The development of an assessment protocol for Reactive Attachment Disorder.

Attachment is a critical issue among children in foster and adoptive settings. It is essential for mental health counselors who work with these children to develop appropriate appraisal skills for diagnosing Reactive Attachment Disorder (RAD), a syndrome associated with extreme attachment problems. However, there is no comprehensive procedure to assess a child for RAD. Thus, we propose a battery of semi-structured interviews, global assessment scales, attachment-specific scales, and behavioral observations to help mental health counselors identify the disorder We provide a case example to illustrate the utility of each assessment process.

**********

Reactive Attachment Disorder (RAD) is a developmental disorder resulting from either severe abuse and/or neglect of a child. Diagnosing RAD according to Diagnostic and Statistical Manual of Mental Disorders (4th ed. Text rev.; American Psychiatric Association, 2000) criteria is often a difficult process due to problems with differential diagnosis as well as disagreement among professionals regarding the etiology of RAD. In order to facilitate accurate diagnosis, an assessment protocol for RAD is necessary. While assessment protocols are controversial in nature, the difficulties faced by mental health counselors assessing children with RAD mandate accurate assessment protocols. We propose a battery of semi-structured interviews, global assessment scales, attachment-specific scales, and behavioral observations to identify attachment-related issues. We review the instruments used in our clinic and provide rationale for the application to RAD.

CURRENT CRITERIA AND COMORBIDITY

In efforts to simplify and create a quick reference, the DSM-IV-TR (2000) criteria for RAD as well as other related disorders are depicted in Table 1. Information provided on the etiology of this disorder is limited, and the prevalence of RAD is reported by the DSM-IV-TR as very uncommon. In fact, criteria and prevalence of RAD are areas of debate for many researchers. Boris, Zeanah, Larrieu, Scheeringa, and Heller (1998) noted that the DSM-IV-TR has been critiqued for its focusing of RAD as a "maltreatment syndrome" with emphasis on problematic parental care and overt social oddities, rather than a diagnostic focus of attachment issues. Consistent with this criticism, the DSM-IV-TR makes a distinction between two subtypes of RAD: (a) inhibited and (b) disinhibited types. The inhibited type focuses on social deficits that result in a child's inability to respond or initiate developmentally appropriate interactions. Disinhibited type describes a deficit in a child's ability to selectively choose an appropriate attachment individual.

Another important concern of the RAD diagnosis is its convenience to account for difficulties in children with maltreatment backgrounds. The diagnosis of RAD, while still considered uncommon, is gaining popularity in diagnosing children with neglectful and abusive histories. Hanson and Spratt (2000) noted that increasing use of this diagnosis among a particular population represents a possible danger of pigeonholing children into a diagnosis that is unwarranted or incomplete. Richters and Volkmar (1994) put forward the notion that the basis for the RAD diagnosis is twofold: (a) to conceptualize the deficits in social development, and (b) maintain a clinical awareness of the pathological familial background.

Conversely, prevalence and diagnosis has been called into question. In a recent article, Sheperis, Renfro-Michel, and Doggett (2003) noted that RAD symptomotology mimics that of many childhood disorders found in the DSM-IV-TR (2000). According to these researchers, not only can we attribute RAD's symptoms to another disorder, but RAD is often overlooked as a possible diagnosis for children who are potentially meeting its criteria. Hanson and Spratt (2000) cite misuse of defining terms such as bonding and attachment in creating diagnostic confusion, as the terms are being applied interchangeably by researchers and mental health counselors. The use of these terms synonymously implies the potential lack of conceptualization of the foundation of RAD that is imperative to its diagnosis.

The results of research conducted by Lynam (1996) supports the need for expansion of criteria for RAD and its potential relationship to other disorders such as conduct disorder, oppositional defiant disorder, ADD/ADHD, and the development of antisocial personality disorder. Lyman also examines the possible relationship between ADHD and conduct disorder as a possible basis for the development of a serious disorder he terms as "fledgling psychopath." Thus, RAD and other conduct-type disorders all reflect severe symptoms within children. These types of behaviors are potentially linked to the most resistant type of disorders to treat such as antisocial personality disorder. Therefore, there is added pressure for mental health counselors as well as researchers to determine adequate criteria so that these disorders can be diagnosed and treated closest to their onset.

In a study of 60 partially hospitalized adolescents, Rosenstein and Horowitz (1996) noted the many factors affecting diagnosis such as: (a) pathological or traumatic familial history, (b) low SES and economic conditions, (c) insecurity in attachment relationships, and (d) early onset of symptomology. All of these factors can contribute to the difficulty in adequately diagnosing and distinguishing RAD from a host of other psychiatric disorders. It can also account for the probability of the emergence of pathology.

DIFFERENTIAL DIAGNOSIS

Differential diagnosis is a critical area for any disorder; but due to the resemblance of other disorders, it is of particular importance to RAD (Muladdes, Bilge, Alyanak, & Kora, 2000). Many disorders are listed in the DSM-IV-TR (2000) that should be differentiated from RAD such as (a) mental retardation, (b) autism, and (c) pervasive developmental and attention-deficit disorders. Lynam (1996) provided other disorders to differentiate from RAD, including oppositional defiant disorder, conduct disorder, and the development of antisocial personality disorder. Lynam cited an overlap between conduct disorders and ADHD as well as concurrent and historical heterogeneity as potential variables maintaining diagnostic uncertainty.

Key components in adequately diagnosing RAD include: (a) differentiating the cognitive and lingual portion of the disorder adequately from other developmental disorders (such as those previously listed); (b) noting the behavioral portions, despite their tendencies to overlap other conduct type disorders; (c) paying particular attention to the assumed origin of the disorder as it relates to symptomotology; and (d) placing special emphasis on careful consideration of these criteria when making the diagnosis. With these considerations in mind, we developed an assessment protocol that aids in the identification of RAD. We provide a case example to illustrate the complex nature of RAD and the function of each instrument in the protocol. All of the identifying information related to the case has been changed to protect the confidentiality of the client.

