Introduction

More than a quarter of a million children entered foster care (i.e., removal from home and placement into an out-of-home setting) between October 1, 2009 and September 30, 2011 (US Department of Health and Human Services [US DHHS], Administration for Children and Families [ACF], Administration on Children, Youth and Families [ACYF], Children’s Bureau, 2012). Between 52 and 80 % of children entering foster care have been found to have at least one mental, emotional, or behavioral health problem (hereafter, mental health problem; Clausen et al. 1998; Glisson 1994, 1996; Halfon et al. 1995; Horwitz et al. 1994; Landsverk et al. 2006; US DHHS, US Department of Education and US Department of Justice 2000; Urquiza et al. 1994).

Children entering foster care are at high risk for mental health problems. They frequently endure numerous adverse experiences prior to entry into foster care. The most common reasons for a child’s entry into foster care are neglect (50–60 %), physical abuse (20–25 %), and sexual abuse (10–15 %; US DHHS, US Department of Education and US Department of Justice 2000). Additional concerns have arisen regarding the impact of in utero exposure to substances or infections through maternal risk behaviors, experiences of witnessed domestic violence and poverty, genetic predisposition to mental illness, and histories of receiving poor preventive health and educational services (Battistelli 1996; Chernoff et al. 1994; Curtis 1999; Halfon et al. 1992; Hurlburt et al. 2004; Needell and Barth 1998; Pilowsky 1995; Stahmer et al. 2005). Upon removal from home and placement into foster care, children also frequently experience sudden loss of families, friends, and communities, which may lead to experiences of trauma, bereavement, and internalized or externalized distress (American Academy of Child and Adolescent Psychiatry [AACAP] and Child Welfare League of America [CWLA] 2003).

Mental health evaluation of children entering foster care is critical both to identify pre-existing mental health needs and those needs arising from the transition to foster care. The identification of these mental health needs and referral to appropriate treatments are essential components to assuring these children’s well-being while in state custody. Professional guidelines to ensure this have been called for since the late 1980’s (CWLA 2007). Until recently, there has been little consistent focus in child welfare policies or protocolsFootnote 1 on children’s mental health care (Romanelli et al. 2009); see Table 1 for terms and working definitions used in this paper.

Table 1 Terms and working definitions used in this paper (in alphabetical order)
Full size table

The Current Context

Several existing federal guidelines and acts have been put into place to expand the definition of, and to regulate, the well-being of adopted children. The Adoptions and Safe Families Act of 1997 extended the purview of state child welfare agencies from merely ensuring the safety and permanency of children in foster care to also addressing well-being, a term often used to refer to children’s overall physical, mental, developmental, and dental health. Recent federal legislation also reflects this increased commitment to well-being, with the enactment of the Fostering Connections to Success and Increasing Adoptions Act of 2008 (hereafter, ‘Fostering Connections 2008’ P. L. 110-351), and the Child and Family Services Improvement and Innovation Act of 2011 (hereafter, ‘Improvement and Innovation Act;’ P. L. 112-34, 2011). The Fostering Connections Act required state child welfare agencies, in collaboration with medicaid departments and pediatric experts, to develop a plan for providing oversight of mental health services. State plans were expected to include a coordinated strategy for mental health evaluations of children in foster care, including a schedule for initial evaluations on entry into foster care and periodic follow-up mental health evaluations (P. L. 110-351, 2008; US DHHS, ACF, ACYF 2012). State plans were also advised to include a description of how the mental health needs identified through screenings would be monitored and treated, how medical information would be updated and appropriately shared, how continuity of health and mental health care would be ensured, and how oversight of prescription medications would be achieved (P. L. 110-351, 2008). The Improvement and Innovation Act (2011) additionally required states to develop plans to identify and monitor the mental health care of children exposed to trauma and to create specific protocols for overseeing the use of psychotropic medication among children in foster care; states receiving federal assistance through Title IV-B were required to submit state plans by June 30, 2012 (P. L. 112-34, 2011). To support the development of these plans, available guidance emphasized the importance of accessing child mental health experts, as well as inter-agency collaborations, specifically highlighting the critical role of partnerships between child welfare agencies, Medicaid departments, managed care organizations, and mental health agencies.

Particular challenges, however, arise in the states’ adoption, dissemination, and implementation of policies/protocols to identify the mental health needs of children entering foster care. First, state child welfare agencies may not have policies or protocols in place regarding mental health evaluations (i.e., screens and assessments) of children entering foster care. Leslie et al. (2003) reported that less than a half (47.8 %) of primary sampling units [PSUs] in 2000–2001 endorsed comprehensive policies to evaluate the mental health for all children entering foster care; an additional one-fifth of PSUs (20.2 %) had endorsed policies inclusive of mental health evaluations to specific subgroups of children entering foster care based on non-clinical characteristics, such as maltreatment type or placement setting. To our knowledge, a more up-to-date survey of the proportion of child welfare agencies with policies/protocols for mental health evaluation has not been published in the research literature.

