Dr. Meyer is associate in psychiatry at both Beth Israel Deaconess Medical Center and the Program in Psychiatry and the Law at Massachusetts Mental Health Center, and assistant clinical professor of psychiatry at Harvard Medical School, all in Boston, MA.

Acknowledgments: The author would like to thank Dr. Robert Simon for his comments and encouragement and Noreen O’Connor for her editorial assistance. The author reports no financial, academic, or other
support of this work.



Neither mental illness nor its treatment is clearly separable into either medical and nonmedical components or on the basis of the different mental health disciplines.  Mental health care with multiple clinicians requires the resolution of inherent clinical and professional ambiguities.  Effective clinical care and risk management requires clarification of the individual and interprofessional duties and effective communication between clinicians.  Patient assessment for anticipatable therapeutic problems with multiclinician therapies can help assure therapeutic efficacy and lower clinician liability risks.



In the current psychiatric world, split treatment has become a fixed part of the landscape. Whether the treatment is termed “split,” “collaborative,” “combined,” “coordinated,” “divided,” or simply “med backup,” the names for this mental healthcare paradigm underscore the ambivalence with which it is often viewed by clinicians.1 At the heart of this ambivalence are the clinical and professional conflicts that pose potential impediments both for quality clinical care and effective risk management. Split treatment is a part of the wider concept of mental health care provided by more than one clinician and across professional divides. Clinicians who practice split treatment must decide how the general conflicts inherent in this treatment paradigm apply to the mental health care of the individual patient under consideration. Only then can the clinician determine whether split treatment is an acceptable clinical option.


Potential Benefits and Problems Associated With Split Treatment

In communities that have a shortage of psychiatrists, split treatment increases patient access to psychopharmacology. For patients with mental disorders that call for multimodal treatment unavailable through a single clinician, split treatment may be the only pragmatic method. Furthermore, some patients feel that split treatment maximizes their choices as healthcare consumers by providing access to the greatest range of mental health treatments and treatment providers.

Physicians are familiar with the concept of delegating a portion of medical care and responsibility to another professional colleague.2,3 Practitioners are expected to confine their practice and responsibility to their individual area of expertise. In most medical settings this concept is supported by the separability of the patient’s illness into component parts that roughly correspond to the expertise of the clinician. In the diabetic patient, the endocrinologist attends to the patient’s insulin levels, the neurologist to the patient’s neuropathy, the ophthalmologist to the patient’s retina, and the nephrologist to the patient’s renal function. The primary care physician has the job of holding the clinical center.

However, mental illness is not similarly divisible. Neither mental illness nor its treatment is clearly separable into medical and nonmedical components or on the basis of the different mental health disciplines.4 While prescription refills are clearly medical, drug compliance is not. Though many drug side effects may be purely medical, a patient with mania secondary to antidepressant treatment requires both medical and psychotherapeutic intervention. Also, a depressed patient whose depression has worsened or who has become actively suicidal requires both medical and therapeutic intervention. Added to the difficulty of parsing out which clinician should be responsible for which aspect of the patient’s care is the fact that the split treatment fraction of the mental health treatment population is the sickest.

We know that a team of two or more clinicians requires more administration and communication than a single-clinician practice. Clinicians comprising a treatment team have less authority individually than if each were to conduct the therapy alone. The question of authority is further complicated by the frequent absence of an internal hierarchy or reporting relationship for the split-treatment team.

Psychiatrists and their mental health colleagues who practice split treatment are faced with clarifying their respective and shared clinical duties with the most severely mentally ill, with an illness that does not lend itself to being divisible along professional lines, and within a team that may lack a formal hierarchy and reporting structure. Another level of complexity may be introduced by the presence or absence of formal policies from a clinic or hospital.

Added to the clinical and professional ambiguities of split treatment is the concern that split treatment may increase liability exposure for the psychiatrist.5 Psychiatrists who participate in split treatment have the potential to see greater numbers of more severely ill patients than they could if they were singly responsible for all facets of the patient’s psychiatric care. Patients in split treatment often have more serious or treatment-resistant mental illnesses.  These two factors taken together would seem to confer a greater exposure to negative clinical outcomes compared to a psychiatrist who does not deal with split treatment.


