With the introduction of new compounds to treat mental illness during the last half of the twentieth century came some resistance to their use within the psychotherapeutic relationship (Karasu, 1982; Klerman, 1991). Fears were expressed that medication would somehow submerge important feelings and conflicts and therefore impede psychotherapy and/or provide the message that the patient was less rewarding or even too ill for more formal psychotherapeutic interventions. Few psychiatrists, psychoanalysts, and other clinicians maintain this position any longer. Instead, most mental health professionals, regardless of discipline, maintain that psychotropic medications, in conjunction with psychotherapy, are enormously helpful to patients and can often provide the following benefits to the psychotherapeutic process.

Pharmacotherapy can reduce uncomfortable levels of anxiety and depression allowing the patient greater access, expression, and understanding of feelings.

• Medications, through the reduction of acute symptoms, may enhance the patient's self-esteem by decreasing feelings of helplessness, futility, and passivity as well as enhancing the acceptability of treatment.

• Medication may increase the safety with the therapeutic relationship permitting more open expression of fantasies, feelings, and fears.

• Pharmacotherapy, for some patients, may have a positive placebo effect allowing a more substantial therapeutic alliance and decreasing the stigma of seeking mental health treatment.

• Medication, from the viewpoint of ego psychology, may improve autonomous ego functions (concentration and recall for example) that allow the mobilization of greater resources for the therapeutic process.

• Improvement from medication. Feelings about medication-related side-effects and pharmacologically unrelated nonspecific medication side-effects (Barsky et al., 2002) often provide invaluable insight into the patient's personality and emotional experience, both conscious and unconscious, and clarify countertransference issues as well, especially in the case of heightened resistance or therapeutic impasse.

• As in psychotherapy, can elucidate the patient's self-defeating conflicts about achievement and success.

• Medications may not only increase the likelihood, but also the speed and magnitude of the response to psychotherapy.

• During times of interruption of treatment, medication can maintain a connection to the treatment relationship.

On the other hand, psychotherapy, when added to an ongoing pharmacotherapy may have the following benefits.

• Psychotherapy promotes improved adaptation and coping.

• Psychotherapy improves compliance with pharmacotherapy (Paykel, 1995).

• Psychotherapy, even in patients with the most severe disorders, decreases the likelihood of recurrence of symptoms (Kay, 2001).

• Psychotherapy decreases relapse when medications are discontinued (Wiborg, 1996; Teasdale et al., 2001).

• Psychotherapy provides a much broader and more comprehensive inquiry into the patient's condition than is the case with medication monotherapy.

The flexibility of adding an additional treatment modality when the initial intervention is unsuccessful or partially successful is a major advantage in caring for patients and has been called sequential or stepped treatment. Some authors have suggested a systematically developed plan to add a second treatment from the outset. For example, Pava et al. (1994) and Fava (1999) have proposed the treatment of the acute phase of major depression with antidepressant medication and reserved the use of psychotherapy (cognitive-behavioral therapy, CBT) for the continuation phase to prevent relapse and improve the quality of life by treating residual symptoms. They argue that this approach utilizes psychotherapy resources in a more efficient fashion and specifies the unique advantage of each treatment. Much research is needed in explicating the advantages and disadvantages of sequential interventions.

There are also disadvantages in combining medication with psychotherapy. A significant concern is that patients may attribute their improvement to medication rather than to the active steps they have taken within the psychotherapy. In many such cases, the patient wishes to minimize the importance of the psychotherapeutic relationship. There may be fears that the patient will become too reliant on the therapist, will experience erotic feelings toward the therapist, or may be frightened of rejection, to name but a few. The devaluation of the psychotherapy and the idealization of pharmacotherapy, especially at the beginning of treatment, can be seen as an attempt by the patient to defend against painful feelings and thoughts that would undoubtedly require exploration. Similarly, many educators insist that their trainees treat patients initially, where appropriate, without medication so they may gain some conviction about the usefulness of a psychotherapeutic approach. For some students, especially at the start of their training, it is less anxiety provoking to believe that medication can ameliorate all psychic pain thereby alleviating them from the doubts and uncertainties of engaging in this type of intense work.

There is also some limited literature on the potential of medications in overly dampening a patient's discomfort that is necessary for engaging in a psychotherapeutic experience. Marks et al. (1993) have noted that combined treatment with benzodiazepines (a class of anxiolytic medications) may adversely affect the outcome for those patients suffering from panic disorder. They noted that, when compared with panic patients receiving psychotherapy alone, those who received medication and psychotherapy demonstrated increased relapse rates. Moreover, patients with panic and other anxiety disorders such as posttraumatic stress disorder must repeatedly reexperience in psychotherapy painful memories or feelings and begin to appreciate that some symptoms are not as fragmenting, catastrophic, or dangerous as initially experienced. Theoretically, if they are overmedicated, however, the necessary process of increasing insight and the ability to provide new understanding of symptoms may be decreased. A more recent study demonstrated that as needed anxiolytic treatment with benzodiazepines and psychotherapy for patients with panic disorder and agoraphobia was associated with poorer outcome compared to group cognitive behavioral therapy alone (Westra et al., 2002).

Whether they are aware of it or not, patients ascribe some psychological meaning to the taking of medication. These feelings may be about the agents themselves, about the prescribing and nonprescribing professionals or, as is often the case, both. Table 39.1 summarizes the feelings a patient may have about the medication and the treatment relationship.

