The Multimodal Functional Modelâ€“Advancing Case Formulation

The Multimodal Functional Model–Advancing Case FormulationBeyond the “Diagnose and Treat” Paradigm: Improving Outcomes

and Reducing Aggression and the Use of Control Procedures inPsychiatric Care

Richard H. HunterClinical Outcomes Group, Inc., & Southern Illinois

University School of Medicine

Sandra WilknissThresholds Institute

William I. GardnerUniversity of Wisconsin–Madison

Steven M. SilversteinUniversity of Medicine and Dentistry of NewJersey–University Behavioral HealthCare, and

Robert Wood Johnson Medical School

Improving clinical outcomes, treating aggression, and reducing the use of restraints andseclusion are major concerns to providers of psychiatric services. Devising precisetreatments directed by causal hypotheses that address the functional properties ofmaladaptive behaviors are essential for reducing aggression, increasing understandingof causal factors, and enhancing individual control and adaptive choices. This articledescribes the Multimodal Functional Model (MFM), a sophisticated case formulationand treatment model that moves beyond the “diagnose and treat” paradigm andprovides interventions that address causal factors, reduce aggression, and prevents theuse of restraints and seclusion in psychiatric settings.

Keywords: functional behavioral assessment, case formulation, treatment planning, restraints andseclusion, inpatient treatment, multimodal functional model

The Multimodal Functional Model (MFM)evolved as a sophisticated case formulation andtreatment model from a rich history of behavioranalytic research aimed at assessing and under-standing underlying causes of unwanted behav-iors. O’Brien and Haynes (1993) stated:

Behavioral assessment, with its empirical and prag-matic emphasis on identifying important causalfunctional relationships between setting events andbiobehavioral responses, is the most appropriate and

responsible means of meeting the specialized assess-ment demands of contemporary inpatient psychiatry.These specialized demands require that a comprehen-sive assessment be conducted to identify the compo-nents of disordered and/or adaptive behaviors and thevariables that control them. The assessment informa-tion can then be used to (1) design interventions thatbring about a modification of patient behavior and (2)monitor treatment effectiveness. (p. 39)

Several terms have been used in the researchliterature as these concepts developed. Haynes(1990) used the terms “functional analysis,” Per-sons (1989, 1992) discussed “case formulation,”Wolpe and Turkat (1985) described “behavioralcase formulation,” and Wincze (1982) wroteabout “functional behavioral analysis.” These ap-proaches to client assessment, case formulation,and treatment have had an extensive history in thefield of developmental disabilities and those du-ally diagnosed with mental illness and mental retar-dation. The comprehensive MFM evolved from thework of Gardner and Cole (1987), Gardner andHunter (1995, 2003), Gardner and Sovner (1994),and Griffiths, Gardner, and Nugent (1998).

Richard H. Hunter, Clinical Outcomes Group, Inc., andDepartment of Psychiatry, Southern Illinois UniversitySchool of Medicine, Springfield, Illinois; Sandra Wilkniss,Thresholds Institute, Chicago, Illinois; William I. Gardner,Emeritus Professor, Rehabilitation Psychology Program,University of Wisconsin-Madison, Madison, Wisconsin;Steven M. Silverstein, Department of Psychiatry, Universityof Medicine & Dentistry of New Jersey-University Behav-ioral HealthCare, and Robert Wood Johnson MedicalSchool, Piscataway, New Jersey.

Correspondence concerning this article should be ad-dressed to Richard H. Hunter, Clinical Outcomes Group,Inc., 10202 Briggs Road, Marion, IL 62959-5844. E-mail:[email protected]

Psychological Services Copyright 2008 by the American Psychological Association2008, Vol. 5, No. 1, 11–25 1541-1559/08/$12.00 DOI: 10.1037/1541-1559.5.1.11

11

There are many functional assessment ap-proaches and instruments (see Bellack & Hersen,1998; O’Neill et al., 1997; Sturmey, 1996). TheMFM utilizes any number of these approachesas it may apply to a specific individual,situation, or context. The major contribution ofthe MFM is that it has been designed to inte-grate a comprehensive array of functional as-sessment procedures and observation and datacollection techniques into an integrated, com-prehensive, biopsychosocial case formulationand treatment plan. It is not focused merely onunwanted behaviors. Case formulation using theMFM approach requires attention to both inter-nal (medical, biological, genetic, perceptual, cog-nitive) areas as well as external (situational, envi-ronmental, social, contextual) areas of analysis.The MFM uses empirical analysis to understandthe entire array of biological, psychological,and social experiences and integrates this infor-mation into an empirical, hypothesis-testing,longitudinal approach that utilizes functional(causal) hypotheses to both drive and evaluatespecific treatment interventions.

The major contrast is not between the MFMand the functional assessment/analysis ap-proaches to case formulation but rather theMFM and a typical “diagnose and treat” para-digm that is so pervasive in the medical fieldtoday. Many clinicians believe that case formu-lation is complete when assigning a Diagnosticand Statistical Manual of Mental Disorders,Fourth Edition (DSM–IV) diagnosis and identi-fying the general interventions that are com-monly associated with that particular diagnosis.When the patient does not respond to the ge-neric interventions, then consideration is givento finding another diagnosis that permits anadditional array of generic interventions. Con-tinual attention to finding “the right” DSM di-agnosis and being satisfied with this gross levelof etiological information, diverts attentionfrom more sophisticated analytic approachesthat can actually lead to the discovery of caus-ative factors that lead to precise interventionsdirected at those factors. When the MFM isused, the paradigm changes from one that usesgeneric and primarily pharmacological ap-proaches that manage symptoms, to specific,causally directed interventions that have the po-tential to lead to enduring change. It is impor-tant to understand the difference in these caseformulation approaches because at least 40% of

people experiencing serious mental illness arenot substantially helped with the generic andchemical interventions derived from DSM diag-noses (Ciompi, 1980; Harding & Strauss, 1985;Kane, 1989; Silverstein et al., 2006; Stephens,1978; Torrey, 2001).