Second, even when adopted, there may be a lack of awareness of these policies or protocols among providers serving children entering foster care. For example, Halfon et al. (2002) found that just over two thirds of respondents from state and county mental health agencies serving children entering foster care were not aware whether the child welfare agencies endorsed professional standards of care for mental health, including evaluations.

Third, administrators of mental health evaluations may not have adequate training or experiences in working with children entering foster care. For example, in one study, general community-based practices identified mental health problems among children entering foster care at a rate lower than specialized health care providers of children in foster care (Horwitz et al. 2000).

Fourth, particular challenges may arise in aligning the implementation of policies or protocols with a rapidly expanding evidence-base to inform the provision and implementation of child mental health services (Aarons and Palinkas 2007; Palinkas et al. 2011). Due to potentially high and negative media attention and litigation, child welfare agencies have been described as “passive-defensive cultures,” which “require extensive documentation, supervisory approval, and conformity as protection against intense public criticism, administrative sanctions, and frequent litigation” (Hemmelgarn et al. 2006, p. 76). Such a culture may generate specific challenges in proactively incorporating an emerging evidence-base into policies or protocols.

Fifth, the implementation of a policy or protocol may be challenging since both identification and treatment may require coordinated efforts across child-serving institutions (e.g., child welfare, primary care medicine, Medicaid, early intervention services, mental health services). Moreover, additional challenges exist when agencies share knowledge both within a single institution and between institutions (Easterby–Smith et al. 2008). The hierarchical and bureaucratic nature of public institutions may inhibit effective exchange of information about children within and across these institutions (Liebowitz and Megbolugbe 2003). Overcoming these challenges is critical for children entering foster care due to the frequently limited medical and mental health history available at entry (American Academy of Pediatarics [AAP] Committee on Early Childhood, Adoption, and Dependent Care 2002), and given that these children receive services from multiple public sector agencies, such as Medicaid, early intervention programs, and mental health departments, prior to and during entry into foster care (Allen et al. 2012). The limited availability of medical and mental health histories may be the result of a lack of consistent and routine contact with one health care provider prior to entry into foster care and/or the inability of social workers or mental health clinicians to obtain a comprehensive history from the birth parents at time of entry into foster care (American Academy of Pediatrics Committee on Early Childhood, Adoption, and Dependent Care 2002). Legal mandates, such as the Health Insurance Portability and Accountability Act of 1996 [HIPAA] Privacy and Security Rules, and the Family Educational Rights and Privacy Act of 1974, may also limit coordination in their efforts to support confidentiality. In facilitating information exchange between institutions, differences in organizational missions, mandates, and standard operating procedures may also inhibit the creation of shared policies/protocols (Allen et al. 2012).

Finally, although legislated, Fostering Connections (2008) and the Improvement and Innovation Act (2011) are not prescriptive. The Federal government has allowed considerable latitude for state and county governments to determine how to meet the mental health needs of local communities. In this respect, state and county child welfare agencies serve as “laboratories of democracy,” experimenting in their respective approaches (Holahan et al. 2003). A complicating factor for state and county child welfare agencies is a lack of consistency in the evidence around best practices for identifying, treating, and monitoring the mental health needs of children with complex behavioral and emotional issues as a result of abuse and neglect, including what tools to use to identify their mental health needs. As highlighted below, professional and advocacy organizations share several common elements in their recommendations to state and county governments regarding the best approach to overseeing the mental health of children in foster care, but there are also some differences.

Selected Professional Guidelines for Mental Health Evaluation

In light of the emerging evidence-base on the importance of mental health evaluation for children entering foster care and these implementation challenges, professional mental health organizations and expert consensus panels have set standards for the identification and treatment of the mental health needs of children entering foster care. Guidelines published by professional organizations highlight that mental health screening and assessment are imperative to ensure timely and appropriate mental health services for this vulnerable population of children (AACAP and CWLA 2003; AAP Committee on Early Childhood, Adoption, and Dependent Care 2002; AAP District II, & New York State Task Force on Health Care for Children in Foster Care 2005; Romanelli et al. 2009). Professional organizations have sponsored panels to garner expert consensus and to issue guidelines for policies/protocols to identify mental health needs of children entering foster care (AACAP and CWLA 2003; AAP Committee on Early Childhood, Adoption, and Dependent Care 2002; AAP District II, & New York State Task Force on Health Care for Children in Foster Care 2005). In addition, advocacy organizations, including the Casey Family Programs, the Annie E. Casey Foundation, and the REsource for Advancing Children’s Health (REACH) Institute, convened the Child Welfare–Mental Health (CW-MH) Best Practices Group (Romanelli et al. 2009), comprised of leading child welfare and mental health researchers, clinicians, policy-makers, and parent and youth advocates. Together, the CW–MH Best Practices Group published guidelines in 2009 specific to the mental health needs of children in child welfare.

