NADD Bulletin Volume IX Number 6 Article 3

Complete listing

A model of treatment to optimize behavioral health outcomes for individuals with cognitive disabilities

Dina McFalls, M.S. 
Carol Persons, M.D.
Amy Nemirow, Ph.D.
Philadelphia Coordinated Health Care

A model of treatment to optimize behavioral health outcomes for individuals with cognitive disabilities

What if we were able to conceptualize the elements that would be needed to create better mental health outcomes for people with cognitive disabilities?  Since the deinstitutionalization movement began to take shape and people with cognitive disabilities moved to less restrictive community homes, the community mental retardation/developmental disabilities system has struggled with how to provide support to people whose challenges not only include cognitive disabilities but also mental illness.  Models of support that exist include approaches such as Applied Behavioral Analysis and its component Functional Assessment, the use of psychoactive medications, Positive Approaches, Person Centered Planning, and Biography, as well as a variety of team approaches.  Much has been written about each approach and the successes and failures.

Further, approaches that have been shown to be most effective are those that combine the expertise of all the supporting and treating professionals.  All of these interdisciplinary approaches hold as their basic tenant the need for communication between supportive team members.  Lew, Zaslow-Crème, and Lepler (1990), noted the brainstorming of direct line staff and their consultant as being essential to good management, as well as willingness to adjust strategies based upon data and important anecdotal information about relevant factors.


At the core of accurate diagnostic assessment for people who have cognitive disabilities as well as the suspicion of mental illness is the precise definition of psychiatric symptoms, the collection of accurate data regarding frequency of the target symptoms, the understanding of the diagnostic shadowing that may cloud symptom identification, and the impact of neurological deficits. (Sovner & Pary, 1993)  During the 1980s, many studies were undertaken to look at the accuracy of medication prescription for individuals living in institutional setting using the multidisciplinary process.  

However, even with acknowledgement of the need for the interdisciplinary approach, community mental retardation/developmental disabilities and mental health systems have been slow to adopt formalized approaches to supporting people with mental illness who also have cognitive disabilities.  While Sovner and Hurley (1992) as well as Sovner, Beasley, Hurley, and Silka (1995) have put forth specific workable models, no approach has been adopted broadly by the community mental health system, and thus treatment of the individuals we support has been more art than science, more intuition than algorithm, and more guess work than accurate identification of target symptoms and good data collection.

We would like to suggest an approach that includes all of the elements of treatment and accepted medical practice as well as behavioral supports.  This is a schema that is based upon the medical principle of the need for a match between presenting symptoms, the diagnosis, and the chosen treatment methodology.  This model is also based upon the need for team communication as a way to reach optimum outcomes for the individual.  Through the use of this model we find that each element of support plays an important role, but all must work in concert.  Further, while we recognize that a form cannot dictate good treatment in and of itself, the use of the form we will describe allows teams and professionals to look at all elements of the individual's supports and how they work together to support the person.  This form can clearly highlight what each individual's needs are and how to improve team support.

Background and history

In 1994, the City of Philadelphia was found to be in contempt of the 1972 Court Order that stipulated not only the deinstitutionalization of some 1200 individuals from Pennhurst State School and Hospital, but also the oversight of class member well being once in the community (Terri Lee Halderman et al v. Pennhurst State School and Hospital et al. Civil Action No. 74-1345). In 1990, the Quality Enhancement & Monitoring Group of Temple University submitted a report on the Pennhurst class members to the Pennhurst Implementation Team of the Philadelphia Office of Mental Retardation that related concerns about the "prevalence of prescriptions."  Forty-two percent of class members were taking psychotropic medications at that time.  The concerns of this report centered on the percentage of people who were taking psychotropic medications as well as the instances of the use of multiple medications (11%). The 1994 order from the Court to Philadelphia required, among many other things, that physicians treating Philadelphia class members, "certify at least every ninety days that the continued use of such medication is not excessive or unnecessary according to accepted standards of medical practice and that the prescription otherwise is in accordance with accepted standards of medical practice."   In addition, to address the Courts' concerns, Philadelphia was ordered to review and evaluate the medical records of each Philadelphia class member for whom psychotropic and/or anti- seizure medication is prescribed through the use of an independent physician reviewer.   The independent physician was specified to certify in writing: (a) the reasons for the medication, and (b) that each prescription is in accordance with accepted standards of medical practice. Thus began the journey of Philadelphia Coordinated Health Care (PCHC) and the creation of a diagnostic and treatment formulation similar to and derived from that espoused by Robert Sovner. 

