IDENTIFYING SUBSTANCE ABUSE AMONG CHILD WELFARE CARETAKERS : AN AMERICAN NATIONAL STUDY

American Child welfare staff has long recognized that substance abuse is common in families they serve. The unanswered questions are how well do child welfare social workers identify substance abuse in caretakers; and, what client and agency variables complicate this assessment? Chaos theory has much to offer social work in dealing with the uncertainty and complexity that characterize families who experience child abuse, neglect, and substance abuse. This exploratory study analyzed the National Study of Protective, Preventive and Reunification Services Delivered to Children and Their Families dataset. This dataset was a national sample of 2,109 opened child welfare cases (National Data Archive on Child Abuse and Neglect). Caretakers included birth mothers and fathers, stepparents, extended family members; such as aunts, uncles, and grandparents, and adoptive parents. Analysis explored the frequency of abuse and neglect substantiation, mental health problems and substance abuse problems identified by caseworkers. Three multivariate logistic regression models were developed to predict whether independent variables were associated with; the dichotomous dependent variable was identified caretaker substance abuse. Twenty-nine percent of caretakers were identified as substance abusers. This study found child welfare cases with identified substance abuse were significantly more chaotic than other families. Identified caretakers’ problems were twice that of non-identified caretakers. Substance abusing cases are opened more than twice the number of years and have twice as many caseworkers. This study found that neglect cases are fifty percent less likely to be identified and abandonment are more than ninety percent less likely to be identified.


Introduction
Child welfare staff has long recognized that substance abuse is common in the families they serve and numerous studies have shown that parents with substance abuse problems are more likely to maltreat their children (English, Marshall, Coghlan, Brummel, & Orme, 2002;Marcenko, Kemp, & Larson, 2000;Markward, Dozier, Hooks, & Markward, 2000;Wolock, Sherman, Feldman, & Metzger, 2001;Zelenko, Lock, Kraemer, & Steiner, 2000;Marsha, Smith, & Brunic, 2011).The prevalence of substance abuse in child welfare parents varies.Further, the more intervention required for a family higher rates of parental substance abuse are found (Canfield, Radcliffe, Marlow, Boreham, & Gilchrist, 2017).Parental substance abuse is lower for families who reported for child abuse rather than substantiated; but is higher for the more restrictive interventions, such as foster care (Sun, 2000).However, studies find most parents who are reported to child welfare agencies are not screened or assessed for substance abuse (Rittner & Dozier, 2000).
The research rapidly skips ahead from identification to interventions, that is, what to do about the substance abuse problem without sufficient exploration of the assessment problem (Canfield, Radcliffe, Marlow, Boreham, & Gilchrist, 2017;He, 2017;Young, 2003;Washington, 1999).In child welfare literature, substance abuse is seen as one possible cause of child abuse and neglect (Semidei, Feig Radel, Nolan, 2001;English et al., 2002).In the addiction literature, substance abuse is seen as a primary problem evidenced by multiple and progressively more severe problems that may or may not include in child abuse (Marcenko & Spence, 1995;Olsen, 1995).Research in both fields focus on linear, simple explanations of causality.
In lieu of a more complex examination of how child welfare workers identify substance abuse, research seems to focus on the sequence of problems in these chaotic families (Canfield, Radcliffe, Marlow, Boreham, & Gilchrist, 2017;He, 2017;Marsha, Smith, & Brunic, 2011).Studies argue that mental illness comes first followed by substance abuse and child abuse (Gomez, Primm, Tzolova Iontchev, Perry, Vu, & Crum, 2000;Manteuffel, Stephens, & Santiago, 2002).Some find that shame and guilt precipitate child abuse precipitating more shame precipitating substance abuse (Amodeo & Jones, 1997;Kellogg & Hoffman, 1997).Research has argued that substance abuse came first followed by mental illness and child abuse (Besinger et al., 1999;Weinman, Smith, & Buzi, 2002).The combinations of client problems researched are studied in a linear way; that is problem A causes problem B causes problem C. Are these linear causality arguments helpful to social workers charged with identifying substance abuse in these cases?
Social workers in child welfare and substance abuse fields know neither are child abuse or substance abuse are simple problems (Canfield, Radcliffe, Marlow, Boreham, & Gilchrist, 2017;He, 2017).These cases often include unemployment, homelessness, poverty, mental illness, domestic violence, and criminal behavior (Semidei, Feig Radel, Nolan, 2001;Carten, 1996;Zelenko et al., 2000).In a word, child welfare and substance abusing families are chaotic and family problems seem to multiple exponentially (Sun, 2000).. How child abuse and substance abuse are connected, in what order the behaviors occur and how to untangle the two are extremely complicated (He, 2017).Research of human behavior has a strong preference for simplistic analyses, ending up in policies and practices, which are also, included in the social workers' belief systems (Bolland, 1999;Hudson, 2000;Marsha, Smith, & Brunic, 2011).Child welfare policies focus on training workers in substance abuse, moving substance abuse workers into child welfare offices, or encouraging collaboration between the agencies (Canfield, Radcliffe, Marlow, Boreham, & Gilchrist, 2017).If research only touches on assessment then moves rapidly to explain interventions, is it not logical that in practice social workers assessment of the problem easily takes a backseat to substance abuse interventions?
To find any simplicity, research should first look on the level of complexity case details and client problems.Comparing caretakers who are identified with substance abuse to those not identified may lead to a deeper understanding of identification (Canfield, Radcliffe, Marlow, Boreham, & Gilchrist, 2017).A deeper understanding requires the analysis of more variables; including multiple and more severe client problems, examining caretaker substance abuse, child abuse and mental illness.Relationships that are more complex require controlling for types of child abuse or neglect, sheer number of identified client problems, and agency dynamics, such as length of time the case has been opened and number of caseworkers assigned to the family (Canfield, Radcliffe, Marlow, Boreham, & Gilchrist, 2017).Analysis of more variables and their relationship to identifying substance abuse is absent in the literature and is needed.More complex research concerning identification of substance abuse can expect better explanations that can apply to many different clients, multiproblem or chaotic families.