Clinical and Risk Management With Multiple Mental Health Clinicians

Mental illness and treatment are not separable into respective professional specialties. Therefore, a psychiatrist and psychotherapist must create a treatment plan in which each clinician can answer the following questions:
• What is my clinical competence and duty to the patient?
• What is my relative clinical autonomy?
• What are the areas of clinician interdependence?
• Is there any specific agency or legal statute that has bearing on the treatment being offered to the patient?

Answers to these questions should incorporate available guidance from published reviews,4,6-8 professional ethical standards,2 and, where available, legal decisions or statutes pertaining to psychiatric practice.

Psychiatrists are always responsible for determining the scope of the psychiatric examinations that they perform. Patients, collaborating clinicians, and healthcare agencies may each have their own reasons to circumscribe initial clinical contacts. However, a psychiatrist cannot delegate the responsibility for that decision to anyone.2 It is unethical to medicate a patient whom the psychiatrist has not directly examined.9,10 There may be patients with whom a more circumscribed examination is clinically justified; a psychiatrist can weigh the immediate clinical benefits against the risks of not discovering what he/she may never have the opportunity to know or ask.

The clinical agency within which a mental health treatment has occurred, such as a hospital or clinic, can also have enormous impact on the liability risks for clinicians who are collaboratively providing mental health treatment. Agencies typically have policies that affect the professional credentialing, supervision, quality assurance, standard clinical assessment, and emergency clinical assessment for the care of patients. Violation of the agency’s accepted protocol is disastrous for risk management. In view of this, policies should not set unrealistically high standards that in practice cannot be followed.11


The Consultation

Perhaps the most narrowly delimited request for interprofessional collaboration is a psychiatric consultation. In this case, the patient is being treated by a therapist who is qualified through training, experience, and independent licensure. The psychiatrist is responsible for making a clinical assessment, collecting an adequate clinical database, and forming and communicating a professional opinion to both therapist and patient.

The psychiatrist should believe that the therapist is capable of handling the patient’s treatment. Psychiatrists should not perform consultations for therapists who do not have the necessary training, experience, and license, even if this conflicts with a patient’s choice of provider (eg, alternative therapy). It is unethical for any physician to participate in treatment with a clinician whom he or she believes to be unfit to manage a patient.12

Importantly, the psychiatric consultant does not assume any ongoing care for the patient or supervision of the treatment. The consultant’s only authority is over the opinion offered, and not whether its recommendations are followed. Both patient and therapist need to understand that the consultation does not represent continuing participation in the patient’s care. A consultation is the least ambiguous collaboration between clinicians because there are no shared clinical duties. A consultation is a time-limited event, not an indeterminate clinical process.


Ongoing Collaboration Between Independently-Licensed Clinicians

The most common form of split treatment is a collaboration involving “a mutually shared responsibility for the patient’s care.”2 As in a consultation, the psychiatrist should have a basis to believe that the nonphysician therapist is qualified by training, experience, and licensor to make the anticipated clinical judgments and interventions. Both clinicians have a responsibility to ask for and provide this professional information. It is also advisable for both clinicians to verify that each has adequate malpractice insurance. In the event of a bad clinical outcome and one underinsured or uninsured colleague, a liability action will seek the deeper pocket.

The two clinicians must agree upon the basics of the patient’s diagnosis, the anticipated therapies, and the risks that derive from the patient’s diagnosis and treatment. The major sources of clinical risk—suicide, violence, other types of impulsivity, worsening of the patient’s clinical condition, diagnostic uncertainty, and deterioration of the relationship with the patient—should be discussed to anticipate what response, if necessary, would be initiated and by whom.

Several of these issues have been the subject of excellent reviews.13-15 It is important to remember that risk assessment is an ongoing process concurrent with clinical care. It is not a one-time event. Most patients in split treatment have diagnostic factors that increase their statistical risk of suicide. Some of these patients have multiple risk factors such as a diagnosis of an affective or psychotic disorder and concurrent chronic pain, substance abuse, or history of (self) destructive behavior. Feelings of paranoia, humiliation, anxiety, or agitation are emotional springboards to behavioral enactment for patients already at risk.

