Back to Basics
When managing treatment for patients diagnosed as having mood disorders that are complex from the start or that become complex as they unfold, we recommend a back-to-basics approach. That is, treatment management begins with a comprehensive history that incorporates family, personal, medical, psychiatric, premorbid, and present illness elements into a case formulation. Although this formulation serves as a starting point, the formulation must remain dynamic and open to change over time as additional information is gathered from the patient, collateral sources, and clinical observation. Whereas this exercise serves most obviously as a route to diagnosis, it is also therapeutic in the most fundamental way. It is central to the unifying process of psychotherapy that was perhaps best articulated by Jerome Frank—the thorough assessment becomes a form of role induction for the patient to consider all aspects of life as potentially salient, establishes a therapeutic alliance in the act of the patient entrusting the clinician with his or her life story, and the commitment of time and effort to the endeavor instills hope and confidence by the patient in the clinician’s commitment and ability to help (
47). The value of building alliance, both with patients and families, cannot be overemphasized; it has been shown to predict clinical outcomes across a range of treatment approaches (
48–
50). Especially for patients with complex conditions, who have already endured a diagnostic odyssey, treatment management and benefit often begin with the clinician listening to the patient’s story and making a diagnosis, well before other therapies are applied.
Four essential points can be derived from these observations about assessment. First, and most simply, we have highlighted the established goal for all patient-provider interactions to be in some sense psychotherapeutic. Second, we have emphasized that being heard and understood can be profoundly meaningful, and this may be particularly true for patients who have been misdiagnosed and so in that way, misunderstood. Third, this approach serves as a reminder that patients come to clinical attention asking questions about diagnosis, prognosis, and how to proceed and sometimes need to be encouraged to ask these questions directly. Simply answering these questions, including what we do not know, can be deeply meaningful and therapeutic (
51–
53). Clinicians also need to keep in mind that it is common for patients who have not achieved remission or adequate improvement in other settings to remain untrusting. Treatment management for such patients should be simple and carried out in an unfailingly straightforward way. Finally, treatment management for all patients, including those with the most complex conditions, necessarily involves some form of psychotherapy. This practice is, of course, not only deeply rooted in the humanistic goal of therapy—to be present with those who are suffering—but it has also been empirically shown to work in synergy with other treatment modalities (
48,
54–
56). A somewhat similar caveat about the use of medications also applies to psychotherapy approaches. That is, the severity of illness (and to some extent the complexity of the patient’s clinical issues) should be inversely correlated to the complexity of therapy. That is, for patients with the most difficult and complicated conditions, it is often best to start by simply providing support, validating their distress, and prioritizing focus on functioning over an intensive search for deeper insights (
1,
53,
57,
58). Moreover, insight-oriented or collaborative therapeutic styles (e.g., psychodynamic, interpersonal, and cognitive behavioral therapies) often founder or can be counterproductive for patients whose severe depression prevents them from seeing anything in a hopeful light. Experienced therapists using such techniques know when to temporarily adjust or narrow the focus (e.g., in the immediate aftermath of the loss of a loved one).
With regard to medications, a good basic assessment entails digging as deep as necessary to develop a thorough review of prior medication trials. Ideally, this review will produce more than a list of past and current medications and will at least incorporate timing, dosage, duration of trials, side effects, and quality of therapeutic response. Once the history is clarified, and a decision is made to initiate or adjust medications, the psychotherapeutic alliance can aid in instilling hope in a patient who has been repeatedly disappointed or frustrated by treatment. We particularly believe information about duration of trials and response (even if partial) is important to know, because many patients and providers describe 4 weeks as adequate. In outpatient settings, we typically set 8 weeks at a usually therapeutic dose as the point at which no response or only the most limited hint of a response indicates that a trial has failed. Trials of a month or less have not had a reasonable chance to work. When embarking on trials that ask for such sustained patience, a trusting alliance—in which the patient is confident that he or she is getting competent care—often allows the patient to better bear the frustrations that are sometimes inevitable with complex conditions.