There are a number of ways to inquire about meaning that patients attribute to medication and to those that prescribe them or treat them in psychotherapy. All, however, are predicated upon the clinician's willingness to recognize and explore a patient's expectable ambivalence about the treatment situation. This ambivalence may manifest itself on a continuum from severe suspiciousness to overidealization of one or more components in the treatment plan. For example, the clinician must recognize a patient's delusional thinking about the toxicity of medications as being representative of paranoid feelings about the treatment experience. The precise manner of the patient's thinking therefore, requires exploration to assess patient resiliency and cohesiveness. Patients who ascribe overly positive or unrealistic qualities or powers to medication at the exclusion of acknowledging their self-experience, also must be questioned about their views on their illness. In these situations it often helpful to inquire about previous relationships with healthcare professionals to ascertain the presence of long-standing characterlogical difficulties with those who are in authority, or in the case of a treatment relationship, requiring a trusting relationship.

As is the case in any psychotherapy, the patient will readily appreciate their professional's discomfort with psychological pain. For example, the practitioner who does not ask about the experience of psychological discomfort but rather focuses exclusively on phenomenology to arrive at a diagnosis, is more likely to consider medication as a monotherapy and the patient will undoubtedly feel on some level that he or she is dismissed. On the other hand, a clinician that fails to recognize the components of a specific disorder whose symptoms are quite treatable may leave the patient with doubt about the minimization of their painful symptoms. Both of these require that the professional observes the patient's response to the clinician and attention to countertransferential issues. Ultimately, regardless of theoretical orientation, clinicians must attend to the distortions that patients bring to the treatment situation. In the case of psychoanalytic psychotherapy, this is termed transference. These phenomena in CBT are called beliefs and automatic thoughts. At the initiation of treatment, for most patients these transferences and beliefs and automatic thoughts are outside of awareness. Moreover, regardless of the patient's level of psychopathology, these issues must be understood and brought to light for treatment to be effective. This is true for the continuum of psychotherapy from predominantly supportive to expressive or insight-oriented modalities.

Because as many as 60% of all patients do not take their medications as prescribed, appreciating the reasons for noncompliance becomes a powerful tool in the therapeutic armamentarium (Baso and Rush, 1996; Ellison and Harney, 2000). A recent comprehensive review of the prevalence of and risk factors for medication nonadherence in patients with schizophrenia noted that nearly 50% of these patients did not take their medications as prescribed (Lacro et al., 2002). Demyttenaere et al. (2001) studied depressed patients treated in primary care settings who dropped out of continuation treatment. They found that nearly 30% of patients stopped treatment because they worried about becoming drug dependent, felt uncomfortable taking medications, or were concerned that they were relying inappropriately on medication to solve their problems. Similarly, a study of 155 depressed patients in primary care revealed that 28% had stopped taking their antidepressants by the first month and 44% had done so by the third month of treatment (Lin et al., 1995). American and Canadian researchers studied why patients may drop out from mental health care (Edlund et al., 2002). This study examined 1200 patients from the US and Ontario, Canada in the early, 1990s and found that the dropout rates from treatment were, 19.2% and 16.9% respectively. This difference was not statistically significant despite the fact that mental health insurance is a major problem for US subjects, whereas Canadians have access to unlimited care. Reasons for dropping out of treatment included: belief that mental health treatment is ineffective, embarrassment about seeking help, and being offered only medication or only psychotherapy instead of combined treatment. Only Americans endorsed not having insurance as an important reason for discontinuing treatment. Lastly, respondents who had received combined treatment were less likely than their counterparts offered only monotherapies to leave treatment prematurely.

Noncompliance behaviors are associated with automatic thoughts about the particular medication, about the psychiatrist (or other physician and therapist), about the illness, and about oneself and others (Beck, 2001). Table 39.2 provides a summary of typical beliefs associated with noncompliance.

The unanticipated prevalence of public antimedication beliefs was illustrated in one public opinion poll of approximately 2200 adults in Germany that found that attitudes toward psychotropic medication were much more negative than those associated with cardiac drugs (Benkert et al., 1997). Even in the case of schizophrenia, 76% felt that psychotherapy was the treatment of choice and only 8% advocated medication. As most respondents were not knowledgeable about these medications nor did they know many people with mental disorders, the authors attributed their findings to lack of information and negative reports from the mass media. Similarly, Jorm et al. (1999), studied the Australian public and mental health professionals regarding the treatment of depression and found that the former frequently believed that antidepressants were potentially addictive and, along with electroconvuslive therapy, considered to be harmful.

It is not surprising therefore, that adherence problems with medications are ubiquitous. Rush (1988) has argued that until proven otherwise, every clinician should assume that noncompliance is present in each patient they treat. There are a number of questions that can be asked of the patient in anticipating medication noncompliance (Beck, 2001). Above all, the clinician must not shy away from exploring particular facets of medication beliefs and behaviors. For example, patients should be asked directly if they believe the medication that is being prescribed will be effective and are they willing to take the medication exactly as instructed. They should be prompted to consider the advantages and disadvantages of following treatment recommendations. Additional questions should attempt to elicit any problems with the purchase of medication and the ability to remember to take the medication at appropriate times. Assessing the impact of family beliefs about taking medication is critical. Patients will not follow medication regimens, especially with psychotropic drugs, when the prevailing belief by influential family members is that these medications are ‘crutches’ for the weak and dependent or required by those with only the most severe and chronic of mental illnesses. Patients should routinely be asked their fantasies about taking medication. Beck has further advocated that covert rehearsal in which patients are instructed to visualize how and when they would take their medication as well as appreciating any negative feelings about this activity. This technique will provide the basis for interventions that will strengthen compliance behaviors.