The guidelines developed by these experts emphasize the importance of mental health evaluation, including both an initial mental health screen to identify urgent mental health concerns (e.g., suicidality, substance abuse, medication monitoring), as well as a comprehensive mental health assessment of children within a prescribed timeframe following entry into foster care. Some differences exist across the endorsed standards with respect to the timing of evaluations, the professional qualifications of the person who administers the evaluations, and recommended tools (see Table 2).

Table 2 Comparison of existing mental health evaluation guidelines
Full size table

All three reviewed guidelines (see Table 2) emphasize the role of an initial mental health screening immediately following entry into foster care. As illustrated in the guideline jointly issued by AACAP and CWLA (2003), a mental health screen is a “process for identifying the immediate/current mental health and use of alcohol and other drugs needs of the child. For children who are removed from their family, the screening process should assess the internalized and externalized level of distress of the child regarding the separation” (AACAP and CWLA guidelines footnote (1). The reviewed guidelines also highlight the importance of a mental health assessment. A mental health assessment, as defined by the AACAP and CWLA (2003), offers “a thorough look at the child in all life domains as well as the strengths and needs of the child and family” (AACAP and CWLA guidelines footnote (2). The CW-MH guidelines emphasize that an assessment is a thorough investigation of mental health and substance abuse concerns as well as a general examination of psychosocial risk factors related to the child’s environment (Romanelli et al. 2009). These assessments “may be based on structured interviews with the foster parents and caseworker, the results of standardized tests of development, or a review of the child’s school progress” (AAP Committee on Early Childhood, Adoption, and Dependent Care 2002, p. 538). Furthermore, guidelines suggest that the administrators of any mental health assessment possess mental health qualifications (AACAP and CWLA 2003; AAP District II Task Force on Health Care Children in Foster Care 2005; Romanelli et al. 2009). However, the guidelines differ with respect to the recommended timeframe for provision of evaluations (see Table 2).

While the importance of mental health evaluations has been recognized by federal legislation and published guidelines, the evidence base for universal tools to evaluate the mental health of all children entering foster care is lacking (Jee et al. 2011). Studies have suggested that the ongoing use of validated mental health evaluation tools by the provider may improve the detection rate of mental health problems among children in the general population (Sheldrick et al. 2011) and in foster care (Jee et al. 2010). However, there are ongoing debates in the field regarding whether evidence-based tools for children require additional study when applied to specific sub-populations (e.g., children in foster care) compared to children in community settings (Hoagwood et al. 2001). Moreover, increased recognition of the complexity of trauma-related symptoms manifested when a child enters foster care has led to a growing call for mental health evaluation tools to address trauma histories in all children entering foster care regardless of mental health symptoms (Greeson et al. 2011; Griffin et al. 2011).

Thus, while similarities exist across published guidelines, variation exists in the recommended mental health evaluation tools. The CW–MH guidelines highlight that the mental health assessment used should be evidence-based (Romanelli et al. 2009), whereas the other guidelines stress that the tools should address issues pertinent for children entering foster care (AACAP and CWLA 2003; AAP District II Task Force on Health Care Children in Foster Care 2005). Recently, experts have generated lists of specific tools validated with and potentially applicable to children in child welfare settings (Romanelli et al. 2008). The California Evidence-Based Clearinghouse for Child Welfare (CEBC) developed a rating system that reflects reliability and validity of certain available evaluation tools, both in terms of their overall evidence-base for use among the general pediatric population and applicability to children in foster care (CEBC 2006). We detail these in Table 3.

Table 3 Identified evidence-based assessment tools for use in foster Care and ratings based on a published expert guideline and the California evidence-based clearinghouse for child welfare
Full size table

Rationale for This Study

Considerable variation has been previously documented among geographic areas and child welfare agencies in the endorsement of policies or protocols for mental health evaluations of children entering foster care (Leslie et al. 2003; Mackie et al. 2011). However, no research specifically documents the qualitative variation in approaches to mental health evaluation nationally. Our study aims to: (1) describe the mental health evaluations conducted for children when initially entering foster care across a national sample of 47 states and the District of Columbia (D.C.; hereafter, ‘states’); (2) analyze whether policies and protocols around mental health evaluations in foster care are consistent with professional guidelines; and (3) examine whether these evaluations employ standardized mental health tools validated for children in foster care. As states refine their plans in response to Fostering Connections (2008) and the Improvement and Innovation Act (2011), they will benefit from an examination of current policies or protocols for evaluating the mental health needs of children in foster care. In addition, an analysis of the state of the field will be important as the implementation of these plans is developed, initiated, and evaluated by child welfare agencies, payers (e.g., Medicaid), and other stakeholders, including parents, practitioners, and policymakers.

