DRAFT – DO NOT CITE WITHOUT PERMISSION
Re-Inventing Treatment Planning
Gordon Harper, MD
Treatment Planning for the 21st Century: An Emerging Agenda
Treatment planning converts the patient‟s complaint and the clinician‟s
assessment into a plan for intervention. A good treatment plan is a synthesis of
clinical art and science, transforming a statement of human distress into a care
plan. But for some nowadays, treatment planning means making DSM diagnoses
– little more. While criterion-based diagnoses (DSM-IV) have improved
diagnostic specificity, facilitated epidemiological and clinical research, and
enhanced the dialog between psychiatric clinicians and other medical specialties,
they do not, by themselves, provide an adequate basis for treatment planning.
Nor is treatment planning based on categorical diagnoses consistent with
emergent values in mental health such as empowerment, transparency, and
strength-based treatment. As these values shape systems of care, the need for
a new approach to treatment planning has become clear. One response has
been to throw over entirely the language of professionals in favor of a language
limited to the point of view of parents and parent-advocates, not using the
expertise of professionals.
Contemporary advice on treatment planning takes three approaches.
1) Casebooks from APPI Press. The American Psychiatric Press has produced a
series of casebooks illustrating how to make and use DSM diagnoses (Spitzer et
al 2002; Spitzer et al 2004). The cases cover all diagnostic categories and age
groups and are rich in clinical detail. But this approach focuses on disorders not
on the patient‟s overall situation and leaves discourse in the hands of the
2) Wraparound Principles. Efforts to enhance child mental health services have
given rise to principles (Table I) derived from the CASSP program and the
systems-of-care movement (Stroul & Friedman 1986, Grimes 2004). These
principles promote participation of family and child, demand cultural competence,
build on strengths, and require services that are continuous. But the principles
don‟t specify how a treatment plan is generated. They give the impression that
clinical formulation is easy, “once we get the right people with the right attitude.”
These principles also fail to bring together the two main streams, frequently
divergent, in contemporary child mental health – the medical-psychiatric stream
and the non-medical psychosocial stream.
3) The Formulation. A third approach (Jellinek, McDermott 2004; Tasman 2002)
recalls earlier descriptions of the psychodynamic formulation (Perry, Cooper,
Michels 1987; Shapiro 1989), in advocating clinical sophistication, using data
derived from interviews by experienced clinicians, and going beyond what can be
learned from symptom check-lists. But this approach does not translate easily
into multimodal treatment and gives little voice to the patient and family and to
less-experienced members of the team.
The present paper outlines an approach to treatment planning that builds on these approaches and provides an operational guide for all parties, regardless
of level of experience. This description of this approach is organized around
three propositions reflecting core values in mental health:
? Formulations should be explicit, transparent, and consensus-based;
? Treatment should respond to the problem that led the patient to seek
? Treatment should be based on strengths and enhance adaptation.
In order to free up a discourse that has gotten “stuck,” a new language is needed that complements categorical diagnoses. In this paper we demonstrate
such a language and present an alternative to the conventional etiquette of
treatment planning (who speaks, who listens, who decides) that facilitates not
just interdisciplinary but consensus-based transactional discourse.
How we Talk, How we Act
Examining the relationship between the language we use and the care we provide may seem out of date. The adoption in the „80s and „90s of the DSM led
to a decrease in interest in the examination of the languages we speak as
clinicians (Harper 2001). The relationship between discourse and action, which
engages many outside Psychiatry (for example, Kegan & Lahey 2000), is pursued
within Psychiatry by those in areas like psychoanalysis and narrative therapy,
outside the Psychiatric mainstream (Kestenbaum 2003, DelVecchio Good et al
The idea of a reflection-free Psychiatry would surprise psychiatrists of earlier decades (Mezzich, Kleinman, Fabrega 1996), heirs of a tradition that
differentiated the several languages of clinical practice and examined the
relationship between professional discourse, the care we provide, and patients‟
behavior (Lazare 1973, Havens, Stanton & Schwartz 1955, Henry 1973; for work
in a more critical spirit, cf Laing 1969 or Szasz 1969).
Proposition One: Treatment should be explicit, transparent, and consensus-
Prominent among the goals of the systems of care movement (Grimes
2004) has been the empowerment of parents, who have too often experienced
child mental health as an adversarial, parent-blaming system. A necessary tool
in this effort is the development of a language of treatment planning that is
shared, explicit and transparent. Treatment planning based on categorical
diagnoses easily becomes a privileged domain in which the professionals, as
experts, have the dominant (or the only) voice and parents are reduced to the
role of onlookers (“consumers”).