CASE EXAMPLE

Reason for Referral

A state adoption specialist referred Joe Smith for a psychological evaluation. The evaluation was requested in order to determine the effect of physical and emotional factors on adoption placement. Prior to the assessment, Joe's adoptive parents reported that Joe demonstrated withdrawal, a high degree of fidgety behavior, difficulty sustaining attention, excessive talking, excessive distraction by extraneous stimuli, lack of boundary management, a degree of forgetfulness, and excessive daydreaming. He had formal diagnoses of Attention Deficit Hyperactivity Disorder (314.01) and Auditory Visual Processing Disorder (315.2) from a previous psychological evaluation conducted by an independent psychologist. Mental health medical interventions at the time of assessment included Adderall and Paxil.

Family History

Both the targeted adoptive parents and biological mother supplied psychosocial information prior to the date of assessment. The adoption specialist also supplied detailed information about Joe's history surrounding adoption placement. Joe, whose primary language was English, had some difficulty supplying historical information. He had problems recalling information about the reasons for his placement in foster care or details about his life with his birth parents. Joe had been in the custody of DHS for over 2 years at the time of assessment. Joe's sister Lisa, 5, who was categorized as developmentally delayed, was also in DHS custody. Joe also had two older brothers. The eldest brother, whose name and age were uncertain, was autistic and was in the custody of the paternal grandmother. Joe's next older brother, Bob, 13, was also developmentally delayed and had been diagnosed with Bi-polar Disorder (mixed type). Bob was removed from the birth parents' home on two separate occasions. Within the birth family, there were detailed reports of domestic violence, multi-generational abuse, alcohol addiction (father), low SES, poor living conditions, mental illness, and mental retardation (mother). Both Joe and Lisa were brought into state custody due to a substantiated sexual and physical abuse report.

Although detailed reporting of the abuse history is not requisite to a case example, it is important to identify some of the abuse patterns and to establish developmental time-lines for the onset and duration of psychological symptomology. One important facet to the substantiated abuse is that it was reported to occur throughout Joe's early childhood years. In fact, Joe's birth mother reports that the abuse began prior to age one, thus indicating a 7-year period of chronic pathogenic care. Reports from the children, the birth mother, and the paternal grandfather indicated that the birth father built cages in closets of the home. The children were often kept in the cages and fed raw meat through the bars. The children were also reportedly exposed to pornographic materials. Although the sexual abuse of Lisa was clearly documented through medical examination, reports were not clear about the extent of sexual abuse suffered by Joe. Some evidence of this abuse surfaced during placement in a foster home. Joe reportedly touched himself and asked his foster parents to spank him while he touched himself. Joe claimed that this was something that his father used to do to him.

From the time Joe was placed in the custody of the state, he had been in five different foster care placements including the Smith's home, which was the adoption setting. Since entering the Smith home, Joe had referred to himself as Joe Smith. Joe reportedly had few behavioral problems in any of the homes. However, one of his former foster parents reported that Joe appeared to confuse fantasy and reality. This behavior occurred in conjunction with video games. This concern remained for the Smith family. Another concern centered on Joe's refusal or inability to demonstrate emotional expression. When upset, his behavior is escalated by attempts at consolation and comforting from foster parents.

Assessment Instruments

Currently there is no single, comprehensive tool to assess a child for RAD. In the absence of such a tool, we have developed a protocol utilizing several reputable assessment tools to use as a battery to assess for RAD. In our clinic, we conduct standard evaluations of intelligence and aptitude and gather a comprehensive psychosocial history. When assessing for RAD, we conduct child and parent clinical interviews. We also use several global rating scales, attachment-specific scales, and behavioral observations (see Table 2 for a summary of each instrument). Below is a review of the global rating scales and attachment-specific scales.

The Child Behavior Checklist (CBCL), created by Achenbach (1991), was designed to assess children and adolescents for abilities and behavior problems in a standardized format. The CBCL was created to differentiate between children with no mental health diagnoses and children who have been receiving mental health services due to behavioral problems. We use the CBCL to provide information on the general behavior of the identified child, specifically to see if the child appears similar to a child with clinically significant internalizing problems (i.e., anxious/depressed, withdrawn, or somatic complaints) or externalizing problems (i.e., delinquent behavior or aggressive behavior). If RAD can be considered a construct composed of several related behaviors, then the presence of internalizing or externalizing problems proves useful in attempting to identify either type of the disorder. If the child scores similar to children within the clinically significant range, or perhaps even the borderline range, the child is likely to have clinically significant problems. Joe's responses on the CBCL yielded clinically significant scores in the areas of Thought Problems and Attention Problems. Critical items included "stares blankly" and "displays nervous movements or twitching."

The Behavior Assessment System for Children (BASC), developed by Reynolds and Kamphaus (1992), was designed to assess children and adolescents for emotional and behavioral disorders in order to develop treatment plans. We use the BASC in order to gather more broad information about the child such as thoughts about himself or herself, feelings toward parents, perceptions toward school, and evaluation of adaptive behaviors. The utility of the BASC for a RAD assessment is in determining if the child does exhibit differences in thinking. The BASC can distinguish a child's pathological thinking from normal peers' thinking on a range of topics (e.g., school, parents, and feelings about self). Clinically significant scores as well as borderline results indicate the child may indeed think pathologically and warrants further probing to gather information to determine if the child meets specific criteria for a diagnosis. Joe's responses to items on the BASC yielded scores in the At-Risk range in the area of Anxiety. The remaining scales and subscales fell Within Normal Limits. Critical items included "uses medication" and "has eye problems."