There is an unfortunate and simplistic view that is held by many prescribers that arriving at the correct diagnosis and providing the appropriately evidence-based medication guarantees that a patient will improve. The ability of a patient to follow medication plans is strongly predicated upon the establishment of a solid therapeutic alliance. Just as the strongest predictor for a positive psychotherapy outcome is the strength of the therapeutic alliance, it is also the strongest predictor for successful pharmacological outcome (Krupnick et al., 1996). No treatment will succeed without a safe, uncritical, empathic, and educative working relationship. There are additional behavioral techniques that some clinicians find helpful. These include:

• asking the patient to call at regularly scheduled times to assess if adherence is problematic;

• meeting with family members to defuse antimedication beliefs and inviting them to assist the patient in following the medication regimen;

• requesting that the patient complete a written record indicating when they took their prescription;

• the use of written coping cards that the patient carries with him that remind him of his unhelpful automatic thoughts about taking medication (Beck, 2001).

An important meta-analysis of 13 studies comparing psychodynamic psychotherapy with other types of psychotherapies and combined treatment demonstrated that, although there were no significant differences between types of psychotherapy, combined treatment was clearly more effective than any monotherapy (Luborsky et al., 1993). In reviewing the growing body of evidence that supports the helpfulness of employing medication and psychotherapy in the treatment of psychiatric disorders, there are three points to keep in mind. First, the literature on using psychotherapy and medication, while growing, is limited. Second, not all studies have found that combined treatment is superior to either monotherapy with medication or with psychotherapy in depression except in the case of severe disorders (Hollon et al., 1992; Manning et al., 1992; Wexler and Chicchetti, 1992; Antonuccio, 1995). Third, while this very brief review will focus on randomized controlled trials (RCT), there remains controversy about their generalizability to everyday clinical practice as these studies often have involved homogeneous patient populations who are without comorbid disorders, employ tightly supervised manualized treatments, and fail to report exclusion rates (Westen and Morrison, 2001). Nevertheless it is important for the clinician to have familiarity with recent research supporting the advantages of providing both medication and psychotherapy to patients.

To provide the reader with some appreciation of the evidence for combined treatment, it would be helpful to review selectively a few studies in major depression. There is, however, strong evidence for the helpfulness of combined treatment in patients with schizophrenia (Falloon et al., 1982; Leff et al., 1985; Hogarty et al., 1991, 1997a, b; Kuipers et al., 1998; Tarrier et al., 1998; Sensky et al., 2000; Granholm et al., 2002; McQuaid et al., 2002; McGorry et al., 2002) but only limited data in the treatment of personality disorders (Bateman and Fonagy, 2001), substance abuse (McLellan et al., 1993; Woody et al., 1995; Feeney et al., 2001), eating disorders (Walsh et al., 1997; Ricca et al., 2001), anxiety disorders (Spiegal et al., 1994; Wiborg and Dahl, 1996; Bruce et al., 1999; Barlow et al., 2000; Stein et al., 2000; Whittal et al., 2001; Kampman et al., 2002), and bipolar disorder (Miklowitz et al., 2000; Fava et al., 2001). For an in-depth review of these studies the reader is referred to Kay (2001) and Grech (2002).

An additional word is in order about the treatment of severe mental illness for which clinicians in many countries believe psychotherapy is ineffective. Researchers from the UK have recently made exciting advances in integrating CBT with medication for patients with acute and chronic schizophrenia.

There is mounting evidence that combined therapy with CBT has been noted to:

• improve medication compliance

• improve hallucinations and delusions among medication-resistant patients

• improve recovery from acute psychotic and first episodes

• decrease relapse and rehospitalization.

The continuing emphasis on combined interventions has underscored the importance of providing comprehensive psychosocial treatment in treating a severe, chronic, and often disabling disorder, which is arguably the most expensive of mental disorders (Knapp, 1997). Moreover, patients with schizophrenia and their families value psychotherapy as a very helpful intervention (Coursey et al., 1995; Hatfield et al., 1996; Kuipers et al., 1998). In treating psychotic disorders as is true with personality, substance abuse, eating, and many mood disorders, medication as monotherapy produces modest effects only and is rarely as effective as combined treatment. Clinicians should be skeptical of a biomedical orientation that reduces psychiatric and emotional disorders to phenomenology and therefore encourages unidimensional treatment approaches. On the other hand, nonprescribing professionals would be most unwise and shortsighted to dismiss the potential for pharmacotherapy in helping those who seek treatment.

There are a number of attractive properties in the one clinician model.

First, this approach counters the prevailing conceptual mind-body split that has so dominated modern Western medicine. In psychiatry for example, this is most apparent in the dichotomization of treating patients with pharmacotherapy for ‘brain-based disorders’ and treating concerns of the mind with psychotherapy. Important neurobiological and neuroimaging research within the last decade has clearly demonstrated the untenability of this position as it is now clear that all mental processes are ultimately products of brain activity (Kandel, 1998, 1999; LeDoux, 2001). In particular, neuronal plasticity (neurogenesis and synaptogenesis) and genetic transduction are central features of learning and memory, which not only determine our knowledge of ourselves and world views, but also how psychotherapy is likely to work (Liggan and Kay, 1999; Lehrer and Kay, 2002). In many ways, this is the most exciting of times because neuroscience has been able to substantiate that psychotherapy can change both brain structure and function. There are a number of comparative studies employing neuroimaging that have illustrated similar effects when patients with obsessive-compulsive disorder or depression are treated with psychotherapy and a second group with medication (L. Baxter et al., 1992; Brody et al., 2001; Martin et al., 2001). Indeed, it may be that these two treatments act on similar pathways (Sacheim, 2001).