Following the court ordered reviews and the refinement of both the data collection tool as well as the review process we found, "that interdisciplinary teams can achieve significant improvement in the match between diagnosis, symptoms and medications when standard protocols are developed to guide clinical practice.  The review process also improved the communication between the psychiatrist, care providers and behavior specialists and resulted in a substantial reduction in the use of polypharmacy" (Franczak & Persons, 1999), Further, in an NIMH grant application, it was asserted that the Sovner and Pennhurst Models, now known as the PCHC Model, have in common the interdisciplinary treatment methodologies and treatment professionals using standardized treatment protocols (McFalls, 1999).

Pennhurst Reviews  

To accomplish the semiannual review of approximately 240 of the 600 

or so member class (47% of the class were known to have prescribed psychotropic medications) a form was devised to collect data.  This initial form, developed by Philadelphia's MH/MR Services Medical Director and Mental Retardation Services Associate Medical Director, asked some basic questions including verification of tardive dyskinesia screening, listed medications and diagnoses, and verified that each individual had a DSM-IV psychiatric diagnosis.  The original format was based upon the principles put forth previously by Sovner and colleagues, as described previously.  However, as the process of review by psychiatrists continued, it became clear that residential agencies and reviewers alike were concerned that the form did not paint a complete picture of the individual and their supports, did not demonstrate team communication, and it did not meet licensing requirements, necessitating duplicative form completion.  Therefore, a committee of residential agency staff was convened with the goal of creating a form that captured the court mandated information relative to good treatment, meet licensing requirements with regard to the documentation of the use of medications, meet requirements to document team communication, and meet agency requirements for medical visits.  The resultant document became known as the 90-day form and was approved by the local psychiatric community, the practicing behavioral specialists as well as the court. Most recently, the 90-day form has morphed into the Behavioral Health Team Review Form (BHTRF) to allow it to be used by community psychiatrists and outpatient providers as their reporting/billing document.  Relatively minor changes were needed to accommodate this use and the integrity of function was maintained.

Triangular PCHC Model

Over the years varying models for the assessment and treatment of people with cognitive disabilities and mental illness have emerged.  In 1992, Robert Sovner and Ann DesNoyers Hurley put forth the model of diagnosis and treatment that forms the basis for much of the treatment formulation that exists today.  One issue raised by Sovner and Hurley relates to the difficulty of sorting out symptoms of mental illness from behaviors that serve other functions for the person with cognitive disabilities. Some researchers have enumerated some of these difficulties by parsing the issue into the following issues:  (a) diagnostic overshadowing, (b) intellectual distortion, (c) cognitive disintegration, and (d) baseline exaggeration.  However we slice it, the origin of any specific behavioral symptom is going to be difficult to clarify.  This difficulty is at the heart of the ongoing concern in targeting psychiatric treatment and behavioral support.

The model we are describing bases much of its foundation upon the Sovner and Hurley (1992) model as well as a good medical practice model.  The PCHC Model, formerly described as the Pennhurst Model (McFalls, 1999), which could be depicted for those of us with a visual learning style in the shape of a triangle, solidifies the existing medical approach of a match between an individual's target symptoms, the diagnosis, and the treatment (medication in the case of psychiatric illness).