The importance of starting with a comprehensive formulation, attending to the patient as a whole, and striving for the simplest effective regimen is also paramount in the effort to focus on syndromes (e.g., depression, mania, or psychosis) and not merely to chase after symptoms (e.g., impaired concentration, which commonly occur in each of the previously noted syndromes) and thus to eschew “artisanal polypharmacy” that leaves patients using medications that have not been established as beneficial. Too often, polypharmacy arises out of an effort to manage side effects of other medications. Such polypharmacy is certainly appropriate when the side effect is extrinsic to the problem being treated, as would be the case with adding an anticholinergic for Parkinsonian side effects of a neuroleptic. But prescribing a stimulant, say, to arouse someone who is over sedated from the medication they are taking to promote sleep, is counterproductive—use less sedative in those instances.
Although this may seem like a simplistic part of standard practice, it is worth noting that psychiatric prescribers are more likely than other providers to use polypharmacy, patients with psychiatric disorders are more likely than patients with nonpsychiatric disorders to be prescribed multiple medications, and the use of polypharmacy for managing mood disorders has expanded over time (
7,
16,
59,
60). Furthermore, at least for patients with bipolar disorder, prescribers move more quickly to add a new agent than to change the initial medication, and discontinuation is rare (
61). There are times when targeting an individual symptom is the only way to quickly bring relief to a patient, at least until the overall therapeutic strategy takes hold. Thus, it is worth the risk of polypharmacy in some instances to provide, for example, a tranquilizing drug to alleviate severe anxiety or difficulty falling asleep, especially when these symptoms directly obstruct functional recovery. Additionally, for some patients, it is necessary to curate regimens that are highly individually tailored. However, there is little in the way of evidence-based medicine to guide complex combination approaches (
7,
59). Therefore, polypharmacy should be managed thoughtfully on an individual basis with frequent review of the current benefit of each medication.
As a final point, some have suggested that an increasing focus on residual symptoms (e.g., lingering impairments in vital sense, motivation, or sleep), as opposed to focusing on function or syndrome-level severity, may at least partially explain expanding polypharmacy in psychiatry (
16,
61). This focus would seem in line with the evolving goal of treating patients with even the most difficult cases to remission (
62,
63). However, it is our view that all treatment efforts should focus on the basic goals of functioning, quality of life, and meaning making, even if definitive diagnosis and remission of symptoms remain elusive. For many patients with complex and chronically disabling conditions, it may be reasonable to establish 80% improvement, 80% of the time, as a realistic intermediate goal (
64). For patients who have been severely and persistently ill for years or sometimes decades, this level of treatment is consistent with much more ability to function. We have seen such patients experience additional improvement after finding renewed interest or fulfillment in work and relationships. Further disease-focused treatment should not be neglected, but may sometimes be delayed until a period of better functioning provides a boost in confidence or resilience, because each new medication trial risks new failures and side effects.
Augmented Treatment
As just alluded to, perhaps the most accessible strategy for treatment augmentation in the management of mood disorders is the mixing of different classes of medication, and there are now several well-studied applications of medication combinations. Although providing a comprehensive list here is beyond the scope of this article, such strategies have been previously reviewed (
65–
71). Instead, we note that, although many combinations are often tried and are sometimes effective, relatively few medication augmentation strategies have been validated. Of those that have been validated, the most thoroughly studied are second-generation neuroleptics and lithium.
After the addition of a second antidepressant, second-generation neuroleptics represent the most common augmentation strategy (
6). Although the utility of neuroleptics in the management of mood disorders had long been clinically recognized (
72), their wide use over the last 20 years has followed approval by the U.S. Food and Drug Administration of second-generation antipsychotics for the treatment of mood disorders (
68,
73). Since then, a number of trials have supported their use in augmentation for treatment-resistant mood disorders (
74–
76). Their rapid onset of action, usefulness in both unipolar depression and bipolar disorder, and ability to treat psychotic mood states have been noted as advantages over other strategies, although low tolerability and side effects remain problematic with intermediate and long-term maintenance use (
8,
65,
66,
76).
Lithium, in combination with tricyclic antidepressants, was one of the first augmentation strategies to be empirically supported (
77,
78). Since then, a number of studies have supported the use of lithium in combination with antidepressants (
71). Only a few studies have specifically assessed the utility of lithium for patients diagnosed as having complex mood disorders; the data are mixed but overall support its use (
14,
79,
80). Interestingly, several studies have found that whereas lithium augmentation has a measurable sustained effect, its greatest benefits come early, in the first few weeks (
71), perhaps providing an additional reason for its use for patients who have especially severe or refractory illness. Specifically, there is abundant evidence across mood disorders demonstrating lithium’s antisuicidal effects (
81). These observations have led to international guidelines advocating for the use of lithium for patients with mood disorders (
82,
83), and some have recommended lithium augmentation as a first-line approach for patients with difficult-to-treat illness (
82,
84). Despite the benefits of lithium, recent practice instead has been to augment with additional antidepressants and neuroleptics (
6). We certainly find this prescribing practice to be true in our own clinical experience as consultants, consistently observing that lithium and tricyclic antidepressants (especially nortriptyline with therapeutic level monitoring) are frequently among the treatment options not yet tried.
