Like most medical complaints, a depressed mood in and of itself represents a symptom, and not a diagnosis. Indeed, depression is a relatively nonspecific complaint, that could reflect anything from the most debilitating psychiatric illness to a normal response to stress or loss (Fig. 3-1). The differential diagnosis for depression among psychiatric disorders alone is broad, requiring a thorough history, examination, and often laboratory tests to discern the underlying etiology (or at least exclude potential confounders). The assessment of mood disorders in the medically ill poses yet several additional challenges, especially since many core, neurovegetative symptoms of depression – such as fatigue, anergia, insomnia, anorexia, weight loss, and pain – often result directly from medical illnesses, themselves. Among patients with severe medical disease, assessment can be confounded by “sickness behavior,” a state of decreased motivation resulting from systemic inflammation, and characterized by malaise, anorexia, insomnia, fatigue, as well as fever.¹ Similarly, delirium, especially the hypoactive subtype, can present with prominent mood symptoms, making patients appear dysphoric. Adding to the complexity in assessment, many commonly used treatments, such as steroids or interferon-α, can also lower mood as a side effect. Proper diagnosis remains critical, as the recommended treatment approaches do vary widely, depending on the etiology.

In this chapter, we review approaches for screening and diagnosing depression in the medically ill. We begin in what is likely familiar territory: the common, primary psychiatric diagnoses that present with a low mood, and their implications for medical patients. We then discuss approaches to the clinical interview, available screening tools, the potentially confounding role of delirium in this population, and the assessment of suicidal risk. Finally, we review cultural and ethnic considerations that inform the assessment of depression.

Along with anxiety, a low mood represents a common, typical response to stressors in life. As is the case with bereavement, a certain degree of depression is considered normal, even adaptive, and does not rise to clinical attention. This phenomenon is of special relevance for individuals with medical illness, all of whom face some degree of stress, loss, or burden. Physical suffering, functional limitations, hospitalizations, frequent medical appointments, social isolation, concern about the future, reliance on caregivers, and the financial burdens are just a few examples of illness-related stress. Feeling at least somewhat discouraged is to be expected in these situations, so how do we know when depression requires clinical intervention? The DSM addresses this issue primarily by requiring the presence of “significant distress or impairment in social, occupational, or other important areas of functioning” as diagnostic criteria for a mood disorder.² That said, even the authors concede that normal grief and illness can coexist and that clinical judgment is ultimately required. For depression arising from a stressor (social or health-related), that crosses the line into clinical significance, the diagnosis of adjustment disorder with depressed mood represents the first point beyond uncomplicated grief or normal coping. According to the DSM 5, adjustment disorders are pervasive in general medical hospitals, with a prevalence of up to 50% among inpatients. The key feature of an adjustment disorder is the presence of a stressor bearing a clear relationship to a depression that appears to be out of proportion to the situation and/or results in functional impairment. The typical course of adjustment disorder with depressed mood is contrasted with that of major depression in Figure 3-2. Adjustment disorder with depressed mood is often seen when medically ill patients receive new diagnoses, have a recurrence or worsening of their illness, or require new types of treatment, such as supplemental oxygen or chemotherapy. Notably, an exacerbation of an underlying, pre-existing mental illness, even in the context of stress, does not constitute an adjustment disorder, underscoring the importance of careful history taking in making the diagnosis.

The diagnosis of adjustment disorder is insufficient for depression that causes severe impairment, or that arises without a clear relationship to stressors, expanding the diagnostic possibilities considerably (Fig. 3-3). The first task is to exclude direct medical causes for a depressed mood and ensure their adequate treatment. (In DSM 5, the formal diagnosis “mood disorder due to a general medical condition” was changed to “due to another medical condition, reflecting the notion that primary depressive disorders are medical.) Examples of mood disorder due to the pathophysiology of another medical condition are found throughout this book, and can arise from a variety of diseases. The initial evaluation of depressed mood includes the exclusion of drug- or alcohol-induced mood disorders as well as of common medical illnesses causing low energy and fatigue, such as anemia and thyroid dysfunction. Taking a relatively prevalent example, hypothyroidism can present with prominent complaints of depressed or irritable mood, apathy, anergia, anhedonia, and amotivation. While clinical hypothyroidism is found in less than 4% of patients with depression, subclinical hypothyroidism is present in up to 40% of patients, and can be important to recognize.³ The diagnosis of depression due to another medical condition also encompasses any drug-related side effects that produce a low mood. Common examples of drugs that can induce depression include isoniazid, prednisone, and interferon.

