5b:  Treatment of Major Depression, Volume 2: Clinical Practice Guideline Number 5

Rationale for Guideline Development—The Cost of Depression

Up to one in eight individuals may require treatment for depression during their lifetimes; up to 70 percent of psychiatric hospitalizations are associated with mood disorders (Secunda, Katz, Friedman, et al., 1973). The direct costs of treatment for major depressive disorder, combined with the indirect costs from lost productivity, are significant. According to one study based on a 1980 population base, the total number of cases of major depressive disorder among those aged 18 or older in a 6-month period would be 4.8 million; in addition, 60 percent of suicides could be attributed to major depressive disorder (Stoudemire, Frank, Hedemark, et al., 1986). This translates to more than 16,000 suicides or 7 deaths per 100,000 annually. In 1980, mood disorders accounted for more than 565,000 hospital admissions, 7.4 million hospital days, and 13 million physician visits annually. Office-based psychiatrists' and psychologists' costs were $453 million, pharmaceutical costs were $138 million, the cost of home and institutional care was $141 million, and total direct costs were more than $2.1 billion. Indirect morbidity costs were estimated to be $10 billion; total mortality costs due to lost productivity, $4 billion. The total cost of mood disorders to society was roughly $16 billion.

Patients with major depressive disorder experience substantial pain; suffering; and psychological, social, and occupational disability during the depression (Johnson, Weissman, and Klerman, 1992; von Korff, Ormel, Katon, et al., 1992; Wells, Golding, and Burnam, 1988a, b). If depressive conditions accompany selected nonpsychiatric medical conditions (e.g., coronary artery disease, diabetes), the outcome of these concurrent general medical disorders is likely to be worse than if depression were not present (Carney, Rich, Freedland, et al., 1988; Carney, Rich, teVelde, et al., 1987a, b; Keitner, Ryan, Miller, et al., 1991; Lustman, Amado, and Wetzel, 1983; Lustman, Griffith, and Clouse, 1988; Lustman, Griffith, Clouse, et al., 1986).

In addition to economic costs, depression can carry great personal costs because of the social stigma associated with the diagnosis and treatment of a "mental" illness. This stigma likely plays a large role in patients' reluctance to seek, accept, adhere to, and continue treatment. Evidence compatible with the notion of stigma includes the following:

• Those with mental illnesses are often required to report them when applying for driver's licenses, security clearances, jobs, and other routine purposes, while those with other medical conditions are not. (The recent Americans with Disabilities Act has rectified this problem somewhat.)

• Third parties reimburse providers for treatment of mental illnesses at reduced rates and with marked capitations on total costs, although there are very few such restrictions on treatment for general medical disorders.

• A recent preliminary survey of physicians revealed that many cases of depression may not be coded in medical records, partly because of a desire to protect patients from stigmatization (Rost and Smith, 1992).

• Only about 20 percent or fewer patients with depression are treated in the mental health sector (Hough, Landsverk, Karno, et al., 1987; Shapiro, Skinner, Kessler, et al., 1984).

• Only 12 percent of people are willing to take medication for clinical depression, while more than 70 percent would take medication for a headache (Roper Reports, 1986).

As a result of stigma, depressed patients often incorrectly believe that they have caused their own illness or are solely responsible for its cure. They fear (sometimes correctly) subsequent discrimination in hiring, promotion, and other occupational opportunities. It is logical to infer that until depression is dealt with on an equal footing with nonpsychiatric medical disorders (attitudinally, economically, socially, occupationally, and politically), it will be underreported and national health statistics will likely remain misleading.

Given this stigmatization, it is vital to educate patients (and their families, if appropriate) about the nature, prognosis, and treatment of depression to increase adherence to treatment, relieve unnecessary guilt, and raise hope. Patients need to know the full range of suitable treatment options before agreeing to participate in treatment.

This Clinical Practice Guideline was developed with support from the Agency for Health Care Policy and Research (AHCPR) by the Depression Guideline Panel to assist both patients and primary care providers (e.g., general practitioners, family practitioners, internists, nurse practitioners, registered nurses, physician assistants, and others) in the treatment of major depressive disorder. The general principles embodied in these guidelines should also provide a framework for others who treat depressed persons. A detailed description of the development of these guidelines is found in Depression in Primary Care: Volume 1. Detection and Diagnosis (AHCPR Publication No. 93-0550, 1993) and in Chapter 1 of the Depression Guideline Report (Depression Guideline Panel, forthcoming).

These guidelines are not intended to render selected procedures reimbursable or not reimbursable. That decision logically falls to third-party payors. Likewise, the guidelines do not specify which professionals should conduct which procedures, an issue addressed by licensing/privileging bodies. Should the recommended steps in the treatment of depression fall outside the expertise of the practitioner or be impractical, given constraints of time or availability of appropriate resources, the practitioner should consult with, or refer the patient to, someone knowledgeable in these matters.

Methodological Background

This Clinical Practice Guideline is an abbreviated version of a far larger document, the Depression Guideline Report, to which the reader may refer for further detail. These treatment guidelines focus on outpatients with major depressive disorder, particularly those seen in primary care settings. They do not address the treatment of children or adolescents, bipolar disorder, or depressive symptoms insufficient to meet the criteria for major depressive disorder. The panel members believe that the steps and procedures proposed here form a reasonable general treatment plan in many cases; they have attempted to be as explicit as possible in making recommendations, such as how often to see a patient, when to increase a medication dosage, and when to change to a different treatment. However, the treatment of an individual patient requires adaptation of both the general (strategic) and specific (tactical) recommendations to suit the patient's particular situation.

• Good research-based evidence, with some panel opinion, to support the guideline statement.

• Fair research-based evidence, with substantial panel opinion, to support the guideline statement.

• Guideline statement based primarily on panel opinion, with minimal research-based evidence, but significant clinical experience.

In applying these guidelines, several caveats are in order. First, the reader should not confuse the absence of studies with the absence of efficacy. In certain situations, there is little direct evidence for the efficacy of various treatments, based on randomized controlled trials (for example, with patients who have major depressive disorder and concurrent general medical conditions). Furthermore, several commonly used treatments, such as supportive psychotherapy, have not been subjected to randomized controlled trials. In the absence of evidence, no scientifically based statement about the efficacy of a treatment can be made. However, case reports, case series, clinical experience, and logical inference form a basis for selecting treatments for particular patients.

Second, the primary care practitioner is cautioned not to persist with extensive, multiple medication trials or with prolonged psychotherapy to which the patient is not responding because the longer the patient's depressive episode lasts, the more difficult it may be to treat (Bielski and Friedel, 1976; Rush, Hollon, Beck, et al., 1978). For example, patients who have not remitted with one or two well-conducted antidepressant medication trials (with or without psychotherapy) or, for less severe cases, a trial of psychotherapy not exceeding 12 weeks are likely candidates for consultation or referral.

Third, a consultation or referral to a specialist is always an option throughout the patient's management. It may be called for immediately, for example, if the patient is suicidal, or later. Consultation or referral to a specialist is particularly appropriate in the following instances:

• The practitioner is not skilled in making the diagnosis.

• The diagnosis is unclear.

• The practitioner is not skilled in providing the recommended treatment.

• Treatment does not result in successful outcome.

• These recommendations cannot be adapted or followed for practical, clinical reasons.

Specific Methodological Strategies

This summary of the methods used by the panel to evaluate treatments considered in this document provides the reader with a better understanding of the scientific basis and limitations of these guidelines, as well as an appreciation of the current state of the depression literature and the methodological and research issues that need further investigation. (See Depression Guideline Panel, forthcoming, for more detail.)

The Depression Guideline Panel's charge was to determine the efficacy of various treatments for major depressive disorder in patients likely to be seen by primary care providers. The ideal evidence on which to base conclusions would be randomized controlled trials of the various therapies conducted in primary care settings with outpatients who have major depressive disorder. A perusal of the literature reveals few such studies. The panel, therefore, constructed an indirect model, incorporating information from non-primary care settings, and extrapolated this information to the setting of interest. The general strategy was to identify all relevant literature, summarize the results in evidence tables, combine results across each study using meta-analysis, and compare the efficacy of alternative therapies.

Meta-Analysis

Methodology and Limitations

The panel used meta-analysis following the confidence profile method (CPM) to calculate summary statistics that describe the likely effects of each treatment considered. (Details can be found in Eddy, Hasselblad, and Schacter, 1990.) Using a hierarchical Bayesian random-effects model, the panel calculated the probability distribution describing the results that would be expected if a hypothetical additional study, similar to the ones included in the analysis, were performed. By taking into account the heterogeneity of study results, this type of analysis depicts the range of results that practitioners can expect should they use the treatment in their own practice settings.

Figure 2. Example of a meta-analysis result

   Figure 2. Example of a meta-analysis result

Each meta-analysis produces a probability distribution depicting the likelihood that the parameter of interest falls within any particular range of values. For example, the meta-analysis result depicted in Figure 2 indicates the difference in success rate for two alternative therapies. The curve represents the percent difference in success rate between the therapies. It can be said that the mean difference between them is 5 percent (.05), with a standard deviation of 6.6, that 95 percent of the area under the curve (the Bayesian equivalent of a 95 percent confidence interval) lies between -8 and 17.7 percent, or that there is a 22.5 percent chance that the actual difference is less than zero. Since Figure 2 is a probability distribution, it is easy to determine the probability that the true difference in the effect of the treatment is greater than, less than, or equal to any selected value. The standard deviation suggests the shape of the distribution; distributions with small standard deviations relative to their mean are tall and narrow, indicating a high degree of certainty regarding the result.

Several factors can compromise the internal validity of the meta-analyses. First, while the random effects model accounts for among-study variations and, therefore, for random bias, it cannot account for any systematic biases that occurred in all of the studies. Second, to be included in the meta-analysis, studies had to include sufficient data to permit calculation of the percent response for each treatment, based on a modified intent-to-treat analysis that used either the HAM-D or the CGI (medication trials) or the BDI (psychotherapy trials). If studies without sufficient data to permit inclusion were fundamentally different from those that were included, summary statistics may be biased. Similarly, a variety of publication biases (particularly the tendency to publish only those studies with positive findings) may result in biased summary statistics.

On the other hand, the hierarchical random effects model is robust. Sensitivity analyses reveal that it would take a huge number of very large studies to change the results in any important way.

Threats to the external validity of the meta-analysis relate primarily to the generalizability of the study populations. While most studies entered a well characterized group of patients with major depressive disorder, others included small, but unspecified, numbers of patients with bipolar disorder or other psychiatric co-morbidities. It is, therefore, difficult to state with certainty the patient populations that the meta-analyses describe. Evidence suggests that patients with different types of depression have different prognoses and react differently to treatment. For example, the placebo response rate for nonpsychotic major depressive disorder is on the order of 25 percent, while the placebo response rate for psychotic depressions is only about 10 percent (Schatzberg and Rothschild, in press). In the absence of knowledge about the exact case mix included in the meta-analysis, caution is necessary to avoid overstating the degree of certainty that is assigned to the results.

Limitations in Comparisons Across Trials

After the meta-analyses have been performed, there is a great temptation to make direct comparisons between summary statistics for medication and psychotherapy trials. Such comparisons should be approached with caution because of the following limitations:

• Patients who agree to participate in studies that provide both medication and psychotherapy treatment cells may differ from patients who participate in medication trials with two active medication cells or in psychotherapy trials that do not offer a medication cell.

• Psychotherapy studies typically last longer, which, on the one hand, increases the chances of a spontaneous remission and prolongs exposure to treatment but, on the other hand, increases the opportunity for dropout.

• The pill placebo controls used in medication studies are not equivalent to the wait-list controls employed in psychotherapy trials. The pill placebo controls receive the same support and case management that the active medication group receives. Many medication trials contain such a pill placebo arm. What is the appropriate control for a talking therapy? Only one psychotherapy trial containing a pill placebo control has been conducted to date (Elkin, Shea, Watkins, et al., 1989). More commonly, patients kept on a wait-list constitute the control group. Most proponents of talking therapies would argue that it is not just the talking, but the content of the talking, that makes the therapy effective. Therefore, the appropriate control should be some type of nonspecific talking that is devoid of the "active" content of the therapy in question so that controls still receive the support and case management that active psychotherapy patients receive. The appropriateness of the placebo becomes important in attempting to compare the treatment-placebo difference across treatment types. A medication that is 30 percent better than its pill placebo control may be more effective, less effective, or equally effective than a psychotherapy that is 40 percent better than its wait-list control. However, one must suspect that those in the pill placebo group received more support, guidance, and reassurance than did those placed on a wait-list, making the comparison between medication and placebo more stringent than that between psychotherapy and wait-list.

• The BDI and other self-reports commonly used in psychotherapy trials may be less sensitive (or slower to show improvement) than the clinician ratings (i.e., HAM-D, CGI) typically used in medication trials.

• Medication trials usually require greater symptom severity for study entry than do psychotherapy trials. Thus, less severely ill patients are typically studied with psychotherapy, while more severely ill patients are studied in medication trials.

Given these limitations, comparisons of medication and psychotherapy trials should not be given undue importance and should be considered hypothesis-generating rather than hypothesis-testing.

Generalizability Issues

Although randomized trials provide the best evidence for the efficacy of a treatment in a specific type of patient, their applicability to the community at large may be limited by the trials' stringent enrollment criteria, unique treatment settings, and unrepresentative clinical procedures. While researchers are working to design methods that address this problem (see Cross Design Synthesis: A New Strategy for Medical Effectiveness Research, US GAO B244808, 1992), these methods were not yet sufficiently developed for use in this project. For these reasons, the panel restricted its analysis to clinical trials. Two caveats must be emphasized, however, one related to patient populations and the other related to treatment quality.

The purpose of these guidelines is to make recommendations regarding the treatment of major depressive disorder in primary care settings. Analyses rely on the results of clinical trials that were typically performed in academic psychiatric settings and involved patients without other medical problems. This methodology raises concerns. First, clinical trials require patient consent and participation. Many patients decline a protocol for various reasons (e.g., too severely ill to accept the additional measurement procedures, fear of being placed on a placebo or wait-list, desire for medication, desire for psychotherapy). Thus, patients who enroll in trials may not be representative of the population of interest. Second, clinical trials of medications are often sponsored by pharmaceutical companies seeking Food and Drug Administration (FDA) approval for their products. To optimize the chances of a good result and to avoid problems with human subject protection committees, these studies typically exclude patients with important medical co-morbidities. Because many patients treated in primary care settings have such co-morbidities, the population of interest is different from the population summarized by the meta-analysis. It is not possible to determine, from the available evidence, how depressed patients with medical co-morbidities will respond to the various therapies. While the few studies done in primary care settings demonstrate response rates similar to those found in psychiatric settings, they are few in number, and they, too, have generally excluded patients with medical co-morbidities.

Randomized controlled trials are often conducted in research settings and always follow a prespecified protocol. Psychotherapy is often specified in a written "manual," with therapists restricted to particular procedures in research trials. For these reasons, the type of treatment provided as part of a clinical trial may differ substantially from that provided in routine practice. Thus, inferences from research trials to day-to-day practice remain tentative, as protocol-driven treatments may perform better or worse than do their community practice counterparts.

Potential Problems with Meta-Analysis Interpretation

Any time a new method for generating numerical output -- in this case, meta-analysis -- becomes available, there is potential for misunderstanding and abuse of the numbers. The panel offers some warnings regarding the meta-analysis presented in this Clinical Practice Guideline.

Figure 3. Comparison of meta-analysis results for treatments (more...)

   Figure 3. Comparison of meta-analysis results for treatments A and B

It is important not to attach undue significance to small differences. Figure 3 depicts the results of two meta-analyses, one for treatment A, with a success rate of 34 percent (SD 12), and one for treatment B, with a success rate of 28 percent (SD 11). While comparison of the means reveals that A is 6 percent better than B, there is about a 34 percent chance that B is actually better than A. Therefore, it would be improper to conclude with any certainty that A is superior.

It is also important not to make improper inferences regarding numbers that are the same. In many cases, the data show that several treatments have similar response rates. These data could lead to the assertion that it is necessary only to use the least expensive agent. This assertion is true if the same patients respond to each treatment. However, there is strong evidence for biologic and psychological heterogeneity among patients with major depressive disorder, which is evident in the differential response of patients to medication (Goodwin and Jamison, 1990; Rush, Cain, Raese, et al., 1991). Thus, for a particular patient, drug A may be ineffective, while drug B may be quite effective; in another patient, the opposite may be true. Furthermore, some research suggests that certain medications may be effective earlier in the longitudinal course of recurrent mood disorders, while others may be better in more longstanding cases (Post, 1992). Similarly, clinical pharmacology studies and clinical experience provide evidence that patients differ in the nature, likelihood, and severity of the side effects that they experience with a medication. One patient may become sedated on a drug, another may develop insomnia, while others have no sleep difficulties. This heterogeneity suggests that more than one agent must be available to ensure adequate treatment of all patients.