The Eyberg Child Behavior Inventory (ECBI; Eyberg, 1999a) and the Sutter-Eyberg Student Behavior Inventory--Revised (SESBI-R; Eyberg, 1999b) are also used as part of our global assessment for RAD. The ECBI was constructed to determine the severity of conduct-problem behaviors in adolescents and children between the ages of 2 and 16 years of age, and the SESBI was developed to determine the severity of conduct-problem behaviors in the classroom of children between the ages of 2 and 17 years of age. The ECBI and SESBI-R can be used in order to examine and quantify the behaviors exhibited by a child suspected of RAD. Both of these measures can identify the presence of problematic and inappropriate behaviors that are exhibited by children with RAD such as problems with compliance, attention problems, or aggression. Because the rating scales focus on the intensity of specific problematic behaviors, the instruments may also be used to measure change due to treatment. According to Mr. and Mrs. Smith's responses to the SESBI-R and Joe's responses to the ECBI, Joe received Intensity and Problem Scores which were within normal limits. Items scored as problems included "dawdles in getting dressed," "whines," "is easily distracted," "has a short attention span," and "fails to finish tasks."

The Randolph Attachment Disorder Questionnaire (RADQ; Randolph, 2000) is a 30-item self-report instrument designed to aid in the identification of Attachment Disorder (AD) in children between the ages of 5 and 18. It should be noted that this instrument is designed to assess Randolph's theoretical formulation of attachment disorder and not the DSM-IV-TR (2000) diagnosis of RAD. The RADQ does not assess subtypes of insecure attachment styles such as attachment styles of middle class, low income, and maltreated infants; nor does it have any correlation with Ainsworth's classifications of attachment behavior. Although this instrument is not designed to assess RAD, the key purpose of the RADQ is to discern children with attachment problems from those with a disruptive behavior disorder. Thus, we have included the RADQ as part of our assessment protocol. In a validity study of the RADQ using only 35 participants who were all patients, Randolph discovered significant correlations between the Delinquent Behavior (r = .36) and Aggressive Behavior (r = .32) subscales of the Child Behavior Checklist and the RADQ. Randolph recommends that the RADQ not be used as the sole diagnostic instrument for AD. In contrast, the RADQ should be administered as a condensed screening instrument to initiate a more detailed inquiry of a prospective client. Randolph defines Attachment Disorder as a clinical syndrome that includes behaviors exhibited by children with an assortment of psychiatric disorders including Attention Deficit Hyperactivity Disorder, Conduct Disorder, Oppositional Defiant Disorder, Bipolar, Psychotic Disorder Not Otherwise Specified, Schizoaffective Disorder, Major Depression, Dysthymia, Intermittent Explosive Disorder, and Post-Traumatic Stress Disorder. Because this instrument has not been subjected to independent analysis, we use it with a degree of caution. According to Mr. and Mrs. Smith's responses to the RADQ, Joe obtained a Clinically Significant total score and displayed behaviors consistent with the Avoidant Subtype.

Another attachment specific instrument that we use in our clinic is The Reactive Attachment Disorder Questionnaire (Minnis, Pelosi, Knapp, & Dunn, 2001). Developed by Helen Minnis and her colleagues in Scotland, this instrument was normed in Europe and thus may present problems with generalizability to the U.S. population. The Reactive Attachment Disorder Questionnaire is a 17-item parent questionnaire. The instrument is used to assess the presence of RAD and subtypes of RAD, which include the Disinhibited Type and Inhibited Type. Like the RADQ, The Reactive Attachment Disorder Questionnaire has not been subjected to independent analysis. However, because this instrument is based on the DSM-IV-TR (2000) criteria, we have included it as part of our protocol. According to Mr. and Mrs. Smith's responses to the Reactive Attachment Disorder Questionnaire, Joe met criteria for the Inhibited Type of RAD.

Direct Behavioral Observations

Molar measures such as structured and semi-structured interviews, behavioral rating scales, and standardized tests are vital components of a comprehensive assessment package and provide valuable nomothetic comparisons needed for diagnosis and pre- and post-treatment assessment. In contrast, direct behavioral observations provide idiosyncratic comparisons allowing the mental health counselor to obtain baseline and on-going treatment information that is unique to the individual characteristics of each client or family. As such, behavioral observations have been described as the hallmark of behavioral assessment (Kratochwill, Sheridan, Carlson, & Lasecki, 1999).

When working with families who have fostered or adopted a child with RAD, the mental health counselor needs to obtain important pre-treatment information about familial interactions and the display of current problematic or target behaviors in need of intervention while obtaining other forms of assessment information. Doing so will not only aid in a proper diagnosis but will also facilitate the development of therapeutic behavioral goals and objectives in a cost-effective manner. Because many families warrant intervention immediately upon referral and most third-party payers limit the amount of funding applied to assessment activities, mental health counselors may find the utilization of pre-developed assessment conditions facilitative in establishing baseline levels of functioning.

For children under the age of 8, we propose the use of analogue observations in evaluating familial interactions. Because Joe was 10 at the time of assessment, we did not conduct the following observations with him. However, we believe that it is important to include a detailed discussion of the conditions in order to illustrate additional methods of assessment for RAD. The conditions used in our clinic have been developed from the empirical literature in both functional behavioral assessment (e.g., Carr & Durand, 1985; Cooper et al., 1992; Moore, Edwards, Wilczynski, & Olmi, 2001;Taylor & Carr, 1992;Taylor, Ekdahl, Romanczyk, & Miller, 1994) and behavioral parent management training (e.g., Hembree-Kigin & McNeil, 1995). Functional behavioral assessment is a process for investigating the function or purpose for the performance of problem behavior (Watson, Gresham, & Skinner, 2001). Parent management training has a literature base spanning over 30 years and typically promotes a problem-solving approach designed to teach parents to be behavioral change agents in order to maintain therapeutic gains in natural environments (Allen & Shriver, 1998). Although not specifically designed for the assessment of RAD, these conditions have been validated for use with other populations who exhibit similar problematic behaviors (e.g., noncompliance, tantrums, aggression, self-injurious behavior, social avoidance) and for other comorbid disorders (e.g., Attention Deficit Hyperactivity Disorder, Oppositional Defiant Disorder, Conduct Disorder, Developmental Delay). Furthermore, these conditions have been utilized with children who have experienced situations involving verbal abuse, physical abuse, and neglect. Finally, these conditions have been developed to directly assess the occurrence of social avoidance and task avoidance often seen in children diagnosed with RAD.