Assessment of psychological and environmental mitigators of risk are equally important. Patients can be asked to articulate what events or relationships they view as important positive factors or attachments in their future. These factors can help a patient persevere through a serious illness. They can also be reassessed during the progress of  the treatment to clarify whether they still retain the same positive meaning for the patient.

Many clinicians will indicate that their patient has “contracted for safety”; sometimes the patient has already signed an actual document. Contracting for safety evokes the concept of a legally-binding agreement or promise between patient and clinician, and offers a false sense of security to both individuals. A patient may have believed what he/she signed, only to not believe it after leaving the doctor’s office. A patient’s belief that he/she is safe is simply one factor of, and not a substitute for, risk assessment.

Risk assessment involves a judgment of the patient’s vulnerability at the time of the examination and a hypothesis about the patient’s anticipated level of function in the immediate and intermediate future. In formulating the hypothesis, the clinician needs to evaluate the patient’s capacity to self-monitor and  communicate information to treating clinicians. Initiating this discussion can begin with asking the patient, “How would you know if things were getting worse?”

Most worrisome are historically impulsive patients with little capacity to notice whether they are deteriorating and those who are psychologically inhibited or unmotivated to collaborate with others about their distress. Risk management for these patients involves helping them foster conscious awareness of high-risk situations, premonitory signs of deterioration, and potential ameliorating responses. As an interim measure, clinicians may find it helpful to articulate concretized potential indicators of psychological deterioration and then construct a series of predetermined responses for the patient.

Certain issues of risk management are unique to split treatment. Both clinicians need to be aware that they have independent and interdependent duties for ongoing risk assessment. Psychotherapists will have more frequent opportunities. Psychiatrists and psychotherapists will be privy to different types of information from the same patient, partly based on the different relationships and partly on the different focus of inquiry for psychopharmacology and psychotherapy. Both therapists have the responsibility for sufficient direct examination of the patient consistent with the patient’s clinical status. Both therapists have a responsibility to let the other clinician know about any substantive change in the patient and/or treatment.

Clinicians should verify with each other their customary practices for emergency coverage during and after office hours and on weekends, holidays, and vacations. Both clinicians can increase their clinical efficacy and risk management by clearly showing the other clinician that they are easily accessible and want to be contacted for significant developments. Successful split treatment involves the development of a responsive system for sharing information. Clarity is the best counterweight to destabilizing clinical and interprofessional ambiguities.


When Split Treatment May Not Be Appropriate

Split treatment is not for every patient. Methods of clinical risk management notwithstanding, part of good clinical practice and good clinical risk management involves patient selection. Therapists who participate in split treatment can review the kinds of clinical issues that may make an alternative treatment modality preferable for a particular patient’s care. Some patients with severe character pathology are especially intolerant of strong or ambivalent emotions. Their capacity for projection, splitting, and projective identification may be unwittingly fostered by having two therapists between whom they can parse out polarized but unintegrated pieces of their experience. Patients with this sort of character problem will be better served with a single individual who possesses the right composite of therapeutic skills. The same can be said of patients who have had a history of strong negative transference reactions.

Rapid-cycling bipolar patients may pose a particular strain for split treatment due to the expectable reassessments and interventions that the two therapists will need to discuss and implement. The split-treatment team may find itself unduly burdened having to allocate large amounts of unreimbursed time for interprofessional communication. Split treatment of bipolar depression may pose some of the same risks because many of the drug treatments involve the risk of the patient switching over into mania.

Other clinical issues can increase the amount of patient oversight and discussion, thereby increasing the burden to the split-treatment clinicians. Treatment resistance, dual diagnosis, multiple or off-label drug regimens, or a history of impulse disorder could be relative contraindications to split treatment.