Attention to diagnosis as the end all of caseformulation has led to an overdependence onpsychiatric medication. Many hospitals todayuse a “drugs first” approach to all patients re-gardless of that person’s history related to med-ication response. Hunter (1999, 2000, 2001)described hospitals whose primary interventionmodel consists of “drugs and TV therapy.”Many people exposed to these limited case for-mulation and diagnostic practices experiencepoor clinical outcomes, high relapse rates, andunnecessary long-term suffering.

Using medication as the only focused inter-vention to “treat” psychiatric patients with co-occurring behavior disorders leads to poor out-comes as well. Attempting to resolve behavioralproblems with drugs also diverts attention frommore comprehensive behavioral assessments,analysis and treatment of the actual causativefactors.

A common error in the treatment of behaviorproblems in individuals with psychotic syn-dromes is to assume a direct link between thepsychiatric diagnosis, its primary symptoms,and the disruptive behaviors of concern. Thiscase formulation error is an illustration of diag-nostic overshadowing, assuming a major diag-nosis or symptom is the direct cause of othertreatable problems, conditions, or symptoms andthat treatments for the major symptom (e.g.,psychosis 3 antipsychotic drugs) will some-how be expected to resolve the remainingproblems, skill deficits, or unwanted behav-iors. Another variant of diagnostic overshad-owing involves overlooking symptoms ofother treatable disorders (e.g., depression) andassuming that those indicators (e.g., with-drawal, sadness, agitation, sleep disturbances,etc.) are components of another, more promi-nent disorder (e.g., mental retardation) (Reiss,Levitan, & Szyszko, 1982; Reiss & Szyszko,1983). This case formulation error directs atten-tion away from the assessment of the uniquefunctional properties of the behaviors of con-cern, the communication intent of the behav-iors, and the interactions between the personand the social and physical features of his or her

12 HUNTER, WILKNISS, GARDNER, AND SILVERSTEIN

environment. The “diagnose and treat” para-digm directs clinical attention to psychiatric di-agnoses and/or symptoms that may only beminimally linked to the expression of theaggressive or dangerous behaviors that were thecause of the hospitalization or that keep theperson from discharge into the community. Fur-thermore, it has come to represent a system ofmanaging or suppressing symptoms primarilywith medications and disregarding the treatmentof causal factors and specific antecedents whichwould result in learning, individual control andempowerment, and a more enduring recovery.Although medication may be critically impor-tant in the treatment of one or two of the majorpsychiatric symptoms, the symptoms most re-sponsive to the medication may not be thosemost strongly linked to a person’s challengingbehaviors and reasons for admission in the firstplace (Citrome et al., 2001). Hence, it is possi-ble for inpatient programs that rely primarily onmedications for treatment of behavior dysfunc-tion in people with serious mental illness toperpetuate a therapeutic regimen that is inade-quate for the person. Furthermore, settings thatare guided by this restrictive “diagnose andtreat” paradigm commonly over rely on restric-tive behavioral management and control proce-dures that include the inappropriate use ofchemical and/or physical restraints and seclu-sion (Donat, 1998, 2002; Paul & Lentz, 1977).Such programs experience unnecessary patientand staff injuries, rely on high-dose medicationregimens that produce adverse side effects anddiscomfort, and have high patient readmissionrates (Hunter, 1995, 1999, 2000, 2001).

Appropriate treatment of behavior dysfunc-tion in a person with serious mental illnessrequires sophisticated behavioral and psycho-logical interventions designed to understand thepurpose and meaning of the behavior in thepersonal and environmental contexts in which itoccurs. This involves identifying the stimulusconditions (both internal and external) associ-ated with the behavior. Resulting case formula-tion strategies would inform both patient andstaff of conditions and circumstances that in-crease the likelihood that the patient wouldchoose aggression or inappropriate behavior asa means of expression. Formal study of theinteractions between antecedent conditions(both internal and external) and consequenceconditions, and their relationships with psychi-

atric symptoms, communication styles, medicalissues, preferences, and other considerationsleads to informed clinical hypotheses about thepurpose and meaning of the challenging behav-iors and the conditions that both prompt andmaintain it. These diagnostic insights offer di-rection to staff and focus intervention strategies(psychiatric, psychological, social, rehabilita-tion, etc.) on causative factors. This article sug-gests that the absence of widespread use ofsophisticated psychological assessment and in-tervention tools designed specifically to identifyand treat behavior dysfunction leads to under-treatment and to maintenance of dysfunctionalbehaviors, that in turn often result in unneces-sary distress and to the misuse of restrictiveprocedures in psychiatric, correctional, andother residential settings (e.g., intermediate carefacilities, institutions for mental diseases, com-munity living centers).

Medication selected for treatment of symp-toms of mental illness must not inhibit memory,learning potential, or other cognitive functionsthat mediate responsiveness to rehabilitative in-terventions (Liberman, Corrigan, & Shade,1989; Liberman, Falloon, & Wallace, 1984;Storzbach & Corrigan, 1996) and should becontinued only when objective evidence reportspositive impact. Furthermore, medicationshould address only those components of a per-son’s illness responding to that medication asindicated by objective measures of progress andoutcome, and not be used to sedate people or tochemically restrain functioning.