Benefits and Harms of Treatment

Formal treatments for major depressive disorder fall into four broad domains: medication, psychotherapy, the combination of medication and psychotherapy, and ECT. (Light therapy is also a treatment option for mild to moderate seasonal depressions.) Each domain has benefits and risks that must be weighed carefully in the selection of a treatment option for a given patient.

Clinical Management

Treatment for major depressive disorder may include three phases: acute, continuation, and maintenance. The overall aim of all three phases is the attainment of a stable, fully asymptomatic state and full restoration of psychosocial function (a remission). Acute treatment aims at removing all depressive symptoms. If the patient improves with treatment, a response is declared. A remission may occur either spontaneously or with treatment. If the symptoms return and are severe enough to meet syndromal criteria within 6 months following remission, a relapse (return of symptoms of the current episode) is declared. Continuation treatment aims at preventing this relapse. Once the patient is asymptomatic for at least 6 months following an episode, recovery from the episode is declared. At recovery, continuation treatment may be stopped. For those with recurrent depressions, however, a new episode (recurrence) may occur months or years later. Maintenance treatment aims at preventing a recurrence. Recurrences are expected in 50 percent of cases within 2 years after continuation treatment (NIMH, 1985). For well established, recurrent depressions, the rate may approach 75 percent (Frank, Kupfer, Perel, et al., 1990).

Figure 4. Phases of treatment

   Figure 4. Phases of treatment

Figure 4

• A patient's full treatment plan must be based on his or her history of illness.

• The simple attainment of remission following acute treatment may be followed by relapse if continuation treatment is not provided.

• Not all, but a selected subgroup of patients, will require maintenance treatment to prevent recurrences.

Figure 5. Overview of treatment for depression

   Figure 5. Overview of treatment for depression

Figure 5

For virtually all patients, the practitioner who provides the medication (or other treatment) also provides support, advice, reassurance, and hope as -- well as side-effect monitoring (including vital signs for those on medication), and dosage adjustments. This "clinical management" is exceptionally important, especially with depressed patients, whose pessimism, low motivation and energy, and sense of social isolation or guilt may lead them to give up, not adhere to treatment, or even to drop out of treatment.

Declaring a Treatment Response or Failure

Guideline: A 4- to 6-week trial of medication or a 6- to 8-week trial of psychotherapy usually results in at least a partial remission (50 percent symptom reduction), and a 10- to 12-week trial usually results in a nearly full response (minimal or no symptoms) to treatment. However, full restoration of psychosocial function often takes longer. (Strength of Evidence = B.)

Once selected, the initial treatment should be applied for a sufficient length of time to allow a reasonable assessment of the patient's response. Switching treatment too early provides no benefit, inappropriately discourages the patient, or leads to an erroneous conclusion that the treatment is ineffective. On the other hand, persisting over a prolonged period without any response is costly in pain and suffering and may unnecessarily prolong the episode.

Table 4. Number of randomized controlled trials of (more...)

Table 4. Number of randomized controlled trials of medication in patients with major depressive disorder [*1,*2]

Medication	Adult				Geriatric
	Acute	Cont	Maint	Acute	Cont	Maint
Tricyclic Antidepressants (TCAs)
Amitriptyline	101 [45]	3 (2)	3	8 [3]	0	0
Desipramine	22 [7]	0	0	0	0	0
Doxepin	19 [7]	0	0	8 [4]	(1)	0
Imipramine	115 [66]	(5)	2	11 [7]	(1)	0
Nortriptyline	6 [5]	0	0	5 [2]	(1)	1
Protriptyline	1 [0]	0	0	0	0	0
Trimipramine	12 [5]	0	0	0	0	0
Heterocyclic Antidepressants
Amoxapine	19 [11]	0	0	1 [0]	0	0
Bupropion	16 [10]	(4)	0	2 [0]	0	0
Maprotiline	28 [14]	(1)	0	1 [1]	0	0
Trazodone	25 [13]	(3)	0	4 [3]	0	0
Selective Serotonin Reuptake Inhibitors (SSRIs)
Fluoxetine	32 [28]	(4)	1	3 [2]	(1)	0
Fluvoxamine*3	18 [12]	0	0	1 [1]	0	0
Paroxetine	10 [5]	0	1*4	1 [1]	0	0
Sertraline	2 [2]	0	1*4	1 [1]	0	0
Monoamine Oxidase Inhibitors (MAOIs)
Isocarboxazid	14 [9]	0	0	0	0	0
Phenelzine	30 [20]	(2)	1	1 [1]	(1)	1
Tranylcypromine	11 [6]	0	0	0	0	0
Anxiolytics
Alprazolam*5	22 [9]	0	0	0	0	0
Buspirone*5	2 [2]	0	0	0	0	0
Chlordiazepoxide*5	4 [2]	0	0	0	0	0
Diazepam*5	11 [2]	0	0	0	0	0

*1 Numbers in brackets show the number of cells that were meta-analyzed for each drug.

*2 Numbers in parentheses indicate the number of double-blind, nonrandomized, extension trials within continuation or maintenance phases.

*3 Not FDA-approved for use in the United States.

*4 Completed after 1990.

*5 Not FDA-approved as an antidepressant medication.

Note: Cont = Continuation. Maint = Maintenance.

The basis for this guideline rests on a careful review of 528 randomized controlled trials of medication (Table 4, page 47) and the 46 randomized controlled trials of psychotherapy (Table 11, page 75) in adult and geriatric out patients (Depression Guideline Panel, forthcoming). A partial response at 6 weeks may indicate the need for further medication dosage adjustments or a longer trial of the same treatment (i.e., medication, psychotherapy, or the combination, see pages 113-115). After 12 weeks, a partial response suggests the need for adjunctive psychotherapy, adjunctive medication, a switch to a different treatment, or consultation/referral.

Objectives of Acute Phase Treatment

Table 1. Objectives and effects of different treatments (more...)

Table 1. Objectives and effects of different treatments

Objective	Medication	Treatment		ECT
		Psychotherapy	Combined
Symptom reduction	Direct effect	Direct effect (cognitive) (behavioral) (interpersonal)	Direct effect	Direct impact
Improved function	Indirect effect (secondary to less depression)	Direct effect (marital)	Indirect and direct effects	Indirect effect (secondary to less depression)
Recurrence prevention	Direct effect (maintenance medication)	Direct effect (continued therapy) Indirect effect (learned skills following therapy)	Direct effect (maintenance medication or psychotherapy)	Direct effect (maintenance medication after ECT)

Note: ECT = Electroconvulsive therapy.

Indications for Acute Phase Treatment

Effective treatment rests on accurate diagnosis. The practitioner must first determine whether the patient has a clinical depression or is simply suffering normal sadness and distress. This distinction is analogous to others made in general medicine. When the patient's condition is primarily sadness, supportive discussions and/or the passage of time may be all that is necessary to resolve the symptoms. On the other hand, if a formal mood syndrome is present, specific treatments are usually indicated because there is clear evidence for their efficacy, because untreated major depressive episodes exact a high cost in pain and disability, and because the long-term prognosis for untreated major depressive disorder is poor (NIMH, 1985; Prien and Kupfer, 1986).

For patients who have very mild cases of major depression or whose diagnosis is unclear (e.g., major depression versus adjustment reaction with depressed mood) and who are not in immediate danger or are not suffering significant functional impairment, the practitioner may want to schedule one to two additional weekly evaluation visits to determine whether symptoms will abate without formal treatment or to discuss treatment options with the patient. There is evidence that clinical management leads to remission in 20 to 30 percent of cases (Elkin, Shea, Watkins, et al., 1989). However, several cautions are in order regarding extended evaluations:

• If the patient has a history of previous major depressive episodes, early intervention is recommended to prevent pain and suffering, once the presence of an episode has been clearly established (NIMH, 1985; Prien and Kupfer, 1986).

• Extended evaluations are not recommended for moderate, severe, or psychotic patients or for those with significant functional impairment.

• Although a reduction in symptoms often occurs with extended evaluation, this does not indicate that no treatment is needed. A full remission is the objective of both nonspecific and formal treatments.

• All patients who remit completely with clinical management (extended evaluation) alone should be followed up carefully (roughly two to three visits over the next 6 to 12 months), as evidence suggests that a large percentage may suffer a recurrence (Shea, Elkin, Imber, et al., 1992).

Given the evidence to date, it is appropriate to treat patients with moderate to severe major depressive disorder with medication whether or not formal psychotherapy is also used. For milder cases of major depressive disorder, there is some (albeit less clear-cut) evidence for the efficacy of medication versus placebo. Medication is administered in dosages shown to alleviate symptoms. The specific medication choice is based on side-effect profiles, history of prior response, family history of response, and type of depression. Typically, no one antidepressant medication exceeds the others in efficacy; some patients respond well to one, while others respond to a different treatment.

In general, the objectives of the formal psychotherapies in the treatment of major depressive disorder are similar to those of medication: symptom remission, improved psychosocial functioning, and prevention of relapse/recurrence. Most of the limited available data (see Table 12, page 76) have established that formal psychotherapy as the sole acute treatment for major depressive disorder is more effective than a wait-list control in outpatients with mild to moderate, nonpsychotic major depressive episodes.

Most psychotherapy efficacy studies have focused on symptom amelioration either by direct, time-limited, symptom-targeted psychological treatments, such as cognitive or behavioral therapy, or by time-limited treatments targeted at resolution of current interpersonal difficulties (interpersonal psychotherapy) or psychological conflicts (brief dynamic psychotherapy) assumed to act as vulnerability or precipitating factors or to maintain the syndrome once it has been established.

Preferred psychotherapeutic approaches are those shown to benefit patients in research trials, such as interpersonal, cognitive, behavioral, brief dynamic, and marital therapies. Because untested therapies are not, by definition, known empirically to be either effective or ineffective, these guidelines recommend choosing tested therapies over untested therapies, when available. The therapy should be limited to 20 sessions, since efficacy research on longer forms of therapy is not available and since strong evidence for the efficacy of medication with clinical management is available.

A second objective of formal psychotherapies is to address the patient's associated psychosocial problems, even if symptom control is largely accomplished with medication. In these patients, it is important to identify the objectives of therapy before selecting the specific treatment. Often, these associated problems are consequences of the depressive episode itself. If the depressive episode is effectively relieved with medication alone, the associated psychosocial problems often abate without additional psychotherapy (Mintz, Mintz, Arruda, et al., 1992). Thus, a reassessment of the patient's condition is advisable once symptom relief has been obtained with medication. The continued presence of associated psychosocial problems provides a reasonably strong rationale for augmenting treatment with formal psychotherapy aimed at the residual problems, even though sequential, randomized trials to support this stepwise approach are largely lacking to date.

Treatment Selection

Several general principles guide the selection of the first acute treatment:

• The patient and, where appropriate, the family need to be explicitly informed of the nature of the illness, all relevant treatment options, prognosis, and overall treatment plan.

• The ultimate selection of the first and subsequent treatments should, whenever possible, be a collaborative decision between practitioner and patient. Such shared decision making is likely to increase adherence and, therefore, treatment effectiveness.

• Treatments with established efficacy are preferred initially over less well tested or untested interventions.

• The initial treatment must be used in optimal fashion for a sufficient length of time to determine whether it is effective for the specific patient.

• Visits should be sufficiently frequent to optimize adherence.

• Outcome should be carefully assessed by interview for criterion signs and symptoms of depression. (This may be facilitated, but not replaced, with a self-report rating scale.)

• If the patient does not respond, it is logical to switch to a second treatment, as a wide range of effective treatments is available and no one treatment is effective for every patient.

• If the initial one or two treatments fail, a consultation with or referral to a psychiatrist or other mental health care professional trained in the diagnosis and treatment of depression is indicated.

• Consultation, if needed, must not be overly delayed, since longer episodes of major depressive disorder are more difficult to treat with either medication or psychotherapy (Bielski and Friedel, 1976; Goodwin and Jamison, 1990; Rush, Hollon, Beck, et al., 1978) and those with more recurrent forms of the illness may require more complex pharmacologic treatment (Post, 1992).

• The presence of a competent practitioner is a prerequisite for the selection of a specific treatment. It has been shown, for example, that the competency of the psychotherapist affects treatment efficacy (Shaw, Elkin, Vallis, et al., unpublished manuscript, 1993; Shaw and Olmsted, 1989). There is variability in the quality of administration of all treatments, and the quality of administration affects outcome for both medication and formal psychotherapies.

• For those with milder, less chronic, less recurrent, nonpsychotic major depressions, an extended evaluation (two to three visits) during which symptoms are monitored may usefully identify those who will remit fully with clinical management alone. However, careful followup is called for even in these patients since, in many, symptoms will return.

Table 2. Strategic choices in the acute treatment of (more...)

Table 2. Strategic choices in the acute treatment of major depressive disorder

Medication[1]	Define Treatment Phases, Objectives, and Options with Patient (and Family, Where Appropriate)		ECT[4]
	Formal Psychotherapy	Combined Treatment
More severe	Less severe	More severe	Psychotic
Chronic	Less chronic	Chronic	Severe or extremely severe
Recurrent	Nonpsychotic	Partial response to either alone	Prior positive response
Psychotic	Prior positive response	Availability	Failure on several medications or combined treatment trials
Melancholic	Availability	Personality disorder[3]	Need for rapid response
Prior positive response	Medical contraindication to medications	Patient preference	Medical contraindication to medications
Family history	Patient preference[2]
Patient preference
Failure to respond to psychotherapy

[1] Medication is always combined with clinical management.

[2] Patient preference applies more if depression is milder, nonpsychotic.

[3] This recommendation has not been empirically tested. It rests solely on clinical experience.

[4] Electroconvulsive therapy (ECT) is very rarely required for patients seen in primary care settings. It is reserved nearly always for those who have severe, often chronic, often psychotic depressions that have not responded to several trials of different standard medications.

Indications for Acute Phase Medication

Table 3. Considerations for acute phase medication

Table 3. Considerations for acute phase medication

Indication	Strength of Indication
Melancholic symptoms	Very strongly recommended
Psychotic symptoms	Very strongly recommended
Severe symptoms	Very strongly recommended
Moderate symptoms	Strongly recommended
Maintenance treatment planned	Very strongly recommended
Previous positive response to medications	Strongly recommended
Recurrent (> three episodes)	Very strongly recommended
At least two episodes with:
Poor interepisode recovery	Strongly recommended[1]
Family history of depression	Strongly recommended[1]
Atypical symptoms	Recommended[2]

[1] Because a recurrent form of depression is likely (and, therefore, the need for maintenance medication).

[2] Because there are several randomized controlled trials indicating medication is more effective than placebo, but psychotherapy has not been studied to date by such a trial in this group.

• For severe and psychotic depressions, there is strong evidence for the efficacy of medication and little or none for the efficacy of psychotherapy alone.

• There have been no randomized controlled trials of psychotherapy in patients with major depressive disorder with atypical symptom features, but placebo-controlled randomized controlled trials do indicate that antidepressant medications are effective (especially the monoamine oxidase inhibitors [MAOIs]).

• Maintenance medication clearly prevents recurrences, while, to date, maintenance psychotherapy does not (Frank, Kupfer, Perel, et al., 1990).

For virtually all patients, the practitioner who provides the medication also provides support, advice, reassurance, and hope, as well as monitoring side-effects (including monitoring of vital signs), adjusting the dosage, or switching the medication, if needed. This "clinical management" is critical with depressed patients, whose pessimism, low motivation, low energy, and sense of social isolation or guilt may lead them to give up, not adhere to treatment, or drop out of treatment. Clinical management of medication has been detailed in the form of a manual (Fawcett, Scheftner, Clark, et al., 1987) and may well be effective, either alone or with a placebo, in mildly to moderately depressed outpatients with major depressive disorder. For this reason, clinical management is viewed not only as a method to optimize dosage and patient adherence to medication, but also as a treatment, albeit nonspecific, in its own right. Common sense and a body of studies on adherence suggest that appropriate clinical management improves adherence and, therefore, patient response. In addition to routine clinical management, specific sessions aimed at increasing adherence (adherence counseling) may be indicated in the situations noted in Chapter 2.

Evidence for Efficacy

The search for randomized controlled trials evaluating the efficacy of medication treatment for adult and geriatric patients with major depressive disorder was restricted to English language literature published from 1975 to the present, as the diagnostic taxonomies in use before that time are incompatible with the present system. However, studies of MAOIs carried out between 1960 and 1975 were included because they represent an important body of work and used sufficiently well defined patient groups to make inferences to the present system. The vast majority of studies were conducted on outpatients with moderate to severe, nonpsychotic major depressive disorder seen in non-primary care settings.

Table 5. Meta-analyses of antidepressant medications (more...)