Insert Figure 1 about here

This triangular model works with the principles laid out by Sovner and Hurley (1992) in the need to clearly and objectively describe the behavioral symptoms that are the manifestation of psychiatric illness.  The PCHC Model builds on the idea of clearly illuminating the rationale for the prescription of medication based upon the target symptoms.  Crucial to the PCHC Model is the need, to the greatest extent possible, to separate target symptoms of mental illness from behaviors that serve other functions. (Lowry  & Sovner, 1991. Continuing with our triangular description, if any one of the three triangle points (symptoms, diagnosis and/or treatment) is incorrectly identified, the treatment outcome for the individual becomes less than optimal.  In addition, should any member of the team fail to communicate with other members, treatment becomes less effective.

Finally, the underpinning for this model of treatment is good team communication.  The prevailing model of all treatment and support of individuals with cognitive difficulties has been based on the concept of team support.  In working with people who have the additional challenge of mental illness, the concept of the treating clinicians working with the residential and familial team to achieve the best treatment has been explored by Silka and Hauser (1997), in which they assert that psychiatric care of individuals with mental retardation is most effectively rendered when the psychiatrist uses an interdisciplinary team model. Further clarifying how important team communication can be, they state that a successful treatment outcome may depend in part how well the psychiatrist can bridge the gap between different conceptual models (medical versus habilitative), clinical languages, and organizational styles. The PCHC model tries to do just that: bridge the gap and create opportunities to present information from the team to the psychiatrist in an organized format that relies on both habilitative and medical information.

What is a target behavior?

The clear identification of the target symptoms of mental illness is one of the pivotal points in evaluating the effectiveness of treatment for people with cognitive disabilities and mental illness.  Without clear identification of the target symptoms, the treating physician cannot make an educated guess as to the diagnosis; without accurate identification of target symptoms the behavioral specialist cannot accurately devise a support plan, and without clear identification of the target symptoms, the direct support staff cannot collect data on frequency and report on progress toward healthy outcomes.  In other words, without good team communication, psychiatric treatment becomes pure guesswork and outcomes are less than optimal.

For example, the identification of target symptoms seems like it should be a simple proposition.  After all, the person may be exhibiting disruptive behavior, so that means the person needs some medication, right?  However, in the population of people with cognitive disabilities, there are many who cannot identify verbally what their sensory experiences are and what may underlie the behaviors that are observed.  Does head banging mean the person is hearing voices, has a headache, is frustrated or any number of other postulations?  Does aggression mean the person does not want to be bothered, is responding to internal stimuli, hates the person next to them, doesn't feel well, or is just having a bad day?  How do we sort out what is a target behavior of a mental illness and what is just a behavior?  Further, if we do not sort out target behaviors of mental illness from behavior that is a result of medical illness or environmental factors, or just plain part of the person's personality, we, too, will be guilty of chemical restraint rather than treatment of mental illness. 

The first step is getting clear on target symptoms by ruling out medical issues that require treatment, ruling out environmental issues that require address, and ruling out those annoying behavioral habits that have been acquired over time but that are not symptoms of mental illness.  Whatever the behavioral approach, target symptoms of mental illness will not improve without the use of appropriate medical treatment.  Conversely, some behavioral approaches may well address behaviors that are a result of environmental issues or that are just behaviors that are bothersome to staff or the rest of society, but that are not caused by internal psychiatrically-driven stimuli.  It is a team problem to clarify just what exactly a behavior is communicating.  It is a team challenge to explore all possible causes of a behavior before applying the label of a target symptom.

Diagnosis and treatment

Once the team has clearly identified the target symptoms of mental illness, the psychiatrist must then use family and medical history as well as current status to determine a working diagnosis.  The diagnosis then will dictate a treatment: in this case, medication.  If the diagnosis is correct and the medication chosen is correct and in the correct dosage, the target symptoms should decrease.  If the target symptoms being tracked by the team do not decrease, then the treating physician will postulate a different diagnosis, a different medication, or a dosage change in medication.  Thus the testing of a diagnosis is based upon the accurate identification of the target symptom and data collection about the frequency, intensity, and duration of a target symptom.  Likewise, if the diagnosis postulated is incorrect, its concomitant medication is incorrect; the team will see no amelioration in the target symptom.  Behaviors that are not target symptoms will not be effected by medications unless medication sedates the individual.