Disempowerment of parents is not limited to systems centered on the
DSM. In some child therapy, protecting the confidentiality of the child has been
taken to mean not listening to or speaking with parents. The principles of
informed consent have been violated when it is left to the nurse on the ward to
tell the child she will be staying in the hospital. In addition, some adolescents
who have been in therapy report that they have little idea about the kind of
treatment they received. In all these settings, paternalistic professionals stay “in
charge” through limiting dialog about the nature of the problem and the
treatment being undertaken.
In such situations, it would be inaccurate to say that the professionals
have all the information; parents, the child, and those working directly with the
child have essential information, too, the on-the-ground equivalent of what
political scientists call “local knowledge” (cf Salmon et al 2004). But conventional,
diagnosis-based treatment planning excludes such information, producing an odd,
“upstairs-downstairs” two-language Babel. Treatment planning needs both kinds of knowledge and needs to facilitate communication between the different kinds
Such “transactional knowledge” arises from a more complex process than
conventional treatment planning, which attempts to follow a linear sequence
from chief complaint through history and examination to diagnosis and treatment.
The conventional sequence may be represented thus:
Chief Complaint -> History -> Examination -> Diagnosis -> Treatment.
It is fair to say that conventional treatment planning “attempts” to follow
this sequence because treatment planning in child mental heatlh often takexs a
confusing form, a kind of fragmented treatment dialog. This diaglog can be
represented as three circles, overlapping to varying degrees. In one sphere,
psychiatrists discuss Axis I diagnoses and treatments, especially
pharmacotherapy. In another, psychologists and social workers carry out
psychotherapeutic interventions. In the third, nurses and child care workers
practice behavioral and other “milieu treatment.” These several endeavors may
be co-located but not interacting with each other (figure).
Creating a shared language that can integrate these several endeavors
starts with the definition of the problem requiring treatment (here called the
Focal Problem) and proceeds through several steps. This process makes explicit
elements that are intrinsic, though unnamed, in all treatment planning. These
elements are the definition of a focal problem (the actual problem requiring
treatment) and goal, the identification of factors judged to contribute to that
problem, the selection of some of those factors for intervention, and the tailoring
of interventions to those selected factors. This sequence can be represented as
follows: Chief Complaint -> Definition of Focal Problem and Goal ->
Identification of Contributing Factors -> Selection of Factors for Intervention ->
. Choice of Interventions
In this approach, each element is openly discussed and a consensus
definition sought. Eliciting the child or family‟s reason for seeking help allows a
search for language that “works” for all parties. It is futile to try to analyze, let
alone intervene, in a problem without consensus as to the problem.
The power of this method can be seen in case discussions when the
definition of the problem by a professional elicits disagreement from a colleague
or a parent. Such disagreement is the starting point of negotiating a consensus
Example: When an adolescent was admitted to hospital, the doctor spoke
of a “manic break.” The boy‟s parents and outpatient doctor disagreed;
for them, the diagnostic question was still open. The problem was
redefined as “A period of agitation and excitement, precluding attendance
at school, and without consensus as to diagnosis.” The hospital doctor
was now better informed as to the community of opinion in which the
patient had been trying to manage his disorder.
Replacing a planning discourse in which the professionals are “in
possession” of the language of diagnosis with a shared language can produce a
process that is explicit, transparent, and consensus-based.
Proposition Two: Treatment should respond to the problem that led to
treatment being sought.
i) Chief Complaint, Diagnosis, and Focal Problem
The power of modern medicine lies in its ability to make diagnoses,
especially diagnoses based on etiology. This power, unimaginable in earlier eras,
makes diagnosis central to care. There are clinical encounters in which
everything depends on the diagnosis.
But in many clinical encounters the diagnosis does not adequately
describe what is going on and what is needed. People seek help for problems
that are incompletely described by their diagnoses. People with longstanding
disorders may seek help not for the disorder per se, but for impaired functioning.
In such cases one must first define the impaired functioning. As professionals,
we don‟t treat disorders, we help with troubles to which disorders contribute. To so do, we intervene in factors that contribute to the impaired functioning.
The “focal problem” describes the symptomatic impairment that leads a
person to seek care (Harper 1989). The defined focal problem may incorporate
the chief complaint in the patient‟s words. It may also include relevant
diagnoses. But such diagnoses are a facet of the focal problem. Focal problem
subsumes diagnosis, not vice versa. Example: A suicidal patient is hospitalized. Saying that the problem is the
diagnosis, “Major Depression,” would direct effort to the factors
contributing to the depression. But help has been sought for a suicidal
crisis, to be sure in a patient with a major depression. The focal problem,
accordingly, is better stated, “Acute suicidality in a patient with major
depression.” This definition of the problem directs attention to all the
factors, individual and contextual, that contribute to the suicidality.
The focal problem, in short, states the real reason a person is at this level of care
(not just a diagnosis, though it may include one), relative to a better level of
functioning for this person, in everyday language that all parties can agree on.