Table 5. Meta-analyses of antidepressant medications for patients with major depressive disorder (intent-to-treat samples)

Drug	Drug Efficacy		Adult Drug-Placebo		Drug-Drug
	Inpt	Outpt	Inpt	Outpt	Inpt	Outpt
Tricyclics
Amitriptyline	54.0% (12.7) [13]	60.1% (14.1) [32]	31.3% (15.5) [2]	21.8% (5.7) [9]	1.3% (11.3) [11]	-0.2% (4.2) [32]
Desipramine	47.5% (10.7) [1]	52.1% (7.7) [6]	N/A	33.5% (19.5) [2]	-26.0% (15.1) [1]	-7.3% (15.4) [2]
Doxepin	37.5% (13.4) [1]	54.2% (11.6) [6]	N/A (5.7)	28.9% (19.0) [1]	-25.0% (6.1) [1]	0.1% [7]
Imipramine	45.4% (10.7) [13]	47.7% (9.0) [53]	21.3% (10.8) [6]	18.0% (6.9) [33]	0.0% (10.3) [13]	-3.0% (7.2) [48]
Nortriptyline	50.6% (19.4) [3]	44.5% (9.9) [2]	N/A	8.6% (8.6) [1]	-1.6% (21.4) [4]	1.5% (11.6) [2]
Protriptyline	N/A	N/A	N/A	N/A	N/A	N/A
Trimipramine	9.6% (12.9) [2]	63.2% (17.2) [3]	N/A	N/A	20.1% (17.8) [2]	0.3% (13.6) [1]
Total	50.0% (6.5) [33]	51.5% (5.2) [102]	25.1% (11.5) [8]	21.3% (3.9) [46]	-4.0% (7.4) [32]	-0.8% (3.4) [92]
Heterocyclics
Amoxapine	54.8% (14.2) [3]	58.9% (11.6) [8]	N/A	27.5% (9.5) [1]	4.3 (14.9) [3]	-1.4% (6.9) [8]
Bupropion	51.9% (9.7) [4]	66.6% (15.3) [6]	39.5% (10.2) [3]	16.8% 96.6) [1]	N/A	-5.0% (7.3) [5]
Maprotiline	62.8% (12.5) [2]	63.2% (6.0) [12]	N/A	-5.9% (29.1) [2]	19.2% (19.9) [2]	4.7% (5.3) [12]
Trazodone	60.2% (11.8) [[4]	59.8% (10.0) [9]	38.0% (13.5) [4]	23.6% (16.9) [3]	7.3 (19.4) [4]	7.9% (10.4) [9]
Total	55.1% (4.8) [13]	62.3% (11.0) [35]	39.3% (8.3) [7]	16.5% (9.9) [7]	8.7% (9.7) [9]	2.1% (4.5) [34]
Tricyclics
Amitriptyline	N/A (11.3) [3]	44.4% N/A	N/A	N/A	2.4%	(10.9) [3]
Desipramine	N/A	N/A	N/A	N/A	N/A	N/A
Doxepin	80.0% (12.1) [2]	31.7% (10.0) [3]	46.7% (19.2) [1]	26.7% (16.0) [1]	-10.0% (15.1) [1]	-7.9% (8.3) [3]
Imipramine	94.4% (7.2) [1]	40.7% (9.2) [6]	63.2% (17.1) [1]	17.4% (15.4) [5]	11.1% (13.8) [1]	-10.7% (15.3) [4]
Nortriptyline	N/A	50.2% (16.6) [2]	N/A	37.5% (18.7) [2]	N/A	0.9% (13.8) [1]
Protriptyline	N/A	N/A	N/A	N/A	N/A	N/A
Trimipramine	N/A	N/A	N/A	N/A	N/A	N/A
Total	83.2% (14.3) [2]	40.4% (6.7) [14]	53.5% (21.9) [2]	22.0% (14.0) [6]	1.0% (20.7) [2]	-3.9% (6.8) [11]
Heterocyclics
Amoxapine	N/A	N/A	N/A	N/A	N/A	N/A
Bupropion	N/A	N/A	N/A	N/A	N/A	N/A
Maprotiline	N/A	36.7% (8.8) [2]	N/A	N/A	N/A	25.3% (13.2) [1]
Trazodone	13.3% (8.5) [1]	36.7% (8.8) [2]	N/A	19.9% (13.1) [1]	-42.9% (14.7) [1]	10.9% (10.6) [3]
Total	13.3% (8.5) [1]	43.3% (12.1) [3]	N/A	19.9% (13.1) [1]	-42.9% (14.7) [1]	13.4% (11.8) [4]
Selective Serotonin Reuptake Inhibitors (SSRIs)
Fluoxetine	67.2% (8.2) [1]	46.6% (11.7) [27]	N/A	21.7% (9.7) [16]	0.0% (11.6) [1]	2.7% (9.2) [14]
Fluvoxamine	51.3% (12.1) [6]	42.5% (14.1) [6]	36.9% (13.6) [1]	14.0% (7.4) [4]	4.4% (11.2) [6]	-1.5% (5.6) [6]
Paroxetine	N/A	59.2% (23.1) [5]	N/A	21.3% (23.6) [2]	N/A	8.0% (11.9) [5]
Sertraline	55.9% (11.7) [1]	51.7% (4.1) [1]	14.2% (16.3) [1]	18.9% (5.6) [1]	-27.5% (15.0) [1]	-6.0% (5.7) [1]
Total	54.0% (10.1) [8]	47.4% (12.5) [39]	25.5% (21.7) [2]	20.1% (7.8) [23]	0.5% (10.4) [8]	1.7% (6.8) [26]
Monoamine Oxidase Inhibitors (MAOIs)
Isocarboxazid	56.7% (10.5) [4]	60.1% (7.1) [5]	15.3% (12.6) [4]	41.3% (18.0) [3]	-14.1% (27.5) [2]	1.9% (10.0) [2]
Phenelzine	49.5% (14.0) [7]	57.8% (7.4) [13]	22.3% (30.7) [5]	29.5% (14.6) [7]	-24.1% (16.6) [6]	7.2% (9.0) [11]
Tranylcypromine	58.6% (10.8) [3]	52.8% (13.4) [3]	N/A	21.1% (25.4) [3]	33.3% (23.5) [3]	10.1% (23.6) [2]
Total	52.7% (9.7) [14]	57.4% (5.5) [21]	18.4% (22.6) [9]	30.9% (17.1) [13]	-3.1% (26.1) [11]	6.2% (7.4) [15]
Selective Serotonin Reuptake Inhibitors (SSRIs)
Fluoxetine	56.7% (12.4) [1]	48.7% (5.6) [1]	N/A	N/A	46.0% (14.7) [1]	0.6% (7.9) [1]
Fluvoxamine	N/A	57.4% (8.4) [1]	N/A	34.0% (13.5) [1]	N/A	19.0% (12.1) [1]
Paroxetine	N/A	64.6% (7.4) [1]	N/A	N/A	N/A	-6.6% (10.2) [1]
Sertraline	N/A	52.2% (3.9) [1]	N/A	N/A	N/A	2.2% (6.8) [1]
Total	56.7% (12.4) [1]	54.2% (6.7) [4]	N/A	34.0% (13.5) [1]	46.0% (14.7) [1]	2.4% (8.4) [4]
Monoamine Oxidase Inhibitors (MAOIs)
Isocarboxazid	N/A	N/A	N/A	N/A	N/A	N/A
Phenelzine	N/A	57.7% (9.5) [1]	N/A	50.4% (10.3) [1]	N/A	0.9% (13.1) [1]
Tranylcypromine	N/A	N/A	N/A	N/A	N/A	N/A
Total	N/A	57.7% (9.5) [1]	N/A	50.4% (10.3) [1]	N/A	0.9% (13.1) [1]
Anxiolytics
Alprazolam	-5.0% (13.1) [1]	56.6% (9.0) [10]	N/A	27.2% (6.1) [6]	N/A	2.6% (6.2) [12]
Buspirone	N/A	59.7% (5.2) [2]	N/A	21.4% (14.1) [2]	N/A	N/A
Chlordiazepoxide	N/A	46.6% (8.5) [2]	N/A	11.7% (28.2) [2]	N/A	-1.6% (16.3) [2]
Diazepam	N/A	42.4% (9.8) [2]	N/A	4.3% (8.3) [1]	N/A	-9.3% (14.4) [4]
Total	-5.0% (13.1) [1]	54.3% (8.4) [16]	N/A	20.8% (4.0) [11]	N/A	-0.4% (7.1) [18]
Anxiolytics
Alprazolam	N/A	N/A	N/A	N/A	N/A	N/A
Buspirone	N/A	N/A	N/A	N/A	N/A	N/A
Chlordiazepoxide	N/A	N/A	N/A	N/A	N/A	N/A
Diazepam	N/A	15.3% (5.1) [1]	N/A	N/A	N/A	-17.6% (9.7) [2]
Total	N/A	15.3% (5.1) [1]	N/A	N/A	N/A	-17.6% (9.7) [2]

Note: The percentage shown in the Drug Efficacy column is the anticipated percentage of patients provided the treatment shown who will respond. The Drug-Placebo column shows the expected percentage difference in drug versus placebo in patients, based on direct drug- placebo comparisons in trials that included at least these two cells. The Drug-Drug column shows the percentage difference between the named compound and all other listed antidepressant medications studied in randomized controlled trials that included at least these two cells. A negative (-) sign before the percentage means the named compound fared less well than did the "other" to which it was compared. The numbers in parentheses are the standard deviations of the estimated percentage responders. The bracketed numbers are the number of studies on which these estimates are calculated. However, the Drug Efficacy calculations include all cells in all studies for which meta-analysis was feasible (i.e., if a study had two cells of the drug at two different dosages, efficacy trials included both cells). For drug-placebo comparisons, only those trials that contained both of these cells and that reported outcome in a manner that allowed for an estimate of the percentage of randomized patients who responded were included.

N/A means no information that allowed for this type of meta-analysis was available, Inpt = Inpatient, Outpt = Outpatient.

• The oldest tricyclic medications are the most extensively studied, largely as a result of industry-sponsored studies that use older drugs as standards. In general, medications with the largest number of randomized controlled trials have efficacy equal to those with fewer studies.

• The efficacy of medications studied to date in geriatric patients is similar to that seen in younger adults (nearly all geriatric patient studies have been conducted on depressed, but otherwise medically fit, patients with major depressive disorder).

• The MAOIs have a larger number of studies reported, in part because of the longer period of research reviewed compared to that of other compounds.

Over 30 randomized controlled trials of antianxiety agents used for patients with depression were found to fulfill the panel's inclusion/ exclusion criteria. Of these, 22 investigated alprazolam; 11, diazepam; and 4, chlordiazepoxide. Two additional studies of the non-benzodiazepine anxiolytic, buspirone, were also found. Diazepam, alprazolam, and chlordiazepoxide are not FDA-approved as antidepressant medications, although they are occasionally used in patients whose general medical conditions represent a special risk for the use of FDA-approved antidepressant medications.

A few studies that were not included in the evidence table merit comment. Several studies performed about 20 years ago using the combination of chlordiazepoxide and amitriptyline (Limbitrol, an FDA- approved antidepressant) suggested a faster onset of action, but no greater overall efficacy than amitriptyline alone (Feighner, Brauzer, Gelenberg, et al., 1979; Rickels, Gordon, Jenkins, et al., 1970). These studies were suggestive, but generally did not use sufficiently rigorous methodology or sufficiently high medication dosages to be clearly interpretable.

Two recent studies, however, have addressed the issue of combining an anxiolytic and an antidepressant (Feet, Larsen, and Robak, 1985; Kravitz, Fogg, Fawcett, et al., 1990.) A double-blind comparison of imipramine and the combination of imipramine plus 10 mg of diazepam per day in outpatients with nonagitated depression did not demonstrate any benefit from the addition of diazepam to imipramine (Feet, Larsen, and Robak, 1985). One study found that the combination of desipramine and alprazolam was associated with a more rapid response than was desipramine alone (Kravitz, Fogg, Fawcett, et al., 1990). In fact, this study, as well as several studies of alprazolam alone in the treatment of depressed outpatients, suggests that most of the clinical benefit derived in depressed patients who respond at all to alprazolam occurs during the initial 7 to 14 days of treatment (Overall, Biggs, Jacobs, et al. 1987; Rush, Erman, Schlesser, et al., 1985). Whether this finding applies to other benzodiazepines is unstudied.

In summary, these data suggest that benzodiazepines, with the exception of the triazolo-benzodiazepine, alprazolam, should generally not be used to treat major depressive disorder, even in combination with other antidepressants. Moreover, alprazolam is not recommended for routine clinical use because no continuation or maintenance phase trials have been published and because problems associated with discontinuing this compound in remitted, depressed patients have not been fully studied. However, the presence of certain concurrent general medical conditions for which standard antidepressants are contraindicated may necessitate consideration of this agent in selected patients because of its cardiovascular safety, quick onset of action, and generally low side-effect profile.

Table 6. Meta-analyses of primary care antidepressant (more...)

Table 6. Meta-analyses of primary care antidepressant medication trials

Drug	Drug Efficacy	Drug-Placebo	Drug-Drug
Tricyclics
Mindham (1977) (amitriptyline) (versus maprotiline)7)	66.2%	N/A (11.0)	13.2%
Rickels and Case (1982) (amitriptyline) (versus trazodone)(6.0)	52.9% (7.9)	25.7% (8.4)	2.8%
Rickels, Chung, Csanalosi, et al. (1987) (imipramine) (versus alprazolam)	45.2% (6.8)	3.7% (9.9)	(9.7) -13.3%
Winsauer and O'Hair (1984) (doxepin) (versus amoxapine)	62.5%	N/A	-13.3%
Total	55.5% (8.2) [4]	15.0% (19.0) [2]	1.5% (9.3) [4]
Heterocyclics
Mindham (1977) (maprotiline) (versus amitriptyline)	52.7% (8.1)	N/A	-13.5% (11.2)
Moon and Davey (1988) (trazodone) (versus mianserin)	90.5% (6.3)	N/A	-0.4% (8.6)
Richards, Midha, and Miller (1982) (trazodone) (versus mianserin, diazepam)	60.0% (7.2)	N/A	24.6% (11.7)*
Rickels and Case (1982) (trazodone) (versus amitriptyline)	50.0% (6.0)	22.9% (8.0)	-2.8% (8.4)
Winsauer and O'Hair (1984) (amoxapine) (versus doxepin)	76.6% (7.4)	N/A	13.7% (11.3)
Total	62.8% (13.9) [5]	22.9% (8.0) [1]	7.0% (11.8) [6]
Selective Serotonin Reuptake Inhibitors Corne and Hall (1989) (fluoxetine) (versus dothiepin)	51.0% (7.0) [1]	N/A	-21.1% (9.3) [1]
Grand Total	57.8% (8.9) [10]	18.2% (12.1) [3]	2.2% (8.5) [11]

* Two cells.

Note: The numbers in parentheses are the standard deviations of the estimated percentage of responders. The bracketed numbers are the numbers of studies on which these estimates are calculated.

Antidepressant Medication Selection

Side effects occur in a selected number of patients taking any medication and are typically dependent on dosage and blood level. Many side effects are more likely to occur at the initiation of treatment or within a short time following dosage increases, and patients often adapt to side effects over time. The reader is urged to review current publications for data on the actual incidence of various specific side effects of each medication (AMA, 1990; Physician's Desk Reference, 1992; USP, 1992).

Table 7. Side-effect profiles of antidepressant medications (more...)

Table 7. Side-effect profiles of antidepressant medications

Drug	Side Effect[1]
	Central Nervous System			Cardiovascular		Other
	Anticholinergic[2]	Drowsiness	Insomnia/Agitation	Orthostatic Hypotension	Cardiac Arrhythmia	Gastrointestinal Distress	Weight Gain (Over 6 kg)
Amitriptyline	4+	4+	0	4+	3+	0	4+
Desipramine	1+	1+	1+	2+	2+	0	1+
Doxepin	3+	4+	0	2+	2+	0	3+
Imipramine	3+	3+	1+	4+	3+	1+	3+
Nortriptyline	1+	1+	0	2+	2+	0	1+
Protriptyline	2+	1+	1+	2+	2+	0	0
Trimipramine	1+	4+	0	2+	2+	0	3+
Amoxapine	2+	2+	2+	2+	3+	0	1+
Maprotiline	2+	4+	0	0	1+	0	2+
Trazodone	0	4+	0	1+	1+	1+	1+
Bupropion	0	0	2+	0	1+	1+	0
Fluoxetine	0	0	2+	0	0	3+	0
Paroxetine	0	0	2+	0	0	3+	0
Sertraline	0	0	2+	0	0	3+	0
Monoamine oxidase inhibitors (MAOIs)	1	1+	2+	2+	0	1+	2+

[1] 0 = absent or rare, 1+ , 2+ = in between, 3+, 4+ = relatively common.

[2] Dry mouth, blurred vision, urinary hesitancy, constipation.

If a patient suffers concurrently from both major depressive and obsessive-compulsive disorder, for example, the practitioner may be well advised to choose a medication, such as clomipramine or fluoxetine, that has demonstrated efficacy for both conditions.