Methods

Study Design

This study employed a mixed methods approach, consistent with calls for greater mixed methods research in mental health services (Palinkas et al. 2011). We employed both key informant semi-structured interviews and document review of state child welfare policies/protocols. The interviews and the document review were completed to gain a better understanding of the mental health policies/protocols of state child welfare agencies and partnering youth-serving systems among the states. The key informant interviews were hour-long telephone interviews conducted between March 2009 and January 2010. Given both the limited research on state approaches to mental health evaluation for children in foster care and the results of our pilot for this study, we employed an open-ended qualitative interview to ascertain the variation in approaches to mental health evaluation across states. The document review and analysis took place during this same timeframe and extended to June 2010.

Additional details about the methodology for the overarching study are available in a publication that examines state psychotropic medication oversight policies for youth in child welfare custody (Mackie et al. 2011).

Survey Sample

We received data from 47 U.S. states and the D.C. (response rate: 94.1 %). In cases in which the key informant for a state or district was not knowledgeable about a specific question(s), we requested a referral to interview the appropriate colleague. Accordingly, a total of 58 informants participated in these interviews. The informants held a wide variety of roles within the child welfare or partnering youth-serving system, such as program officers (n = 16), medical or mental health directors/specialists (n = 12), and deputy or associate directors of policy and/or practice (n = 6).

Document Review

We reviewed child welfare policies obtained from publicly available documents on state child welfare agency websites or through keyword searches, including foster care, policy, and mental health assessment, on internet search engines, as well as documents provided by key informants. Documents provided by key informants were issued by an array of relevant entities, including the state legislature, the child welfare agency, the Medicaid office, the state mental health authority, and the judicial system.

Measures

The key informant interviews employed a semi-structured guide, with 61 open- and close-ended questions about the practices and policies for mental health evaluation of children entering foster care. The interview included 10 questions specific to mental health evaluation that addressed the following five domains: (1) policies or protocols for screening and assessment; (2) eligible populations for evaluations (e.g., universal, specific subpopulations defined by categories, such as age); (3) timeframe and qualifications for the individual administering the evaluation; (4) standardization and types of mental health evaluation instruments; and (5) use and type (if applicable) of monitoring mechanisms used to track the status of mental health evaluations for children entering foster care. Key informants were asked to think broadly about mental health evaluation and include any policies/protocols related to tools (e.g., screening questionnaires, structured interviews, standardized tests), and the administrators (e.g., case workers, primary care providers, mental health professionals) who provided the mental health evaluation for children entering foster care.

The document review included a systematic review of the same domains for policies or protocols regarding mental health evaluation. Discrepancies between the data obtained from the interview and the reviewed document were noted in two cases, and in this these cases, the information was coded based on the document review. This approach was chosen because the purpose of this study was to examine what the existing state policies or protocols for mental health evaluations were rather than the knowledge of key informants or compliance of the states with these policies or protocols.

Analysis

All quantitative data were entered into an Excel spreadsheet and imported into STATA 11.0™ (College Station, TX, USA) for analysis. Summaries of qualitative responses for each state were recorded in a standardized template. For the purposes of this study, information extracted by state included (1) general approach to mental health evaluation (i.e., both screening and assessment) of children entering foster care, including specification of sub-populations to be evaluated (See definitions in Table 1); (2) timeframes for completion and administrators of the evaluations; (3) names of screening and assessment instruments and whether the instruments were standardized; and (4) whether a monitoring mechanism existed for these evaluations. Data on key questions pertinent to this study were extracted from the standardized template and tabled in Excel. For each state, these data were also extracted from the documents reviewed. Discrepancies between information provided by key informants and in documents were noted. Decisions about the appropriate coding were based on additional information obtained from state child welfare administrators, document review, or Internet searches. If the key informants from specific states did not know or indicate the answer and the document review failed to identify any relevant materials, the relevant variable(s) were identified as “missing” for the analysis. Accordingly, our results report both the total number of states with the approach or component over the number of states with valid data for the variable, expressed as a percentage. Therefore, the change in the value of the denominator reflects that some key informants did not know or indicate the response to the reported variables.

If a state reported having an approach to mental health evaluation for a specific age group, the specific approach was coded under the category of the mental health screen or assessment, as reported by the respondent, while noting that the approach was age-specific. We also categorized whether components of state approaches to mental health evaluation were consistent with professional guidelines issued by the AAP (2005), the AACAP and CWLA (2003), and the CW-MH (Romanelli et al. 2009); we specifically examined the timeframes for completion of these evaluations and required or recommended qualifications of the providers conducting the mental health evaluation (AACAP and CWLA 2003; AAP 2005; Romanelli et al. 2009). If a state required two mental health screenings or assessments, the timeframe referenced the initial screen or assessment employed.

Mental health screens and assessments were classified as standardized if either a tool or a structured clinical interview was routinely used and required. The various mental health evaluation tools were categorized as screening tools or assessment tools based on how the respondent characterized the use of these tools to evaluate the mental health of children. The administrator of the mental health screening was labeled as a primary care provider when a state identified using the Early and Periodic Screening, Diagnosis, and Treatment [EPSDT] program for the mental health screening of children entering foster care. The decision to label primary care providers as the mental health administrators in these cases was derived from the Health Resources and Services Administration [HRSA] maternal and child health description of EPSDT screening services as “a comprehensive health and developmental history, an unclothed physical exam, appropriate immunizations, laboratory tests, and health education” (US DHHS, Health Resources and Services Administration, Maternal and Child Health n.d.).