Second, the integrated model in many cases allows for closer attention to medication adherence and side-effects. Some have argued that medication noncompliance should be anticipated in all psychiatric patients and vigilance to this issue is a hallmark of successful treatment (Basco and Rush, 1996). It may be easier to appreciate the meaning attached by the patient to medication and medication side-effects (especially sexual side-effects as in the first vignette) in the one clinician model, which does not require the input of a second professional, and for some patients, permits a more secure and safe treatment experience. The same could be said of the ability to understand the meaningfulness of side-effects as they express issues, such as transference, within the therapeutic dyad. Undoubtedly one of the most important qualities of integrated treatment is the likelihood of deeper therapeutic relationships that permit a more in-depth treatment experience through appreciating the subtleties of transference, countertransference, and resistance phenomena.

Third, although it remains to be established, there may be a number of disorders and clinical situations in which the one-person model should at least be considered by psychiatrists who are adept at both the psychotherapeutic and pharmacoptherapeutic models. These include patients with severe medical disorders where a physician may more likely appreciate the interplay between the psychosocial and biological factors, including drug–drug interactions. In addition, perhaps some patients with so-called primitive personality disorders (narcissistic and borderline disorders) who tend to polarize their helping relationships, have a significant propensity for self-harm, and often require hospitalization, can be provided with a more continuous type of care than is possible in the split treatment model. Gunderson and Ridolfi (2001), however, believes strongly that treatment for patients with borderline personality disorder should always have a least two professionals working together to minimize frustration and burnout. He argues that these patients should also receive at least psychotherapy and medication.

There are many psychiatrists who prefer to treat patients with serious disorders, such as schizophrenia and bipolar disorder, in an integrated model. This approach in skilled hands permits closer monitoring not only of symptoms but provides significant opportunities for psychoeducation about the nature of the chronic illness, the importance of medications, and the role of social and family contributions (Gabbard and Kay, 2001).

At least in the US, risk and liability issues are less complicated in an integrated treatment setting. The oversight of a treatment by one professional requires less time, collaboration, and of course paperwork. McBeth (2001) has elucidated specific risks with the split treatment approach and has noted that seeing greater numbers of patients less frequently carries a greater statistical risk for malpractice suit.

The most central challenge in delivering effective integrated treatment is the physician's obligation to master two complicated approaches to the patient. This treatment model requires that a physician be able to integrate the biological with the psychosocial in a moment to moment process. The capacity to ‘shift gears’ in listening to a patient is a skill that must be mastered for the delivery of effective care. A second challenge in the provision of integrated treatment is the obligation to keep abreast of the burgeoning field of psychopharmacology. It is difficult to underestimate the commitment necessary to stay current not only with new medications but also the increasing awareness of long-term side-effects in some new compounds. Coupled with need to continue to grow in one's psychotherapeutic skills, the responsibility for continuing education is significant.

There are of course many important principles in the provision of medication to all patients. This discussion will be limited, however, to those that will assist the physician in providing integrated treatment (Kay, 2001). The centrality of the therapeutic relationship has been discussed in detail. Some clinicians though believe that arriving at the correct diagnosis guarantees the success of the working alliance. Making the correct diagnosis and providing the most up to date information on medication does not ensure an empathic, nonjudgmental rapport with patients. Without the capacity to establish a safe and secure therapeutic relationship, adherence problems are bound to be more plentiful, including dropouts from treatment. As Frank et al. (1995) have written, a sound philosophy of care should focus on alliance, not compliance.

Along with safety, consistency, and predictability are critical components of the treatment relationship. Technical mistakes and boundary violations are more recognizable when a consistent manner of conducting the psychotherapy has been established. This is also true regarding how the clinician addresses pharmacotherapy issues within an ongoing psychotherapy. Although there is no one correct approach to this issue, the significant element is to establish a routine for such investigation. The examination of when clinicians deviate from their routine is exceptionally helpful in detecting subtle transference or countertransference issues. Some clinicians prefer to address medication concerns at the very beginning of a session, others wait until the end of the session. There are virtues in both methods. In the former case, the clinician will have the entire session to explore the issues surrounding the medication concerns and how it undoubtedly reflects on the therapeutic relationship. The limitation to this approach is that it may artificially set an agenda for a session and derail some of the patients pressing or immediate concerns. Leaving the medication inquiry to the end of a session alleviates the issue of steering the content of the session but may not provide sufficient time to address critical medication-related issues. Still others prefer to address medication-related topics whenever they arise in a psychotherapy session. Regardless of the chosen approach, deviation from the routine will often lead the psychiatrist to question the presence of countertransference. As an example, a beginning psychiatry resident was presenting to his supervisor a challenging and anxiety provoking treatment with a difficult patient. Immediately following the patient's verbalizing her strong sexual feelings for the therapist, he asked her if the medication she had been prescribed was helpful. With the supervisor's assistance, the trainee was able to appreciate that he became anxious about the patient's expression of her erotic longings and switched the subject to medication as an attempt to combat his anxiety.