The MFM provides an example of how treat-ment planning becomes a more specific assess-ment and case formulation process involving anempirically iterative process that results in thegeneration of causal hypotheses, followed bycausally directed interventions, followed bymeasurements of progress, followed by furtherfunctional assessments leading to revised andmore sophisticated causal assumptions, leadingto treatment interventions specific to these hy-potheses, and so on until measurable progressand improved outcomes are achieved. This on-going multimodal functional case formulation,intervention, and measurement process directlyinvolves the client, associated family members,and all staff interacting in an exploration andinvestigation aimed at uncovering the multipleand interacting causes of complex behaviors orsymptoms and developing and testing interven-

13THE MULTIMODAL FUNCTIONAL MODEL

tions, skills training, and other supports re-quired to assist the person in becoming empow-ered and achieving desired self-agency and suc-cessful entry into community life.

The MFM

The MFM provides a structure for clinicalcase formulation and the implementation of in-tervention strategies to assist patients and clin-ical teams in designing treatment for behaviorproblems and other symptoms associated withmental illness and/or mental retardation (Bedell,Hunter, & Corrigan, 1997; Gardner & Hunter,1995, 2003; Gardner & Sovner, 1994; Griffiths,Gardner, & Nugent, 1998). The approach re-flects an integrative biomedical-psychologi-cal-socioenvironmental perspective thatstands in contrast to the “diagnose and treat”model typically used in most hospitals today.The emphasis on multimodal, as opposed tomultidisciplinary, focuses attention on the var-ious biomedical and psychosocial modalities ofinfluence contributing to behavioral and emo-tional symptoms rather than on the discrete pro-fessional disciplines involved. Functional con-notes the assumption that a patient’s symptomsare best viewed as the “final common pathway”of an individual’s functional adaptations to en-vironmental, physical, and social demands aswell as disturbances in neurochemical and phys-iological functioning. Under the model, the ini-tial step toward designing interventions to treatsocially inappropriate adaptations and to re-place them with more normalized ones involvesidentification and assessments of these multiplefunctional relationships. That is, it is first nec-essary to identify preceding instigating condi-tions, and consequences of a behavior from theperspective of each modality, and to determinethe relative influence of biological, psychoso-cial, and environmental factors on the initiation,severity, variability, and persistence of the pa-tient’s current behavior problems or symptoms.The resulting interventions will shift from asymptom management (i.e., symptom suppres-sion) focus to one that directly addresses caus-ative factors, including biomedical factors, cog-nitive, perceptual, psychosocial skills deficits,and ongoing tertiary or vulnerability influencesthat contribute to the patient’s symptoms andbehaviors. The steps in the model also apply todetermining functional antecedents for proso-

cial and replacement behaviors as well. Spacedid not allow for descriptions of this aspect orfor providing charts of prosocial responses inthe case example. Clients are active in makingchoices about replacement behaviors and in for-mulating causal hypotheses for both unwantedand chosen behaviors.

MFM Diagnostic Formulations1

Step 1: Assessing triggering and setting con-ditions. The treatment planning process be-gins with the identification of the specific set oftarget symptoms and/or behaviors of concern.These symptoms and behaviors must be opera-tionally defined so all staff and the patient canreadily observe them when they occur and thenrate their frequency, intensity, duration and/orvariability. Thereafter, the case formulationprocess focuses on the development of specificcausal hypotheses about the functional signifi-cance or meaning of the symptoms and behav-iors within the larger context in which theyoccur. Specifically, all staff should be involvedin identifying current external (e.g., loss of aloved one; realization that housing optionshoped for are not currently possible) and inter-nal (e.g., anxiety, chronic anger, iatrogenic se-dation) stimulus conditions. Each stimulus con-dition is considered to serve either a primary orsecondary instigating function (Gardner &Sovner, 1994). Primary influences are those thatmust be present for symptoms to occur (neces-sary conditions). That is, they are discriminativeevents or “trigger” stimuli that are observed toimmediately precede (and significantly increasethe likelihood of) the behavior. Examples in-clude a task demand in a harsh tone (an envi-ronmental psychosocial stimulus) that may setoff an episode of aggressive behavior. A loudnoise (a stimulus condition in the physical en-vironment) may trigger an anxiety attack; aseizure (a personal/medical condition) may trig-ger screaming in a patient. Secondary instigat-ing influences are those whose presence eitherincreases or decreases the likelihood that asymptom or behavior will occur in the presence

1 A shortened version of this section previously appearedin Schizophrena by Steven M. Silverstein, William D.Spaulding, & Anthony A. Menditto, ISBN 978-0-88937-315-0, 2006, pp. 30–31 © 2006 by Hogrefe & HuberPublishers.


of the primary instigator. There are many con-ditions that serve to “set up” the person, orincrease the probability of the behavior given asufficient “trigger” stimulus. Sleep deprivation,for example (personal/medical condition), mayincrease the likelihood that a minor trigger stim-ulus would set off the behavior. The patient maynot become aggressive when given a task de-mand in a harsh tone if they were not sleepdeprived. Headaches, physical pain, noise, height-ened anxiety, irritability, anger, temperature,and boredom are examples of potential settingstimulus conditions that if controlled or elimi-nated may reduce the probability of the expres-sion of the maladaptive behavior.

Step 2: Assessing the purpose of the behav-ior. Following determination of the externaland internal stimulus conditions under which theproblem symptoms/behaviors occur, staff focusassessment on the purposes or functions beingserved by the symptoms/behaviors (e.g., modulat-ing pain, avoiding rejection, communicating aneed). It should be noted that some symptoms(e.g., command auditory hallucinations) directlyreflect underlying neurobiological abnormalities.However, even these symptoms often begin toacquire functional features as they becomeassociated with distinct reinforcing conse-quences, and one can observe a change infrequency, duration or intensity related tospecific consequences (e.g., getting attention,being left alone, etc.).

