 3). Sedlak and colleagues (2010) also contained families who were not reported to child welfare but a community professional reported to the study team that the child was maltreated. Three studies reported prevalence rates among cases in which in-home services were provided to the family by child welfare (Jones, 2004; Murphy et al., 1991; USDHHS, 1997). Finally, six studies reported prevalence rates among cases in which the children were removed and placed in foster care (AFCARS, unpublished; Besinger et al., 1999; Famularo et al., 1989; Famularo et al., 1992; GAO, 1994, 1998).

An external file that holds a picture, illustration, etc.
Object name is nihms-790387-f0003.jpg

Figure 3

Prevalence Rates by Type of Child Welfare Involvement

Prevalence rates of parental substance use disorders were observed to vary based on the type of child welfare involvement of the sample. Among all families reported to child welfare (n = 4), prevalence estimates ranged from 3.9% 11.4% (M= 8.93, SD= 3.43). Among families receiving in-home services (n = 4), estimates ranged from 26% to 68% (M= 46.75; SD = 17.39). In families with children placed in foster care (n = 60), estimates ranged from 5.2% to 79% (M = 36.00, SD = 17.26).

Methods of Data Collection

The method of data collection for each study was examined (see Figure 4). Each statistic was classified based on the method of data collection used. Therefore, a study by Gibbons and colleagues (unpublished), which collected data using more than one method, was listed in two categories with a different statistic reported in each category. Six studies assessed for the presence of parental substance use disorders through case record reviews (Besinger et al. 1999; Famularo et al., 1989; Famularo et al., 1992; Jones, 2004; Murphy et al., 1991; USGAO, 1994). Parental substance use disorders were assessed through caseworker reports in five studies (AFCARS, unpublished; Gibbons et al., unpublished; Sedlak et al., 2010; USDHHS, 1997; USGAO, 1998) and parent report in one study (Gibbons et al., unpublished). In the Sedlak and colleagues (2010) study, both child welfare caseworkers and community sentinels were reporters for the study. The Sedlak and colleagues study was grouped with the caseworker report studies because data reported by both child welfare caseworkers and community sentinels were combined and separate results could not be obtained. Prevalence estimates varied based on the source of information about the parent's substance use disorder. Only one study provided estimates based on the caregiver's self-report (3.9%). Caseworker report of substance use (n = 57) produced estimates from 5.2% to 74.3% (M= 32.02, SD= 14.94). Case record reviews resulted in the highest prevalence estimates of parental substance use disorders, ranging from 16% to 79% (M= 55.34, SD = 20.42).

An external file that holds a picture, illustration, etc.
Object name is nihms-790387-f0004.jpg

Figure 4

Prevalence Rates by Method of Data Collection

Discussion

Despite the vast differences among the studies, some generalizations can be made. The family's type of child welfare involvement has some impact on the prevalence rate of parental substance use disorders in the child welfare population. Rates of substance use disorders were consistently lower in samples of reported cases (M = 8.93, SD = 3.43) than in families receiving in-home services (M= 46.75; SD = 17.39) or families with children placed in foster care (M = 36.00, SD = 17.26). Although it was anticipated that rates of substance use would be higher in the foster care sample than the in-home services sample, the prevalence of substance use disorders was higher in the in-home services group than the foster care sample. The high level of variability within the AFCARS data (5.2% to 62.2%) was one possible reason for this relationship. When the AFCARS data were removed from the foster care sample, the anticipated relationship was found with the foster care sample having a higher mean prevalence (M = 59.10, SD = 21.63) than the in-home services group (M = 46.75, SD = 17.39). The AFCARS data only reports alcohol and drug use that was associated with the reason for the child's removal, likely excluding cases in which parents were engaged in substance use but the child was reported as removed for a different reason. In addition, some states collect only the primary reason for removal and not factors associated with the removal.

The method of data collection assessing for parental substance use disorders is also an important factor. Case record reviews consistently found higher rates than parent self-report or caseworker report. Case files may provide the most comprehensive way to examine information about families. Within case files, parental self-report data, caseworker perceptions, and additional corroborating information, including drug screens and assessments for substance use disorders, can be examined in one place. However, being a comprehensive source of information does not necessarily indicate that information found during a case file review is entirely accurate. Additional information is needed to confirm the best method for measuring the prevalence of parental substance use disorders.

A key finding of this analysis is that parental substance use disorder prevalence rates are based on extremely old data that may no longer be representative of current trends. Of the nine published studies, one was published in the 1980s, six were published in the 1990s, and only two were published since 2000. Due in part to the length of time publication can take, the data used in the analyses were often much older than the publication date. Of the six studies published in the 1990s, only four studies reported any data from the 1990s, with the remaining studies reporting data from the 1980s. Interestingly, the study by Gibbons and colleagues and the AFCARS data were unpublished, but contained some of the most current data, collected between 1999 and 2013. These findings indicate the need for more current data collection and analyses. Federal, state, and local policy-makers should explore opportunities to update and refine these data, ensuring allocation of resources that adequately respond to the federal requirement in child welfare funding that reasonable efforts be made to prevent placement and to reunify families. The lack of current information on the scope of the problem calls into question the appropriate service mix available to families.