A much smaller number of trials, each involving relatively few patients, have investigated other augmentation strategies specifically for patients with difficult-to-treat conditions. Among them include studies supporting augmentation with bupropion or buspirone (
85), pindolol (
70), thyroid hormone (
86,
87), mirtazapine (
88), lamotrigine (
89), and several psychostimulants (
66). Although these strategies remain promising, and many are widely used, much work remains to determine the patients and circumstances for which these approaches are most effective.
Even when these various efforts result in inadequate response, many alternative pathways for treatment, and reasons to continue to instill hope in patients, remain. Perhaps the most underused, but also most promising approach, is neurostimulation. Whereas neurostimulation has been historically viewed as an effort of last resort, guidelines are increasingly advocating for its use earlier during treatment. Given the abundant evidence demonstrating benefit, safety, and limited contraindications, electroconvulsive therapy (ECT) is now considered a preferred therapy for patients with difficult-to-treat conditions (
90). In fact, several major groups have advocated for use of ECT as an overall first-line treatment for some patients diagnosed as having mood disorders with psychotic features or acute suicidality (
82,
90,
91). For patients with lower acuity who have not responded to at least one antidepressant trial, repetitive transcranial magnetic (rTMS) stimulation offers another validated option (
92–
95), although the superiority of ECT compared to rTMS has been repeatedly demonstrated (
90). Other promising neurostimulation treatments, such as transcranial direct current stimulation and vagus nerve stimulation, are increasingly available and supported by evidence (
90) but are underused. ECT perhaps provides the most striking example of underuse. Despite decades of evidence supporting ECT as one of the most effective treatments for mood disorders, including remission rates of about 50% among patients with treatment-refractory conditions (
96), only about 0.25% of patients with mood disorders receive ECT (
97), even though approximately 20% of mood episodes remain refractory after 1–2 years of medication management alone (
15). Although patient preferences and availability of neurostimulation play some role in this discrepancy, ironically clinicians’ enduring negative views also contribute (
98).
More recently, there have been exciting developments regarding several novel medication-based strategies for the treatment of patients with severe and difficult-to-treat mood disorders. Among them, ketamine and its derivatives, the progesterone metabolite brexanolone, and psychedelic drugs such as psilocybin, have gathered the most attention and are beginning to have wider clinical use. Although promising, these approaches remain either entirely experimental or require unique clinical considerations and should be used in coordination with specialized consultants.
Expanding the Treatment Team
Although medication management in combination with psychotherapy represents the mainstay of outpatient treatment for complex mood disorders, as already noted, many patients continue to experience residual symptoms and require a higher level or additional components of care. Fortunately, several other specialized approaches can be integrated into routine practice, highlighting the value of the multidisciplinary and multimodal care models that have emerged across psychiatry (
99–
102). Disciplines that now have well-defined roles within psychiatric care include social work, occupational therapy, case management, vocational counseling, and psychiatric nursing (
99). Although studies of each individual discipline as applied specifically to patients with complex mood disorders are limited or lacking, the integration of these roles into higher levels of care for patients with severe mental illness (e.g., assertive community treatment) has been supported (
103–
106). A related resource that should be considered is family engagement. Although routinely a part of child, adolescent, and young adult care, the value of family involvement is likely underappreciated in adult psychiatry. Family engagement has demonstrated benefits for treatment retention, medication adherence, patient satisfaction with care, and reduction of depressive symptoms (
107–
110), so when feasible, it becomes a critical component of care for patients with complex conditions.