Step 3: Assessing Vulnerability Influences

The next hypothesis-generating step involvesobservations regarding how primary and sec-ondary instigators are combined with tertiary orvulnerability influences. Tertiary stimulus con-ditions are ongoing challenges or deficits thatinteract with primary (trigger) and secondary(setting) stimuli that increase the probabilityof the expression of the behavior. Examplesinclude sensory impairments, personalitycharacteristics, cognitive impairment, limitedcommunication skills, limited anger manage-ment, impulse control, or conflict resolutionskills, chronic low frustration tolerance, andso forth. These tertiary conditions serve astargets for rehabilitation efforts that focus onteaching skills and coping strategies to im-prove resilience to environmental stressors.Understanding the interplay among primary,

secondary, and tertiary stimulus conditionsacross both external and internal variablesleads to a detailed understanding of how ab-normal behaviors emerge and provides causalhypotheses that direct the design of interven-tions focused on outcomes associated directlyto these causal hypotheses (see Figure 1).

Collecting data. Collaboration and ongoingdialogue with the patient concerning his or herperceptions of the significant influences on tar-get symptoms and behaviors along with his orher goals and motivations are central to thefunctional assessments, the causal hypothesesgenerated, and the measurement process. In ad-dition, collecting baseline data using objectivemeasures is critical to determining instigatingevents and consequences as well as the dimen-sions of the behavior (i.e., frequency, intensity,duration, and variability). Any one of severaldata collection formats may be useful to thisend.

Recording target behaviors across time inter-vals throughout each day, over time (Patternsand Trends Data Sheet, available from first au-thor), results in very helpful cues for studyingand understanding many behaviors. Observingbehavior over time allows for many functionalhypotheses related to cycles, specific days (e.g.,structured days vs. unstructured days), periodswhen certain staff are present or absent, therelationship between the behavior and any num-ber of environmental or programmatic changes,and patterns of expression that may be linked tovarious biological cycles (e.g., hunger, sleepdeprivation, medication changes or administra-tion schedules, premenstrual periods).

There are several other methods and formatsfor data collection that are useful for any givensituation. Clinicians experienced in behavioraltechniques should assume responsibility for de-signing effective data collection systems, train-ing direct care staff in collecting data, and an-alyzing and graphing data for distribution to thepatient and other members of the treatment teamfor further analysis. These data are then used inconducting more formal functional assessmentsand in the generation of causal hypotheses con-cerning the interplay of antecedent and conse-quence conditions on the behavior.

Functional assessments/analysis. There area variety of functional assessment techniquesfrom simple to complex that improve hypothe-sis formulation and enrich implementation strat-


Fig

ure

1.M

ultim

odal

func

tiona

lca

usal

hypo

thes

esw

orks

heet

.


egies (Haynes & O’Brien, 1990; Nezu & Nezu,1989; O’Neill et al., 1997). In the MFM, thecausal hypothesis-intervention formulation pro-cess is case specific and, therefore, does not relyon prior determined assessment protocols thatare completed on all clients by separate disci-plines. Rather, assessment activities focus onunderstanding symptoms/behaviors as they re-late to the unique personal and environmentalinfluences relevant for that individual. Data col-lection and functional assessment lead to theformation of functional causal hypotheses thatare linked to specific intervention strategies. Inthe MFM, the Multimodal Functional CausalHypotheses Worksheet (see Figure 1) allows forcausal hypotheses to be formulated for bothantecedent conditions (instigating influences)and consequence conditions (consequences/functions) as they may influence both external(environmental) and personal (internal) vari-ables. Understanding how consequences (posi-tive reinforcement, negative reinforcement, sen-sory feedback and other) are linked to maladap-tive behaviors also helps in designing focusedintervention strategies. Moreover, utilizing theCausal Hypotheses Worksheet promotes con-sideration and formulation of testable hypothe-ses concerning the interplay of these factors.

Intended use of the Causal HypothesesWorksheet is as follows. For each symptom orbehavior, the multidisciplinary team member(s)most skilled in evaluation of a specific modalityof influence (e.g., medical, neurological, psy-chological, socioenvironmental) will develophypotheses about the instigating and exacerbat-ing conditions specific to that modality pur-ported to be related to the symptom/behavior. Inaddition, hypotheses about the level of influenceof these conditions (i.e., primary, secondary,and tertiary) and the functionality of the pre-senting symptom/behavior are developed. Forexample, a patient’s violent outburst may bedue to the combined instigating influences ofanother patient’s excessive staring (primaryevent) and persecutory delusions that others aretrying to recruit him against his will to becomea hit man (secondary setting event). In this case,the functional purpose of the violent outburstmay be an adaptive response to scare othersaway from forcing him to be a hit man. Finally,the worksheet is used to formulate ideas aboutpossible interventions (categorized as treatment,management, and control procedures) (Gardner

& Cole, 1987; Gardner & Sovner, 1994; Hunter,2000) related to the clinical impressions.