All mental health evaluation tools standardized across a state child welfare agency were also categorized according to the level of evidence supporting their use among children with maltreatment histories. The CEBC (2006–2012) rating scale was used to categorize mental health evaluation tools from a range of “A,” indicating the highest level of validity and reliability, to “C,” indicating unacceptable rates of validity and reliability (see Table 3). Using the scale developed by Romanelli et al. (2008), we categorized the evaluation instruments as: (1) not validated among children; (2) validated among children in clinical settings and potentially valuable for use in child welfare settings; and (3) validated for children in child welfare settings (see Table 3). While we recognize that additional evidence may exist to support the use of the mental health evaluation tools referenced by state child welfare agencies, we relied on these two available rating systems as they employed standardized methodologies.

For quantitative analyses, descriptive statistics were used to examine the presence and characteristics of mental health evaluations in state child welfare agencies, the frequency of adherence with published guidelines from AAP (2005), AACAP and CWLA (2003), and the CW-MH group (Romanelli et al. 2009), and the frequency of standardized mental health evaluation tools categorized by validation type.

Results

Mental Health Evaluation Approaches for Children Entering Foster Care

Our study revealed three approaches to the mental health evaluation of children entering foster care (see Fig. 1). The majority of the 48 states (n = 35/48, 72.9 %) employed the first approach, mandating mental health screenings for all children entering foster care, with a more comprehensive mental health assessment conducted for children when indicated by the screen. A quarter of the states (n = 12/48, 25.0 %) in our sample utilized the second approach, requiring a mental health assessment without requiring an initial mental health screen for all children entering foster care. Five of the 48 states (10.4 %) explicitly reported using a hybrid of the first and second approaches to mental health evaluations. These states used the first approach to mental health evaluations (i.e., screening then assessment as necessary) for most of the children entering foster care, but used the second approach to mental health evaluations (i.e., assessment for all children) for specific situations. For example, in some states a screen was bypassed and a full assessment was conducted if a child entering child welfare custody was <5 years of age and had a history of abuse or neglect (n = 1/48, 2.1 %), or if a child was <3 years old (n = 1/48, 2.1 %), or <3 years old and had a history of abuse or neglect (n = 2/48, 4.2 %). One of the 48 states (2.1 %) conducted a full assessment if a child was either <5 years old or if caregivers, school, or caseworkers identified any ‘red flags’. For data presentation purposes, these five hybrid states were incorporated in the description of the first approach below. Finally, only one of the 48 states (2.1 %) utilized a third approach, requiring both an initial mental health screen and a comprehensive mental health assessment for all children entering foster care.

Fig. 1
figure 1

Approaches to mental health evaluation for children entering foster care

Full size image

Approach 1: Mandated Screen for All Children and Assessment for Children as Indicated by Screen

Screening Timeframe and Administrator

35 of the 48 states (72.9 %) employed the first approach. The timeframes within which the mental health screens were completed varied across the 35 states. Seven of the 35 states (20.0 %) either did not have or did not mention requiring a specified time constraint regarding the completion of the mental health screen, while the majority (n = 28/35, 80.0 %) set a specific timeframe for which the screen should be conducted, ranging from 24 h to 90 days of entry into foster care. Two of the 35 states (5.7 %) required a second screen to assess for ongoing mental health issues, one within 30 days and one within 90 days of entry into foster care. The professionals required to complete the mental health screens were primarily caseworkers (n = 12/21, 57.1 %) and primary care providers (n = 8/21, 38.1 %).

Screening Instrument

Over half of the states employing the first approach (n = 19/34, 55.9 %) required a standardized screening instrument. The most frequently used standardized screening tools included the Pediatric Symptom Checklist-17 (PSC-17; n = 6/29, 20.7 %; Gardner et al. 1999), the Child and Adolescent Needs and Strengths-Mental Health (CANS-MH; n = 5/29, 17.2 %; Lyons 1999), the ages and stages questionnaire (ASQ; n = 2/29, 6.9 %; Squires and Bricker 1999), and the North Carolina Family Assessment Scale (NCFAS; n = 2/29, 6.9 %; Lee and Lindsey 2010). Two states (n = 2/31, 6.4 %) had age-specific screening requirements. For example, one state (n = 1/31, 3.2 %) only performed mental health screening on children older than 3 years of age, and one state (n = 1/31, 3.2 %) only performed mental health screening on children older than 4 years of age.