Specific symptom clusters may suggest the appropriateness of a particular drug class for treatment. For example, patients whose conditions have atypical features appear to fare better on MAOIs (Liebowitz, Quitkin, Stewart, et al., 1988; Quitkin, Harrison, Stewart, et al., 1991; Quitkin, McGrath, Stewart, et al., 1990; Quitkin, Stewart, McGrath, et al., 1988; Thase, Carpenter, Kupfer, et al., 1991) or SSRIs (Reimherr, Wood, Byerley, et al., 1984) than on standard TCAs. Similarly, given the available data, the presence of psychotic features points more strongly toward TCAs combined with neuroleptics (three open trials or retrospective analyses [Charney and Nelson, 1981; Frances, Brown, Kocsis, et al., 1981; Minter and Mandel, 1979b] and five prospective randomized controlled trials [Kaskey, Nasr, and Meltzer, 1980; Minter and Mandel, 1979a, b; Moradi, Muniz, and Belar, 1979; Nelson and Bowers, 1978; Spiker, Weiss, Dealy, et al., 1985]), ECT (Avery and Lubrano, 1979; Avery and Winokur, 1977; Brown, Frances, Kocsis, et al., 1982; Charney and Nelson, 1981; Davidson, McLeod, Kurland, et al., 1977; DeCarolis, Gilberti, Roccatagliata, et al., 1964; Frances, Brown, Kocsis, et al., 1981; Glassman, Kantor, and Shostak, 1975; Lykouras, Malliaras, Christodoulou, et al., 1986a, b; Minter and Mandel, 1979a, b; Moradi, Muniz, and Belar, 1979), or possibly, based on one study, amoxapine (Anton and Burch, 1990; Anton and Sexauer, 1983).

If the patient is considered likely to take an overdose, the practitioner is advised to dispense only 1 week's supply of potentially lethal antidepressants (e.g., the tricyclics or MAOIs), to see the patient at least weekly, and to be readily available by telephone. In such situations, certain heterocyclic agents (bupropion or trazodone) or SSRIs, which appear safer in cases of potential overdose, may be preferred.

The likelihood that it will be necessary to monitor therapeutic blood levels of the drug is also a consideration in the selection of an antidepressant. In a patient who is older, who has other medical conditions, or who is taking medications that affect the metabolism of antidepressants, determining the blood level of the medication to gauge the minimal therapeutic dosage for that patient may be particularly helpful. Similarly, patients with complex general medical conditions that put them at particular risk for side effects from standard antidepressant medications may need blood level monitoring to ensure that the lowest possible dosage is used. Thus, an antidepressant with better established therapeutic and toxic levels, such as nortriptyline, may be preferred over another for which such levels are less well studied.

There is evidence, although not randomized controlled trials, that response to different classes of medication may run in families. (For a review, see Stern, Rush, and Mendels, 1980.) Thus, if the patient has a first-degree relative who responded well to a compound, a drug from the same class may be preferred. Whether this suggestion applies to the newer compounds, such as the SSRIs, has not been studied.

A history of failure to respond to a truly adequate trial of a drug in one class, such as the TCAs, strongly suggests that it would be appropriate to try a medication from a different class rather than another drug from the same class. The evidence for this recommendation is strongest for switching patients who have failed to respond to a TCA or an MAOI. Evidence is suggestive for switches among other classes of antidepressant medications (Depression Guideline Panel, forthcoming).

If the medication is ineffective or not tolerated, then the dosage can be adjusted or the medication changed. In milder cases, medication may be stopped and a trial of time-limited psychotherapy substituted. For patients with partial responses to medication, the dosage is adjusted or the blood level monitored to ensure that it is adequate. In some patients, adjunctive psychotherapy may be indicated, depending on the medication type. In others, a switch to another medication is logical. In yet others, adjunctive medication (augmentation) with agents such as lithium may be preferable.

Table 8. Pharmacology of antidepressant medications (more...)

Table 8. Pharmacology of antidepressant medications

Drug	Therapeutic Dosage Range (mg/day)	Average (Range) of Elimination Half-Lives (Hours)[1]	Potentially Fatal Drug Interactions
Tricyclics
Amitriptyline (Elavil, Endep)	75-300	24 (16-46)	Antiarrhythmics, MAOIs
Clomipramine (Anafranil)	75-300	24 (20-40)	Antiarrhythmics, MAOIs
Desipramine (Norpramin, Pertofrane)	75-300	18 (12-50)	Antiarrhythmics, MAOIs
Doxepin (Adapin, Sinequan)	75-300	17 (10-47)	Antiarrhythmics, MAOIs
Imipramine (Janimine, Tofranil)	75-300	22 (12-34)	Antiarrhythmics, MAOIs
Nortriptyline (Aventyl, Pamelor)	40-200	26 (18-88)	Antiarrhythmics, MAOIs
Protriptyline (Vivactil)	20- 60	76 (54-124)	Antiarrhythmics, MAOIs
Trimipramine (Surmontil)	75-300	12 (8-30)	Antiarrhythmics, MAOIs
Heterocyclics
Amoxapine (Asendin)	100-600	10 (8-14)	MAOIs
Bupropion (Wellbutrin)	225-450	14 (8-24)	MAOIs (possibly)
Maprotiline (Ludiomil)	100-225	43 (27-58)	MAOIs
Trazodone (Desyrel)	150-600	8 (4-14)	----
Selective Serotonin Reuptake Inhibitors (SSRIs)
Fluoxetine (Prozac)	10- 40	168 (72-360)[2]	MAOIs
Paroxetine (Paxil)	20- 50	24 (3-65)	MAOIs[3]
Sertraline (Zoloft)	50-150	24 (10-30)	MAOIs[3]
Monoamine Oxidase Inhibitors (MAOIs)[4]
Isocarboxazid (Marplan)	30- 50	Unknown	For all 3 MAOIs: Vasoconstrictors,[5] decongestants,[5] meperidine, and possibly other narcotics
Phenelzine (Nardil)	45- 90	2 (1.5-4.0)
Tranylcypromine (Parnate)	20- 60	2 (1.5-3.0)

[1] Half-lives are affected by age, sex, race, concurrent medications, and length of drug exposure.

[2] Includes both fluoxetine and norfluoxetine.

[3] By extrapolation from fluoxetine data.

[4] MAO inhibition lasts longer (7 days) than drug halflife.

[5] Including pseudoephedrine, phenylephrine, phenylpropanolamine, epinephrine, norepinephrine, and others.

• Well Established Therapeutic Ranges Nortriptyline - 50-150 ng/mL.Amitriptyline - 80-250 ng/mL of amitriptyline plus nortriptyline.Desipramine - 125-300 ng/mL.Imipramine - 150-250 ng/mL of imipramine plus desipramine.

• Less Well Established Therapeutic Ranges Clomipramine - Up to 700 ng/mL of clomipramine plus desmethylclomipramine.Doxepin - 150-250 ng/mL of doxepin plus desmethyldoxepin.Fluoxetine - 200-700 ng/mL of fluoxetine plus norfluoxetine.Trimipramine - 150-250 ng/mL.Amoxapine - 200-600 ng/mL of amoxapine plus 8-hydroxyamoxapine.Maprotiline - 200-600 ng/mL.

Therapeutic ranges have generally not been established by randomized controlled trials with patients on predesigned fixed medication levels in which group efficacy was measured. Rather, the suggested ranges are derived from laboratory/clinical interaction such that most patients who responded did so with dosages that produced the therapeutic levels indicated. Furthermore, most of these levels were established on inpatients. Their applicability to primary care outpatients is largely unstudied.

Nortriptyline has the best defined therapeutic window (upper and lower limits). The lower limits for the remaining medications are reasonably well established; upper limits are far less well established. The routine use of therapeutic blood level determinations generally is not needed, although their selected use in particular clinical instances can be of value.

Table 9. Selecting among antidepressant medications (more...)

Table 9. Selecting among antidepressant medications for depressed outpatients

First- and second-line choices: Secondary amine tricyclics (e.g., nortriptyline, desipramine).[1] Bupropion. Fluoxetine. Paroxetine. Sertraline. Trazodone. Alternative agents for patients with special presentations or needs: Tertiary amine tricyclics (e.g., amitriptyline, imipramine): Special considerations: Absence of serious medical illnesses, including cardiac disease, that preclude use. Need for rapid sedation. Monoamine Oxidase Inhibitors (MAOIs): Special considerations: Nonresponse or intolerance to at least one tricyclic and one heterocyclic. Family or personal history of MAOI response. Atypical symptom features. Selected Anxiolytic Medications:[2] Special considerations: Medical contraindications to FDA-approved antidepressant medications. No adverse cardiovascular effects. Low side-effect profile. Substantial withdrawal with long-term use. Limited exposure time expected (<3 months). Patient has no history of substance abuse. Quick action needed.

[1] Other first- and second-line choices are recommended for patients with arrhythmias, cardiac conduction defects, ischemic heart disease, cardiomyopathy, or cardiac valve disease.

[2] Evidence is clearest for alprazolam. Not recommended in severe depressions as studies reveal reduced efficacy. Not recommended for prolonged care as no studies longer than 12 weeks are available. Not recommended when FDA-approved antidepressant medications can be used safely. For buspirone, efficacy is suggested in those with primary anxiety disorders and mild associated depressive symptoms.

Note: Evidence for efficacy with severely depressed inpatients is more abundant for the standard tricyclics than for newer agents.

Although their efficacy is equal to that of other standard antidepressant medications, the MAOIs are usually not first-line treatments because of the required dietary restrictions and potentially fatal interactions with other medications. However, patients who have major depressive disorder with atypical features may optimally benefit from such agents. As for second-line treatments, there is substantial evidence that patients who fail to respond fully or partially to tricyclic medications may benefit substantially from MAOIs and modest evidence that these patients may benefit from the newer SSRIs.

Frequency of Visits

Since providing patient support and education optimizes adherence, facilitates dosage adjustments, minimizes side effects, and allows monitoring of clinical response, careful followup is essential. The panel recommends that patients with more severe depressions be seen weekly for the first 6 to 8 weeks of acute treatment. Once the depression has resolved, visits every 4 to 12 weeks are reasonable. Patients with less severe illness may be seen every 10 to 14 days for the first 6 to 8 weeks or more frequently if required to ensure adherence.

Medication Dosage Adjustments

Several suggestions can be made concerning dosage adjustments for frequently prescribed medications in primary care. For TCAs, patients generally begin with a low dosage (e.g., 25 to 50 mg/day of desipramine) administered at bedtime. The dosage may be increased in increments to the full therapeutic dosage over 1 to 3 weeks. There is no pharmacologic rationale for prescribing most TCAs in divided doses since most have half- lives approximating 24 hours. Once-daily dosing, usually at bedtime, often minimizes side effects. The full therapeutic dosage is maintained until the patient has either responded or clearly failed to respond. The major consideration in dosage increases is the emergence of side effects. Because side effects decrease with time, the patient should be told that they may occur, but are likely to decrease within several weeks of beginning medication.

Fluoxetine may be initiated at 20 mg/day, usually given in the morning, and should be maintained at a full therapeutic dosage (20 mg/day) for at least 4 to 8 weeks before the dosage is increased. Some patients may require less than 20 mg/day, while others may require more. Some patients feel restless or hyperalert with fluoxetine or develop initial insomnia. If significant side effects occur within 7 days, the dosage should be lowered or the medication changed, as the drug at a fixed oral dosage will continue to increase in plasma concentration for 4 weeks (until steady state is reached). Bupropion must be given in divided doses up to 450 mg/day. Dosages above 450 mg/day are associated with an increased risk of seizure.

Figure 6. Steps in medication management of major depressive (more...)

   Figure 6. Steps in medication management of major depressive disorder

Figure 6

Antidepressant Drug Blood Levels

Determinations of antidepressant drug blood levels can establish that patients are receiving a therapeutic dosage, help in the evaluation of adherence, or exclude toxicity. Logically, patients cannot be declared to be nonresponsive to a treatment unless the steady-state serum level is within the therapeutic range for at least 2 to 4 weeks. While the presence or absence of side effects can be helpful indicators of trial adequacy, serum levels are more accurate; patients differ widely in their individual side- effect sensitivity and in the blood levels achieved with a fixed oral dosage. Blood level determinations also enable clinicians to adjust the oral medication dosage for differences in metabolism due to age, race, drug interactions, concurrent medical conditions, or erratic adherence (Amsterdam, Brunswick, and Mendels, 1980; Bourin, Kergueris, and Lapierre, 1989; Glassman, Schildkraut, Orsulak, et al., 1985; Preskorn, Dorey, and Jerkovich, 1988; Preskorn and Kent, 1984; Voris, Morin, and Kiel, 1983).

Four antidepressant medications - nortriptyline, desipramine, imipramine, and amitriptyline - have more consistent evidence of a minimal therapeutic blood levels. Most have established toxic ranges, and a few have established upper therapeutic ranges. The best established therapeutic blood levels are for nortriptyline

Toxic blood levels can occur in patients who initially exhibit no clinical signs of tricyclic toxicity (Tamayo, Fernandez de Gatta, Gutierrez, et al., 1988). Similarities between depressive and toxic symptoms in patients with deteriorating clinical status often make it difficult for clinicians to assess a patient's condition correctly without quantitative blood level information (Appelbaum, Russell, Orsulak, et al., 1979; Preskorn and Simpson, 1982; Preskorn, Weller, Jerkovich, et al., 1988). In addition to high antidepressant medication dosages, risk factors for developing toxic levels include advanced age and serious concurrent, general medical illnesses. For these high-risk patients, determinations of medication blood levels provide a safe, effective method of attaining an optimal dosage while avoiding toxicity.

Patients for whom blood level monitoring may be particularly helpful include the following:

• Nonresponders or partial responders to therapeutic dosages of medication.

• Those who show symptom breakthrough following full response to medication.

• Those with symptoms of medication toxicity.

• Elderly or pregnant patients (because of altered metabolism).

• Those of African or Asian descent (because of slower metabolism).

• Those who are likely to take overdoses.

• Those who are potential nonadherers.

• Those with general medical illnesses that either alter metabolism or necessitate minimal, but therapeutic, levels of medication.

• Those taking medications that affect antidepressant drug metabolism.

• Those requiring quick assurance of an adequate medication trial.

• Those with particular medical risks for antidepressant side effects such that minimal effective levels are highly desirable.

• Those who develop new medical conditions or are prescribed additional medications that may change called-for blood levels.

Failure to Respond

Ongoing (but undisclosed) substance abuse or underlying general medical conditions causing the depression are two common diagnostic pitfalls. If either is found, treatment focuses on the relevant potential cause. In cases of substance abuse, a substance-free state for 6 to 8 weeks is usually sufficient to determine whether the depressive syndrome will remit without additional formal treatment for the depression. In some cases, however, the abusive and depressive disorders may require treatment simultaneously. Furthermore, some patients may have another psychiatric condition not initially disclosed to the practitioner, so reevaluation for both psychiatric and nonpsychiatric illnesses should be strongly considered for partial responders.

In addition to reassessment of the diagnosis, reevaluation of the adequacy of treatment is indicated. Medication underdosing is a common cause of nonresponse. Increasing dosages should be given in the first few weeks for most antidepressants. The major exception is fluoxetine, for which the starting dosage (20 mg/day) is often the full therapeutic dosage.

At 6 weeks (range, 4 to 8 weeks), if medically safe, a further dosage increase may be called for, especially in those without significant side effects. In selected clinical situations and with selected medications, a blood level determination may reveal suboptimal antidepressant levels in poor adherers, rapid metabolizers, or those concurrently taking medications that alter the metabolism of antidepressants (e.g., anticonvulsants). If the therapeutic range is well established, and the patient's blood levels are not within that range, adjust the dosage appropriately.

Some evidence from randomized controlled trials suggests a poorer response to TCAs alone in patients with the following conditions:

• Depression with atypical features (Davidson and Pelton, 1986; Liebowitz, Quitkin, Stewart, et al., 1984; Liebowitz, Quitkin, Stewart, et al., 1988; Quitkin, Stewart, McGrath, et al., 1988; Thase, Carpenter, Kupfer, et al., 1991).

• Depression with psychotic features (Chan, Janicak, Davis, et al., 1987; Charney and Nelson, 1981; Glassman, Kantor, and Shostak, 1975; Kroessler, 1985; Kupfer and Spiker, 1981; Nelson and Bowers, 1978; Scott, 1989; Spiker, Weiss, Dealy, et al., 1985).

• Depressions complicated by severe personality disorder (Akiskal, 1985; Frank and Kupfer, 1990; Soloff, George, Nathan, et al., 1987).

• Severe (melancholic) depressions in the elderly (Abou-Saleh and Coppen, 1983; Brown, Sweeney, Frances, et al., 1983).

While these conditions may be associated with a lower response rate to standard tricyclic medications in some studies, the response rates are still nearly always greater than those obtained with placebo. Typically, the initial strategy is to use a single standard antidepressant medication before moving to more complex regimens, even for these subgroups. The exception is for psychotic depression, where the combination of a neuroleptic and an antidepressant or ECT may be optimal initial treatments.

Figure 7. Six-week evaluation: responders to medication (more...)

   Figure 7. Six-week evaluation: responders to medication

Figure 8. Six-week evaluation: partial responders or (more...)