Method
This study analyzed the National Study of Protective, Preventive and Reunification Services Delivered to Children and Their Families dataset.This dataset was a national sample of 2,109 opened child welfare cases (National Data Archive on Child Abuse and Neglect).The dataset utilized a two-stage stratified random sample design covering all 50 states and the District of Columbia.The collection of data involved a one-time telephone interview with child welfare caseworkers.Questions included were relevant to case history, number, and types of services.
This study focused on caretakers of the identified children represented in the dataset.Caretakers included birth mothers and fathers, stepparents, extended family members, such as aunts, uncles, and grandparents, and adoptive mothers and fathers.Analysis explored the caretaker data to find the frequency of abuse and neglect substantiation and mental health problems and substance abuse problems as identified by caseworkers.Bivariate analysis explored the differences between caretakers identified with substance abuse concerns and caretakers who are not identified.Independent variables included caretaker age, race, and education; category of substantiated abuse or neglect; mental health problems; and the number of years the case was opened, number of workers assigned to the case, and total number of caretaker problems identified by the caseworkers.Bivariate Chi-Squares and ttests were conducted as appropriate.
Three logistic regression models were developed to predict whether independent variables still are associated with the dichotomous dependent variable of identified caretaker substance abuse.Model 1 independent variables were substantiated child abuse, including physical, sexual, and emotional abuse; neglect, including physical and emotional neglect; and abandonment, accounting for lack of supervision, failure to thrive, and desertion.Model 2 independent variables controlled for more chaotic client problems.Caretaker mental health problems, caretaker partner or spouse with substance abuse problems, and multiple substantiations of abuse and neglect was added to individual variables of substantiated abuse, neglect and abandonment.Model 3 controlled independent variables of number of years the child welfare case was opened, total number of caretaker problems identified, presence of a spouse or partner with substance abuse problems, caretaker with mental health problems, and number of caseworkers assigned to the family.
The dataset is based on complex survey designs and the sample is heavily weighted.WesVar 4.0 was used to compute this study's statistics to allow all aspects for the complex sample design, including the weighting process, and reflected in the estimated standard of errors (Mohadjer, Morgan, Chu and Rhoads, 1986).