Linking functional causal hypotheses to in-terventions. In the MFM, it is important tolink each intervention (including medication) tospecific hypotheses. The Specific DisciplineHypothesis-Intervention Formulations Work-sheet (see Figure 2) provides the format for thatprocess. Each formulation (hypothesis) devel-oped from previous worksheets is listed in col-umn A. In column B, the specific intervention islisted that derives from that hypothesis. Forexample, if one hypothesis is that aggressivebehavior is a result of command hallucinationsresulting from schizophrenia, then the interventionstrategy might be the use of an antipsychotic med-ication to reduce or eliminate the hallucinations. Inthis phase of the case formulation process, onewould list command hallucinations in column A,the antipsychotic medication in column B. Sinceall formulations are actually clinical causal hy-potheses, all must be tested. In this example, theclinician ordering the medication would be askedto state what would be a successful outcome of themedication regimen within a designated time pe-riod. The expected change (column D) might be astatement that successful elimination of the hallu-cinations would reduce those aggressive episodesinstigated by the hallucinations, for example, thefrequency of aggression should decline from anaverage of 7 incidents a week to zero incidentswithin 6 weeks. This would become the metric forassessing the “hallucination/medication” hypothe-sis. A data collection system would be designedwith staff responsible for recording. The next step(column F) would involve someone being as-signed responsibility for analyzing the data andinforming the team within the 6-week period ofthe outcome. If no change were observed or ifaggression increased over the course of the eval-uation period (6-weeks), the hypothesis would benot be supported and the medication stopped (un-less it provided other measurable benefits associ-ated with other clinical hypotheses).

During the completion of the Hypothesis-Intervention Formulations Worksheet, severalhypotheses may be developed with several spe-cific intervention strategies. Each would have ameasure of outcome associated with it (often thesame data system is used to evaluate any num-ber of different interventions). Staff, during theplanning session, would agree to a staging plan(column C) whereby they would prioritize


Fig

ure

2.Sp

ecifi

cdi

scip

line

hypo

thes

is-i

nter

vent

ion

form

ulat

ions

wor

kshe

et.


which interventions would be implemented andin what order. As the team identified hypothesesthat were not useful (i.e., the accompanyingintervention strategies did not produce positiveoutcomes for the individual), other more refinedcausal hypotheses with more focused interven-tion strategies would be selected. This evalua-tion sequence provides valuable informationabout when to stop an intervention, and can pre-vent the practices of either adding interventionswithout knowing what the previous interventionscontributed, or of continuing interventions that areineffective.

The Multimodal Integrated Intervention Plan(see Figure 3) represents a summary compila-tion from the previous worksheets and includesthe multimodal causal hypotheses, hypothesis-based interventions, staging plan, expectedchange for each intervention, an ongoing datacollection system for monitoring change, and areview schedule for each intervention selected.As this system is deployed, there is an ever-increasing database about what works, howmuch is added or lost with each subsequentintervention, and a record of what has beenpreviously tried along with its effects. This caseformulation and intervention model results instreamlined treatment programs that, over time,continue only those interventions that result inmeasurable benefits to the individual. Further, theinexcusable practice of continuing marginallyeffective or ineffective interventions for years ischallenged by the continuous review of out-come data and the team’s ongoing hypothesisformulation process.

Case Example

A summary of the case of Gary,2 a pseud-onym, represents a patient treated at the SecondChance Program at Weill Medical College ofCornell University/New York PresbyterianHospital, Westchester Division (Silverstein etal., 2006).

Behavioral description. Gary, a 28-year-old African American man with a history ofpsychosis (DSM–IV diagnosis of Schizoaffec-tive disorder), history of polysubstance abuse/dependence (cannabis dependence, alcohol andcrack cocaine abuse), and a major movementdisorder (uncontrollable choreiform-like move-ments of all four limbs, trunk, neck, and tongue,and dystonia), was referred to the Second

Chance Program. It would be helpful here to seeFigure 1 and view the first two columns that listpersonal and environmental attributes (Column1) and the domains (Column 2) mentioned insubsequent paragraphs concerning relevantstimulus conditions.

Gary was transferred from the state hospitalwhere he had spent 2 years on an inpatienttreatment unit (Environment–Physical, second-ary setting condition; see Figure 1, Columns 1and 2) and during which he had little contactwith family which he desperately wanted (En-vironment–Psychosocial, secondary settingcondition). Gary’s psychotic symptoms were inremission and affective symptoms and move-ment disorder were partially managed by thefollowing medication that were prescribed on en-try to the Second Chance program: Olanzapine15 mg qhs (psychosis), Methadone 10 mg, bid(for pain due to movement disorder), Gabapentin,600 mg bid (mood stabilizer), Baclofen 200 mgqhs (muscle relaxant), Trihexyphenidyl (move-ment disorder) 22.5 mg qam, Sertraline 100 mqqhs (depression), Lorazepam 2 mg bid (behav-ior control).

Gary’s behavior dyscontrol was operationallydefined as loud moaning, yelling, crying, crawl-ing and writhing on the floor, indiscriminantgrabbing and hugging of anyone near him, re-moving clothing, and begging others to helphim remove or adjust clothing. In the earlyassessment phase, it was unclear what primaryevents would trigger this set of behaviors. How-ever, it quickly became clear that occurrences ofthese behaviors appeared to be highly correlatedwith the severity of his movement disorder(Personal–motor tertiary condition) which, inturn, appeared to cause him more distress (Per-sonal–motoric secondary setting). Gary dis-played these behaviors immediately on arrivalon the unit, and approximately 7 hours per dayfor the first several days. Initially, he demon-strated such severe writhing, yelling, and cryingthat he was in restraints or in seclusion much ofhis first two days on the unit. These behaviors,though initially frightening and confusing toother patients and staff, resulted in mobilizing

2 Details of this case history were previously published inSante Mentale au Quebec as: Wilkniss, S. M., Hunter,R. H., & Silverstein, S. M. (2004). Traitement multimodalde l’agressivite et de la violnce chez des personnes souffrantde psychose. Sante Mentale au Quebec, 29, 2, 143–175.


Fig

ure

3.M

ultim

odal

inte

grat

edin

terv

entio

npl

an.


others to attend to his needs in attempts toreduce his distress (e.g., patients would helpadjust his clothing and get him water, staffwould talk to and attempt to soothe him). Thus,social attention, which he appeared to enjoy wasprovided as a consequence of his behavior.