Another opportunity to improve patient engagement comes with implementation of measurement-based care—a systematic process of data collection and integration to monitor progress and inform decision-making. Although recognized as beneficial and central to evidence-based psychiatry for more than a decade, routine use of measurement-based care has been largely confined to research and large practice settings, with only about 5% of providers regularly using any standardized progress measure (
111). Now, with wide adoption of electronic records, the process of collection, analysis, and interpretation of measurements is becoming less burdensome for providers and patients. The approach, which has been associated with improved alliance, also provides objective data that can facilitate diagnosis and treatment progress (
112–
117), which would benefit patients with complex conditions for whom a definitive formulation remains elusive or when benefit from treatment is uncertain. Additionally, for patients with mood disorders, measurement-based care has been associated with improved insight, more rapid and greater treatment response, and better assessment of functional improvement (
115,
116). For patients who have medically complex conditions, the approach seems to facilitate communication among multiple members of a care team, which has also been associated with improved control of both general medical and psychiatric problems (
118).
Several other measures that may aid in diagnosis and/or treatment are also now routinely available. Perhaps one of the more common and most broadly useful is neuropsychological assessment to identify potential complicating or comorbid conditions, such as personality vulnerabilities, learning and memory impairment, executive functioning problems, and functional deficits. Such testing may be useful in managing treatment for patients with complex adversities. In fact, neuropsychological testing has been successfully used for patients with difficult-to-treat depression to distinguish between subjective report of symptoms and objective cognitive and functional performance (
119,
120), which can be challenging to distinguish in typical clinical encounters. An additional benefit is that neuropsychological assessment can be personalized, with a battery of instruments tailored to the individual presentation and the provider’s questions. Brain imaging techniques represent another set of widely available tools that can be complementary to routine psychiatric evaluation. Whereas a variety of structural abnormalities have been reported among patients diagnosed as having mood disorders, with evidence of differences between unipolar depression and bipolar disorder (
121–
124), the changes are often subtle and far from universal. Therefore, structural imaging is not currently diagnostic for mood disorders per se but can be useful when considering other brain disorders, such as those resulting from traumatic injury, dementing illnesses, or inflammatory processes as complicating factors. More recently, functional brain imaging has identified changes among individuals with mood disorders, and there is some suggestion that functional techniques may be useful in predicting responsiveness to different medication classes (
125), although this possibility remains in an investigational stage of development. Among many other techniques developed to empirically aid medication choice and dosage, genetic and pharmacogenetic testing appear to be the most rapidly developing. Pharmacogenetic testing, currently based on predicting pharmacokinetics and identifying individuals with uncommon polymorphisms for whom certain medications or drug combinations could carry increased risk, may be useful for patients who have unexpectedly little benefit or adverse reactions to a medication trial (
126,
127). Although testing is not currently able to predict individual treatment response to a specific psychiatric medication, one study (
128) suggests that such testing may be useful for patients with severe affective disorders who are new to treatment. Use and integration of these tools into routine practice may be hampered by relatively limited exposure of clinicians to genetic testing, interpretation, and counseling during psychiatric training (
129).
In many cases, and especially when the patient’s presentation defies a clear formulation, it is advisable to seek opinions from a colleague or specialty consultation team. Although the potential goals of doing so are many, we believe the most important point is that expert consultation should offer more than a diagnosis and list of treatment recommendations. Instead, the process should provide a reconsideration of the entire history and formulation through a different lens, critical assessment of past treatment trials, options and rationale for future treatment management, and the opportunity for ongoing guidance. However, despite the best efforts of treatment teams, patients with the most complex conditions may be too acutely ill to be safely treated in a typical outpatient setting. For these patients, there are several options for higher levels of care. Depending on the individual situation, psychiatric rehabilitation programs, intensive outpatient programs, partial hospitalization programs, residential programs, or assertive community treatment may be appropriate and, ultimately, acute inpatient hospitalization may be required. In fact, there is some evidence that the need for psychiatric admission is rising, with psychiatric stays being the only major type of hospitalization to increase from 2004 to 2014, and mood disorders now ranked seventh among all causes of hospitalization (
130). Although the patient trajectories and ultimate rationales for inpatient admission vary widely, we view the objectives of admission as relatively narrow. Certainly, safety is always the primary objective. Other objectives may include acute stabilization in a therapeutic milieu, diagnostic clarification and accelerated medical workup, intensive longitudinal assessment, accelerated medication change with close monitoring for high risk patients, or initiation of neurostimulation. Meanwhile, when patients are struggling not only with their mood symptoms and other complicating factors, but also the distress and frustration that can come with multiple ineffective trials and escalating care needs, it becomes increasingly important for providers to rely on the therapeutic alliance and the patient’s support network to remain present during times of patient distress and steadfastly committed to an appropriately optimistic future.