   Figure 8. Six-week evaluation: partial responders or nonresponders to medication

7

8

If the patient continues to have psychosocial problems following a good symptomatic response to medication, psychotherapy may be added. The rationale for delaying the use of formal psychotherapies to ameliorate associated psychosocial problems rests on recent evidence that, for outpatients with major depressive disorder who respond to medication, psychotherapy, or the combination during acute treatment, return to the premorbid level of psychosocial function often occurs, but is delayed by weeks to months following significant symptomatic improvement (Mintz, Mintz, Arruda, et al., 1992).

Figure 8

Objectives and Indications

Clinical experience and logic suggest that this recommendation may particularly apply if (a) the patient expresses a preference for psychotherapy as the initial treatment, (b) medication is contraindicated for the particular patient, or (c) the patient exhibits prominent psychosocial difficulties or evidence of a personality disorder.

Symptom removal and functional restoration are the goals of therapies targeted at symptom relief (e.g., cognitive or behavioral therapy) or current psychological problems (e.g., interpersonal or brief dynamic psychotherapy) that theoretically cause or maintain symptoms. It is hoped (though not established) that acute phase psychotherapy will prevent subsequent relapses and recurrences once treatment ends. (For recent reviews, see Hollon, DeRubeis, and Seligman, 1992; Persons, in press; Rehm, in press). Psychotherapies in combination with medication may also be used to augment symptom relief or to address collateral issues, such as adherence or secondary psychosocial problems (Rush, 1986).

Many commonly used forms of psychotherapy have not been subjected to randomized controlled trials in patients with major depressive disorder. That is, their efficacy has not been proven or disproven. Whether all psychotherapies are equally effective and whether they are differentially effective in different patient groups is unknown. The efficacy of psychotherapies studied to date is more similar than different, though only a few studies have included two psychotherapy cells to allow a direct comparison. In addition, most psychotherapies have many common features, although they may differ in specific procedures, tactics, or techniques. Whether these differences result in differential efficacy or whether they are of therapeutic value for some, but not other, patients is not known.

Randomized controlled trials of time-limited psychotherapies that have been codified in manuals reveal a significant benefit over wait-list controls for patients with mild to moderate forms of major depressive disorder (Depression Guideline Panel, forthcoming). However, there is minimal evidence that psychotherapy alone is effective in severe or psychotic forms of depression. There is strong evidence for the efficacy of medication (or ECT) for severe or psychotic depressions (Depression Guideline Panel, forthcoming).

A key problem in fully understanding the role of psychotherapy in major depressive disorder is identification of an adequate placebo control. Formal therapies nearly always exceed the effects of no treatment, minimal contact, or wait-list controls. Whether efficacy can uniquely, specifically, and with certainty be attributed to the particular therapy still remains an open question. That is, whether the beneficial outcome with therapy would have occurred with nonspecific interpersonal interactions of equal time and frequency has not been fully investigated.

On the other hand, studies directly comparing medication (as the "gold standard") to psychotherapy alone have generally found equal efficacy in acute phase treatment in mild to moderate depression (Beck, Hollon, Young, et al., 1985; Covi and Lipman, 1987; Hersen, Bellack, Himmelhoch, et al., 1984; Murphy, Simons, Wetzel, et al., 1984; Roth, Bielski, Jones, et al., 1982; Rush and Watkins, 1981). However, equivalent response rates do not imply that the same patient would have responded to these two different treatments. It is logical to consider that some would have responded to one, while others would preferentially respond to the alternative treatment. For this reason, a time-limited treatment trial is recommended for acute phase treatment with psychotherapy, just as it is for medication.

If some patients do respond to psychotherapy, while others respond to medication, then patients' recent treatment history will influence response rates. Yet, to date, only one study has attempted to evaluate this variable (Blackburn, Bishop, Glen, et al., 1981). This omission limits the utility of these studies in developing algorithms for which patients should receive which treatment. Furthermore, time-limited psychotherapies codified in manuals may not be particularly representative of actual clinical practice, which limits generalizability of the trials performed to date (Persons, 1991).

The psychotherapies that have been tested rarely differ from each other in efficacy. Failure to find that psychotherapies have different comparative response rates is not evidence that all are either ineffective or effective. In addition, most such studies have used small patient groups so that the ability to detect group differences is limited.

During the course of therapy, criteria-based symptoms should be reviewed at 6 and 12 weeks to gauge response. Self-report or interviewer-rated symptom scales are useful in following patient progress.

The possibility that psychotherapy prevents relapse/recurrence deserves special comment. There is suggestive evidence from some post-acute treatment followup studies that those who respond to therapy, particularly cognitive therapy, as acute treatment have a lower relapse rate over the ensuing 6 months to 2 years than do those who respond to medication (see Hollon, DeRubeis, and Seligman, 1992; Depression Guideline Panel, forthcoming). However, methodological limitations make interpretation of these findings quite tentative. First, all studies to date are naturalistic, meaning that the patients' treatment in the followup period is not controlled. Second, it is possible that those who respond to medication have a poorer prognosis, if untreated, than do those who respond to therapy. Thus, there may be a sample bias favoring a better prognosis for responders to psychotherapy. (For a recent review, see Hollon, DeRubeis, and Seligman, 1992.) Third, not all these followup studies reveal a better prognosis in those receiving acute phase psychotherapy versus medication. Given these methodological limitations, interpretive difficulties, and inconsistent results, the panel concludes that acute phase psychotherapy has not yet been shown to be of prophylactic value for recurrent major depressive disorder.

Evidence for Efficacy

Peer-reviewed publications that report patient diagnoses using a specified descriptive system and a specified form of psychotherapy constitute the evidence for psychotherapeutic efficacy. Adult and geriatric studies were pooled for this synopsis because the number of geriatric studies that met criteria was small and there is no evidence yet for differences in adult and geriatric efficacy rates.

Table 11. Number of randomized controlled psychotherapy (more...)

Table 11. Number of randomized controlled psychotherapy trials in patients with major depressive disorder

Therapy	Adult			Geriatric
	Acute	Cont	Maint	Acute	Cont	Maint
Cognitive	19 [10]**	1* [0]	0	3 [2]	0	0
Behavioral	11 [8]	0	0	2 [1]	0	0
Interpersonal	2 [1]	1*** [0]	1 [0]	0	0	0
Brief dynamic	5 [5]	0	0	3 [1]	0	0
Marital	1 [0]	0	0	0	0	0

* Not randomized (open trial of responders to acute phase).

** Number in brackets is the number of cells for which meta-analysis was possible. Studies published only as dissertations or in non- peer-reviewed sources were not included.

*** Trial included responders to acute phase interpersonal psychotherapy plus imipramine.

Note: Cont = Continuation. Maint = Maintenance.

Cognitive Therapy

For depression, cognitive therapy aims at symptom removal by identification and correction of the patient's distorted, negatively biased, moment-to-moment thinking and theoretically aims at prevention of relapse/recurrence by identifying and correcting silent assumptions (personal beliefs or schemas) (Beck, Rush, Shaw, et al., 1979; Dobson, 1989; Shaw, 1989; Wilson, 1989).

Twenty-two acute phase randomized controlled trials were conducted with cognitive therapy alone in adult or geriatric patients (Beck, Hollon, Young, et al., 1985; Blackburn, Bishop, Glen, et al., 1981; Covi and Lipman, 1987; de Jong, Treiber, and Henrich, 1986; Elkin, Shea, Watkins, et al., 1989; Gallagher and Thompson, 1982; Hogg and Deffenbacher, 1988; Murphy, Simons, Wetzel, et al., 1984; Neimeyer and Feixas, 1990; Neimeyer, Heath, and Strauss, 1985; Neimeyer and Weiss, 1990; O'Leary and Beach, 1990; Pecheur and Edwards, 1984; Ross and Scott, 1985; Rush, Beck, Kovacs, et al., 1977; Rush and Watkins, 1981; Scott and Stradling, 1990 [two trials]; Selmi, Klein, Greist, et al., 1990; Steuer, Mintz, Hammen, et al., 1984; Thompson, Gallagher, and Breckenridge, 1987; Turner and Wehl, 1984). Nearly all were conducted on outpatients. Certain pivotal studies (e.g., Blackburn, Bishop, Glen, et al., 1981) could not be included in the meta-analysis because the number randomized to each cell was not reported. In other cases, outcome was not reported in a categorical form. When outcome was reported categorically by a measure other than the BDI, the panel included the study in the analysis. Based on analysis by means of the confidence profile method (CPM), the overall efficacy of cognitive therapy alone was 46.6 percent. For adult outpatients, efficacy was 46.9 percent; for geriatric outpatients, it was 51.3 percent. In the only inpatient study available for CPM meta-analysis, the response rate was 58.3 percent (de Jong, Treiber, and Henrich, 1986).

Table 12. Meta-analyses of psychotherapy trials in (more...)

Table 12. Meta-analyses of psychotherapy trials in outpatients with major depressive disorder

Therapy[1,2]	Overall Efficacy	Therapy vs. Wait List	Therapy vs. Placebo	Therapy vs. Other Therapy	Therapy vs. Drug Alone
Behavioral therapy alone[3]	55.3% (9.3) [4] [10] [5]	17.1% (34.0) [5]	N/A	9.1% (19.9) [6]	23.9% (11.6) [2]
Brief dynamic psychotherapy alone	34.8% (17.8) [6]	N/A	N/A	-7.6% (14.6) [8]	8.4% (21.3) [2]
Cognitive psychotherapy alone	46.6% (6.9) [12]	30.1% (22.0) [2]	9.4% (8.3) [1]	-4.4% (16.9) [6]	15.3% (26.1) [3]
Interpersonal psychotherapy alone	52.3% (6.1) [1]	N/A	22.6% (8.4) [1]	13.3% (8.6) [1]	12.3% (8.6) [1]
Marital psychotherapy alone	N/A	N/A	N/A	N/A	N/A
Totals	50.0% (5.3) [29]	26.0% (23.5) [7]	15.7% (13.0) [2]	4.7% (8.5) [21]	14.0% (11.2) [8]

[1] Intent-to-treat sample.

[2] Adult and geriatric patient studies are combined.

[3] Behavioral therapy includes one cell with behavioral therapy plus placebo (Hersen, Bellack, Himmelhoch, et al., 1984).

[4] The numbers in parentheses are the standard deviations of the estimated percentage of responders.

[5] The bracketed numbers are the numbers of cells on which these estimates are calculated.

Table 12 shows a comparison of cognitive therapy alone and alternative acute phase treatments. Overall, cognitive therapy was similar in efficacy to all other psychotherapies taken together. Cognitive therapy exceeded pill placebo with clinical management by only 9.4 percent (Elkin, Shea, Watkins, et al., 1989). Compared to medication alone, cognitive therapy had a slight advantage.

These analyses combined group and individual cognitive therapy. Separating these two treatment formats, the overall efficacy of group cognitive therapy was 39.2 percent, and the overall efficacy of individual cognitive therapy was 50.1 percent. However, these response rates come from different trials. The one small study that directly compared these two formats found both to have equal efficacy (Rush and Watkins, 1981). Whether the group format is less effective than is individual cognitive therapy remains an open question that is particularly germane to the issue of cost containment. Of special note is a study that showed the efficacy of computer-assisted cognitive therapy to equal that of standard individual cognitive therapy (Selmi, Klein, Greist, et al., 1990). Further research on this potentially cost-effective approach is needed.

The prophylactic effects of cognitive therapy once acute phase treatment has been discontinued have not been established. They have not yet been fully evaluated because of methodological problems, such as brief followup periods, lack of controls, difficulties in defining relapse/ recurrence, inclusion of acute phase nonresponders in followup, naturalistic rather than controlled followup, and interpretive limitations. (For a recent review, see Hollon, DeRubeis, and Seligman, 1992.) Seven studies included a 1-year followup (Beck, Hollon, Young, et al., 1985; Gallagher and Thompson, 1982; Kovacs, Rush, Beck, et al., 1981; O'Leary and Beach, 1990; Ross and Scott, 1985; Scott and Stradling, 1990; Simons, Murphy, Levine, et al., 1986), one included a 2-year followup (Gallagher-Thompson, Hanley-Peterson, and Thompson, 1990), and one included an 18-month followup (Shea, Elkin, Imber, et al., 1992). All followups were naturalistic. Three studies showed that acute phase cognitive therapy was followed by fewer depressive symptoms at followup than was wait-list (Gallagher-Thompson, Hanley-Peterson, and Thompson, 1990; O'Leary and Beach, 1990; Ross and Scott, 1985). When compared to patients treated with acute phase pharmacotherapy without a formal continuation or maintenance medication phase, patients treated with cognitive therapy showed fewer self-reported depressive symptoms (Kovacs, Rush, Beck, et al., 1981) or a lower relapse rate (Simons, Murphy, Levine, et al., 1986) at followup. In the 18-month followup of patients treated in the multisite NIMH collaborative study, all four treatments tested (cognitive-behavioral therapy, interpersonal psychotherapy, imipramine, and placebo plus clinical management) were associated with relatively high relapse rates (Shea, Elkin, Imber, et al., 1992). No one acute phase treatment was associated with a better prognosis in this naturalistic followup. Of all patients who entered the acute study (intent-to-treat sample), only 15 to 28 percent suffered no further major depressive disorder and needed no further treatment in the 18-month followup period.

Selection of a Psychotherapy

The exception to this guideline may be brief dynamic psychotherapy, which seems to have a lower overall response rate (34.8 percent, based on meta-analyses) in outpatients with major depressive disorder. However, different individuals may differentially benefit from one or another treatment (Persons, 1991), which mitigates strong inferences from the available data.

When formal psychotherapy is selected as the sole treatment for less severe episodes of major depression, the following general principles may be useful:

• The psychotherapy should generally be time-limited, focused on current problems, and aimed at symptom resolution rather than personality change as the initial target.

• Since it has not been established that all forms of psychotherapy are equally effective in major depressive disorder, if one is chosen as the sole treatment, it should have been studied in randomized controlled trials.

• The therapist should be experienced and trained in use of the therapy with patients who have major depressive disorder.

• Assessment of symptom response (as with medication trials) is very useful in planning the next steps, should the patient not respond fully.

• If the patient fails to show any improvement in depression by 6 weeks, or nearly full remission by roughly 12 weeks, a reevaluation and potential switch to, or addition of, medication should be considered.

Medication and formal, short-term therapies or the combination of both are effective treatments for major depression. Long-term therapies have not been studied in randomized controlled trials. Therefore, long-term therapies used alone are not recommended as acute phase treatments for major depressive disorder.

Some forms of short-term psychotherapy not specifically designed for treatment of depression (e.g., marital therapy) may contribute to symptom improvement if the psychosocial situation indicates that focused therapy would be helpful.

Only a few forms of therapy as solitary acute treatments for major depressive disorder have been studied. In these cases, the therapists were highly trained in the particular modality, followed a manual, and often received ongoing supervision. Therefore, the question of the effectiveness of these therapies in everyday practice has not yet been fully addressed.

Frequency of Visits

There are no randomized controlled trials comparing different psychotherapy session schedules. In nearly all of the published trials of psychotherapy, sessions occur once (most often) or sometimes twice a week (for cognitive therapy) (see Beck, Rush, Shaw, et al., 1979). In primary care settings, therapy is likely to be provided by a consultant/therapist who sees the patient one to two times a week. This schedule facilitates the assessment of the severity of depressive symptoms, seems sufficient to promote the therapeutic alliance between therapist and patient, and allows frequent enough visits to ensure adherence. Based on the available randomized controlled trials, at least a partial response is often present by 6 weeks, and remission occurs in most patients by 12 weeks.

Failure to Respond

No crossover trials to either another form of psychotherapy or medication have been conducted on patients who have not fully responded to psychotherapy alone. A switch to or addition of antidepressant medications, which are thought to work by different mechanisms of action and have been well demonstrated to be effective in major depressive disorder, should be considered. Medication should be started in those who do not respond at all. Formal psychotherapy may be continued or discontinued.

As with medications, a significant number of patients (20 to 50 percent) may drop out of psychotherapy shortly after its initiation for various reasons (Persons, Burns, and Perloff, 1988; Depression Guideline Panel, forthcoming). The primary care practitioner will find it helpful to provide the patient with the option of returning for consultation or alternative therapy if the patient remains depressed, even before the 6-week reassessment visit.

Objectives and Indications

It is recommended that medication be added to (or substituted for) acute phase psychotherapy if:

• There is no response at all to psychotherapy alone at approximately 6 weeks.

• There is only a partial response to an adequate 12-week trial of psychotherapy.

• The patient worsens with psychotherapy alone.

• The patient requests medication during or following psychotherapy and symptoms are appropriate and sufficient to warrant medication.

There are few data to guide clinicians in determining when to choose the combination of medication and a formal psychotherapy. While it is common psychiatric practice to use such a combination in the treatment of major depressive disorder, the need for this combination in primary care patients is not well established. Available research does not allow for strong recommendations in this area for several reasons:

• The efficacy of combined treatment has not been tested in primary care and has been infrequently tested in tertiary care settings.

• The evidence to date in tertiary care settings does not consistently support the notion that combined treatment is more effective than medication alone, at least in terms of symptom reduction, in many patients with major depressive disorder.