What is the Behavior Health Team Review Form?

The Behavioral Health Team Review Form (BHTRF) is a three-part form that grew out of the work with the Pennhurst class of individuals.  The BHTRF, formerly known as the 90-Day form, documents the status of the individual from the perspective of the living situation provider (family or residential), and the behavioral support consultant as well as the treating physician or psychiatrist.  Because of its unique orientation, this three-part form can clearly show good or poor team communication; it can clearly demonstrate the relationship between the target symptoms identified, the related psychiatric diagnosis and the treatment being provided (medication and behavioral support). 

Page one is completed by the staff who care for the individual and enumerates changes in health status, current medications, and diagnoses, as well as any changes in the person's life or environment.  Page one also includes tardive dyskinesia screening and allergies.  The behavioral support person completes Page two.  This page clearly lists the target symptoms of the mental illness that exist for this person and includes data collected by the team as to the frequency of those symptoms.  Page three, completed by the treating physician after having reviewed pages 1 & 2, includes the physician's current diagnoses, progress towards health, any medication changes, and signatures.

If any portion of the BHTRF is not completed, it can be assumed that a leg of the triangle of support is missing.  What seems clear, after looking at the data from ten years of use of the form and working with local teams, is that all three legs of support must be in place and working in concert to provide a comfortable seat for an individual coping with cognitive disabilities and mental illness.

How does this optimize mental health outcomes?

The ultimate question of the effectiveness of this model rests not only in the data results of testing for a match between target symptoms, diagnosis, and treatment, but in looking for improved outcomes in terms of functioning.  To that end, we will be working with Temple Episcopal Hospital to test the effectiveness of the model described.  It is hoped that by using the medical principle of the match between the symptom, diagnosis, and the treatment as well as improved team communication, that outcomes for people with cognitive disabilities and mental illness can improve. 



Franczak, M. and Persons, C.  (1999). Independent review of Pennhurst Class Members Receiving Psychotropic Medications.  Philadelphia, PA: Author.

Hauser, M.J. (1997). The role of the psychiatrist in mental retardation. Psychiatric Annals, 27,170-174. 

Lew, M., Zaslow-Crème, B., & Lepler, S. (1990). Effective consultation with community programs. The Habilitative Mental Healthcare Newsletter, 9, 65-69.

Lowry, M., & Sovner, R. (1991).  The functional existence of problem behavior: a key to effective treatment.  The Habilitative Mental Healthcare Newsletter, 10, 59-63.

McFalls, D. J. (1999). Proposal for Exemplary Practice. Philadelphia, PA: Philadelphia Coordinated Health Care.

Quality Enhancement & Monitoring Group of Temple University. (1990). Evaluation of the Well-Being of Pennhurst Class Members Living in the Community in 1988-89, Results of Temple Monitoring in Philadelphia County. Philadelphia, PA: Temple University. 

Silka, V.R., & Hauser, M.J. (1997). Psychiatric Assessment of the Person with Mental Retardation, Psychiatric Annals, 27, 3.

Sovner, R., and Pary, R. (1993). Affective disorders in developmentally disabled persons. Psychopathology in the mentally retarded. New York: Allyn and Bacon.

Sovner, R., Hurley, A.D., Beasley, J., & Silka, V. (1995). Commentary: Fifteen-minute medication follow up visits. The Habilitative Mental Healthcare Newsletter, 14, 63-65.

Sovner, R. & DesNoyers Hurley, A. (1992). The Diagnostic-Treatment Formulation for Psychotropic Drug Therapy.  The Habilitative Mental Healthcare Newsletter. 11, 81-86. (reprinted in Mental Health Aspects of Developmental Disabilities, 9, 34-40.)

Terri Lee Halderman et al v. Pennhurst State School and Hospital et al. Civil Action No. 74-1345.  March 28, 1994.