The potential therapeutic richness of exploring medication side-effects has been discussed. It is imperative that the clinician pay close attention to the patient's questions about side-effects, changes in type and dosage, and to the discontinuation of medication. Often, frequent complaints about improbable side-effects can illustrate a patient's resistance in the psychotherapy. The prescribing of medication for the first time in an ongoing and challenging expressive psychotherapy may signal a growing frustration with a therapeutic impasse.

The termination phase of treatment is often the most overlooked. Novices frequently become overly concerned with the reappearance of symptoms at the very end of psychotherapy. Some attempt to treat this well known phenomenon by increasing or prescribing new medications. Most often, therapists have ambivalent feelings about termination, especially in forced terminations, and can feel conflicted about insufficiently helping their patients. Similarly, a patient's request for additional medication during the termination phase can represent an important entry into the patient's mixed feelings about ending a very meaningful relationship. Lastly, the question of continuing medication and who should monitor this after termination is frequently a challenge and mandates exploration of both the patient's and therapist's feelings.

One advantage frequently cited about the practice of split treatment is that it promotes the use of the unique talents of more than one mental health professional (Balon, 2001) and therefore provides the potential for the patient to receive a more sophisticated and comprehensive treatment experience. Second, many have proposed that it is more cost-effective and affords patients greater access to clinical care. Third, some have argued that more clinical information becomes available for more refined treatments. Fourth, there may be a greater opportunity for patients to be treated by therapists of similar ethnicity to that of the patient. Fifth, there may be greater professional and emotional support for each of the professionals. Sixth, some patients, such as those with severe personality disorders and or histories of overwhelming abuse, establish very intense relationships in treatment and can be enormously taxing to one clinician. The opportunity for sharing treatment responsibility can be protective for the collaborators in terms of decreasing the intensity of feelings on the part of the patient for each. Lastly, there may be an opportunity for collaborators to strengthen their clinical skills through a mutual education process. This is especially true when the result of a collaborative treatment experience provides great insight to the patient's fears and dynamics, thereby presenting a more comprehensive clinical understanding of the patient's plight. On the other hand, the effectiveness of medication has the potential to illustrate some of the biological bases of some disorders for the psychotherapist and demonstrate the usefulness of medication in addressing target symptoms in the areas of impulsivity, affective lability, and cognitive and perceptual limitations.

The collaborative, or two-person model, has other advantages as well. In working with patients with intense transference reactions, these can be somewhat diluted and more easily addressed within the treatment. Similarly, collaborative treatment will decrease a patient's opportunity to spend all or most of the sessions discussing medication at the expense of addressing psychological concerns when integrated treatment is employed.

It is not always possible for collaborators to know about each other's qualifications as well as the quality of care routinely provided. Such a situation may leave either or both clinicians anxious about the reliability of the collaborator. This doubt can be readily appreciated by a patient in split treatment and will undoubtedly cause the patient to feel less secure and for some patients, encourage splitting. There are other patients, who as children experienced significant disagreement between their parents. They may, attempt to repeat an important childhood behavior to pacify their parents and diminish conflict when they perceive it between the professionals. This can be accomplished through obsequious behavior or even acting out in an attempt to unite those overseeing the treatment.

A second challenge in the provision of split treatment is the inappropriate prescribing of medication by the physician when he or she is unaware of the process taking place in the psychotherapy (Balon, 2001) or when the prescriber provides medication as a practice routine without careful assessment of the patient's symptoms or feelings about medication. Similarly, a prescriber may insist on medication to dampen the patient's intense feelings in his or her relationship without appreciating the impact on the overall treatment experience. This is often the case for an example, when a physician, without consultation with his collaborator, impulsively begins medication. As well, there may be a wish to provide medication because one collaborator feels they are being ineffective with the patient or cannot tolerate the emotional pain experienced by the patient. If this is the case the patient invariably experiences that one collaborator wishes to become less involved. Also, it may be challenging to collaborate with a prescriber who attempts to treat every symptom experienced by the patient with a different medication because of a lack of an overarching theoretical understanding. This frequently appears as a problem in the treatment of women have been sexually abused because there is a failure by the pharmacotherapist to appreciate that affective lability, perceptual distortions, self-destructive behavior, and hopelessness about life, to mention just a few, are consistent with a traumatic disorder and readily treatable through a psychotherapeutic approach. Other prescribers will give patients multiple medications by rationalizing that they are treating comorbid disorders such as major depression and still others will fail to appreciate the potential for transference reactions to the prescriber in that they are giving the patient something that may be experienced as a significant and highly affect laden gift. Should the medication prove to be ineffective over time, as is often the case in many patients with personality disorders, for example, the physician should not be so quick to ascribe this to drug failure.

A nonprescribing professional my feel it is a violation of their beliefs to request the use of medication. Introducing medication can be experienced as an attack on a theoretical system. When this occurs, the patient is placed in a no-win situation as loyalty to one professional will demand disloyalty to the other.

Despite good intentions, the reality is that effective collaboration takes time and is not accomplished without a strong commitment from both care providers. Another persistent challenge is addressing the propensity for splitting by the patient, especially those with significant character disorders. Most frequently in this situation, one clinician is viewed by the patient as admirable, the other in highly negative terms. When this idealization or de-idealization is expressed in the treatment relationships it can be uncomfortable and difficult to address. Frequently, it may take the form of negativity about one collaborator. That is, some patients will be critical of the physician for only prescribing medication and others will feel similarly about the psychotherapist for not prescribing. Often patients will complain to the nonphysician collaborator that the doctor merely prescribes and is disinterested in any other aspects of the patient's life. Fourth, without consistent and effective collaboration, mental health professionals cannot appreciate that their patient has been providing very different information to each clinician.