• There is evidence that effective treatment of major depression with medication and clinical management ameliorates associated psychosocial problems in many patients without formal psychotherapy to redress these problems (Mintz, Mintz, Arruda, et al., 1992).

On the other hand, some research supports the ideas that:

• Some patients either respond more completely or a larger number of patients respond when both psychotherapy and medication are used, especially in psychiatric, but not necessarily in primary care, settings (Blackburn, Bishop, Glen, et al., 1981).

• The combined approach addresses a wider domain of disabilities (symptoms plus psychosocial functioning) (Corney, 1987; Friedman, 1975; Weissman, Prusoff, DiMascio, et al., 1979).

• For some, exposure to formal therapy may improve the prognosis compared to that associated with medication plus clinical management, once treatment is stopped (Frank, Kupfer, Perel, et al., 1990). (See Chapter 5 of this guideline and Depression Guideline Panel, forthcoming.)

There are no well established predictors of which patients preferentially benefit from combined treatment. The following suggestions are based on logical inference from the few studies that have searched for predictors of who is best served by combined treatment, from the clinical experience of the panel, and from knowledge about the course and complications of major depressive disorder.

Psychotherapy as an adjunct to medication can be useful in addressing associated psychosocial problems, such as marital difficulties (Friedman, 1975). The structured psychotherapies also tend to be specific, targeted, and time-limited. Some practitioners believe and some evidence suggests that the therapies are more effective if patients are less severely symptomatic and, therefore, better able to participate in the therapy.

Psychosocial difficulties may well improve during the 1 to 3 months following symptomatic improvement in depressed outpatients (Mintz, Mintz, Arruda, et al., 1992). For example, should marital problems persist well past the reduction of the depressive symptoms, marital therapy may be more clearly indicated.

One problem is to decide when the combined treatment should begin. In primary care settings, given the limited time and need to educate the patient (and family, where appropriate) about the treatment and prognosis of major depression, the likely optimal procedure in the treatment of patients who are to receive medication and a formal psychotherapy is to (1) begin medication, (2) provide support and education, (3) optimize adherence, (4) adjust the dosage, and (5) gain symptom relief and resolution of psychosocial problems before starting formal therapy. Once medication has reduced symptoms, a reassessment for continuing psychosocial or chronic interpersonal problems may identify those patients who may benefit by adding psychotherapy to the medication. This approach has the added advantage of allowing the practitioner time to develop a close alliance with the patient and to explore briefly in the course of general clinical management ongoing difficulties that may be exacerbating the depression.

Table 13. Considerations for combined treatment

Table 13. Considerations for combined treatment

Consider combined treatment as an initial option more strongly if: History reveals a partial response to a full trial of either treatment alone. Current episode of major depression is longer than 2 years. Patient has a history of two or more episodes with poor interepisode recovery. Significant psychosocial difficulties are present that interfere with adherence and indications for medication are present. Patient requests it. Add medication to psychotherapy if: Patient shows poor response to psychotherapy alone after 6 weeks or only a partial response after 12 weeks; if no response at all to psychotherapy, it may be discontinued and clinical management provided. Add psychotherapy to medication (if medication has been used optimally) if: Patient shows partial response to medication and residual symptoms are largely psychological (e.g., low self-esteem). Patient shows partial or complete response to medication and psychosocial problems remain significant. Patient has difficulty with adherence.

Evidence for Efficacy

Table 14. Meta-analyses of combined treatment in outpatients (more...)

Table 14. Meta-analyses of combined treatment in outpatients with major depressive disorder

Combination*1	Combination Efficacy	Combination vs. Wait-List or Placebo	Combination vs. Therapy*2 Alone	Combination vs. Other Therapy Alone*3	Combination vs. Drug Alone
Behavioral therapy plus medication	34.6% (10.9)*4 [2]*5	N/A	-7.4% (22.3) [2]	-2.2% (11.5) [1]	6.2% (11.4) [1]
Brief dynamic psychotherapy plus medication	N/A	N/A	N/A	N/A	N/A
Cognitive therapy plus medication	53.7% (17.3) [5]	N/A	6.4% (15.3) [6]	35.4% (9.4) [1]	39.4% (32.9) [2]
Interpersonal therapy plus medication	N/A	N/A	N/A	N/A	N/A

*1 Intent-to-treat sample.

*2 Therapy contrast is that specified in the combination column (e.g., behavioral therapy).

*3 The combination indicated in the first column was compared to a different form of psychotherapy alone.

*4 Standard deviation is in parentheses.

*5 Number of cells in the meta-analysis is in brackets.

One difficulty in the literature available is the lack of placebo-controlled, primary care psychotherapy, medication, and combined treatment randomized controlled trials. It is clear from the many available studies that some patients with major depressive disorder respond and some even remit fully with placebo and clinical management. Patients with less severe, less chronic forms of illness seem most responsive to such nonspecific treatment. It is logical to suspect that such patients are more likely to be found in primary care settings. For this reason, a two- to three-visit evaluation may be useful to determine which patients require formal treatment (see Chapter 3). There are no studies in primary care settings of less severely ill patients with major depressive disorder given several sessions of clinical management followed, for those who do not remit, by either psychotherapy, medication, or placebo with clinical management. Such information would provide a firmer basis for recommending treatment options for these less symptomatic, less chronic, and less recurrent forms of major depression.

For these reasons, routine use of the combination of medication and a formal psychotherapy as an initial treatment is not recommended for all patients. Some patients who actually need psychotherapy may be lost to treatment if they find the medication intolerable. For others, medication may be especially useful and therapy unneeded. Thus, initial therapy for less severely ill patients may be either a trial of medication or a trial of time-limited psychotherapy, as long as the outcome (symptom remission) is monitored. Should either trial fail to produce remission, the alternative treatment or a consultation are options.

Selection of a Combined Treatment

Practically speaking, the type of psychotherapy used will be substantially dictated by the availability of trained and skilled practitioners. When psychotherapy is added to medication to assist patients with psychosocial problems that may be contributing to their symptoms, the selection of therapy should be based on the specific difficulties identified by the patient (e.g., marital or family therapy for such problems, vocational counseling for those with occupational skill deficits, cognitive or interpersonal psychotherapy for low self-esteem/interpersonal difficulties). Since short-term, targeted therapies in combination with medication have been studied to some degree, while longer term therapies have not, logic and parsimony suggest that psychotherapy be limited in focus and duration when used in combination with medication in acute treatment and that its effectiveness be monitored.

Frequency of Visits

Patients in combined treatment may require greater time for each session than do those receiving either medication or psychotherapy alone. The frequency of visits for medication and psychotherapy is dictated by the same forces as if either were being used alone, such as severity of illness, need for determinations of blood levels, and type of psychotherapy. As symptoms improve, the intervals between visits will also increase. When possible, appointments with the physician and therapist should be coordinated to minimize inconvenience, thereby maximizing adherence.

Dosage Adjustments

Full therapeutic dosages of medication should be prescribed for patients receiving combined treatment. Dosage adjustments follow suggestions in Chapter 4.

Failure to Respond

There is no evidence that changing the form of psychotherapy will alter the course of symptoms. However, changing or augmenting the medication is a strong consideration. For severely depressed patients who are unresponsive to combined treatment, a consultation or referral for alternative medication or ECT may be considered.

The Suicidal Patient

There are approximately seven suicide attempts for every completed suicide, and about 2 to 3 percent of those who attempt suicide go on to complete suicide. In the general population, the single most reliable risk factor for completed suicide is a psychiatric diagnosis. Other risk factors are being Caucasian, male, and elderly. Roughly 25 percent of all suicides occur in the elderly, though they constitute only 10 percent of the general population. Elderly men continue to have the highest suicide rates, even though the rates among male youths are increasing.

As in the general population, the primary risk factor for suicide in clinical populations is a psychiatric diagnosis. Male gender is less significant than in the general population, but being Caucasian remains an important risk factor. The middle years seem to have increased risk in patient populations, while higher rates are found in the elderly and youths in the general population. The period of highest risk is during hospitalization or immediately after discharge.

The suicide literature on both general and clinical populations consistently reveals the risk factors of depression, alcohol or drug abuse, and psychosis (psychotic depression, mania, or schizophrenia). The presence of general medical disorders is a risk factor in the general population, but is not always evaluated in clinical studies. Approximately 70 percent of suicide completers have one or more active or chronic medical illnesses at the time of death. Acquired immune deficiency syndrome (AIDS) has recently been reported to carry a high suicide risk as well. To summarize, the following risk factors have been empirically related to suicide:

• Hopelessness.

• General medical illnesses.

• Family history of substance abuse.

• Depression.

• Substance abuse.

• Male gender.

• Caucasian race.

• Psychotic symptoms.

• Living alone.

• Prior suicide attempts.

Many patients with suicidal ideation can be managed as outpatients if they are not psychotic, if they do not abuse substances to a significant extent, and if they themselves feel that they can control their suicidal ideations or impulses, based on direct discussions about their thoughts of suicide. A few general principles for managing the patient with suicidal ideation are suggested by the panel, based on current clinical practice, logic, panel consensus, and knowledge of risk factors.

The practitioner is advised to:

• Evaluate the patient's condition for relevant risk factors.

• Suspect suicidal ideation in most depressed patients and others with risk factors.

• Inquire directly about the frequency and content of suicidal ideation during both diagnostic and treatment visits.

• Ask the patient about factors that argue against and for attempting suicide.

• Evaluate the patient's access to means of suicide.

The practitioner is advised to consider hospitalization for suicidal patients if:

• Psychosis is present.

• Suicidal ideation is present with significant substance abuse, severe hopelessness, strong impulses to act on the ideas, or specific suicide plans.

• The patient lacks sufficient social support to execute effective outpatient treatment.

• Concurrent general medical conditions make outpatient medication treatment unsafe.

• The patient lacks the ability to participate in outpatient care.

For outpatient management, the practitioner is advised to:

• Explain the nature of depression, its prognosis, and treatment, thereby raising hope; explain that depression itself is often associated with a feeling of hopelessness and thoughts of suicide and that, as the depressive syndrome resolves, so should the hopelessness and suicidal ideation.

• Develop either a formal contract (patient to call practitioner if suicidal feelings become strong) or be very easily available to patients if suicidal ideation or impulses increase.

• Strongly advise patients to discontinue all alcohol and drug use immediately.

• See patients at least weekly while suicidal ideation is significant.

• Avoid prescriptions for more than 1 week of medication until suicidal risk is lower.

• Explain to family members, when appropriate, how to respond to suicidal ideas (e.g., provide reassurance, call the practitioner).

• Advise patients, when appropriate, to remove easily available instruments of suicide (e.g., guns) from the house.

• Provide supportive interactions with the patient in 15- to 20-minute visits in acute phase treatment until suicidal ideation resolves.

• Provide reassurance and support the patient's reasons to live.

• Reassess the patient's condition for hopelessness, suicidal ideation, and substance abuse at each treatment visit.

Geriatric Depression

The diagnosis of late-life depression must be given particular attention. The prevalence of major depressive disorder may be lower in the elderly than in nonelderly adults, which suggests that aging per se is not an etiological factor in depression. In elderly patients with recurrent depression, the possibility of bipolar disorder must be considered. Other psychiatric disorders to be considered in the differential diagnosis include early dementia, delusional disorders, organic (secondary) disorders, and substance-induced mood disorders. Finally, the first onset of primary major depressive disorder past the age of 50 is uncommon and often related to a specific medical etiology. A particularly careful medical evaluation is often needed in these patients.

The efficacy of the various treatments for depression in the elderly is, by and large, equal to that found in adults in general. The differences in dealing with the elderly are the particular practical problems that they and the treatments confront. These problems require more strategic planning and somewhat different tactics (e.g., more likely use of blood level determinations, particular care in minimizing side effects). Treatment of late-life major depressive disorder was reviewed by an NIMH Consensus Development Conference (1991), the general conclusions of which are entirely consistent with these guidelines. The consequences of unrecognized and untreated depression in the elderly include increased health services utilization, longer hospital stays, poor treatment compliance, and increased morbidity and mortality from medical illness and suicide. The costs of treatment are relatively modest and can be minimized by careful monitoring of the patient's clinical status.

In general, major depressive disorder in late life is a treatable illness. The evidence for the specific efficacy of medication is strongly based on randomized placebo-controlled trials. The evidence for the efficacy of psychotherapy alone as a treatment for less severely ill, nonpsychotic outpatients is beginning to accumulate, though this area remains understudied. Electroconvulsive therapy appears to be as effective in geriatric patients with severe or psychotic major depressive disorder as in nongeriatric groups. Evidence for or against the efficacy of combined acute phase treatment is generally lacking in geriatric patients, but studies are currently under way. One preliminary analysis (based on intent-to-treat samples) found that interpersonal psychotherapy and nortriptyline, within the context of a supportive, psychoeducationally oriented milieu, achieved a treatment success rate of around 75 percent in acute phase treatment (Reynolds, Frank, Perel, et al., 1992). Finally, the utility of maintenance phase medication is suggested by a few studies.

Nearly all randomized controlled treatment trials in elderly depressed individuals to date have been conducted in the otherwise medically healthy. These patients are not representative of all depressed elderly. In fact, other nonpsychiatric medical conditions are risk factors for depression. Thus, in recommending treatments, the panel is assuming that those treatments effective in the depressed, but otherwise healthy, elderly will be effective in those with other concurrent medical conditions. However, the presence of such medical conditions appears to be a risk factor for a poorer prognosis in younger depressed patients (Keitner, Ryan, Miller, et al., 1991).

The problems of treating depression in the elderly can be complex and challenging, particularly with the high rate of co-morbid medical disorders. Although treatment appears helpful for the depressed elderly, a pressing need remains for further controlled intervention research for depressions late in life, especially in patients with depressive disorders occurring in association with other nonpsychiatric medical and neurologic disorders.

Table 15. Confounds in the diagnosis and treatment (more...)

Table 15. Confounds in the diagnosis and treatment of depression in the elderly

Concurrent nonpsychotropic medications may: Cause depression. Change antidepressant blood levels. Increase antidepressant side effects. Biochemically block antidepressant effects. Call for modifying the oral dosage. Concurrent medical illnesses may: Cause depression biologically. Reduce the efficacy of antidepressant medication or psychotherapy. Change antidepressant drug metabolism. Impair ability to participate in psychotherapy. Create disability contributing to both chronicity and reduced treatment efficacy. Increase the need for simplified medication dosing schedules (e.g., once daily) . Concurrent nonmood psychiatric conditions may: Cause depression (e.g., early Alzheimer's). Call for different medications. Impair participation in psychotherapy . Reduce response to antidepressant medications (e.g., personality disorders) . Worsen prognosis of the depression (e.g., alcoholism). Other issues: Slower metabolism with age often requires lower dosages. Transportation difficulties may restrict access to care. Increased interview time needed. Fixed income may limit availability of therapy and nongeneric antidepressant medications due to cost.

• Poor adherence, due to greater sensitivity to side effects.

• Medically significant side effects.

• Inadequate family support to continue taking medication properly.

• Intercurrent medical disease that interferes with proper antidepressant dosing.

• Occult self-medication with drugs such as alcohol.

• The need for education of the patient and family about depression and its treatment.

• The risk of death from suicide.

To provide optimal pharmacotherapy in elderly patients taking other medications or who have other nonpsychiatric medical disorders, the practitioner may need to rely on repeated plasma antidepressant drug level determinations, as the other medications and/or disorders can alter antidepressant absorption or metabolism. Because of the natural metabolic slowing that accompanies aging, blood level monitoring may be advisable in otherwise healthy elderly patients, particularly where therapeutic ranges are better established.

Psychosocial difficulties that may interfere with optimal treatment response include:

• Negative intercurrent life events.

• Ongoing severe interpersonal conflicts about dependency and role transitions.

• Prolonged or unresolved grief.

• Social isolation.

General clinical management or a formal psychotherapy may be equally helpful in addressing these psychosocial obstacles. Social casework, particularly during acute treatment, to remove or modify "real life" obstacles to recovery may be helpful. However, the efficacy of social casework in those with major depressive disorder has not been formally tested. Regular contacts with family members during periods of individual crisis may also be useful.

The use of stimulant medications in the elderly was recently reviewed by an NIMH Consensus Development Conference (1991). While no randomized controlled trials were available, 8 to 12 open trials conducted over the last 3 decades, as well as clinical experience of psychopharmacologists who are expert in the treatment of the depressed elderly, suggest some efficacy and a low abuse potential for stimulant medications (NIMH, 1991). The panel does not recommend the routine use of stimulants in depressed elderly patients, given the availability of antidepressants with more predictable and established efficacy and side-effect profiles. However, stimulants may be an option for highly selected elderly patients who have not responded to adequate trials of other treatments. A psychopharmacologic consultation is mandated before selection of this treatment.

Seasonal Depression

Light therapy is a relatively new treatment that has been tested in randomized trials for up to 2 weeks in patients with seasonal mood disorders, nearly always those with winter depressions (Depression Guideline Panel, forthcoming). Its longer term efficacy has not been formally evaluated, though case reports and clinical experience suggest efficacy throughout the winter and in subsequent episodes. In addition, light therapy has not been tested against other potentially active treatments, such as medication, psychotherapy, or the combination, so whether seasonal depressions respond to these alternative treatments is unknown. Case reports suggest some efficacy for selected antidepressants. Since many patients with major depressive episodes occurring in a seasonal pattern have bipolar II or recurrent major depressive disorder, it is logical to believe that medication will be effective.