Although derived globally from the functional behavioral assessment and parent management training literature, the analogue conditions used in our clinic are modified primarily from those developed by Moore, Doggett, Edwards, and Olmi (1999) and promote an evaluation of task avoidance (e.g., refusing to comply with instructions) and social avoidance (e.g., disengagement from social interaction). Each condition lasts for 5 minutes and the child and/or family is observed from behind a one-way mirror. Before the assessment, the mental health counselor meets with the parents and informs them that the 5-minute conditions will allow them to evaluate patterns of family interaction under low and high periods of task demands and social attention. The mental health counselor then thoroughly reviews the instructions for each session during 5 to 10 minute breaks between the conditions. The mental health counselor uses a bug-in-the-ear device to cue the parents to engage in the prescribed behaviors (i.e., demand or social attention) at pre-set intervals during the observation. Mental health counselors who do not have access to such facilities can remain in the room with the family; however, they need to inform them that they will only be there to observe during this period of assessment and will only interact with them by cueing them to engage in the prescribed behaviors to be delivered in each condition. If the mental health counselor does not have a bug-in-the ear device, he or she can simply nod his or her head as a cue for the parents to engage in the specified behaviors.

The first condition involves Free Play (FP). During FP, the child is placed alone in a therapy room and allowed to play freely with developmentally appropriate toys or activities. This condition serves as a control condition for the other situations included in the protocol.

The second condition involves Low Attention (LA). During this condition, the mental health counselor instructs the parent to provide noncontingent attention only twice during the entire session while reading from a magazine or newspaper. At pre-set periods (2 minutes and 4 minutes), the mental health counselor instructs the parent to deliver a neutral phrase to the child (e.g., I see that you are playing) and then provide no other forms of interaction. The parents are also instructed to ignore the child if he or she tries to engage them at any other point during the 5-minute condition. Finally, the parents are directly instructed not to deliver commands, instructions, redirections, or any other form of demand during this condition.

The third condition is Low Demand (LD). During this condition, the parents are instructed to provide demands or instructions to the child that have a high probability (greater than 85% of the time) of being performed (e.g., Look at me.) at 2 minutes and 4 minutes during the 10-minute condition. The mental health counselor simply instructs the parents to have the child do something that he or she is likely to do. This condition is considered to be one of low demand because the parent only issues two instructions of low difficulty during the entire time period.

The fourth condition involves High Attention (HA) and involves delivering high amounts of social attention. The parents are instructed to play with the child with developmentally appropriate toys or activities on the floor or at a table. The parents are cued to provide social attention at 20-second intervals during this condition. However, the mental health counselor does not instruct the parent to use a neutral phrase as in the previous condition. Instead, he or she tells the parents to provide attention in a manner consistent with how they would attend to their child while playing with them at home. This allows the mental health counselor to maintain the integrity of the condition by ensuring that high amounts of attention will be provided but also allows the mental health counselor to evaluate the patterns of play among family members.

The fifth condition is High Demand (HD). This condition involves task demands that have a low probability (less than 25% of the time) of being performed and can include verbal instructions (e.g., Clean up the cars) or academic assignments (e.g., working on a math worksheet). The mental health counselor should identify these demands during the interview or from a behavior log that could be completed by the parents at home prior to this assessment. Such information will allow the mental health counselor to empirically determine which demands have a low probability of being performed. The mental health counselor instructs the parents to provide a demand or instruction at 20-second intervals during this condition. If the child is working on a worksheet, the parent is instructed to ask the child how he or she is performing on the task at 20-second intervals. Again, the mental health counselor should not provide the parents with examples of how to provide effective instructions or ask appropriate questions so that he or she can evaluate the manner in which the parents try to obtain compliance from their child.

The final two sessions of the assessment include a replication of the conditions that had the lowest and highest occurrences of problematic behavior. This replication allows the mental health counselor to have more confidence in the results and bolsters the internal validity of the assessment procedures. In addition, the length of each session (e.g., 5 minutes) makes it practical for the mental health professional to conduct the analysis in one 50-minute therapy session

While each family presents with their own set of interaction styles and referral concerns, the mental health counselor should have a specific set of behaviors that he or she is observing. For example, the mental health counselor will want to record data on the occurrence of social initiations made by the child in each condition. Based on guidance from Moore et al. (1999), social initiations are defined as any attempt made by the child to start an interaction or communicate with the parent. The second behavior that warrants investigation is compliance. Generally, compliance is defined as the child's initiating compliance with a parental request within 10 seconds of the request. The third behavior under investigation includes engagement and is defined as the child directing his or her eyes toward activities or materials and/or manipulating materials associated with a given task demand. The final set of behaviors in the assessment should be specific to the referral concerns presented by the family and could include behaviors such as tantrums, aggression, interrupting, inability to play independently, whining, crying, or any other problematic behavior that is causing a deterrence to positive family functioning. A 15-second partial interval coding system (10-second observe, 5-second record) is used to record the occurrence or nonoccurrence of target behaviors, in cassette tape specifically designed to signal observe-and-record intervals is used in our clinic. However, mental health counselors who do not have access to such technology can be cued by another observer or can use a wristwatch or stopwatch to cue the intervals.