Assessment Timeframe and Administrator

Most of the states employing the first approach (n = 20/26, 76.9 %) did not have a set timeframe for completing a follow-up mental health assessment. The administrator of these assessments was required to be a mental health professional (n = 21/35, 60.0 %), or either a mental health professional or a primary care provider (n = 13/35, 37.1 %). One of the 35 states (2.9 %) specified that in addition to a mental health professional or a primary care provider, a caseworker may complete the mental health assessment. Two of the 35 states (5.7 %) required the administrator of the assessment to specifically have experience in child welfare systems.

Assessment Instrument

The states that employed the first approach also varied in terms of the type of mental health assessment instruments used; more than a quarter of the states (n = 9/34, 26.5 %) utilized standardized assessment instruments. The most commonly used assessment tools reported were the Child and Adolescent Functional Assessment Scale (CAFAS; n = 3/33, 9.1 %; Hodges 1990, 1994a), CANS-MH (n = 3/33, 9.1 %; Lyons 1999), Ohio youth problems, functioning, and satisfaction scales (n = 2/33, 6.1 %; Ogles et al. 2000), and the Child and Adolescent Service Intensity Instrument (CASII; n = 2/33, 6.1 %; AACAP 2007). Two of these states (n = 2/33, 6.1 %) specified the use of different standardized tools based on the child’s age; these states used the CAFAS (Hodges 1990, 1994b) for children older than 8 years old and the Preschool and Early Childhood Assessment Scale (PECFAS; Hodges 1994a) for younger children. Among the states without a standardized mental health assessment, key informants endorsed multiple factors influencing the type of assessment provided, including provider discretion (n = 17/34, 50.0 %), county or regional policies or protocols (n = 10/34, 29.4 %), the identified needs of the child (n = 9/34, 26.4 %), the type of health care insurance (n = 3/34, 8.8 %), and the age of the child (n = 2/34, 5.9 %).

Monitoring Receipt of Mental Health Evaluation

More than half of the states using the first approach (n = 20/34, 58.8 %) had a definite mechanism for tracking whether these mental health screens and assessments occurred. The Statewide Automated Child Welfare Information System (SACWIS) database was used to track the mental health evaluations in ten states (n = 10/34, 29.4 %), partnering agencies or liaisons were used to track evaluations in four states (n = 4/34, 11.8 %), and Medicaid systems were used to track data in one state (n = 1/34, 2.9 %). Child welfare staff either individually, as part of quality improvement teams, or as qualitative case review, were responsible to track the information in five states (n = 5/34, 14.7 %).

Approach 2: Mandated Assessment for All Children

Assessment Timeframe and Administrator

Twelve of the 48 states (25.0 %) employed the second approach to mental health evaluation (i.e., performing mental health assessments on all children without an initial screening). Of these twelve states, the majority (n = 11/12, 91.7 %) had set timeframes for completion of the assessments. 8 of the 12 states (66.7 %) required that assessments be conducted within 30 days; whereas, one state (n = 1/12, 8.3 %) required that they be conducted within 60 days. One state (n = 1/12, 8.3 %) required a comprehensive assessment be done within 24 h and repeated within 45 days. Finally, the administrator of these assessments was required to be a mental health professional in the majority of states (n = 8/12, 66.7 %). Three states (n = 3/12, 25.0 %) required that the administrator of the assessments have child welfare experience.

Assessment Instrument

Out of these 12 states, seven (58.3 %) used standardized assessments. Two of the 12 states (16.7 %) required child welfare to conduct an assessment only if a child was older than a specified age (i.e., older than 5 years old or older than 4 years old). Examples of the most frequently used tools were the CAFAS (n = 2/12, 16.7 %; Hodges 1990, 1994b) and the Child Behavior Checklist (CBCL; n = 2/12, 16.7 %; Achenbach 1991).

Monitoring Receipt of Mental Health Evaluation

Almost all of the 12 states (n = 10/12, 83.3 %) reported using a specific mechanism for tracking these assessments. The SACWIS database was used to track the mental health evaluations in seven states (n = 7/12, 58.3 %). Three states (n = 3/12, 25.0 %) tracked mental health evaluations through partnering agencies, Medicaid systems, and case review by child welfare supervisors.

Approach 3: Mandated Screen and Assessment for All Children

Screening and Assessment Timeframe and Administrator

Of the 48 states, only one state (2.1 %) utilized the third mental health evaluation approach (i.e., requiring all children entering foster care to be evaluated by both a mental health screening and a mental health assessment). The mental health screening and assessment were mandated to occur within 24 h and 21 days, respectively, of a child’s entry into foster care. This state required that a caseworker performed the screen and a mental health professional with experience in child welfare systems performed the assessment.

Screening and Assessment Instruments

This state required that the mental health assessment be standardized. This state did not specify the screening instrument used, but reported the use of CANS-MH as a required assessment tool.

Monitoring Receipt of Mental Health Evaluation

Finally, this state utilized SACWIS as a method for tracking the completion of both routine screenings and assessments.