Given the current modest knowledge about light therapy and continuing research into the optimal method of administration, long-term safety, and other practical issues, the panel suggests the following principles:

• Light therapy is a logical consideration only for well documented seasonal, nonpsychotic, winter depressive episodes in patients with recurrent major depressive or bipolar II disorders or milder seasonal episodes.

• It should be administered by a professional with experience and training in its use who deems it suitable for the particular patient.

• It may be a second-line treatment option after the patient has failed to respond to an adequate medication trial.

• It may be a first-line treatment for these patients if they are not severely suicidal and if there are medical reasons to avoid antidepressants, if the patient has a history of a positive response to light therapy and no negative effects, if the patient requests it, or if an experienced practitioner deems it indicated.

Since the underlying mechanisms implicated in the seasonal pattern are not likely to be remedied by psychotherapy alone and there is no evidence to date for its efficacy, formal psychotherapy is not recommended as a first-line approach for truly seasonal depressions. Logically, light therapy should not be used as an adjunct to medication until either one alone has been optimally used. Light therapy can be useful to augment the response (if partial) to antidepressant medication and vice versa.

As with any treatment, the patient's response should be closely monitored. Response to light therapy can be rapid (4 to 7 days), but for some, response may be delayed to 2 weeks. However, the placebo response rate may be significant as well. Therefore, one or several "extended evaluation" visits may be useful in identifying those in whom symptoms persist. Caution is urged in the use of light therapy with patients with specific ophthalmologic or other conditions (Depression Guideline Panel, forthcoming). Since safety and efficacy have not been fully established beyond 2 weeks, consultation with a specialist may be helpful in determining specific risks and benefits for particular patients. Further information is available for interested practitioners (see Oren and Rosenthal, 1992; Terman, Williams, and Terman, 1991).

Depression and Other Nonpsychiatric Medical Disorders

Figure 9. Relationship between major depressive and (more...)

   Figure 9. Relationship between major depressive and other current general medical disorders

Figure 9

There are insufficient studies to recommend one medication over another solely on the basis of efficacy data. The side-effect and pharmacologic profiles of the antidepressant, patient age, prior response to specific antidepressants, family history of response to an antidepressant, and drug-drug interactions are among the many factors that must be weighed by a clinician when choosing a particular medication. The efficacy of psychotherapy is suggested by some open studies (see Watson, 1983, for a review) and by 20 randomized controlled trials for patients with cancer and depression (Depression Guideline Panel, forthcoming), but predictors of response are not well defined. Those with low emotional support and pessimism who are widowed or divorced may particularly benefit from psychotherapy (Weisman and Worden, 1976- 77). An empirical treatment trial with careful assessment of outcome (just as with major depressive disorder not associated with a general medical illness) is recommended. Further research in this area is sorely needed to determine which patients should be treated with which therapy and to establish with greater certainty the efficacy of various treatments.

Depression and Other Psychiatric Disorders

Figure 10. Relationship between major depressive and (more...)

   Figure 10. Relationship between major depressive and other current psychiatric disorders

Figure 10

The reasons to treat major depression as the primary target for those with concurrent generalized anxiety disorder (GAD) are:

• Randomized controlled trials of medication or psychotherapy have generally not excluded concurrent GAD.

• The presence of GAD has not been shown to reduce the efficacy of treatments for depression.

• Many patients with major depression have anxiety symptomatology of sufficient severity to be diagnosed as GAD, but the anxiety symptoms abate with remission of the depression.

• The Diagnostic and Statistical Manual of Mental Disorders (American Psychiatric Association, 1987) recommends that the diagnosis of GAD not be made if it occurs only in the presence of a mood disorder.

Available data on treating major depression when it is associated with a personality disorder suggest the following:

• Nearly all randomized controlled trials of treatment for major depressive disorder do not exclude personality diagnoses.

• It is difficult to assess personality disorders reliably in the presence of a major depressive episode.

• While a personality disorder may complicate treatment or slow response, it does not prevent a response in many patients.

• What may appear to be a "personality disorder" may actually be dysthymia concurrent with major depressive disorder, which, by its chronicity, affects interpersonal function. However, when the depression is specifically treated, the "personality disorder" abates.

The rationale for treating concomitant obsessive-compulsive disorder, bulimia, anorexia, or substance abuse as the primary treatment target includes the following:

• Randomized controlled trials indicate that, when these conditions are present with depressive symptoms or major depression and the conditions are effectively treated, the depression remits (Depression Guideline Panel, forthcoming).

• As a general principle, it is prudent when selecting a medication to consider initially those compounds for which efficacy is established or suggested for both conditions; for example, the compounds with established efficacy for obsessive-compulsive disorder also have established antidepressant efficacy.

• As always, careful followup and repeat assessment are essential to producing the optimal result.

When panic disorder and major depression co-occur, the rationale for determining which disorder is to be the initial target of treatment rests on the following:

• Panic disorder often presents with depressive symptoms or even major depression but, when the panic disorder is treated effectively, the depressive symptoms may remit.

• Many patients with more severe depression have panic attacks, but only during the depressive episode. For these patients, the major depression is the primary target.

• The clinician can further gauge which disorder may be primary by determining which disorder causes the greatest disability and reviewing the patient's family history. If the family history is positive for depression, the practitioner treats depression; if positive for panic, the practitioner treats panic.

• Some antidepressant medications (e.g., TCAs and MAOIs) appear effective in panic disorder, especially if combined with behavioral therapy, in which case they may be the preferred medication options. Case reports and case series of SSRIs in panic disorder suggest that lower dosages than those used for major depressive disorder may be preferred.

Objectives and Indications for Continuation Treatment

The recommendation for continuation medication treatment rests on a few randomized controlled trials with medication in continuation phase treatment, a substantial number of open continuation trials, and a more substantial number of such trials in maintenance phase studies. In general, it is preferred that acute phase response to medication be followed by continuation treatment with the same medication at the same dosage. Early discontinuation is followed by a roughly 25 percent relapse rate within 2 months (Maj, Veltro, Pirozzi, et al., 1992; NIMH, 1985).

Unless maintenance treatment is planned, antidepressant medication is discontinued at 4 to 9 months or tapered over several weeks (TCAs should be tapered). Patients are followed during the next several months to ensure that a new depressive episode (recurrence) does not occur. If a major depressive episode does return after discontinuation of medication, the patient is likely to respond to the same medication at the same dosage effective previously, which should then be continued for 4 to 9 months; in other words, the depression should be treated as a new episode. Thereafter, discontinuation may be attempted unless the patient has become a candidate for maintenance treatment.

No studies have evaluated acute response to medication followed by continuation phase psychotherapy or vice versa. In all cases, continuation phase treatment should be monitored for symptom breakthrough.

Continuation phase psychotherapy following acute treatment response to psychotherapy has not been studied in randomized controlled trials to date. Suggestive, albeit indirect, evidence for better, longer term outcome is available from two open continuation studies of cognitive therapy (Blackburn and Bishop, 1983; Jarrett, Ramanan, Eaves, et al., 1992) and one randomized controlled trial with interpersonal psychotherapy (Weissman, Prusoff, DiMascio, et al., 1979). Thus, there is only scant evidence to confirm or disconfirm the utility of continuation psychotherapy alone at once or twice a month for 6 to 8 months following acute treatment response.

Once the depressive symptoms have remitted with medication, therapy aimed at residual or persistent psychosocial problems is logical and reasonable. However, the efficacy of these types of intervention in this sequence has not been well studied. A notable exception is one controlled continuation/maintenance phase trial (32 weeks) of high versus lower contact with a focus on interpersonal function following response to acute phase medication alone (Weissman, Prusoff, DiMascio, et al., 1979). This study revealed that medication, but not placebo (with or without psychotherapy), prevented relapse/recurrence. This finding suggests that acute phase medication responders should be continued on medication and not switched to psychotherapy alone as a continuation/maintenance treatment. On the other hand, psychotherapy as opposed to no psychotherapy was associated with better social adjustment 1 year later.

Continuation medication treatment is also used following acute response to ECT and generally reduces the relapse/recurrence rate. By extrapolation from medication trials, patients should be followed up for relapse/recurrence of the disorder and should be maintained on antidepressant medication for at least 8 months. However, in some patients who have not responded to trials of many medications, but respond to ECT as acute treatment, continuation ECT is a treatment option; some of these ECT-responsive, medication-resistant patients will also not respond to continuation medication (Sackeim, Prudic, and Devanand, 1990; Sackeim, Prudic, Devanand, et al., 1990). There are no randomized trials of continuation ECT, but case reports suggest a potential role for this treatment. Psychotherapy alone following acute phase ECT has not been studied; given the apparent prophylactic efficacy of medication and continuation ECT in particularly refractory patients, continuation phase psychotherapy alone following acute phase ECT is not recommended.

Objectives and Indications for Maintenance Treatment

Patients who have had three episodes of major depression have a 90 percent chance of having another (NIMH, 1985). In fact, the World Health Organization (WHO) recommends considering maintenance treatment for patients who have experienced two depressive episodes within a 5-year period (Coppen, Mendelwicz, and Kielholz, 1986). Clinical issues that affect the decision for instituting maintenance medication include:

• Time between prior episodes.

• Severity of the episodes.

• Risk of suicide during the prior episodes.

• Acuteness of episode onset.

• Patient preference.

• Practitioner comfort in managing patients with recurrent depressions without maintenance medication.

Table 16. Considerations for maintenance medication (more...)

Table 16. Considerations for maintenance medication

Considerations	Strength of Indication
Three or more episodes of major depressive disorder	Very strongly recommended
Two episodes of major depressive disorder and
(a) Family history[1] of bipolar disorder	Strongly recommended
(b) History of recurrence within 1 year after previously effective medication was discontinued	Strongly recommended
(c) A family history of recurrent major depression[1]	Strongly recommended
(d) Early onset (before age 20) of the first episode	Strongly recommended
(e) Both episodes were severe, sudden, or life-threatening in the past 3 years	Strongly recommended

[1] A family history is a positive, clear-cut history in one or more first-degree relatives.

The panel's recommendations are based on the following considerations:

• Those with modest rather than full symptom control appear to have a poorer prognosis (Prien and Kupfer, 1986).

• Optimal symptom control requires optimal dosing.

• Lower imipramine maintenance dosages (roughly half the acute treatment dosage) are associated with poorer maintenance results (Prien, Balter, and Caffey, 1978) than is a full dosage (Frank, Kupfer, Perel, et al., 1990). However, the issue of maintenance dosing has not been fully evaluated for all antidepressants.

• Preferred maintenance medications are logically those with fewer long-term side effects, since adherence remains an issue.

Based on the evidence that patients with a family history of recurrent major depressive or bipolar I or II disorder are more likely themselves to have recurrent major depressive disorder (Winokur, 1991; Winokur and Kadrmas, 1989; Winokur and Wesner, 1987), it is logical to consider those with two episodes of major depressive disorder and such a family history as potential candidates for maintenance medication. Earlier age at onset of the initial episode of major depressive disorder is associated with a greater familial incidence of the disorder (Strober, 1984; Strober and Carlson, 1982; Winokur, 1979), as well as an earlier and more likely recurrence once treatment has ended (Giles, Jarrett, Biggs, et al., 1989; Giles, Jarrett, Roffwarg, et al., 1987).

Maintenance psychotherapy (at least in once-a-month sessions) does not appear to be effective in ultimately preventing a recurrence, though it may delay by months the onset of the next episode (Frank, Kupfer, Perel, et al., 1990). The efficacy of more frequent maintenance psychotherapy has not been evaluated. For selected patients with recurrent depression, such as those undergoing surgery; those who are pregnant or trying to become pregnant; and those in athletic competitions or occupational roles that preclude medication for a fixed, limited time period, maintenance psychotherapy may have an important role if the patient has had a thorough acute and continuation phase response to medication plus psychotherapy.

Continuation/Maintenance Phase Management with Medication

Continuation/Maintenance Phase Management with Psychotherapy

The objective of continuation psychotherapy is the same as that for medication to maintain the patient in an asymptomatic state. The efficacy of continuation psychotherapy alone as a treatment is not well established.

Continuation/Maintenance Phase Management with the Combination of Medication and Psychotherapy

Continuation phase combined treatment is not uncommon practice, especially for patients with the more chronic or complex forms of depression that are often seen in tertiary care settings. The rationale for combined continuation treatment is to maintain symptom remission and to improve psychosocial function. It is unclear which patients specifically benefit from this combination. The general principles outlined for continuation or maintenance management with either medication or psychotherapy hold for combined treatment.

A Second Opinion or Referral

Depression Guideline Panel Members: Biosketches

• A. John Rush, MD, Chair

• Betty Jo Hay Distinguished Chair in Mental Health

• Professor and Vice Chairman for Research,

• Department of Psychiatry

• University of Texas Southwestern Medical Center

• Dallas, Texas

Dr. Rush received his BA from Princeton University and his MD from the College of Physicians and Surgeons of Columbia University.

He is currently the Director of the Mental Health Clinical Research Center, an NIMH-funded center studying the biology, psychology, pharmacology, and psychotherapy of mood disorders. Dr. Rush has published extensively on both the psychology and biology of depression. He has received several NIMH grants to study depression and has helped develop and study the efficacy of cognitive therapy in treatment of depressed outpatients. His research has sought to identify biologic and psychological predictors of specific treatment responses, as well as relapse and recurrence. He serves as the Chair of the DSM-IV Work Group on Mood Disorders for the American Psychiatric Association.

Dr. Rush is a Fellow of the American College of Psychiatry, the American Psychiatric Association, and the American College of Neuropsychopharmacology. He is Past President of the Society for Psychotherapy Research.

• William E. Golden, MD

• Director, General Internal Medicine

• Associate Professor, Department of Medicine

• University of Arkansas for Medical Sciences

• Little Rock, Arkansas

Dr. Golden received his AB from Brown University (1975) and his MD from Baylor College of Medicine (1978). He completed his internal medicine residency and chief residency at Rush-Presbyterian St. Luke's Medical Center in Chicago (1983). Dr. Golden is Director of the Quality Assurance Research and Education Center, Director of the Division of General Internal Medicine, and Associate Professor of Medicine at the University of Arkansas for Medical Sciences. He is currently Chairman of the Quality Assurance Committee of University Hospital. He has had several funded projects in quality assurance and has expertise in perioperative care and medical informatics. He has authored more than 40 journal articles and book chapters on perioperative care, medical education, and quality improvement.

Dr. Golden is a trustee of the American Society of Internal Medicine and a member of the board of directors of the American Medical Review Research Center. He serves on the American Medical Association Council on Medical Education and has recently been appointed to the Liaison Committee on Medical Education. He is a member of many internal medicine societies and has served numerous roles in educational activities and policy-making matters for these organizations.

• Gladys Walton Hall, PhD, MSW

• Associate Professor, School of Social Work

• Howard University

• Washington, District of Columbia

Dr. Hall received her BS from Morgan State University (1966), MSW from the University of Connecticut (1971), and PhD from the University of Maryland (1982). She recently completed a post-doctoral fellowship at the National Institute of Mental Health (1990).

Dr. Hall teaches clinical social work methods courses and is a licensed clinical social worker with experience in the private and public sectors. Her private practice is in a comprehensive medical setting and includes the treatment of depressed women. She has authored several publications in the area of depression (childhood depression) and served on various professional boards, both national and local (including the D.C. Mental Health Association). Her research focus is on the psychosocial factors related to the co-morbidity of depression and conduct disorder among children.

• Col. Moses Herrera, MD

• Chief, Primary Care Clinic

• Robins Air Force Base Hospital

• Robins Air Force Base, Georgia

Dr. Herrera is a Fellow of the American Academy of Family Practice. He has many years of experience in the practice of family medicine, including obstetrics, geriatrics, pediatrics, and adolescent medicine, as well as in short-term psychotherapy and in the diagnosis and treatment of depression. He has served on the Mental Health Committee of the American Academy of Family Practice. With his experience, Dr. Herrera can speak with particular regard to the utility of guidelines as they may be applied to family practice in military medicine.

• Artie Houston

• Consumer Representative

• Fort Worth, Texas

Mrs. Houston represents patients (consumers). She has lived successfully with manic-depressive illness, which was diagnosed in 1968. She played a prominent role in the establishment of the National Depressive and Manic-Depressive Association (NDMDA) and served as its executive vice president and president in 1988-89. Mrs. Houston founded the first NDMDA chapter in Tarrant County, Texas, and served as its president for 2 years. She has a background in business and public relations. She served for 15 years as the business director of a large blood center in Fort Worth, Texas. Mrs. Houston has an extensive history of volunteer work in the areas of depression, manic-depressive illness, and hemophilia.