These conditions can help facilitate the diagnosis of RAD when combined with the diagnostic interviews, rating scales, and standardized tests. From a diagnostic perspective, these conditions have assisted in identifying the purpose of the problem behavior and developing treatment goals for children who display problematic behaviors similar to the behaviors displayed by children diagnosed with RAD (Moore et al., 1999; Taylor & Carr, 1992; Taylor et al., 1994). During previous research, some children have exhibited more noncompliance and problematic behavior in the high demand conditions than the low demand conditions. Similarly, they exhibited low amounts of problem behavior in both of the attention conditions with the exception of trying to initiate interactions with the parents more frequently in the low attention condition. Such patterns of responding are consistent with those behaviors displayed by children diagnosed with the Disinhibited Type of RAD. In contrast, other children have been labeled as socially avoidant and engaged in more problematic behavior during high attention condition in an effort to escape social interactions. Furthermore, they often demonstrated high levels of engagement during the low attention condition and rarely interacted with the parents. Finally, they often demonstrated high levels of compliance with academic demands during the demand condition in an effort to avoid social interaction with the parents or other adults (Moore et al.). This pattern of responding is representative of the behaviors displayed by children diagnosed with the Inhibited Type of RAD.

In addition to assisting in the diagnostic process, the assessment conditions also serve to provide direct targets for intervention that are unique to the child and family. For example, the free play condition not only serves as a control for the other conditions, it also allows the mental health counselor to obtain information about the child's individual play behaviors and ability to entertain himself or herself when alone. The attention conditions allow the mental health counselor to observe the patterns of familial interaction that may be supporting the display of problematic behavior. For example, in our clinic during the attention conditions, we often observe the delivery of questions (e.g., "What are you doing?") or criticisms (e.g., "No! That is not the way you do it."), which are often viewed as demands by the child and set the stage for the display of problem behavior. This delivery of questions often serves as a cue to us that the parent needs to be instructed in providing forms of social attention that are not perceived as demands. During the demand conditions, we often observe parents providing vague and ineffective instructions (e.g., "Put that over there."), asking inappropriate questions (e.g., "What are you supposed to be doing right now?") or delivering rapid fire instructions (e.g., "Come on! Hurry up! Come on! Come on! Do it!") that often promote the display of problematic behavior. From a therapeutic perspective, such information can be extremely helpful to the mental health counselor in establishing goals and objectives toward increasing familial interaction and attachment.

We want to express caution in the utilization of these conditions for several reasons. First, they are developed from research literature in two empirically sound areas of functional behavioral assessment and behavioral parent management training for children displaying behaviors similar to those displayed by children with RAD. However, we are still in the process of validating these conditions specifically for the diagnosis and treatment of RAD. Second, these conditions are developed from a behavioral perspective not shared by all members of the therapeutic community. Mental health counselors who are not trained in or who do not espouse the use of behavioral procedures will probably not implement the procedures with the level of integrity needed to derive empirically valid results. Third, these conditions are specifically designed to increase the occurrence of problematic behavior. Parents need to be properly informed of the purpose of the conditions and prepared for such increases in behavior. In addition, the mental health counselor needs to be trained in de-escalating problematic situations and should develop standardized protocols for managing severe problem behavior. Fourth, these conditions may not be representative of all situations faced by the family. For example, we often add in other conditions if there are other children in the home in an effort to evaluate interactions among the siblings. We have also had to implement separate conditions with the mother and father in order to evaluate paternal and maternal interaction patterns with the child. In conclusion, these assessment conditions can be valuable in providing assessment and therapy services to families with children with RAD; however, they must be performed with integrity and evaluated within the context of a multi-modal assessment and treatment package.

Clinical Interview Material

The observations that we did conduct for Joe indicated that he was fully oriented to time, place, and person. He was able to give the proper date, month, and year. He was aware of the physical location. He was able to give correct answers concerning his street address, city, and state. He was able to give his own name and knew his adoption status. Joe was asked about his memory functioning and reported that he had some difficulty recalling incidents that resulted in his placement in foster care. He stated that people told him that bad things happened but that he doesn't remember them. When asked about his relationships with individual birth family members, he provided vague details and became extremely agitated. Observations by the examiner during the assessment found him to be vague in responding. His short-term memory, however, was found to be normal. Delayed recall after interference was assessed by asking Joe to recall three words after a 5 to 10 minute delay with other questions interspersed. He recalled all three words, suggesting no impairment of delayed recall. Recent memory appeared to be good. He was able to describe what he had eaten for breakfast, and he could recall what his activities had been during the previous evening. His remote memory was intact. He knew his birthday, where he was born, and the highest grade that he attained in school. Behavioral observations indicated that Joe's inability to recall details about his birth family and placement history may have been due to evasiveness rather than amnestic or fugue qualities.

During the psychosocial history interview, Joe's targeted adoptive parents reported that he was easily distractible with fleeting attention to task and had often engaged in daydreaming. During the interview he was observed to show lapses of attention and to focus on non-threatening topics such as video games. During the interview he was also seen to have a great deal of tic-like behavior and exaggerated gross motor movements.

Throughout the assessment, Joe's amount and productivity of speech was seen as normal and the coherence and progression of his speech was appropriate when answering questions. However, Joe appeared to be pre-occupied by video games and often displayed a flight of ideas. Joe was further seen to have reluctance to expand on questions related to his birth family. He displayed a lack of insight into the relationship between his adoption process and problems within his birth family. His trends of thought were scattered. However, his general intellectual abilities were seen to fall within the above-average range. His judgment was considered to be appropriate.