Mental Health Evaluations as Compared to Professional Guidelines

In reviewing similarities of state policies/protocols with the AACAP and CWLA (2003), AAP (2005), and the CW–MH group (Romanelli et al. 2009), we report in Table 4 on the number of states that specified, (1) a timeframe for completion of the mental health evaluation; (2) the administrator of the mental health evaluations; and (3) standardized instruments, in accordance with recommendations from any of the guidelines. Various components of states’ approaches aligned with recommendations from professional guidelines. For example, one state’s policy or protocol followed the AAP guidelines (2005) with respect to timeframe for screening, but followed the CW–MH guidelines (Romanelli et al. 2009) with respect to the administrator of the evaluations. Overall, only one state had most closely approximated an approach consistent with the guidelines.

Table 4 Characteristics of mental health evaluation approach as a frequency of the sample (percentages in parentheses)
Full size table

Use of Validated Tools across Mental Health Evaluation Approaches

Of the 48 state representatives interviewed in this study, half (n = 24,50.0 %) required the use of mental health evaluation tools that have been developed for use in children; our results show that the applicability of some of these tools to children in foster care is still unknown. Based on the CEBC ratings, less than half of the states used at least one screening and assessment tool that has been rated as an A or B (n = 12/48, 25.0 % and n = 8/48, 16.7 %, respectively; see Table 3 for list of tools and ratings). When compared to guidelines by Romanelli et al. (2009), seven states and nine states of the 48 (14.6 and 18.8 %, respectively) specified the use of tools validated for use in foster care for the mental health screening and mental health assessment of children entering foster care, respectively. Romanelli et al. (2009) further identified mental health assessment tools that have been validated in clinical settings and are likely applicable to foster care; none of the states required these tools for mental health screening, and only three of the 48 states (6.3 %) required the use of these tools for mental health assessment.

Discussion

Federal legislation (e.g., P. L. 110-351; P. L. 112-34) and professional guidelines (e.g., AAP 2005; AACAP and CWLA 2003; Romanelli et al. 2009) call for child welfare agencies to assume responsibility for the well-being of children entering foster care, including their mental health. One of the findings of this study is that all the states (n = 48) have integrated policies or protocols inclusive of mental health evaluations for children entering foster care. As prior examinations found only 47.8 % of primary sampling units endorsed mental health evaluation policies/protocols (Leslie et al. 2003), our findings suggest a drastic increase in the use of mental health evaluations within the last decade. The widespread integration of mental health evaluations into policies or protocols may reflect an increased awareness among child welfare agencies and other entities of the importance of child welfare’s role in providing mental health evaluations for children entering foster care. They may also be responses to legislation and other political pressures, such as class action lawsuits focused on the mental health needs of these vulnerable children (Kutz 2011; Leslie et al. 2010; Mackie et al. 2013).

Three different approaches to the mental health evaluation of children in foster care were identified in our study: (1) a mandated screen for all children and assessment for children as indicated by screen; (2) a mandated assessment for all children; and (3) a mandated screen and assessment for all children. The difference in approaches to the mental health evaluations of children in foster care may be explained by the flexibility provided to state and county child welfare agencies to meet the mental health needs of their communities. However, there is also limited evidence regarding the best approach to identifying the mental health needs of children entering foster care. For example, some argue that the high rates of mental health problems indicate that a screen is unnecessary and that all children should receive a mental health assessment on entry into foster care (Chernoff et al. 1994). Others suggest that the traumatic events surrounding entry into foster care could skew the results of a mental health assessment done at entry (Halfon et al. 2002); thus, an initial screen at entry is necessary to detect urgent problems, but should be followed by an assessment done at a later time by a mental health professional to identify ongoing mental health problems. Without specific guidance from the federal government or research evidence, states have developed their own approaches to identifying mental health concerns for children in their care.

As with the variation found among states in their approach to identifying mental health needs of children entering foster care, this study also highlights differences in states’ policies/protocols regarding timeframes for the completion of screenings and/or assessments and required qualifications of the providers conducting the mental health evaluations. Some states in this study had very specific timeframes for completion of a mental health evaluation, while others had either broad timeframes or none specified. Some states require a licensed mental health professional to conduct both screenings and assessments, while others use caseworkers or pediatricians to perform initial screenings and then refer to a licensed mental health professional if the need for a full assessment is indicated. These variations may be the by-product of non-prescriptive policy mandates that do not have a clear or consistent evidence base to inform practice. They may also reflect the required training needed for specific tools; for example, CANS may be administered by trained child welfare staff. However, they also reflect real differences in both fiscal and human resources across states to perform mental health evaluations in children entering foster care.