Additional challenges are subsumed broadly under legal and ethical tensions. A frequent problem is the failure to establish clear guidelines for the sharing of clinical information between the collaborators. Similarly, the failure to delineate specific responsibilities for each collaborator can be very problematic. For example, who actually decides whether hospitalization is indicated, and if so, who should follow the patient while he or she is hospitalized? Should a diagnostic evaluation be performed by both professionals, or is it sufficient for either to conduct the assessment? Which professional should secure informed consent?

With respect to potential ethical dilemmas, Lazarus (1999) has noted that many psychiatrists are unclear about their supervisory or consultative responsibilities with a nonmedical therapist. In addition, there are inconsistent state licensing laws, the potential for the physician to delegate medical decisions to the collaborator, and the physician being merely a figurehead with responsibility but without the customary contractual safeguards that exist in most doctor–patient relationships. Lazarus also notes that within many managed care organizations in the US, cost containment is the greatest priority. If this is used as the basis for providing split care, then it is possible that the psychiatrist will feel that he or she is providing less than desirable care. This may result in resentment of the collaborator, which may potentially damage the treatment experience for the patient.

Interdisciplinary issues are often an enduring source of tension in collaborative treatments. When split treatment is mandated by a healthcare organization, it may be experienced as a ‘shotgun wedding’ approach rather than a true collaboration.

Also, competition may be a destructive element in collaborative relationships. Rivalry over inequality in the professional status and reimbursement of physicians may become an unhelpful source of acting out within the psychotherapy. Similarly, some physicians can be quite dismissive of their collaborator's skills and professional backgrounds. At least in the US, communication between the two professionals is rarely done well (Hansen-Grant and Riba, 1995). As a result, one collaborator may not know of a patient's suicidal or homicidal feelings or even when the other professional is out of the office and therefore unavailable to the patient. Some patients will not know whom they should contact if an emergency arises.

In psychoanalytic psychotherapy accepting and understanding transference feelings within the therapeutic dyad is a central, and at times, challenging task. However, this task becomes enormously complicated when there are two clinicians about whom the patient has distinct transference reactions. Consider also that the patient is receiving medication about which he or she may have strong conscious and/or unconscious feelings. To this therapeutic relationship must be added the attendant countertransferences from each of the providers. It is not difficult to imagine that the treatment experience for all participants can become complicated and confusing. The following clinical vignette illustrates nearly every problem (including the failure to ascertain important transference issues and the meaningfulness of medication to the patient) that has been discussed. The frustrating experience for the collaborators, and undoubtedly for the patient as well, can be understood within the context of poorly defined clinical roles, expectations, and professional boundaries.

A number of important conclusions can be drawn from this vignette about conducting effective split treatment. First, collaborative treatment cases should be selected carefully and collaborators should meet to discuss the reasons for referral when therapist and physician have never worked together (Rand, 1999). Once a successful working relationship has been established and the clinicians become comfortable with each other, communication then may be via telephone or written reports. Still, at times of crisis, collaborators may need to meet.

The physician or prescriber and therapist must agree on the responsibilities and boundaries of their collaborative work. Is the prescriber being viewed as consultant, equal partner, teacher, or supervisor? Collaborative treatment does not imply that the physician will supervise the treatment provided by the therapist or vice versa. Failure to clarify roles is the source of much confusion and ill will and may have legal implications. This is a particularly important point in the education of psychiatric and primary care residents when they work in medication clinics that employ split treatment. Beginning professionals often lack confidence in their skills and therefore may feel threatened by clinicians who have had different training and are more experienced. Other responsibilities of the collaborators include, but are not limited to agreement on their: frequency of communication, contact with family members where indicated, coverage when one collaborator is out of town, discussion with insurance personnel, and securing of informed consent. This understanding should be documented. As well, the patient must be educated to the characteristics of split treatment by both clinicians regarding their roles as collaborators and the need for constant sharing of treatment information. If splitting becomes prominent resistance in the treatment, collaborators often should meet in person to discuss their united approach to this problem. Physicians have an obligation to educate therapists about why medication is being prescribed as well as possible medication side-effects and how to report them. Therapists should assist the prescriber in identifying conflicts about medication, compliance problems, and the initial presentation of side-effects. In the eventuality of hospitalization, collaborators should be explicit about the level of responsibility and obligations of each. Collaborators must never use the patient to convey information that should be discussed more appropriately with the providers. Sometimes, the request for split treatment with a challenging referral can be a covert wish on the part of one clinician to either terminate or transfer a patient. It is difficult to overestimate the negative impact of such an issue on the patient as well as the collaborative relationship. Similarly, when treatment is to be discontinued, the decision about termination and follow-up (if required) should be jointly made and explained to the patient by each collaborator. In cases where a patient is not be able to establish a therapeutic relationship with one of the collaborators, both have an obligation to support a change in the treatment relationship. A reflection of an effective collaboration is the willingness of both collaborators to identify a therapeutic impasse or plateau and jointly to seek consultation. Lastly, the physician and the psychotherapist should never place each other in legal jeopardy by refusing to see the patient in crisis. At the initiation of each collaborative treatment relationship, therefore, both parties must stipulate about responsibilities in the event of crisis, coverage on vacations and weekends, and how clinical issues such as suicidality and homocidality will be evaluated. However, the physician should never place a therapist in the position of having to make medical decisions.