SUMMARY AND FUTURE RESEARCH

The assessment and diagnosis of RAD requires extensive information about a client both from the perspective of historical and current functioning. In order to gather this information, it is important to use a variety of assessment techniques. However, because the field research related to the characteristics and prevalence of the disorder is in its infancy, few efforts have been made to develop assessment protocols specific to RAD. The research efforts that have been made have focused on the self-report of parents or guardians. Although the information gathered from these types of protocols is useful in developing a diagnostic picture, the authors of these scales argue that they should not be used in isolation and that assessment should be multimodal and multifaceted. To compound the diagnostic dilemma, there has been little effort to identify procedures and scales to complement the attachment-specific self-reports. We recommend a combination of diagnostic interviews, global assessment scales, attachment-specific scales, and behavioral observations to identify the disorder and provide targets for intervention.

In our clinic, we have identified several procedures and instruments that have enhanced our ability to diagnose RAD and to differentiate it from other behavioral disorders of childhood and adolescence. We begin the assessment process by gathering information from a number of global rating scales--Child Behavior Checklist (Achenbach, 1991), Behavior Assessment System for Children (Reynolds & Kamphaus, 1992), Eyberg Child Behavior Inventory (Eyberg, 1999a), and Sutter-Eyberg Student Behavior Inventory (Revised; Eyberg, 1999b)) that results in information about home and school settings. Each of these instruments has proven useful in identification of problem behaviors. We also use the RADQ and the Reactive Attachment Disorder Questionnaire to gather information specific to attachment-related behaviors. Because time is of the essence for many mental health counselors serving strictly in the practitioner role, we recommend that these behavior rating scales be mailed to the home and school environments prior to the initial clinic appointment. The family can then bring this information with them and provide it to the clinician prior to the interview. Once the family arrives, we begin with semi-structured clinical interviews of both the guardians and the identified client. As part of our interview, we gather extensive psychosocial history data including information about current referral concerns, biological parental history, medical history, developmental history, mental health history, school history, disciplinary practices, legal and victim issues, and expectations of adoptive or foster parents. Next, we use a set of analogue observations in evaluating familial interactions. Finally, we may conduct standard intelligence and achievement testing when indicated (e.g., academic failure).

Through our assessment of Joe, we concluded that he met the criteria for the inhibited type of RAD. According to the psychosocial history, Joe developed a set of symptoms that began between ages 1 and 2. Consistent with diagnostic criteria, the adoptive parents reported that Joe was excessively inhibited in his social interactions and that he demonstrated an inability to exhibit appropriate selective attachments. It was clear from our assessment that Joe did not have a developmental delay or any pervasive developmental disorder that would explain these symptoms Also clear was the fact that Joe and his siblings experienced chronic pathogenic care over a period of several years. According to the birth mother, there was a clear correlation between the development of Joe's symptomology and the onset of abuse.

To further aid in diagnosis, results of the CBCL were clinically significant for internalizing problems related to thoughts and attention. Results of the BASC further suggested internalizing problems. However these results suggested that Joe was at-risk for anxiety problems. The ECBI and the SESBI-R did not reveal any significant behavior problems, a finding that would be consistent with the inhibited type of RAD. Both the results of the RADQ and the Reactive Attachment Disorder Questionnaire indicated that Joe met the criteria for the inhibited type of RAD. In-clinic observation of Joe further supported our eventual diagnosis. The combination of results derived from our clinical interviews, assessment instruments, and observations provided a clear diagnostic picture and provided us with enough information to develop detailed directions for treatment intervention.

Although we have found the proposed assessment procedures to be useful in our clinic, further research needs to be conducted to standardize this process to ensure consistency in diagnostic and assessment criteria. For example, there is overlap in some of the behavioral rating scales that we use. Because reimbursement for assessment activities may indeed be a problem for many mental health counselors, they could choose to limit their administration of these instruments to one or two scales. The CBCL is often used in more clinical settings; whereas the BASC is utilized more often in school settings because of the additional scales addressing school problems, social skills, and adaptive behavior. Therefore, the mental health counselor could choose to eliminate one of these instruments. We recommend that the mental health counselor administer the ECBI as it provides information about specific behavioral problems that can be used as targets for intervention. Furthermore, the ECBI has been proven useful for evaluating treatment gains. Additionally, we recommend that mental health counselors administer the RADQ and Reactive Attachment Disorder Questionnaire as these instruments are specifically designed to address attachment issues. Another way that the process could be streamlined is to eliminate input from the school setting. RAD does not require the presence of symptoms across two environments like other disorders (e.g., Attention Deficit Hyperactivity Disorder); however, such information could assist in the diagnostic and treatment process.

Mental health counselors strained for time could also reduce the number of behavioral observations conducted. For example, three 5-minute sessions could be conducted instead of seven sessions. The mental health counselor could conduct the LA, HA, and HD conditions in order to obtain information about problematic behaviors, engagement, and compliance rates. Although this combination of conditions would not be as rigorous as the seven session combination, the information gleaned from these sessions would still be helpful in formulating a diagnosis and generating treatment goals.

As mentioned previously, future research needs to be conducted to standardize the process for assessing RAD and treating children with the diagnosis. Future studies may reveal that only attachment-related questionnaires and rating scales are needed to diagnose RAD. Furthermore, additional research may reveal that the three session approach to observing parent-child interactions is just as effective as the seven session approach suggested here. Finally, information across two environments (e.g., school, home) may not be needed in order to diagnose or intervene with children with RAD. However, it is extremely important to note that such decisions have not been made and until such time, mental health counselors are ethically, legally, and professionally obligated to conduct thorough evaluations that are multi-modal and multi-informant in nature in order to effectively provide differential diagnosis, evaluate comorbidity, and establish comprehensive treatment plans.