In inpatient treatment, the focal problem refers to trouble where the patient
came from – not to symptoms in the current treatment setting. Accordingly, the
problem requiring treatment is often usefully defined as some dysfunction pre-
admission with risk that it will recur once the patient is there again.”
A consensus-based focal problem has its own kind of validity, different
from a diagnosis generated unilaterally by professionals without patient or family
input or one generated by parents and advocates without professional input.
Consensus-based statements are valid because they integrate the perspectives
of the participants. In this sense, they are transactional statements.
Focal problems may also include an element of context, if the episode of
care would be difficult to understand without such a statement. A Focal Problem
might specify a particular context (e.g., “…suicidal ideation emergent two days
after arriving in a new group home”) or differential functioning according to
context (e.g., “…explosive threatening behavior at home, not at school”).
Here are some examples of focal problems:
• Failure-to-thrive in a 24-month-old boy, without consensus as to
• Explosive and threatening behavior in a 15-year-old with a history of
emotional instability and possible bipolar disorder, 2 weeks after arriving
in a new group home.
• Relapse with increased paranoia, agitation and threats to his wife in a 24-
year-old man with previously well-compensated schizophrenia;
ii) Goal and Future Vision
Once the Focal Problem is defined, the Goal is defined as enough
amelioration of the Focal Problem so that the present level of care is no longer
The Goal should be parsimoniously defined. That is, it should specify
those minimal changes in the Focal Problem needed to help the patient move on
to the next level of care. It answers the question, What does the patient (and
the patient‟s world) have to look like for the patient to move on to a less restrictive setting? Where the focal problem is defined as the risk of recurrence,
the goal is decreased risk of recurrence.
future vision Defining such a goal always depends on having an explicit for the patient. Although the term “future vision” may be unfamiliar, the concept
is familiar to all clinicians. The future vision answers the questions, Where will
the client live? With whom? In what legal and physical custody? With what
symptoms? What new troubles can be anticipated? What new coping will the
patient have? Which services will s/he receive? How will continuing treaters see
Example: Case managers considering hospital admission for a teenager,
say, “This bipolar girl will get one more chance to “make it at home”, then
we‟ll send her to residential treatment.” Threatening, penultimate
language like “one more chance” fails to anticipate the future vision for
this girl who, if she indeed has bipolar disorder, will have future crises,
some of which may require hospitalization.
Making the future vision explicit helps avoid a narrow focus on current
pathology, differentiating the patient‟s experience of illness and the person‟s life.
An explicit statement of Future Vision helps the professionals keep in mind the
larger goals of the patient, namely what improved health will allow him/her to do.
Spelling out the future vision is not always easy. Discussing the patient‟s
life after treatment, or after the current episode of care, may be more distressing
than talking about the vicissitudes of treatment. For example, if leaving inpatient
or residential care requires a choice between living with one parent or the other,
talking about treatment without reference to the future allows the team and
patient to avoid that painful choice.
iii) Formulation, Factors, and Interventions
Starting treatment planning with a focal problem, as opposed to starting
with a categorical diagnosis, allows an open exploration of what is contributing to
that problem. An open-systems approach decreases the pressure to “explain” the problem prematurely, allowing the explicit listing of possible
contributing ”Factors”. The “formulation” consists of the Factors identified as
possibly contributing to the Focal Problem.
Factors may be identified in any domain (for example, Constitutional,
Developmental-Psychological, Family, School-Peers-Social, and Systems).
Making the domains explicit guards against one-factorism. “Systems” factors lie
in the service systems, such as “Non-consensus among treaters as to diagnosis and indicated treatment” or, “Treating clinician about to leave on vacation; no
coverage identified.” Naming such factors requires overcoming inhibitions about
speaking frankly about colleagues, inhibitions that come from equating
professionalism with a code of silence.
The list of contributing factors can be reviewed with several questions in
mind. One of the most useful is, “Can we hear the patient‟s voice in this list of
factors?” That is, the factors must convey the actual experience of the patient.
The same can be said of the experience of the family. Psychiatric treatment
planning can benefit from research on the “patient‟s voice” in fields like primary
care (cf Barry et al 2001).
As the team identifies contributing Factors, they must also identify
strengths. Figure One represents how Factors contribute to and Strengths
mitigate the Focal Problem.
Here are some examples of Contributing Factors:
• Vulnerability to breakdowns with manic features, likely bipolar diagnosis
• [Mostly unrecognized] poor adherence to anti-manic pharmacotherapy
• [Patient is] at an early stage of integrating the fact of chronic illness in his
• Partner is intermittently supportive and harshly critical
In using an open-systems approach, the team must stay alert for
discussion-squelching statements spoken authoritatively by senior clinicians.
These statements may stipulate a diagnosis, a psychological state, or an etiology.