In the UK, where therapist and psychiatrist or primary care physician are often (but not always) both employees of the National Health Service, regular case conferences including the prescriber, psychotherapist, and care coordinator, and often the patient him/herself and family are an essential component of good mental health care. When patients are hospitalized, it is highly advantageous for members of the hospital treatment team to meet with collaborative treatment professionals. As a corollary, the wisdom of this approach is also helpful in the problems of a treatment impasse with difficult patients in any country. Collaborators should always be open to seek consultation when a treatment is proving to be ineffective, severe symptoms reappear and do not respond to intervention, or when tensions arise in the collaborative relationship.

First and foremost, all clinicians should familiarize themselves with the Diagnostic and Statistical Manual of Mental Disorders IV-TR (DSM-IV) published by the American Psychiatric Association (2000) or in Europe, the ICD-10 or International Classification of Disease and Related Health Problems (1992). Every clinician working with psychiatric patients in any capacity should have an appreciation of the nomenclature of mental disorders and their specific criteria. In addition to helping mental health professionals in their day to day clinical work with eliciting key symptoms and appreciating what diagnostic criteria constitute a particular disorder, these classification systems permit clearer communication from one clinician to another about patients. In the US, the ability to document a diagnosis is required by third party payors such as behavioral healthcare organizations. Moreover, these companies insist that mental health professionals provide a service that is likely to assist a particular patient with a particular psychological problem. Having said this, it is important that all clinicians appreciate the limitations of a categorical approach to understanding mental disorders. These include but are not limited to the following:

• each patient with the same diagnosis may not present in the same fashion

• each diagnosis may not always be distinguished from others.

This latter point is certainly the case in the classification of personality disorders, which has many clinicians arguing for a dimensional classification scheme. A dimensional model would attempt to quantify the attributes that characterize a patient rather than placing symptoms within a distinct diagnosis. As an example, a patient with borderline personality disorder may also have a number of features common to the diagnosis of narcissistic and histrionic personality disorders. Moreover, at times of maximal stress, some patients with borderline personality disorder may experience short-lived psychotic episodes. In short, providing a DSM-IV diagnosis for a patient is only one of the important components in developing the biopsychosocial formulation and treatment plan.

The following case example is representative of the issues that should be considered in the treatment of a patient who is depressed in determining whether medication may be of some use in treatment.

This vignette illustrates a number of important steps in conducting this or any other treatment. The clinician should:

• take a formal and in-depth history to secure a thorough appreciation of those factors contributing to his illness

• develop a formulation to provide a working hypothesis regarding the patient's condition

• establish a working diagnosis

• appreciate the severity of symptoms

• develop a treatment plan

• monitor the patient's compliance, his symptomatic and functional status, and response to treatment.

Before presenting the option of medication, the therapist speaks with the referring physician about her observations and the possible helpfulness of combined treatment, and if both clinicians agree, the policies regarding collaboration once medication is started. A thorough discussion with the patient then ensues about the possibility of medications, including highlighting their adjunctive role in treatment, exploring any initial resistance to take medication, encouraging them to ask the physician as much as possible about the medication, the need for the patient to be a collaborator in discussing his responses to the medication with both the social worker and the physician, and the importance of adhering to the medication plan as presented. In the US at least, the patient should sign an informed consent statement that detail the risks and benefits of combined treatment with both clinicians and stipulates that they are free to speak with each regarding the patient's progress in treatment.

Other indications that pharmacotherapy would be helpful for a patient who is treated with psychotherapy alone include the worsening of a patient's disorder, the appearance of a new disorder, and the failure of a patient to respond to the medication currently being prescribed.

As a second scenario, let us assume that Mr Davis presented with the following history: three previous episodes of depression, one requiring hospitalization; persistent thoughts of suicide, a weight loss of 12 pounds over a 6-week period without attempting to diet, and a pervasive feeling of hopelessness. On the Beck Depression Inventory he scores 39. In this instance, the social worker would have noted the presence of major depressive episode of significant disruption, safety concerns, and a significant history of depression that required hospitalization. As her patient's depression is severe, he is in significant psychological pain, she is aware that medication, in general, works more quickly than psychotherapy, and of the evidence to support combined treatment in moderately severe to severe depression, she contacts the referring physician and explores his willingness to collaborate with her in Mr Davis's treatment in which psychotherapy and pharmacotherapy will be initiated simultaneously.

It has been discussed previously that many patients who require medication and with severe medical or surgical illnesses may do better with a one-person model. As an example, a panic-disordered patient with severe inflammatory bowel disease may require, among many drugs, very high doses of a corticosteroid. This type of medication has a propensity for producing significant side-effects that often appear to be like other psychiatric disorders. Such patients may become profoundly depressed, manic, or psychotic. A psychiatrist is often able to monitor this patient's treatment more efficiently than might occur in a two-person model. For a surgeon and or gastroenterologist, they may find consulting with another physician to be more conducive to a collaborative situation as well. Also to be considered is the level of anxiety that a nonphysician might experience in treating a medically unstable patient. Of course, some patients who come to a nonmedical therapist, may decide that they would prefer seeing only one professional. For some patients who strongly evidence splitting and have demeaned the nonprescribing professional and which does not seem lessened by interpretation, they too may do better in an integrative relationship. A persistently suicidal person may unnerve either collaborator, but there may be times when a psychiatrist who has worked extensively with such patients and who has the opportunity to hospitalize, may treat the patient more efficiently. There are some patients with paranoid disorders requiring medication who threaten litigation consistently and the psychiatrist may be more comfortable with these types of behaviors. Also these patients can only maintain a single treating relationship and will always be a challenge but they may be better contained in an integrated situation. It is important to remember that as yet, data are not available to support any of these assumptions.