Table 1
Diagnostic Differences in Childhood Disorders

DISORDER            KEY COMPONENTS            DIFFERENTIAL DIAGNOSIS
                    OF DIAGNOSIS              FACTORS

Reactive            Overall symptomotology    Symptomotology similar to
Attachment          resulting from            other disorders listed,
Disorder of         pathologenic care         but occur due to ongoing
Infancy or Early                              pathogenic care (see DSM-
Childhood 313.89    Inhibited Type-           IV-TR- for definition)
                    Persistent inability to   Social and attachment
                    engage in or respond in   issues present (see
                    appropriate social        differences in two types)
                    relationships

                    Disinhibited Type-        May exhibit language,
                    Little to no discrimi-    behavior, and communica-
                    nation given in           tion deficits, but
                    selecting attachment      typically are a result of
                    figure                    poor care, versus a
                                              developmental origin
Conduct Disorder    Persistent pattern of     May have been in a
312.81 (Childhood   severe behavior that      restrictive emotional
Onset)              violates the basic        environment
                    rights of others
                                              Anti-social and
                                              aggressive behaviors lead
                                              to decreased ability to
                                              develop adequate
                                              social relationships
Oppositional        Pattern of defiant        Behaviors overall are not
Defiant Disorder    behavior toward           intended to harm others
313.81              authority figures
                                              Social difficulties occur
                    Behaviors centered        due to defiant behaviors
                    around deflecting self
                    blame
Attention-Deficit   Pattern of impulsivity    Disinhibited social
Hyperactivity       and hyperactivity         behavior results from
Disorder 314.01     causing disturbance in    impulsivity versus
(Impulsive Type)    functioning at home or    seeking comfort from
                    school                    attachment figure

                    Social deficits that      Behaviors are not being
                    may occur stem from       done to intentionally
                    these impulsive           bother another individual
                    behaviors
Autistic Disorder   Restricted level of       Social impairments stem
299                 interests and             from restricted field of
                    activities                interest and communica-
                                              tion deficits present
                    Normally accompanied by
                    marked disturbance in
                    communication and
                    repetitive, stereotypic
                    behaviors

Note: All information taken directly from DSM-IV-TR (American
Psychiatric Association, 2000)

Table 2--Psychometric Instrument Reference For Assessment Protocol

Instrument       Number     Age            Type of        Key Outcomes
                of Items   Norms        Report/Setting      Measured

Child           100-118    2-18 years   Multiple types    Access
Behavioral      Items *                 of reports.       behavior
Checklist                               Parent, teacher   problems as
(CBCL) (a)                              and self report   well as
                                        forms available   abilities
                                        (see below)
Parent          See        2-3 years;   Parent Report     Home
Report Form     Above      4-18 years   Home              environment
of CBCL                    (2 age-      environment       behavior
                           related                        reference
                           forms)
Teacher         See        6-16 years   Teacher Report    Classroom
Report Form     Above                   Classroom         behavior
of CBCL                                 Setting           reference
Youth Self-     See        11-18        Child Report      Used as a
Report Form     Above      years                          self
of CBCL                                                   inventory
Behavior        109-131    2-1/2-18     Child Report      Classifies
Assessement     Items *    years                          developmen-
System for                                                tally
Children                                                  appropriate
Revised                                                   thinking from
(BASC) (b)                                                pathological
                                                          thinking
Eyberg Child    36 Items   2-16         Parent Report     Determines
Behavior                   years        Home              the severity
Inventory                               environment       of conduct
(ECBI) (c)                                                problems
Sutter-Eyberg   38 Items   2-17         Teacher Report    Determines
Student                    years        Classroom         the severity
Behavior                                Setting           of conduct
Inventory                                                 problems
Revised                                                   in school
(SESBI-R) (d)                                             setting
Randolph        30 items   5-18         Parents Report    Distinguishes
Attachment                 years                          between
Disorder                                                  attachment
Questionnaire                                             problems and
(RADQ) (e)                                                behavioral
                                                          disorders
Reactive        17 items                Parents Report    Used to
Attachment                              Home              identify
Disorder                                environment       RAD symptoms
Question-
naire (f)

Instrument      Application to           Specific Test
                     RAD               Properties to Note

Child           Distinguishes          Multiple forms
Behavioral      between                applicable to
Checklist       internalizing and      many
(CBCL) (a)      externalizing          environments
                problems that can
                be symptoms of
                either type of PAD
Parent          Can identify RAD       Multiple scales
Report Form     like behaviors in      and subscales
of CBCL         multiple               that measure a
                environments           variety of
                                       behaviors
Teacher         Can identify RAD       Same scales as
Report Form     like behaviors in      Parent Report
of CBCL         multiple               Form with
                environments           addition of
                                       Academic
                                       performance
                                       and Academic
                                       Functioning
Youth Self-     Identifies a           Same scales as
Report Form     pathological           Parent Report
of CBCL         view of self           Form excluding
                                       the school portion
Behavior        Identifies             Used to develop
Assessement     pathological           treatment plans
System for      thinking leading to
Children        further evaluation
Revised         for a potential
(BASC) (b)      diagnosis

Eyberg Child    Conduct behaviors      Distinguishes and
Behavior        are often expressed    examines
Inventory       in RAD                 inappropriate
(ECBI) (c)                             behaviors
                                       characteristic
                                       of RAD
Sutter-Eyberg   Can be used to         Also can be
Student         quantify RAD           pre and post
Behavior        symptoms;              testing for
Inventory       Aids in differential   evaluation of
Revised         diagnosis of RAD       progress
(SESBI-R) (d)
Randolph        Aids clinician in      Does not follow
Attachment      making a               DSM-IV-TR;
Disorder        differential           Based on
Questionnaire   diagnosis              Randolph's
(RADQ) (e)                             definition of RAD,
                                       recommended as
                                       a screen
Reactive        Distinguishes          Follows
Attachment      BAD subtypes, as       DSM-IV-TR;
Disorder        inhibited and          European
Question-       disinhibited types     instrument
naire (f)                              generalization
                                       concerns

* Item numbers vary according to forms used.