Once adjustment and substance-use disorders and medical causes are excluded, the key entity to consider in the differential diagnosis is major depressive disorder (MDD). MDD is marked by at least 2 weeks of depressed mood, occurring most days, causing significant distress, and accompanied by at least four other symptoms. Four of the nine diagnostic criteria (weight change, dyssomnia, psychomotor retardation or agitation, and fatigue) are also common in acute and chronic medical disease, which complicate the diagnostic process in the medically ill. In this case, greater reliance on the cognitive and affective symptoms (decreased pleasure in activities, feelings of worthlessness or guilt, helplessness, hopelessness, decreased concentration, and suicidal thoughts) may be required to make the diagnosis. Importantly, major depression may arise in the context of one or more stressors, at first resembling adjustment. Alternatively, it may arise either without a clear precipitant or coincide with the onset of a life stressor. While adjustment disorders generally respond well to supportive psychotherapy, major depression typically requires medication or a combination of treatment modalities, including cognitive behavioral therapy. In severe cases of major depression, psychotic features (typically mood-congruent, negative delusions, and auditory hallucinations) can arise, prompting treatment with antipsychotics as well as antidepressants.

Although major depression is typically believed to be episodic, marked by interspersed periods of normal mood (or euthymia), more chronic forms do exist. For clinically significant depression lasting 2 years or longer, clinicians should consider the diagnosis of persistent depressive disorder (a consolidation of the former diagnoses of dysthymia and chronic major depression, thus superseding the prior distinguishing criteria of symptom severity). These patients typically report depression beginning as early adults, if not during childhood, and have some chronic degree of mood disturbance with fluctuating severity. Earlier onset of persistent depressive disorder is associated with comorbid personality and substance-use disorders.

Evaluating a patient in the midst of a major depressive episode, regardless of duration, is important to determine whether a bipolar diathesis exists. Individuals with an underlying bipolar diathesis, or full disorder, are at risk for developing mania from antidepressants, and may face a worsened course of their mood disorder without an appropriate mood stabilizing agent.^4,5 It is therefore important to screen for a history of mania or hypomania, as well as to identify individuals at risk for developing the disorder but who have yet to manifest manic symptoms (Table 3-1). Unfortunately, the depressive phase of bipolar disorder is generally clinically indistinguishable from that of (unipolar) major depression. While inquiring about a past history of mania or hypomania is a necessary first step, even structured diagnostic interviews fail to identify past manic or hypomanic episodes in a significant number of patients with bipolar depression.⁶ Patients with active major depression may have difficulty recalling any history of (hypo-)mania fully, so collateral information from family or friends is often useful.

Beyond these primary mood disorders, many psychiatric illnesses can present with some symptoms of depression, such as posttraumatic stress disorder, substance-use disorders, anxiety disorders, personality disorders, or eating disorders. It is always important to keep in mind that multiple diagnoses can coexist, and that a depressive disorder alone may not fully explain the clinical picture. In cases where the diagnosis is unclear, a systematic approach to excluding contributing factors, such as outlined in Table 3-2, can be useful.

The psychiatric interview of individuals with active medical illness can present various challenges, depending on the environment and nature of the referral. Certainly for patients in medical inpatient or emergency department settings, small spaces, limited privacy, noisy surroundings, interruptions, and unpredictable time constraints can all affect the quality of the interview. The same may be true for outpatient clinicians embedded in primary care offices or other specialty clinics, where mental health may not be the primary focus of the practice. In these venues, consultations may be unexpected, arising with little to no warning to the patient or provider, and require a flexible, yet efficient approach. Since the timeframe for assessment and management can be relatively short compared to customary intakes in an outpatient psychiatry clinic, it is ideal to establish rapport quickly and address the “critical” issues, such as safety, early. In this sense, operating under the assumption that the interview could be interrupted at any moment can help the clinician focus on the most important elements of the history and examination, before the patient must leave to obtain a laboratory test, or vacate the examination room.

For initial encounters, it is important to keep in mind that many patients seen in general medical settings (or who are referred from these settings), may never have seen a psychiatrist or mental health provider before. They might not have sought out psychiatric care of their own accord, and in some cases may be entirely unaware of the reasons for the referral. This situation may be especially true of medical inpatients, individuals noted to be depressed on routine screenings, and patients seeking psychiatric “clearance” for surgeries or other procedures. Many patients therefore harbor some anxiety about the psychiatric interview and how the findings will be used. Some may have misconceptions or embarrassment about mental illness, or even present with hostility toward the psychiatric provider for thinking they need mental health services. Since a comfortable and productive rapport is so important, a few strategies for managing these issues are reviewed here.

Asking patients if they are aware of the reasons for the referral – and clarifying any inaccuracies – builds the alliance from the outset. This process also gives the clinician some sense of the patient’s insight into their illness, mood, and social interactions. For patients reluctant to entertain a depression diagnosis, it can also be helpful to provide some education about the impact of depression on medical disorders and the benefits of addressing both simultaneously. For patients with chronic depression, especially in the context of health or other social stressors, they may not necessarily view their low mood and related functional impairment as abnormal, or in need of treatment. Indeed, they may have gotten the message from family, friends, and even medical providers, that being depressed is “normal,” appropriate, or expectable given their situation. In these cases, it often helps to portray the initial interview more as an evaluation to see “what could be done to be helpful,” whether it is improving sleep, energy level, or motivation, as opposed to making a diagnosis.