Hypotheses. The initial opinion of severalstaff was that Gary would not benefit from theprogram, was “untreatable” using psychologicalinterventions, and needed to be moved asquickly as possible to a medical unit at anotherhospital to address an unknown neurologicaldisorder (Personal—biomedical tertiary condi-tion). An alternative hypothesis was that he hadboth a severe movement disorder and was dem-onstrating severe anxiety/adjustment reactionsto the change in his residence (Environment–Physical and Psychosocial secondary settingstimulus condition). Thus, the observed symp-toms (including the increase in severity of hisabnormal movements) were mainly psycholog-ical in nature. The latter hypothesis was based,in part, on (a) observations of the patient at hisformer hospital (where the frequency and dura-tion of his behavioral dyscontrol were report-edly less), (b) reports that he recently had spenta day in court without behavioral dyscontrol,and (c) a meeting with several staff in which heremained in good control and lucidly identifiedhis needs and goals. Staff favoring this secondview believed that Gary’s behavioral dyscontrolcould be accounted for by attention-seeking,immature communication strategy, and a lack ofappropriate social skills (Personal—social/coping and communication tertiary stimulusconditions) to get his needs met (e.g., smoking,having contact with family members). It thuswas felt that he could learn to meet his needsthrough more prosocial means. The attentionseeking behavior and poor communication/social skill when under distress seemed inpart to have its origin in his having beenabandoned and his experiencing chaotic cir-cumstances as a child (his mother and fatherwere drug dealers, as he himself later became,and his mother was killed) (Personal–psychosocial/coping tertiary and setting con-dition). His current attention-seeking behav-iors were clearly triggered when his needswere not immediately met (Personal—social/coping primary discriminate events). Thus,the identified treatment goals were (a) to in-crease socially skilled, adaptive means of

gaining attention and getting needs met, (b) todecrease the frequency, duration, and intensity ofbehavioral dyscontrol, and (c) to teach ways ofcoping with the severe movement disorder inadaptive ways. In view of the intensity and longduration of Gary’s episodes of behavioral dyscon-trol, it was hypothesized that his behavior could beshaped by positively and successively reinforcingincreasingly longer periods of good control com-bined with immediate time-out from reinforce-ment at the onset of dyscontrol.

Data collection. Since functional analysisof behavior using the MFM is powered by col-lection of objective data, staff began to collectdata on the frequency and duration of the afore-mentioned behavioral excesses using a Patternsand Trends Data Sheet (available from the firstauthor) that records observations over each24-hr period, across days. This allowed staff todetermine whether there was a discernable pat-tern to the periods and duration of behavioraldyscontrol. The pattern that emerged clearlyindicated that all of the above operationallydefined behaviors occurred together and withthe same intensity in each episode of behavioraldyscontrol (rather than separately) and weretherefore targeted as a whole. It was observedthat the episodes occurred mainly in themorning and evening hours with fewer inci-dents during hours of more structured activ-ity. Additionally, weekly behavior ratings ofover 20 categories of socially intolerable andprosocial behaviors were measured as a stan-dard part of the milieu treatment in the programand allowed for the collection of additional ob-jective data on Gary’s progress.

Integrating treatment and behavioral incen-tives. As part of treatment plan, medicationchanges were made to attempt to reduce theseverity of his movements (Trihexyphenidylwas lowered ultimately to 2.5 mg, as it increasesmovements at the higher dose, Sertraline wasdiscontinued and Nortryptyline added, as Selec-tive Serotonin reuptake inhibitors may contrib-ute to movement disorders in some individuals).According to the second hypothesis, that Gary’sbehavioral dyscontrol could be reduced andprosocial behavior could be increased, a behav-ior contract involving differential reinforcementwas implemented. Subsequent behavior plansreflected Gary’s demonstration of greater con-trol over his behavior by increasing the amountof time in good control required for reward.


This change occurred in small increments ofdecreasing the amount of time allowed for be-havioral dyscontrol, at the rate of approxi-mately 0.5–1 hour each week (i.e., plan 1 al-lowed for 4.5 hours of dyscontrol, plan 2 � 4hours, plan 3 � 3 hours; plan 4 � 2 hours; andplan 5 � 0.5 hours). This succession of behav-ior plans represents a procedure in which theperson’s behavior is shaped toward a specifiedoptimal goal of more socially appropriate be-haviors to replace the maladaptive ones. Afterreaching the optimal goal, the behavior plan isdiscontinued.

Implementation and response. Figure 4shows the change in behavioral excesses (asdefined above) over time as recorded by staffusing the Patterns and Trends Data Sheet overa 6-week period. Note the trend line (dashedline) has a significant negative slope indicat-ing a significant reduction in duration of be-havioral dyscontrol over time. Some variabil-ity is evident which can, in part, be explainedby various “triggering” stimuli (examples fol-

low). Important events (e.g., when behaviorplans were implemented, triggering events)are indicated on the graph. Of note, a docu-mented break in the consistent administrationof the initial contract had a noticeable effecton Gary’s behavior, resulting in a responseburst. Even though Gary’s frequency of inap-propriate behaviors had decreased from ap-proximately seven hours per day to close toone hour per day, nursing staff felt that they“couldn’t take it anymore” during one of hisbehavioral dyscontrol episodes and gave in tohis request for a cigarette (a break in thetreatment contract). As might be expectedunder this condition of intermittent reinforce-ment, Gary’s frequency of inappropriate be-haviors immediately increased to near base-line levels. Similarly, the day of his birthday(also, reportedly the day his mother was shotand killed in front of him), seemed to serve asa triggering event, resulting in a significantincrease in behavioral dyscontrol for twodays.