(a) (Achenbach, 1991). (b) (Reynolds & Kamphaus, 1992). (c) (Eyberg,
1999a). (d) (Eyberg, 1999b). (e) (Randolph, 2000). (f) (Minnis, 1996).

REFERENCES

Achenbach, T. M. (1991). Manual for the Child Behavior Checklist/4-18 and 1991 Profile. Burlington, VT: University of Vermont, Department of Psychiatry.

Alien, K. D., & Shriver, M. D. (1998). Role of parent-mediated pain behavior management strategies in biofeedback treatment of childhood migraines. Behavior Therapy, 29, 477-490.

American Psychiatric Association. (2000). Diagnostic and statistical manual of mental disorders (4th ed. Text rev.). Washington, DC: Author.

Boris, N. W., Zeanah, C. H., Larrieu, J. A., Scheeringa, M. S., & Heller, S. S. (1998). Attachment disorders in infancy and early childhood: A preliminary investigation of diagnostic criteria. The American Journal of Psychiatry, 155, 295-297.

Carr, E. G., & Durand, V. M. (1985). Reducing behavior problems through functional communication training. Journal of Applied Behavior Analysis, 18, 11-26.

Cooper, L. J., Wacker, D. P., Thursby, D., Plagmann, L. A., Harding, J., Millard, T., et al. (1992). Analysis of the effects of task preferences, task demands, and adult attention on child behavior in outpatient and classroom settings. Journal of Applied Behavior Analysis, 25, 823-840.

Eyberg, S. M. (1999a). Eyberg child behavior inventory. Odessa, FL: Psychological Assessment Resources.

Eyberg, S. M. (1999b). Sutter-Eyberg student behavior inventory--Revised. Odessa, FL: Psychological Assessment Resources.

Hanson, R. F., & Spratt, E. G. (2000). Reactive attachment disorder: What we know about the disorder and implications for treatment. Child Maltreatment, 5, 137-145.

Hembree-Kigin, T. L., & McNeil, C. (1995). Parent-child interaction therapy. New York: Plenum.

Kratochwill, T. R., Sheridan, S. M., Carlson, J, & Lasecki, K. L. (1999). Advances in behavioral assessment. In C. R. Reynolds & T. B. Gutkin (Eds.), The handbook of school psychology (pp. 350-382). New York: John Wiley.

Lynam, D. R. (1996). Early identification of chronic offenders: Who is the fledgling psychopath. Psychological Bulletin, 120, 209-234.

Minnis, H. (1996). Reactive attachment disorder: Usefulness of a new clinical category. The Journal of Nervous and Mental Disease, 184, 440.

Minnis, H., Pelosi, A. J., Knapp, M., & Dunn, J. (2001). Mental health and foster career training. Archives of Disease in Childhood, 84, 302-306.

Moore, J. W., Doggett, R. A., Edwards, R. E, & Olmi, D. J. (1999). Using functional assessment and teacher-implemented functional analysis outcomes to guide intervention for two students with Attention-Deficit/Hyperactivity Disorder. Proven Practice, 2, 3-9.

Moore, J. W., Edwards, R. P., Wilczynski, S. M., & Olmi, D. J. (2001). Using antecedent manipulations to distinguish between task and social variables associated with problem behaviors exhibited by children of typical development. Behavior Modification, 25, 287-304.

Muladdes, N. M., Bilge, S., Alyanak, B., & Kora, M. E. (2000). Clinical characteristics and treatment responses in cases diagnosed as reactive attachment disorder. Child Psychiatry and Human Development, 30, 273-287.

Randolph, E. (2000). The manual for the Randolph Attachment Disorder Questionnaire (3rd ed.). Evergreen, CO: The Attachment Center Press.

Reynolds, C. R., & Kamphaus, R.W. (1992). Behavior Assessment System for Children (BASC). Circle Pines, MN: AGS Publishing.

Richters, M. M., & Volkmar, E R. (1994). Reactive attachment disorder of infancy or early childhood. Journal of the American Academy of Child and Adolescent Psychiatry, 33, 328-332.

Rosenstein, D. S., & Horowitz, H. A. (1996). Adolescent attachment and psychopathology. Journal of Consulting and Clinical Psychology, 64, 244-253.

Sheperis, C. J., Renfro-Michel, E., & Doggett, R. A. (2003). In-home treatment of reactive attachment disorder in a therapeutic foster care system: A case example. The Journal of Mental Health Counseling, 25, 76-88.

Taylor, J. C., & Cart, E. G. (1992). Severe behavior problems related to social interaction. I: Attention seeking and social avoidance. Behavior Modification, 16, 305035.

Taylor, J. C., Ekdahl, M. M., Romanczyk, R. G., & Miller, M. L. (1994). Escape behavior in task situations: Task versus social antecedents. Journal of Autism and Developmental Disorders, 24, 331-344.

Watson, T. S., Gresham, E M., & Skinner, C. H. (2001). Introduction to the mini-series: Issues and procedures for implementing functional behavior assessments in schools. School Psychology Review, 30, 153-155.

Carl J. Sheperis, Ph.D., NCC, LPC, is an assistant professor in the Community Counseling Program. R. Anthony Doggett, Ph.D., is an assistant professor in the School Psychology Program. Nicholas E. Hoda is a doctoral student in the School Psychology Program. Tracy Blanchard, Edina L. Renfro-Michel, NCC, and Sacky H. Holdiness are doctoral students in the Community Counseling Program. Robyn Schlagheck is a research assistant. All are with the Department of Counselor Education and Educational Psychology, Mississippi State University, Mississippi State. E-mail: [e-mail]