All psychopharmacology involves psychotherapy. A patient’s feelings of transference do not observe the professional division of duties that a psychiatrist and psychotherapist may have constructed in a split treatment. Psychiatrists who are acting as psychopharmacologists can anticipate that their short sessions with a patient will be sufficient for the development, but not the analysis of transferential feelings. The short sessions can also be an unintended problematic repository for a patient’s splitting off a part of the polarized ambivalent feelings that have developed in a psychotherapy relationship.

The psychiatrist in a split treatment team can be prepared to make comments that can identify and clarify the appearance of transference as a segue to referring the material to the psychotherapist, where it can be more fully explored. Patients who have a history of intensely ambivalent or mercurial transference relationships may not be appropriate for split treatment. These patients, in the midst of experiencing intensely positive or negative feelings toward a healthcare provider, may unwittingly play one clinician against the other to the detriment of their own care.

Psychiatrists may find it useful to ask patients about their experiences in other treatment relationships to anticipate some psychological responses to the current treatment. Patients who have felt shame or anger in response to feeling dependent may view taking medication and having a medication doctor as a similar narcissistic wound. Patients who have a history of idealizing important relationships may bring that same capacity to their experience of psychotropic drugs and having a psychopharmacologist.

Managed care organizations regulate access to treatment benefits. Treatment authorization and appeals processes represent additional unreimbursed administrative requirements for clinicians in a split treatment team who may already feel overextended by the time required for interprofessional collaboration about the patient. Notwithstanding these burdens, clinicians continue to have a duty to the patient’s welfare independent of the actions of a managed care organization.16 In the face of treatment authorization denials and in the absence of being able to make alternative fee arrangements, the clinician must make a good faith effort to arrange for other affordable mental health care for the patient.


When Split Treatment Does Not Work

Sometimes a split treatment proves unworkable. For example, the match between the two clinicians is poor, the patient’s illness is too thorny, the two clinicians have diagnostic or therapeutic differences that cause strategic conflicts in patient management, or the patient’s relationship with one of the clinicians has deteriorated.17,18

Clinicians should not be bound to a treatment that no longer meets their own standard of care regardless of whether or not the treatment conforms to the standards of some practitioners in the community. A clinician should respectfully resign from the treatment in a time frame that allows the patient and the other clinician to make appropriate arrangements. The patient’s best interests should guide which clinical misgivings or disagreements between clinicians are disclosed directly to the patient.8

Patients may want to change their therapist or psychopharmacologist during the course of their treatment. While the patient has that right, patients may be unaware that the remaining treating clinician cannot automatically agree to work with whomever the patient has chosen. The remaining clinician has the right and the responsibility to make an informed choice regarding whether or not the successor clinician will be acceptable in the newly configured split-treatment team.


Supervisory Collaboration Between Clinicians With Dependent Licensure

In some settings a psychiatrist may collaborate with another mental health professional who is licensed to work only with concurrent psychiatric supervision. A psychiatrist who supervises a psychiatric resident, medical student, or a psychology or social work trainee is working with a colleague whose license for clinical activity legally requires the participation of the psychiatrist. For some clinicians with dependent licensure (eg, a clinical nurse specialist), the type and frequency of psychiatric supervision may be specified by law.

In the case of concurrent care of a patient with a mental health professional trainee, the psychiatrist delegates and supervises in accordance with the abilities of the trainee. The psychiatrist retains some responsibility for the patient’s care, including that portion that was delegated. In one of the most famous malpractice suits in psychiatry, Tarasoff v Regents of California,19 the attorney for Tarasoff commented thus:

It is my view that if (the supervisor of the clinic) had personally examined the patient Poddar and made an independent decision that the patient Poddar was not dangerous to himself or his victim—Titiana Tarasoff—there would be no cause of action based upon foreseeability. However, (the supervisor) never saw the patient Poddar and ignored the medical records developed by his staff.

A similar sentiment was articulated in the sentinel case Cohen v New York,20 in which an inpatient under the care of a psychiatric resident left the unit and committed suicide. The court commented:

The treating physician did not, at this point in his medical career, possess the requisite skill or trained psychiatric judgment to, essentially unsupervised, provide ordinary and reasonable care to this decedent. There was not one but many errors of judgment made by a doctor not qualified in an unsupervised status to make a judgment; made by those in a supervisory capacity; and all made without careful examination.