For instance, one might hear, “It‟s depression/ OR PTSD,” in which certainty is
attached, amid controversy, to a diagnosis. Or one might hear, “It‟s
ambivalence/ OR fear of success,” in which certainty is attached, amid
uncertainty, to a psychological inference. Or one might hear, “It‟s the mother/
OR the father‟s death,” in which certainty is attached, amid uncertainty, to a
possible contributing factor.
Statements like these deter others from speaking, cutting off further
inquiry. Premature closure leads to “one-factor treatment,” ill-advised mono-
After identifying Contributing Factors, the next step is to select Factors for
Intervention. Selection is based on two criteria: one can reasonably expect to
make a difference in that Factor and one can reasonably think that the differences made will reduce the Focal Problem.
Factors constitute our best understanding of the child‟s trouble. Factors
should evoke natural processes of growth, healing, and adaptation. How to do
state Factors in such terms is discussed in the next section.
Proposition Three: Treatment planning should be based on strengths and
Contributing Factors should be stated in terms of adaptive processes
which are implied in each part of the patient‟s story. Clinicians can listen to such
stories from multiple points of view. The most familiar perspective is that of
looking for symptoms in order to diagnose disorders. Another is listening for
what life feels like to the patient. Another is listening for adaptive processes
relevant to the strengths and the troubles of patient or family.
Talking about adaptive processes may be new, but the concept is
understood by all clinicians. Adaptive processes are the other side of pathology:
not a catalog of deficits, but the ways the patient and family are coping with
To put the patient‟s story in terms of adaptive processes, take any part of
the story, especially a Contributing Factor. Then ask, How do people normally
cope with this kind of event or condition? What constitutes good adaptation?
Then, compared to that kind of coping, how is this patient doing? Has the
problem been recognized or acknowledged? Regarding clinical services, is
management optimal? The answers to these questions point to interventions.
Defined by such a process, intervention promotes adaptation, rather than
Adaptive processes can be recognized all around us. We convey an
implicit knowledge of adaptive processes in remarks like, “How‟s the new job
going?” “What do you hear from your son at college?” or, “The children are
dealing well with the separation.” Some adaptive processes are familiar in
ordinary life, some draw on our knowledge of biology and psychology. For
instance, we speak of processes triggered by injury, like wound healing; those
triggered by a life stage, like managing the changes of puberty; those triggered
by a life event, like grieving; those triggered by substance abuse, like getting
into recovery; and those triggered by trauma or molestation, like psychological
The narrative of adaptive processes provides an alternative to talking
about pathology. This language is closer to experience than is discourse about
disorders. It is also a positive language, focusing not on adversity but on coping
and mastery. Asking people how they are doing fosters reflection, conveys
appreciation, and fosters self-esteem.
Example: Here are examples of the adaptive processes identified in a
patient‟s story. The details of the patient and family can be inferred.
• Family re-formation after separation
• Grieving for lost special closeness
• Helping a handicapped child develop
• Same, when the child has serious emotional disturbance, too
• Acculturation of an immigrant family
• Living with allegations of sexual misuse
• Recovery from out-of-home placements
• Living with the possibility of domestic violence
To transform information from the language of deficit and pathology to a
language of strength, adaptation and coping sounds easy. It requires
considerable effort to apply in practice.
Figure Two represents the whole treatment planning sequence.
Putting into Practice – Responses to Frequently Asked Questions
A Vessel, not the Contents. This approach to treatment planning does not their require a particular theoretical orientation or explanatory model – for example, of what is going on. Seemingly “incompatible” factors, from best understandingbiochemical, psychodynamic, social learning, or family systems. Rather, the divergent theoretical orientations, can be listed alongside each other, so that the method is theory-neutral and inclusive, inviting all parties to come up with action implications of each can be appreciated. What this method does require is
that hypotheses about what may be contributing to the problem, or about
possibilities for intervention, be made explicit.
The Infrastructure of Treatment Planning. Putting these ideas into practice takes more than talk. Simply sharing this approach, even with support
of leadership on the unit or in the facility, does not produce behavioral change.
Part of the reason is that conventional treatment planning is supported and
reinforced by existing organizational structures. For example, the conventions of
case discussion prescribe roles for presenters, expert-critics, and listeners. In
inpatient or residential facilities, those who do the most talking often have the
most experience in the field, but those who do the most listening often have the
most first-hand experience of the child being discussed. To achieve the kind of
open discussion advocated here, structural re-arrangements are necessary. As
with other complex group clinical processes, dedicated training is necessary. In
addition, it is useful to assign individuals to certain functions in the discussion.
These “listening positions” might include, for instance, listening for clarity and specificity of the definition of the focal problem; listening for explicit statement of
the future vision; and listening to make sure the child‟s point of view is audible in
Treatment planning, easily a scene of divergent languages, can become
the part of clinical service in which a shared language promotes collaboration,
strengths are recognized and children and family are helped to move forward