• Roger G. Kathol, MD

• Professor of Psychiatry and Internal Medicine

• University of Iowa Hospitals and Clinics

• Iowa City, Iowa

Dr. Kathol received his BA from the University of Kansas (1970) and his MD from the University of Kansas School of Medicine (1974). He completed his residency in internal medicine at the University of Iowa (1978) and his residency in psychiatry at the University of Iowa (1980). He completed a 1-year fellowship in endocrinology at the University of Otago in Wellington, New Zealand (1981).

Dr. Kathol is the director of the combined Internal Medicine/Psychiatry Unit and the General Hospital Psychiatry Services at the University of Iowa Hospitals and Clinics. He currently teaches both psychiatrists and internists about the diagnosis and treatment of medical and psychiatric disorders in patients with complex and combined medical/psychiatric difficulties. He has received NIMH and private foundation grants for clinical research on depressive disorders in the medically ill, on endocrine changes in patients with primary depression, and on pharmacokinetic drug interactions. His research has contributed to the understanding of depression in the medically ill, as well as the potential relationship of hypothalamic-pituitary-adrenal axis dysfunction in patients with primary affective disorder.

Dr. Kathol has lectured widely, is on the editorial board of several journals, holds membership on the boards of national organizations, and is published widely in the area of psychiatric pathology as seen in medically ill patients. He is board-certified in both internal medicine and psychiatry, and is a Fellow of the American College of Physicians, the American Psychiatric Association, and the American Academy of Psychosomatic Medicine. He is the current president of the Academy of Clinical Psychiatry and is a founding officer of the Association of Medicine and Psychiatry.

• Wayne Katon, MD

• Professor of Psychiatry

• Chief of Consultation-Liaison Psychiatry

• University of Washington Medical School

• Seattle, Washington

Dr. Katon received his BA from the University of Vermont (1971) and his MD from the University of Oregon (1976). He completed his residency in psychiatry at the University of Washington (1979).

Dr. Katon is Chief of the Division of Consultation-Liaison at the University of Washington and head of the Psychiatry Liaison Service to Family Medicine at University and Providence Medical Centers. Dr. Katon received the American Academy of Family Practice Award for Excellence in Teaching. He has taught medical students, psychiatric residents, and family medicine practitioners in the area of psychiatric disorders in family practice and primary care. He has received independent NIMH funding for the study of depression and chronic tinnitus, a randomized trial of psychiatric consultation for patients who are high medical service utilizers, and a randomized trial of psychiatric consultation in treatment of major depression in primary care.

Dr. Katon has published widely in the area of psychopathology (depression, panic disorders, and somatization) in the primary care setting. He has authored more than 100 journal articles and chapters and a book commissioned by NIMH entitled Panic Disorder in the Medical Setting. His research has sought to identify the psychological and social factors associated with medically unexplained somatic symptoms (chest pain, back pain, irritable bowel syndrome, tinnitus, pelvic pain, and dizziness). He is a recognized national and international authority on psychiatric disorders in family practice and general internal medicine.

• Catherine L. Matchett, MD

• Matchett Medical Center, President

• Grapevine, Texas

Dr. Matchett is a Fellow of the American Academy of Family Practice in private practice, who serves approximately 3,500 families in the North Texas area. She completed a rotating internship and 2 years in a psychiatry residency, passing the written boards in psychiatry before completing a family practice residency.

Dr. Matchett is Vice President of the Dallas Chapter of the American Academy of Family Practice. She has a strong interest in preventive medicine and patient education. She has developed and lectured on the utilization of patient education materials and educational methods to facilitate detection and recognition of a wide variety of common health problems, including several disorders that are commonly missed or misdiagnosed by primary care physicians, including depression, anxiety disorders, premenstrual syndrome, and headache disorders. These materials are used to engage patient participation in the process of identification and translation of symptoms into a proper diagnosis so that recognized treatment strategies may be employed. There is an assumption that patient compliance with treatment increases when the patient is educated and actively involved in the treatment process.

• Frederick Petty, PhD, MD

• Associate Professor, Department of Psychiatry

• Veterans Administration Medical Cente

• University of Texas Southwestern Medical Center

• Dallas, Texas

Dr. Petty received his PhD from the Georgia Institute of Technology (1971) and his MD from the University of Tennessee (1976). He completed his residency in psychiatry at the University of Iowa (1980).

Dr. Petty is the recipient of a Research Career Development Award from the Department of Veterans Affairs. He was the Director of the Consultation/Liaison Service at both the Iowa City and Dallas Department of Veterans Affairs Medical Centers. He conducts both basic and clinical research on the biology of stress and depression, as well as on biologic markers for alcoholism. Dr. Petty has received independent, peer-reviewed funding to conduct his research from the Veterans Administration Merit Review Board, NIMH, and NIAAA. He has lectured widely to family practice and general internal medicine physicians regarding the differential diagnosis and management of mood disorders. Dr. Petty is Director of the Depression Clinic at the Department of Veterans Affairs Medical Center in Dallas.

• Herbert C. Schulberg, PhD

• Professor of Psychiatry, Psychology and Medicine

• University of Pittsburgh School of Medicine

• Pittsburgh, Pennsylvania

Dr. Schulberg received his BA from Yeshiva College (1955), his PhD from Columbia University, and his MS Hygiene from the Harvard School of Public Health (1963). He completed 2 years of pre-doctoral internships in clinical psychology at several Veterans Administration facilities.

Dr. Schulberg is Director of the Primary Care Consultation Program at the Western Psychiatric Institute and Clinic. He is a Fellow of the American Psychological Association, and was previously president of the American College of Mental Health Administration and a Visiting Scientist at the NIMH. He has published more than 125 peer-reviewed articles, chapters, and books that focus on the delivery and evaluation of psychiatric services, as well as the recognition, differential diagnosis, and treatment of depression and other psychiatric disorders in the primary care setting. Dr. Schulberg is a recognized national authority in this area. He has received several NIMH grants that support his research efforts. Presently, Dr. Schulberg is directing an NIMH-funded 4-year study of the clinical efficacy and cost efficiency of various treatments for major depression in primary care practice.

• G. Richard Smith, Jr., MD

• Professor and Director

• Centers for Mental Healthcare Research

• VA HSR & D Field Program for Mental Health

• University of Arkansas for Medical Sciences

• Little Rock, Arkansas

Dr. Smith received his MD from the University of Arkansas (1977), where he completed his residency in psychiatry. He was a Fellow in Psychiatry and Medicine at the University of Rochester (1980-81). He has served as residency director in the Department of Psychiatry at the University of Arkansas. He has published in the areas of alexithymia, somatization disorder, consultation-liaison, and immunology. He has received NIMH and Robert Wood Johnson Foundation research support for studies of somatization disorder and psychosocial disabilities associated with myocardial infarction. He has been developing disease-specific outcome modules for psychiatric conditions.

Dr. Smith currently serves as Chair of the NIMH Initial Review Group on Mental Health Services Research. He is Director of the Centers for Mental Healthcare Research, which include the NIMH Center for Rural Mental Healthcare Research and the Veterans Affairs Field Program for Mental Health.

• Gail Wiscarz Stuart, PhD, RN, CS

• Associate Professor and Chief, Division of Psychiatric Nursing

• Department of Psychiatry and Behavioral Sciences

• Professor, College of Nursing

• Medical University of South Carolina

• Charleston, South Carolina

Dr. Stuart received her MS in psychiatric nursing from the University of Maryland (1973) and her PhD in behavioral sciences from Johns Hopkins University (1985). She is a certified specialist in adult psychiatric and mental health nursing from the American Nurses Association and maintains a private practice of psychotherapy.

Dr. Stuart's clinical and research interests involve the study of depression, anxiety disorders, and mental health care delivery systems. As Chief of the Division of Psychiatric Nursing, in the Department of Psychiatry, Dr. Stuart is responsible for overseeing the clinical inpatient units at the Institute of Psychiatry. She is also the coordinator of the graduate program in psychiatric nursing at the Medical University of South Carolina in Charleston.

Dr. Stuart has received multiple honors and awards for her work from a large number of organizations and is a Fellow of the American Academy of Nursing. She serves on the NIMH Task Force on Psychiatric Nursing and the NIMH Research Resource Panel for the Severely Mentally Ill, among other national appointments. Dr. Stuart has been the principal investigator or coinvestigator on several independent research grants, including studies of the pharmacotherapy and childhood environments of patients with panic disorder, bulimia, and depression. She has written several nationally recognized textbooks on psychiatric nursing and has a substantial list of peer-reviewed journal publications in both the areas of psychiatric disorders and the role of nursing in health care delivery.

Reviewing Consultants for Treatment Issues ¹

Scientific Reviewers ²

• Hagop S. Akiskal, MD

• Senior Science Advisor, Office of the Director

• National Institute of Mental Health

• Rockville, Maryland

• Deborah Allen, MD

• Professor and Chairman,

• Department of Family Medicine

• Indiana University

• School of Medicine

• Indianapolis, Indiana

• Kenneth Z. Altshuler, MD^*

• Professor and Chairman, Department of Psychiatry

• University of Texas

• Southwestern Medical Center

• Dallas, Texas

• Aaron T. Beck, MD^*

• Professor, Department of Psychiatry

• University of Pennsylvania

• Philadelphia, Pennsylvania

• W. Eugene Broadhead, MD, PhD, FAAFP

• Associate Professor, Department of Community and Family Medicine

• Duke University Medical Center

• Durham, North Carolina

• C. Robert Cloninger, MD^*

• Washington University Medical School

• St. Louis, Missouri

• Allen J. Frances, MD^*

• Professor and Chairman, Department of Psychiatry

• Duke University Medical Center

• Durham, North Carolina

• Ellen Frank, PhD^*

• University of Pittsburgh

• School of Medicine

• Western Psychiatric Institute and Clinic

• Pittsburgh, Pennsylvania

• Jack Froom, MD

• Professor,

• State University of New York at Stony Brook

• Stony Brook, New York

• Junius J. Gonzales, MD

• Chief, Primary Care Research Program Services Research Branch

• National Institute of Mental Health

• Rockville, Maryland

• Frederick K. Goodwin, MD

• Director,

• National Institute of Mental Health

• Bethesda, Maryland

• Philip S. Holtzman, PhD^*

• Harvard University

• Cambridge, Massachusetts

• Neil Jacobson, PhD

• Professor, Department of Psychology

• Center for Clinical Research

• University of Washington

• Seattle, Washington

• Lewis L. Judd, MD

• Mary Gilman Marston Professor and Chair, Department of Psychiatry

• UCSD School of Medicine

• La Jolla, California

• T. Byram Karasu, MD

• Professor of Psychiatry

• Albert Einstein College of Medicine

• Bronx Municipal Hospital Center

• Jacobi Hospital

• Bronx, New York

• Donald F. Klein, MD

• Columbia University, College of Physicians & Surgeons

• New York, New York

• Gerald Klerman, MD (deceased)

• Professor, Department of Psychiatry

• Director of Research

• New York Hospital/Cornell Medical Center

• Payne Whitney Clinic

• New York, New York

• Rodger Kobes, MD, PhD^*

• Timberlawn Psychiatric Hospital

• University of Texas

• Southwestern Medical Center

• Dallas, Texas

• Helena Chmura Kraemer, PhD

• Stanford University

• Stanford, California

• David J. Kupfer, MD

• Western Psychiatric Institute and Clinic

• University of Pittsburgh

• Pittsburgh, Pennsylvania

• Jerome Levine, MD^*

• Department of Psychiatry (MPRC)

• University of Maryland

• School of Medicine

• Baltimore, Maryland

• William T. McKinney, MD^*

• University of Wisconsin

• School of Medicine

• Madison, Wisconsin

• Kathleen Merikangas, PhD^*

• Yale University School of Medicine

• New Haven, Connecticut

• Deborah M. Nadzam, PhD, RN^*

• Associate Director of Outcomes Research and Development

• Joint Commission on Accreditation of Healthcare Organizations

• Oak Brook Terrace, Illinois

• Robert M. Post, MD^*

• National Institute of Mental Health

• Bethesda, Maryland

• Lon S. Schneider, MD>^*

• University of Southern California

• School of Medicine

• Los Angeles, California

• Myrna M. Weissman, PhD^*

• Columbia University

• New York, New York

• Thomas N. Wise, MD

• Professor of Psychiatry

• Georgetown University

• Washington, District of Columbia

Organizations and Individuals Providing Peer and Pilot Review ³

• American Nurses Association

• Washington, District of Columbia

• Contact: Karen O'Connor, MA, RN

• Sandra E. Benter, DNSC, RN, CS

• Psychotherapy and Consultation Practice

• Owings Mills, Maryland

• John M. Davis, MD

• Illinois State Psychiatric Institute

• Chicago, Illinois

• David L. Dunner, MD

• Professor, Department of Psychiatry

• University of Washington

• Seattle, Washington

• William H.M. Finney, MD, MPH

• Shepherd's Clinic

• Baltimore, Maryland

• Terry E. Fitzgerald

• National Council, Community Mental Health Centers

• Rockville, Maryland

• Frederick K. Goodwin, MD

• Director,

• National Institute of Mental Health

• Bethesda, Maryland

• Edgar Heim, MD

• Professor and Co-chair

• Psychiatrische Univ. Poliklinik

• Bern, Switzerland

• Shirley Hibbeln

• Education Consultant

• National Depressive and Manic-Depressive Association

• Chicago, Illinois

• Robert M.A. Hirschfeld, MD

• Professor and Chairman, Department of Psychiatry

• University of Texas Medical Branch

• alveston, Texas

• IMCARE Practice Guidelines Network

• IMCARE (Internal Medicine Center to Advance Research and Education)

• American Society of Internal Medicine

• Washington, District of Columbia

• Contact: Bernard M. Rosof, MD, President

• Betty F. King, Executive Director

• Harold A. Kaminetzky, MD, FACOG

• Director, Practice Activities

• American College of Obstetricians and Gynecologists

• Washington, District of Columbia

• Gerald Klerman, MD (deceased)

• President

• Association for Clinical Psychosocial Research

• Don R. Lipsitt, MD

• President, American Association of General Hospital Psychiatrists

• Cambridge, Massachusetts

• Russ Newman, PhD, JD

• Deputy Executive Director for Professional Practice

• American Psychological Association

• Washington, District of Columbia

• Pharmaceutical Manufacturers Association

• Health Outcomes Work Group

• Washington, District of Columbia

• Contact: Hugh H. Tilson, MD, DrPH

• Harold Alan Pincus, MD

• American Psychiatric Association

• Washington, District of Columbia

• John B. Reichman, MD

• American Academy of Clinical Psychiatry

• Pocatello, Idaho

• Lorraine Richter, BS

• Education Chairman

• National Depressive and Manic-Depressive Association

• Chicago, Illinois

• Joseph A. Rogers

• Mental Health Association of Southeastern Pennsylvania

• Philadelphia, Pennsylvania

• Peter M. Silberfarb, MD

• Dartmouth Medical School

• Dartmouth Hitchcock Medical Center

• Lebanon, New Hampshire

• Jeff Susman, MD

• Department of Family Medicine

• University of Nebraska Medical Center

• Omaha, Nebraska

• Robert L. Thomas

• National Association of Private Psychiatric Hospitals

• Washington, District of Columbia

• Joyce E. Thompson, CNM, DrPH, FAAN

• University of Pennsylvania

• School of Nursing

• Philadelphia, Pennsylvania

Organizations and Individuals Providing Additional Scientific, Technical, and Administrative Support

• University of Texas Southwestern Medical Center

• Dallas, Texas

• M. Trivedi, MD, Scientific Assistant to the Chair

• M. White, MM, Project Manager

• L. Arnold

• W. Hendrickse, MD, MRCP

• G. Kramer

• D. Savage

• Agency for Health Care Policy and Research,

• Rockville, Maryland

• J.J. Clinton, MD, Administrator

• K. McCormick, PhD, RN, Director, Office of the Forum

• E. Corrigan

• C. Crofton, PhD

• G. Hernandez, RN

• S. King, MD

• V. Montgomery

• K. Pearson, RPh, MPH

• R. Siegel

• L. Williams

• Editorial Associates,

• Washington, District of Columbia

• G. Martin

• Fast Word, Dallas, Texas

• Health Systems Research,

• Washington, District of Columbia

• MedStat Systems, Inc.,

• Ann Arbor, Michigan

• Mikalix & Company,

• Waltham, Massachusetts

• M. Madison, MPA, and staff

• Moshman and Associates,

• Bethesda, Maryland

• National Institute of Mental

• Health, Bethesda, Maryland

• F. Goodwin, MD, Director

• J. Burke, MD, MPH

• J. Gonzales, MD

• A. Leshner, PhD

• G. Norquist, MD, MSPH

• National Library of Medicine,

• Bethesda, Maryland

• I. Auston, MLS, and staff

• University of California at Los Angeles, California

• D. Schriger, MD, MPH

• University of California at Berkeley, California

• T.W. Hu, PhD

• University of Pittsburgh Medical Center,

• Pittsburgh, Pennsylvania

• M. McDonald, MFA

• Washington Consulting Group, Inc.,

• Washington, District of Columbia

• C.L. Smith