Findings
The subset was 1863 child welfare caretakers.Twenty-nine percent of the caretakers were identified as substance abusers by child welfare caseworkers (n = 539, x² = 3.602, df = 1, p=.000).Caretakers identified as substance abusers were slightly older (t= 44.87, df =44, p=.000).Caretakers identified with substance abuse also have higher rates of multiple findings of child abuse, neglect and abandonment substantiation (x² = 54.056,df = 1, p = .000).Identified caretakers have more problems (t = 9.61, df = 5, p = .000).Caretakers with substance abuse have more child welfare workers (t = 2.484, df = 5, p = .020)and their cases are opened longer (t = 2.166, df = 5, p = .040)than caretakers who are not identified.Three logistic models were tested.Results from Model 1 (F = 27.147,df = 3/22; p = .000)found in neglect, substance abuse is 53% less likely to be identified.In cases of abandonment, substance abuse is 91% less likely to be identified.Model 2 (F = 26.894;df = 6/19; p =.000) found caretakers with mental health problems are three less likely to be identified with substance abuse issues.Caretakers with partners or spouses who are substance abusers are twice as unlikely to be identified.Model 3 (F = 34.728;df = 5/20; p = .005)found for each caretaker problem identified, substance abuse is six percent less likely to be identified.Caretaker substance abuse is thirty-three percent less likely to be identified with every child welfare worker assigned to the case.
This study found, as did Department of Health and Human Services analysis, more chaotic than other families (HHS/CB, 1997).Identified caretakers mean number of problems was twice that of non-identified caretakers.Substance abusing cases are opened more than twice the number of years and have twice as many caseworkers.What the bivariate analysis did not find as significant is also important.There was no significance as to who the caretaker was and the frequency of child abuse.There was no significance as to who the caretaker was and the frequency of substance abuse.That is, mothers, fathers, and other family members, whether biological, step or adoptive, abuse children in their care and abuse substances at the same frequencies as caretakers who have no substantiated child abuse findings or are not identified with substance abuse problems.
The logistic regressions show several concerns toward identifying caretakers with substance abuse.Previous studies found substance abuse is much more likely to be a factor in child neglect than in child abuse (Dunn, Mezzich, Janiszewski, Kirisci, & Tarter, 2001;Hixon, 1992;Wolock et al., 2001).This study found that neglect cases are fifty percent less likely to be identified and abandonment are more than ninety percent less likely to be identified.Numerous studies have reported comorbidity in substance abusers; yet in this sample, caretakers with mental health problems are three less likely to be identified with substance abuse issues (Sanguineti & Samuel, 1993).Studies have also found that substance abusers likely have spouses or partners who abuse substances (Kaufman, 1980;Windle, 1997).Here, caretakers with partners or spouses who are substance abusers are twice as unlikely to be identified.The literature reports chaotic lifestyles and many problems as an indicator of substance abuse (Clark et al., 2001;Kirisci & Tarter, 2001).In these child welfare cases for each caretaker problem identified, substance abuse is six percent less likely to be identified.One study found 35 percent of families with substance abuse problems had a single caseworker, versus 59 percent of other families, and 41 percent had three or more caseworkers, versus 21 percent (HHS/CB, 1997).This study found, with every child welfare worker assigned to the case caretaker substance abuse is thirty-three percent less likely to be identified.This is likely due in part to the fact that cases involving substance abuse were open for longer periods of time, making multiple caseworkers more likely.

Implications
Numerous questions about child welfare substance abuse screening and assessment practices are raised by this study.This study found that neglect cases are fifty percent less likely to be identified and abandonment are more than ninety percent less likely to be identified as substance abusers.Is it possible that child welfare workers see family chaos that stems from caretakers' substance abuse as neglect?Do caretakers who abuse substances lack the ability to care for these children due to their own intoxication and chaotic existence and not only from lack of parenting skills?Do child welfare workers misinterpret the caretakers' obsession with substances as apathy to the children in their care?Do caretakers who abandon children form a bond with the alcohol or drugs instead of bonding with their children?
Are caretakers with mental health problems referred to mental health services and not screened for substance abuse?Is substance abusing behaviors seen as mental illness only?Do partners or spouses who are substance abusers help in hiding the problem from caseworkers?Are individual problems identified by caseworkers rather than identifying chaotic lifestyles as symptomatic of substance abuse?Does each child welfare worker assigned to cases assume previous workers screened families for substance abuse?Is it agency practice to do assessments once, as cases are opened and not to reassess?If interventions are unproductive, is it agency What is evident is identification of substance abuse in child welfare is a not a simple problem.The study found the patterns among caretakers identified with substance abuse problems and those who were not identified.No matter who is the child's caretaker, child abuse and substance abuse are possible.The type of presenting problem is not indicative of substance abuse.Neglect and abandonment are as likely connected to caretaker substance abuse as physical or sexual abuse.Mental health problems often coexist with substance abuse.Substance abuse is likely to exist in the partner or spouse, as well as the caretaker.Chaotic families who are involved with substance abuse will present numerous problems; financial, employment, parenting skills, and homelessness is found in combination.Finally, no matter how long the case has been opened or how many workers are assigned, substance abuse is likely not to have been identified.