Number of Hours of Gary’s Behavior Excesses

0

1

2

3

4

5

6

7

8

day 1

day 3

day 5

day 7

day 9

day 1

1

day 13

day 15

day 1

7

day 1

9

day 21

day 2

3

day 2

5

day 27

day 29

day 3

1

day 3

3

day 35

day 3

7

day 3

9

day 41

day 43

Day

Nu

mb

er o

f H

ou

rs plan 1

smoke

plan 2

birthday

plan 3

plan 4

plan 5

Figure 4. Number of hours of Gary’s behavioral excesses.


After more consistent implementation of theplan and further data collection (Patterns andTrends Data Sheet), it became evident thatGary’s behavioral dyscontrol occurred in themorning hours (previously it was in the morn-ing and early evening). This, along with con-cerns about heavy self-requested use of prnbenzodiazepines led to the hypothesis that hisbehavior might be, in part, due to withdrawalfrom these medications in the early morninghours. The new hypothesis was tested by givinghim a standing dose of 2 mg of Lorazepam uponwaking. In addition, the contract was modifiedto include additional positive reinforcement forcontrolled behavior in the morning in the formof earning rewards for him and his fellow groupmembers by staying in good control. As can beseen in Figure 4 (see sections marked plans 4and 5), this combined intervention significantlyreduced the duration and frequency of morning

behavioral dyscontrol episodes. Moreover, dur-ing this treatment phase, Gary’s behavioral ex-cesses were of significantly reduced intensity;he would either take himself to the quiet roomor engage in other self-soothing behaviors, forexample, relaxation exercises that he was learn-ing from meetings with a psychologist, or inter-acting with peers. Figure 5 depicts this decrease inintensity and frequency of socially intolerable be-haviors over time. Also, observation of Gary’sbehavior revealed that motor dyscontrol washighly correlated with emotional dysregulation. Inthe later phase of treatment (with morning Loraz-epam), Gary’s motor control was dramatically im-proved during the entirety of waking hours.

Summary

Effective multimodal treatment, informed byfunctional properties of individual symptoms

Frequency of Gary’s Socially Intolerable Behaviors since Admission

0

2

4

6

8

10

12

week

1

week

2

week

3

week

4

week

5

week

6

week

7

week

8

week

9

week

10

Week

Fre

qu

en

cy

pe

rW

ee

k

Loud

Socially Inappropriatenon-Bizarre Behavior

Touching Other People

Figure 5. Frequency of Gary’s socially intolerable behaviors since admission.


and behaviors, leads to ongoing refinements inclinical hypotheses and choices of interventions(medical, psychiatric, behavioral, psychothera-peutic, psychoeducational, rehabilitation-related, etc.). Identification of the functionalproperties of behaviors leads to interventionsthat (a) teach the patient adaptive means ofexpression, (b) provide prosocial replacementbehaviors, and (c) encourage development anduse of coping strategies that increase immuni-ties to the ongoing tertiary conditions associatedwith the behavior disorder. These include stressreduction strategies and effective coping mech-anisms related to specific environmental stres-sors. In sum, a case formulation model thatemphasize multimodal comprehensive treat-ment of behavior dysfunction leads to an im-proved understanding of behaviors both for thepatient and the staff, leads to precise interven-tion strategies, including environmentalchanges and modifications, and teaches adap-tive coping mechanisms that may lead to longterm adaptive strategies. Furthermore, an under-standing of the functional properties of behav-iors opens up new venues for approaching treat-ment involving interventions directed at specifictriggering stimulus conditions, secondary “set-ting” stimulus conditions, tertiary influences,and the various reinforcement consequencesthat strengthen behavior.

References

Bedell, J. R., Hunter, R. H., & Corrigan, P. W.(1997). Current approaches to assessment andtreatment of persons with serious mental illness.Professional Psychology: Research and Prac-tice, 28, 217–228.

Bellack, A. S., & Hersen, M. (1998). Behavioralassessment: A practical handbook (4th ed.). Bos-ton: Allyn & Bacon.

Ciompi, L. (1980). Catamnestic long-term study ofthe course of life and aging of schizophrenia,Schizophrenia Bulletin, 6, 606–616.

Citrome, L., Volvaka, J., Czobor, P., Sheitman, B.,Lindenmayer, J. P., McEvoy, J., et al. (2001).Effects of clozapine, olanzapine, risperidone, andhaloperidol on hostility among patients withschizophrenia. Psychiatric Services, 52, 1510 –1514.

Donat, D. C. (1998). Impact of mandatory behavioralconsultation on seclusion/restraint utilization in apsychiatric hospital. Journal of Behavior Therapyand Experimental Psychiatry, 29, 13–19.

Donat, D. C. (2002). Employing behavioral methodsto improve the context of care in a public psychi-atric hospital: Reducing hospital reliance on seclu-sion/restraint and psychotropic PRN medication.Cognitive and Behavioral Practice, 9, 28–37.

Gardner, W. I., & Cole, C. L. (1987). Behaviortreatment, behavior management, and behaviorcontrol: Needed distinctions. Behavior ResidentialTreatment, 2, 37–53.

Gardner, W. I., & Hunter, R. H. (1995). The Multi-modal Functional Model enhances treatment forpeople with serious mental illness. Unpublishedmanuscript.

Gardner, W. I., & Hunter, R. H. (2003). Psychosocialdiagnosis and treatment services in inpatient psy-chiatric facilities for persons with mental retarda-tion: Practice guidelines. Mental Health Aspects ofDevelopmental Disabilities, 6, 1–13.

Gardner, W. I., & Sovner, R. (1994). Self-injuriousbehaviors: Diagnosis and treatment. WillowStreet, PA: Vida Press.