For patients who have never seen a psychiatric clinician in the past, and did not seek out the evaluation on their own, anxiety about the nature of the interview is common. Patients often worry that they will be “analyzed” or somehow “tricked” into discussing things that are private. This concern may be shaped by the experience – positive or negative – of their family or friends who had previously sought mental health care. Individuals being screened for surgery or employment may be especially guarded about revealing depression symptoms or other signs of mental illness. It is not uncommon for patients to ask for some reassurance during the end of the interview; “So I’m not crazy, right?”

One helpful approach is to begin the history by inquiring about more comfortable, familiar information for the patient: their medical illness. Most patients have had more experience discussing their physical complaints. For many, discussing their pulmonary disease or diabetes is familiar and well-travelled territory, while seldom if ever have they broached the related emotional issues. Conveying or reflecting back some basic knowledge about the medical disease in an empathic manner not only serves to build the alliance, it conveys that the clinician intends to recognize and address the medical as well as psychiatric issues in the patient’s care. This approach allays fears that the clinician does not understand the medical illness, or that psychiatric treatment will hinder their medical care.

In addition to increasing the comfort level of the interview, starting with the medical history also gives a sense of the time course and relative severity of the illness, important for determining the chronological relationship to mood symptoms. Transitioning to more “psychiatric” questions is then accomplished by asking how the medical illness, now well reviewed, has affected them. For instance, one might ask, “How have your spirits been holding up with all of this going on?” or “Has all this affected your activities or mood?” From there, more specific inquiries can proceed to assess hopelessness, helplessness, and self-blaming thoughts. Again, early knowledge of the patient’s medical illness helps contextualize these cognitive features of depression; for example, is the illness terminal, or does the self-blame arise from a perceived burden of family? Exploring negative cognitions can also be useful in determining the nature of nonspecific symptoms, such as fatigue, common to both depression and many systemic illnesses. Given the potential time constraints mentioned above, it is also prudent to screen for suicidal thoughts as early as appropriate in the interview, ensuring an adequate opportunity to address any acute safety concerns.

In concert with the history typically reviewed during a psychiatric interview, special consideration should be given to the longitudinal course of the patient’s functional impairment, social context, and health care resources. Social isolation and financial difficulties often grow as medical illness progresses, for instance, compounding the level of depression. For illnesses with a relapsing and remitting course, inquiring in more detail about possible past episodes of mood symptoms – whether formally diagnosed or not – may be helpful in formulating the current disorder. Eliciting a family history of medical as well as psychiatric illnesses similarly provides a basis not only for the patient’s genetic risk but, for diagnoses that run in families, it also identifies any preconceived notions the patient may have about their illness, based on their relatives’ experiences.

Given the time constraints of many medical encounters, there is great value in employing screening tools for detecting possible cases of depression and integrating them into the clinical workflow. Although many validated tools exist, the most common remains the Patient Health Questionnaire (PHQ-9). This nine-item self-report screening instrument rates the presence or absence, and the frequency, of cognitive and physical symptoms of depression as well as suicidal thoughts. The PHQ-9 is well validated in several outpatient medical settings, though the evidence for its use in medical inpatients remains mixed.^7–11 It has also been validated in several languages, including Chinese, Thai, Spanish, French, Swahili, and British sign language.^12–17 PHQ-9 scores range from 0 to 27, with a score over 20 indicative of severe depression. A score of 10 is generally recommended as the threshold for major depression, with a sensitivity and specificity of approximately 88% for both using this cutoff.¹⁸ A more recent meta-analysis of the validation studies, however, showed similar pooled sensitivity and specificity for scores between 8 and 11.^19,20

A two-item version of this tool, the PHQ-2, also exists and inquires solely about depressed mood and anhedonia in the previous 2 weeks. The PHQ-2 is ideally used as the initial screen. Among primary care and obstetrics-gynecology clinic patients, the PHQ-2 score of 3 or more has a sensitivity of 83% and a specificity of 92% for detecting major depression.²¹ A positive score on the PHQ-2 should prompt the administration of the full (PHQ-9) scale. It is important to note that the PHQ-9 (and PHQ-2), cannot differentiate the causes of depression, such as bipolar disorder, adjustment disorder, bereavement, or mood disorders due to substances or a general medical condition. For all of the utility of screening instruments, however, it is important to use clinical judgment and not rely solely on a score to define next steps. A positive screening score always requires further diagnostic clarification; it is critical to review responses with the patient to ensure correct interpretation and avoid misdiagnosis.