It has been noted that medication compliance problems should be anticipated in any treatment relationship be it an integrative or split approach. First and foremost, mental health professionals must understand the characteristics and the natural course of the condition that they are treating. This information must be provided to all patients in treatment. Psychoeducation for the patient, and at times family members or even employers as well, is considered essential to enhancing treatment adherence. For example, some disorders, as in the patient with major depression marked by repeated bouts of his illness, will require lifelong medication. This must be explained to the patient at the initiation of treatment. Often patients who have severe disorders that have been treated effectively in the initial phase will discontinue their medication when feeling better. Some patients, such as those with bipolar disorder, resist taking mood stabilizers because they dislike the dampening of their affect and at times their hyperactivity. These patients not infrequently feel stronger with boundless energy, hypersexuality, and elevated mood.

Patients must be assisted to understand signs of relapse. This is true for the majority of psychiatric disorders. In integrative and split treatment, patients must be educated about the length of time on medication before they experience some relief. This is true for antidepressants, antipsychotics, mood stabilizers, and some anxiolytics. Patients should be warned about discontinuing medication without informing the professional. Many psychotropic medications, if discontinued abruptly, produce rebound effects that patients experience as a worsening of their condition.

Essential to treatment with medication is the appreciation of the occurrence and meaning of side-effects. Nonprescribing professionals with experience generally become familiar with the common side-effects of classes of medication. When in doubt these therapists should routinely consult their collaborator. For common psychotropic medications, the following side-effects are helpful to keep in mind:

• Selective serotonin reuptake inhibitor antidepressants (SSRIs) during the first few weeks or treatment or so can produce, among others, headaches, gastroenterological symptoms, sedation, agitation, and sexual dysfunction. With the exception of the latter two, these side-effects will often cease after the first 2 weeks of treatment. Sexual dysfunction occurs in both men and women at significant rates and often lasts as long as patients are on these medications. The therapist can speak with a collaborator about steps to decrease this side-effect. Patients who become severely agitated during the first week of treatment will be instructed by the prescriber to discontinue the medication and an alternate medication may be provided.

• Typical antipsychotics, such as chlorpromazine and its relatives or haloperidol, have the potential to cause disturbing side-effects such as abnormal involuntary movements and severe dystonic symptoms such as a very painful stiff neck. In addition, patients who are taking these medications should be observed for a side-effect called tardive dyskinesia that presents as oral facial involuntary movements as this is an irreversible side-effect.

• In general, the side-effects of various mood stabilizers used in the treatment of bipolar disorder and to augment severe depression require greater depth of knowledge as these medications differ in their ability to produce specific side-effects.

There a number of rating scales that may assist the psychotherapist in monitoring side-effects. In patients with schizophrenia, for example, abnormal involuntary movements in various parts of the body can be assessed through instruments that do not require extensive training to administer and that can be used repeatedly in monitoring a patient's course. For atypical medications (those prescribed for off label indications) employed in treatment, it is the responsibility of the physician to inform his or her collaborator as well as the patient of side-effects that may occur.

Although every patient ascribes some meaning to their medication, when this meaning interferes with their ability to comply with treatment, this situation demands examination. As discussed earlier in this chapter, patients may hold negative beliefs about their therapist or pharmacotherapist that greatly impede treatment compliance. These beliefs, and specific beliefs about medication, should be anticipated, identified as resistances, and worked through.

Assisting a patient in taking medication can be made easier if clinicians are familiar with certain techniques. Those described by Beck (2001) are illustrative and recommended. Although Beck refers to CBT, her suggestions are helpful to psychotherapy of any theoretical persuasion. To review, she advocates that clinicians identify and address thoughts and beliefs about medication and psychological treatment that frequently interfere with a patient's ability to follow treatment recommendations. In providing any type of psychotherapy, distortions about self, about others, and his or her world view should be appreciated. Direct assessment of likelihood to adhere to treatment is essential. A patient who has failed previous treatments or has not been able to establish a therapeutic alliance should raise concern and requires thorough exploration. Beck suggests that is helpful to ask patients directly if they are likely to follow a medication regimen, whether they believe medication will work, do they have a specific fear about taking a psychotropic medication, would family members be against medication specifically or medication in general, and even if there are transportation or financial difficulties associated with the filling of a prescription. It is often productive to explore if the nonadherence occurs only at certain times. Once a patient's concerns are identified, the clinician in integrated or split treatment situations is obligated to educate and address certain misinformation held by the patient. Beck also notes that frequently in patients with compliance problems that it is facilitating to speak concretely about the advantages and disadvantages of taking medication. Lastly, Beck suggests other formal behavioral techniques may be of some assistance in dealing with noncompliance. These include having the patient accept praise or take credit for complying, visualize their lives if they chose not to comply, using a medication log, and employ coping cards that remind the patient why it is necessary to take the medication and details simultaneously identified resistances.