This review highlights the need for consistent measurement of parental substance use disorders across the field. Although parental alcohol and drug use as a factor in the case is measured in some way in the AFCARS data, this is insufficient. The extreme variation seen in the AFCARS data is likely due to inconsistencies in measurement of parental substance use disorders. For example, assessments procedures vary greatly in the United States, ranging from locales that rely on drug screens to those with substance use disorder assessments conducted by licensed or certified substance abuse treatment assessors. The inconsistencies across states may also be attributed to data-collection methods and data-system requirements rather than extreme variations in the actual prevalence of parental substance use as a reason associated with the child's removal. However, this assumption cannot be evaluated until data on parental substance use is consistently reported to AFCARS in a standardized way across all states.

As Congress and ACF propose updates and changes to the nation's reporting system, the inconsistent and unreliable measurement of parental substance use disorders as factors in child welfare cases within the AFCARS data should be addressed. First, it should be made mandatory in AFCARS for every state to report data on parental substance use disorders to the Children's Bureau. Experts in the field have been making this recommendation for almost a decade (Young et al., 2007), and the revisions to the AFCARS system provide an opportune time to implement this important change. In order for AFCARS data to be accurate, child welfare agencies must examine their internal systems for assessment of substance use disorders and data-tracking systems to accurately record these indicators. Parental substance use disorders should be documented consistently across cases so that this data can be accurately reported in the AFCARS systems.

Even if consistently obtained from each state, data reported to AFCARS would still not account for the prevalence of parental substance use among the majority of families involved with the child welfare system—those in which the children remain in the home (Dolan, Smith, Casanueva, & Ringeisen, 2011). Children who remain in the home with a caregiver affected by a substance use disorder are at risk for numerous negative outcomes, and these risks remain high over time. For example, Seay and Kohl (2015) found that 3 years after a child welfare investigation, children who remain in the home with a mother who is substance-dependent experienced higher rates of neglect, emotional maltreatment, and internalizing and externalizing disorders than children whose mothers had depression and children whose mothers had comorbid depression and substance dependence. More effort is needed to understand the prevalence and effective response to substance use disorders among families involved with the child welfare system whose children remain in the home.

In addition to incorporating substance use data into the SACWIS system, researchers reporting substance use data for their child welfare research samples should clearly define the type of child welfare involvement, understand how severe the substance use issues are, and discuss how the substance use impacted the case decisions or removal of the child from the home. This level of detail will allow the prevalence estimates to be compared across studies. Only Gibbons and colleagues (unpublished) obtained caregiver self-report of substance use. More research is needed to examine the self-reported prevalence of caregiver substance use in the child welfare system. Although it is unknown if parent self-report will produce a more accurate prevalence estimate, it is clear that knowing how many parents would self-report substance use disorders is an important indicator for service provision and possibly for interest in engagement with substance abuse treatment services.

There are a number of limitations to this review. Although some study characteristics were clearly related to the size of the prevalence estimate, it is possible that these trends are present due to the small number of studies and might not be present if a larger number of prevalence studies were available. The prevalence patterns supported by this analysis should be further tested in additional samples. In the reviewed studies, national samples were predominantly families investigated by child welfare and regional samples were predominantly among in-home services or foster care services. Therefore, in this analysis it is difficult to disentangle the impact of type of child welfare involvement from the impact of sample location. It is also likely that some regional or statewide estimates are unpublished and were not located during the search. Insufficient detail was available in the reviewed articles to compare the degree to which substance use disorders were a contributing factor in the case decisions. Information available from this aspect of the review is provided in Table 1.

Originally intended to narrow the prevalence estimates of parental substance use disorders in the child welfare system, this analysis was severely limited due to small sample sizes, geographically limited regional samples, extremely old data, and a wide range in the estimates. Wide variations in samples and data collection methods make it currently impossible to determine the exact prevalence of parental substance use disorders in the child welfare system. However, determining this prevalence is a critical step toward understanding the scope of the problem; strategizing the best way to intervene; and creating policy that ensures treatment, recovery, and parenting services are provided that address the family's needs and the important need to ensure prevention and intervention services for children who are at high risk of developing their own substance use disorder. Despite the reported prevalence ranges, the extensive limitations of and inconsistencies in the available data decrease generalizability. This analysis indicates the great need to examine caregiver substance use disorders in the child welfare system in large regional and national samples.