Psychiatrists who supervise mental health clinicians with dependent licenses can assign their colleague a clinical task in the patient’s treatment, but they cannot assign the responsibility for the supervisory decisions. Those supervisory decisions should be informed by the training, experience, and individual competence of the trainee and by the clinical difficulties of the patient and the treatment.


Educational Supervision of Independently Licensed Clinicians

Psychiatrists who supervise the work of clinicians who are licensed for independent clinical activity have a reduced responsibility for the clinical work they are supervising. Often in mental health, the term “supervision” indicates an educational activity that does not include managerial authority for oversight, direction, and control of those being supervised. A psychiatrist who educationally supervises an adequately trained, independently-licensed clinician can reasonably expect that the therapist, not the educational supervisor, is the ultimate decision maker regarding the patient’s treatment.

In reviewing a complaint of negligent supervision in a claim brought by a patient against her therapist’s supervisor (Schrader v Kohout),21 the court focused on whether a doctor-patient relationship legally existed, commenting, “There could be no liability of malpractice in the absence of physician-patient relationship.” The court noted the following facts about the treatment and its supervision:  the patient and supervisor had never met; the supervisor had been paid by the therapist and not by the patient; the therapist testified that she had made all the treatment decisions; the supervisor had no access to the patient or the patient’s records; and the therapist controlled the flow of clinical information to the supervisor. The court found that there was no physician-patient relationship between the patient and the therapist’s supervisor and thus decided for the supervisor and against the patient.21

The court’s view of this legal question might not have been the same in a somewhat different clinical setting. The supervisor and therapist in Schrader v Kohout were in private practice. Had they both worked for the same outpatient clinic, the supervisor might have had authority to direct the therapy and access to both the patient and her records. Furthermore, if the supervisor were reimbursed by the clinic from funds the patient had paid for her treatment, the supervision might be legally viewed as a part of the treatment. Psychiatrists who supervise should consider these factors as they assess their liability exposure for activities that they may view as educational rather than clinical.


Promoting Interprofessional Clarity and Cooperation

Mental health treatments involving more than one clinician present enormous opportunities for clinical ambiguity and professional misalignment. Comprehensive assessment at the initiation of a split treatment can protect against many downstream clinical and liability risks. The depth of assessment of a colleague’s skills depends on the type of treatment collaboration, the responsibility that one clinician has for the other’s activities, and on the patient. Assessment can be difficult in the absence of reviewing a colleague’s actual work. However, some easily accessible information can aid in the assessment of a colleague’s skills:
• Is the colleague able to clearly represent the clinical data?
• Is the history coherent?
• Can the colleague reconstruct what was directly heard and observed in the office?
• Having been able to represent the clinical data, can the colleague make interpretive hypotheses based on that information?
• Has a phenomenologic diagnosis been made?
• Has a psychodynamic formulation been made?
• Can the colleague identify and understand transference and countertransference as it may manifest itself with the individual patient?
• Has the patient’s treatment and an overview of therapeutic interventions been formulated?

The answers to these questions can often be derived from initial discussions about the patient and can help guide any type of interprofessional collaboration.

Once the initial evaluation of a patient is complete, the greatest enemies of effective interprofessional collaboration may be clinician anxiety and shame and the resultant chilling of effective comminution. All clinicians worry that clinical uncertainty may lead to a loss of clinical authority and professional reputation. Discussion between colleagues of a patient’s negative transference or a therapist’s countertransference frequently raises concerns of revealing too much about one’s own psyche and of being pathologized by the collaborating clinician.22 Members of a split-treatment team can help counter these risks by making themselves easily available to colleagues and creating a climate where clinical uncertainty can be disclosed and discussed.

Clinicians who communicate an attitude of interest and shame-free tolerance toward their colleague’s uncertainty will foster the kind of open communication of new clinical data that is essential to a good working relationship. Clinicians should actively resolve clinical uncertainty or developing interprofessional conflicts. These conflicts are sometimes mistakenly deferred to a later time when it is hoped that the colleagues might have more of a professional working alliance. True alliances may evolve over time if founded on a working relationship that  can contain and resolve interprofessional clinical conflicts rather than avoid them.