One aspect of mental health evaluations where we found the greatest differences within and across state child welfare agencies was in the screening and assessment tools used to identify existing or emerging mental health problems. Although there is a growing evidence-base around the best tools to use to identify mental health problems for children in foster care, the research in this area is still limited (Levitt 2009). Interestingly, in some instances, the same tools (e.g., CANS, ASQ) were reported to be a screening tool in one state and an assessment tool in another. This variation could reflect how the tool is administrated, the qualifications of the persons conducting screenings and assessments, mandates by other child-serving systems (e.g., early intervention programs through the Individuals with Disabilities Education Act of 2004 [P. L. 108-466]), or contract stipulations by state Medicaid, mental health authorities, or managed care plans. More research is needed to identify tools that have a high sensitivity and specificity for detection of mental health problems particular to children entering foster care, including trauma-related symptoms. Attention to variation in the (1) developmental stage of the child, (2) the cultural and linguistic background of the child, (3) the level of expertise needed to administer the tool, and (4) other systemic factors influencing policies/protocols will be critical to future policy and program efforts of state and county child welfare agencies in identifying the mental health needs of children entering foster care.

At the time of this study, the gap between professional guidelines and practice was perhaps most striking in the lack of evidence-based and standardized approaches to the identification of mental health problems among children entering foster care. Although the professional guidelines emphasize the importance of employing standardized and evidence based tools (Romanelli et al. 2009), there remains limited consensus and explicit guidance among these professional groups regarding the specific tools to be used to facilitate best practice. (See Table 2 for illustration.) The limited use of evidence based and standardized tools within and across states undoubtedly partially reflects this lack of consensus. This within and across state variability will persist until a collaborative effort between the child welfare agencies, child serving partners, and all relevant stakeholders is initiated to identify the most accurate, reliable, and cost effective approaches to detecting the mental health problems of children entering foster care, establishing a uniform set of policies/protocols, developing mechanisms for monitoring both process and child well-being outcomes, and permitting flexibility that is responsive to state resources and priorities. Also important is additional clarity on how to assess trauma in children entering foster care, given recent requirements in the Improvement and Innovation Act (2011). Further investigation into the types of evidence most relevant to the efforts of child welfare agencies and youth-serving partners to provide mental health oversight may also help ensure improved alignment of available evidence in children’s mental health with policy or protocol, and ultimately practice.

The results from this study differ from other recent research in the gray literature, specifically research briefs and government reports, on the mental health evaluation of children in foster care (US Government Accountability Office 2009; Allen 2010). This discrepancy is likely explained by temporal and methodological differences, including the dates of data collection, the study sample (i.e., departmental or agency affiliation of key informants), as well as differences in the definitions employed for a mental health screening and assessment (US Government Accountability Office 2009; Allen 2010).

This study has several limitations that are important to note. First, we relied on written policies/protocols for mental health evaluation of children entering foster care to extract much of the data that are presented in this paper. Documents reviewed and information from key informants may, to varying degrees, reflect mental health evaluations routinely conducted by other child-serving agencies, including but not limited to Medicaid, mental health, early intervention, and education. Accordingly, our findings are not exhaustive, but rather reflective of the mental health evaluation approaches for children entering foster care, as described in the documents reviewed and known to the child welfare administrator and youth-serving partners with whom we spoke. Second, although we did compare verbal responses provided during interviews with information found in policies/protocols, the interview was not designed specifically to address discrepancies. Third, the key informants were not asked about the reason the child welfare agencies had established their approach to mental health evaluations, assigned certain timeframes or professionals to administer the evaluations, or chosen specific tools. This information may have helped to explore the degree to which child welfare agencies based their policies/protocols on the available evidence-based research or professional guidelines, available resources, pilot programs in their state, or other factors, such as the training requirements of the instrument and the availability of the instrument in the public domain. Without this information, we are only able to speculate how influential professional guidelines are on the development of state and county policies/protocols for mental health evaluation. A fourth limitation is that this study did not seek to determine whether states had policies for the identification of mental health needs at periodic intervals (e.g., quarterly, annually) of a child’s stay in foster care; our study only pertains to the use of mental health evaluations at the moment of entry into foster care. Since guidelines have suggested the use of periodic mental health evaluations while children are in foster care, this is an important topic for future research. Last, we recognize that the existence of a policy or protocol is by no means indicative of actual practice. Further research is needed to identify and evaluate rates of mental health evaluation receipt, as well as best practices across states and factors that facilitate or hinder their implementation.

The aim of this study was to achieve an understanding of the states’ current policies/protocols for the identification of mental health problems of children entering foster care. Considerable state-to-state variation in approaches, parameters for administration (i.e., type, timeframe, administrator), and implementation of evidence-based tools was noted in the mental health evaluations of children entering foster care. This study illustrates that there has been a significant improvement in establishing policies/protocols for mental health evaluations throughout child welfare agencies within the last decade. More effort and collaboration between the different stakeholders is needed to aid child welfare agencies in promoting the mental health of this vulnerable population by implementing evidence-based practice and recommendations from published guidelines of the AAP (2005), AACAP and CWLA (2003), and CW–MH (Romanelli et al. 2009). In addition, future research may help in understanding the gap between evidence-based knowledge, guidelines derived from expert consensus, and the actual practices and policies for mental health evaluation among child welfare agencies.