References

Besinger BA, Garland AF, Litrownik AJ, Landsverk JA. Caregiver substance abuse among maltreated children placed in out-of-home care. Child Welfare. 1999;78(2):221–239. [PubMed] [Google Scholar]
Dolan M, Smith K, Casanueva C, Ringeisen H. OPRE Report #2011-27a. Office of Planning, Research and Evaluation, Administration for Children and Families, U.S. Department of Health and Human Services; Washington, DC: 2011. NSCAW II Baseline Report: Introduction to NSCAW II. [Google Scholar]
Famularo R, Kinscherff R, Bunshaft D, Spivak G, Fenton T. Parental compliance to court-ordered treatment interventions in cases of child maltreatment. Child Abuse & Neglect. 1989;13(4):507–514. [PubMed] [Google Scholar]
Famularo R, Kinscherff R, Fenton T. Parental substance abuse and the nature of child maltreatment. Child Abuse and Neglect. 1992;16(4):475–483. [PubMed] [Google Scholar]
Gibbons CB, Barth RP, Martin SL. Substance abuse among caregivers of maltreated children. unpublished. [Google Scholar]
Hill SY, Tessner KD, McDermott MD. Psychopathology in offspring from families of alcohol dependent female probands: A prospective study. Journal of Psychiatric Research. 2011;45(3):285–294. doi:10.1016/j.jpsychires.2010.08.005. [PMC free article] [PubMed] [Google Scholar]
Jaudes PK, Ekwo E, Van Voorhis J. Association of drug abuse and child abuse. Child Abuse & Neglect. 1995;19(9):1065–1075. [PubMed] [Google Scholar]
Jones L. The prevalence and characteristics of substance abusers in a child protective service sample. Journal of Social Work Practice in the Addictions. 2004;4(2):33–50. doi: 10.1300/J160v04n02_04. [Google Scholar]
Kessler RC, Andrews G, Mroczek D, Ustun TB, Wittchen HU. The world health organization composite international diagnostic interview short form (CIDI-SF) International Journal of Methods in Psychiatric Research. 1998;7(4):171–185. [Google Scholar]
Murphy JM, Jellinek M, Quinn D, Smith G, Poitrast FG, Goshko M. Substance abuse and serious child mistreatment: Prevalence, risk and outcome in a court sample. Child Abuse & Neglect. 1991;15(3):197–211. [PubMed] [Google Scholar]
Seay KD, Kohl PL. Caregiver Substance Abuse and Children's Exposure to Violence in a Nationally Representative Child Welfare Sample. Journal of Social Work Practice in the Addictions. 2013;13(1):70–90. doi: 10.1080/1533256X.2013.756335. [PMC free article] [PubMed] [Google Scholar]
Seay KD, Kohl PL. The Comorbid and Individual Impacts of Maternal Depression and Substance Dependence on Parenting and Child Behavior Problems. Journal of Family Violence. 2015 doi: 10.1007/s10896-015-9721-y. [PMC free article] [PubMed] [Google Scholar]
Sedlak AJ, Mettenburg J, Basena M, Petta I, McPherson K, Greene A, Li S. Fourth National Incidence Study of Child Abuse and Neglect (NIS-4): Report to Congress. U.S. Department of Health and Human Services, Administration for Children and Families; Washington, DC: 2010. [Google Scholar]
Semidei J, Radel LF, Nolan C. Substance abuse and child welfare: Clear linkages and promising responses. Child Welfare. 2001;80(2):109–128. [PubMed] [Google Scholar]
Spitzer R, Endicott J, Robins E. Research diagnostic criteria: Rationale and reliability. Archives of General Psychiatry. 1978;35:773–782. [PubMed] [Google Scholar]
Staton-Tindall M, Sprang G, Clark J, Walker R, Craig CD. Caregiver substance use and child outcomes: A systematic review. Journal of Social Work Practice in the Addictions. 2013;13(1):6–31. doi: 10.1080/1533256X.2013.752272. [Google Scholar]
U.S. Department of Health and Human Services, Children's Bureau . National Study of Protective, Preventive and Reunification Services Delivered to Children and Their Families. U.S. Government Printing Office; Washington, DC: 1997. [Google Scholar]
U.S. Department of Health and Human Services . Blending perspectives and building common ground: A report to Congress on substance abuse and child protection. Author; Washington, DC: 1999. [Google Scholar]
U.S. Government Accounting Office . Foster care: Parental drug abuse has alarming impact on young children (GAO/HEHS-94-89) Author; Washington, DC: 1994. [Google Scholar]
U.S. Government Accounting Office . Foster care: Agencies face challenges securing stable homes for children of substance abusers (GAO/HEHS-98-182) Author; Washington, DC: 1998. [Google Scholar]
Young NK, Boles SM, Otero C. Parental substance use disorders and child maltreatment: Overlap, gaps, and opportunities. Child Maltreatment. 2007;12(2):137–149. [PubMed] [Google Scholar]