In 2002, mental health clinicians have increasingly sophisticated interventions to treat mental illness. Mental health care that requires coordination of multiple clinicians is being promoted by a variety of forces: consumer choice advocates, public health services, and managed care organizations. Such coordinated care may offer more treatment combinations. Whether or not these combinations ultimately provide clinical and financial benefits, mental health care with more than one clinician has inherent clinical and administrative ambiguities. Those ambiguities need to be considered in light of the individual patient’s needs to clarify whether coordinated mental health care with multiple clinicians will pose insurmountable or soluble clinical and risk management dilemmas.   PP



1.    Goldberg RS, Riba M, Tasman A. Psychiatrist’s attitudes towards prescribing medication for patients treated by nonmedical psychotherapists. Hosp Community Psychiatry. 1991;42:276-280.
2.    Guidelines for psychiatrists in consultative, supervisory or collaborative relationships with nonmedical therapists. Am J Psychiatry. 1980;137:1489-1491.
3.    American College of Physicians. Ethics manual. Fourth edition. Ann Intern Med. 1998;128:576-594.
4.    Meyer DJ, Simon RI. Split treatment: clarity between psychiatrists and psychotherapists. Part 1. Psychiatr Ann. 1999;29:241-245.
5.    Macbeth JE. Legal aspects of split treatment: how to audit and manage risk. Psychiatr Ann. 2001;31:605-610.
6.    Appelbaum PS. General guidelines for psychiatrists who prescribe medication for patients treated by nonmedical psychotherapists. Hosp Community Psychiatry. 1991;42:281-282.
7.    Sederer LI, Ellison J, Keyes C. Guidelines for prescribing psychiatrists in consultative, collaborative and supervisory relationships. Psychiatr  Serv. 1998;49:1197-1202.
8.    Meyer DJ, Simon RI. Split treatment: clarity between psychiatrists and psychotherapists. Part 2. Psychiatr Ann. 1999;29:327-332.
9.    American Psychiatric Association. The Principles of Medical Ethics With Annotations Especially Applicable to Psychiatry. Washington DC: American Psychiatric Association Press; 1998.
10.    APA condemns Kaiser prescribing policy. APA Press Release No. 0012. April 14, 2000.
11. Meyer, D. The psychiatrist as supervisor: risk management issues. Rx for Risk. 1997;3:1-8.
12. Council on Ethical and Judicial Affairs. Code of Medical Ethics. 1998-1999 ed. Chicago, Ill: American Medical Association; 1999.
13. Simon, RI. Taking the “sue” out of suicide: a forensic psychiatrist’s perspective. Psychiatr Ann. 2000;30:399-407.
14.    Jacobs DG, Brewer M, Klein-Benheim M. Guide to Suicide Assessment and Intervention. San Francisco, CA: Josey-Bass; 1999.
15. Miller MC. A model for the assessment of violence. Harvard Rev Psychiatry. 2000;7:299-304.
16. Wickline v State, 192 Cal App3d 1630, 239 Cal Rptr 810 323 (1986).
17.    Ellison J, Smith J. Intertherapist conflict in combined treatment. In: Ellison J, ed. The Psychotherapist’s Guide to Pharmacotherapy. Chicago, Ill: Year Book Medical Publishers; 1989:96-115.
18. Busch FN, Gould E. Treatment by a psychotherapist and a psychopharmacologist: transference and countertransference issues. Hosp Community Psychiatry. 1993;44:772-774.
19.    Slovenko R. Legal issues in psychotherapy supervision. In: Hess AK, ed. Psychotherapy Supervision: Theory Research and Practice. New York, NY: John Wiley & Sons; 1980:468.
20. Cohen v State, 51 AD2d 494, 382 NYS2d 128 (1976).
21. Schrader v Kohout, 522 Ga App 134, SE2d 19 (1999).
22. Jacobs D, David P, Meyer D. The Supervisory Encounter. New Haven, Conn: Yale University Press; 1995.