Griffiths, D. M., Gardner, W. I., & Nugent, J. A.(Eds.). (1998). Behavioral supports: Individualcentered interventions: A multimodal functionalapproach. Kingston, New York: NADD Press.

Harding, C. M., & Strauss, J. S. (1985). The course ofschizophrenia: An evolving concept. In M. Alpert(Ed.), Controversies in schizophrenia (p. 347).New York: Guilford Press.

Haynes, S. N. (1990). Behavioral assessment ofadults. In G. Goldstein & M. Hersen (Eds.), Hand-book of psychological assessment (2nd ed., pp.423–463). Oxford, U.K.: Elsevier.

Haynes, S. N., & O’Brien, W. O. (1990). The func-tional analysis in behavior therapy. Clinical Psy-chology Review, 10, 649–668.

Hunter, R. H. (1995). Benefits of competency-basedprograms. American Psychologist, 50, 509–513.

Hunter, R. H. (1999). Testimony on behalf of theAmerican Psychological Association at the JointCommission on Accreditation of Healthcare Orga-nizations (JCAHO) Public hearings on the use ofseclusion and restraint in psychiatric facilities.Alexandria, VA, April 13, 1999.

Hunter, R. H. (2000). Treatment, management, andcontrol: Improving outcomes through more treatmentand less control. In H. R. Lamb (Series Ed.) & F. J.Frese, III (Vol. Ed.), New directions for mentalhealth services: The role of organized psychology intreatment of the seriously mentally ill (pp. 5–15). SanFrancisco: Jossey-Bass Publishers.

Hunter, R. H. (2001). Improving outcomes requiresmore, not less, from psychology. The BehaviorAnalyst Today, 2, 4–13.

Kane, J. (1989). The current status of neuroleptics.Journal of Clinical Psychiatry, 50, 322–328.

Liberman, R. P., Corrigan, P. W., & Shade, M.(1989). Drug and psychosocial treatment interac-


tions in schizophrenia. Special Issue: Behaviouralpsychotherapy into the 1990’s. International-Review-of-Psychiatry, 1, 283–295.

Liberman, R. P., Falloon, I. R. H., & Wallace, C. J.(1984). Drug-psychosocial interactions in thetreatment of schizophrenia. In M. Mirabi (Ed.),The chronically mentally ill: Research and ser-vices (pp. 175–212). New York: SP Medical &Scientific Books.

Nezu, A. M., & Nezu, C. M. (1989). Clinical deci-sion making in behavior therapy: A problem solv-ing perspective. Champaign, IL: Research Press.

O’Brien, W. H., & Haynes, S. N. (1993). Behavioralassessment in the psychiatric setting. In A. Bellack& M. Hersen (Eds.), Handbook of behavior ther-apy in the psychiatric setting (pp. 39–71). NewYork: Plenum Press.

O’Neill, R. E., Horner, R. H., Albin, R. W., Sprague,J. R., Storey, K., & Newton, J. S. (1997). Functionalassessment and program development for problembehavior: A practical handbook (2nd ed.). PacificGrove, CA: Brooks/Cole Publishing Company.

Paul, G. L., & Lentz, R. J. (1977). Psychosocialtreatment of chronic mental patients: Milieu vs.social learning programs. Cambridge, MA: Har-vard University Press.

Persons, J. B. (1989). Cognitive therapy in practice:A case formulation approach. New York: Norton.

Persons, J. B. (1992). A case formulation approach tocognitive-behavior therapy: Application to panicdisorder. Psychiatric Annals, 22, 470–473.

Reiss, S., Levitan, G. W., & Szyszko, J. (1982).Emotional disturbance and mental retardation: Di-agnostic overshadowing. American Journal ofMental Deficiency, 86, 567–574.

Reiss, S., & Szyszko, J. (1983). Diagnostic overshad-owing and professional experience with mentallyretarded persons. American Journal of Mental De-ficiency, 87, 396–402.

Silverstein, S. M., Hatashita-Wong, M., Wilkniss, S.,Bloch, A., Smith, T., Savitz, A., et al. (2006).Behavioral rehabilitation of the “treatment-refractory” schizophrenia patient: Conceptualfoundations, interventions, and outcome data. Psy-chological Services, 3, 145–169 (p. 146).

Stephens, J. H. (1978). Long-term prognosis andfollow-up in schizophrenia. Schizophrenia Bulle-tin, 4, 25–47.

Storzbach, D. M., & Corrigan, P. W. (1996). Cog-nitive rehabilitation for schizophrenia. In P. W.Corrigan and S. C. Yudofsky (Eds.), Cognitiverehabilitation for neuropsychiatric disorders(pp. 299 –328). Washington, DC: American Psy-chiatric Press.

Sturmey, P. (1996). Functional analysis in clinicalpsychology. New York: Wiley.

Torrey, E. F. (2001). Surviving schizophrenia: Amanual for families, consumers, and providers(4th ed., p. 130). New York: Quill.

Wincze, J. P. (1982). Assessment of sexual disorders.Behavioral Assessment, 4, 257–271.

Wolpe, J., & Turkat, I. D. (1985). Behavioral formu-lation of clinical cases. In I. Turkat (Ed.), Behav-ioral case formulation (pp. 213–244). New York:Plenum Press.

Received May 17, 2005Revision received October 9, 2007

Accepted October 22, 2007 �

E-Mail Notification of Your Latest Issue Online!

Would you like to know when the next issue of your favorite APA journal will be availableonline? This service is now available to you. Sign up at http://notify.apa.org/ and you will benotified by e-mail when issues of interest to you become available!


The Multimodal Functional Modelâ€“Advancing Case Formulation

Documents

Transcript of The Multimodal Functional Modelâ€“Advancing Case Formulation