Diagnosis and Treatment of Depression in the Intensive Care Unit Patient
Edith S. Geringer
John Querques
Meghan S. Kolodziej
Tuesday E. Burns
Theodore A. Stern
Intense emotions are evoked routinely in intensive care units (ICUs), where life-and-death decisions occur daily. In the ICU, depression can be a psychological reaction to an acute medical illness, a manifestation of a primary affective disorder, a mood disorder associated with a specific organic disease or its treatment, or a result of the confusing overlap of somatic symptoms of depression and symptoms of medical illnesses.
In this chapter, the term depression refers not to being transiently sad, discouraged, disappointed, despondent, or grief-stricken but refers to major depressive disorder (MDD), defined in the 4th edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV) [1] as a syndrome of distinct and persistent dysphoria associated with neurovegetative changes and functional impairment. Varied in presentation,
course, and response to treatment, depressive disorders remain a pathophysiological enigma, despite centuries of recognition and more recent investigation of their possible genetic, neurochemical, neuroanatomic, endocrine, and immune underpinnings [2,3].
course, and response to treatment, depressive disorders remain a pathophysiological enigma, despite centuries of recognition and more recent investigation of their possible genetic, neurochemical, neuroanatomic, endocrine, and immune underpinnings [2,3].
Many physicians believe that depression is appropriate in the ICU because severe illness devastates a person’s life. However, we believe that while being dispirited may be an understandable response to critical illness, having a depressive disorder is not; therefore, it is always important to treat the latter. In fact, compelling evidence shows that untreated depression increases morbidity and mortality from cardiac and neurologic conditions and has detrimental effects on other—perhaps all—organ systems.
In this chapter, we focus on the links between depressive and medical conditions and the diagnosis, evaluation, and treatment of depression in critically ill patients.
Links Between Depression and Medical Conditions
Cardiovascular Disease
That depression is associated with the development and the progression of coronary heart disease (CHD), and with worse prognosis in CHD patients, is well established [4]. Not proven thus far is that treatment of depression can improve or prevent these outcomes. After two trials—the Enhancing Recovery in Coronary Heart Disease Patients (ENRICHD) study [5] and the Myocardial Infarction and Depression–Intervention Trial (MIND–IT) [6]—failed to show this, attention turned to isolating those attributes of a depressive episode that portend greater risk. Secondary analyses of these and other trials have examined symptom type, episode onset before or after an index event [7,8], recurrence [9,10], treatment responsiveness [11], and persistence of the cardiotoxic effects of depression [12]. For example, some studies suggested that worse cardiac outcomes are associated with somatic/affective symptoms (e.g., insomnia, fatigability, and diminished libido) more than with cognitive/affective symptoms (e.g., pessimism, self-dislike, and suicidal ideas) [13,14,15].
Possible explanations for the greater rates of cardiac death among patients diagnosed with depression include hypothalamic–pituitary–adrenal axis hyperactivity, elevation in inflammatory markers (e.g., interleukin 6, tumor necrosis factor α), diminished heart rate variability, decreased parasympathetic tone, increased sympathetic tone, and enhanced platelet activation causing more avid platelet aggregation and plaque formation [16]. Interestingly, sertraline decreases platelet and endothelial activation in depressed patients after an acute coronary syndrome (ACS) [17,18]. The Heart and Soul Study, a prospective cohort study of 1,017 patients with stable CHD, found that behavioral factors, especially physical inactivity, were most responsible for the greater rate of adverse cardiac events in patients with depressive symptoms [19].
Cerebrovascular Disease
As with cardiovascular disease, there appear to be bidirectional links between cerebrovascular disease and depressive illness. The Caerphilly Study of 2,201 men found that psychological distress predicted fatal stroke but not nonfatal stroke or transient ischemic attack (TIA) [20]. The Framingham Heart Study of 4,120 men and women found that depressive symptoms were a risk factor for stroke or TIA before, but not after, age 65 [21].
Poststroke depression (PSD) has been extensively studied during the past 3 decades. Robinson [22] pooled the available data and found the mean prevalence of poststroke affective illness to be 19.3% for major depression and 18.5% for minor depression. Risk factors for the development of PSD include stroke severity, extent of physical disability, presence of cognitive impairment, and poor social support [23].
In the early 1980s, Robinson et al. [24,25] reported that the severity of PSD correlated with the proximity of the lesion to the frontal pole in the left, but not the right, hemisphere. This finding has been replicated by some [26,27], but not all [28], researchers; this localization may hold only during the first few months after stroke [29].
Diagnosis of Depression
Important questions for the intensivist are “What is depression?” and “What does a patient experiencing depression look like in the ICU?” To qualify for a diagnosis of MDD according to the DSM-IV, a patient must have five of the nine symptoms listed in Table 199.1, one of which must be either depressed mood or anhedonia, most of the day, nearly every day, for at least 2 weeks. The mnemonic—SIG: E CAPS (where SIG [abbreviation for the Latin, signa] refers to the instructions on a prescription, E refers to energy, and CAPS refers to capsules)—is a helpful guide to remember the eight neurovegetative symptoms associated with depressed mood. The mnemonic—ABCs of depression—portrays more richly the myriad affective, behavioral, and cognitive aspects of the condition (Table 199.2). Each symptom should be asked about, and questions about suicide should be raised directly. If a patient has thoughts of suicide, he or she should be asked whether there is a specific plan; the physician should then make a judgment about the likelihood of the patient’s acting on the plan. If an active plan for suicide exists, psychiatric consultation is imperative (see Chapter 200).
Four of the nine diagnostic criteria (i.e., insomnia, fatigue or loss of energy, diminished ability to think or concentrate, and anorexia or weight loss) are difficult to attribute exclusively to depression in the medically ill patient. However, in terminally ill cancer patients, Chochinov et al. [30] found that inclusion of these somatic symptoms in the diagnostic criteria did not artifactually increase rates of diagnosis, as long as the cardinal symptoms of depressed mood and anhedonia were held to the strict requirement of presence most of the day, nearly every day, for at least 2 weeks.
Table 199.1 Sig: E Caps—A Mnemonic for Diagnostic Criteria for Major Depressive Disorder | ||
|---|---|---|
|
Table 199.2 Abcs of Depression—Affective, Behavioral, and Cognitive Features | |||||||||||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
|
Patients Who are Unable to Speak
It may be particularly difficult to diagnose depression in a patient who is being mechanically ventilated or who has aphasia. However, much can be learned about a patient even when he or she is mute. It is important to watch facial expressions, observe hand gestures and other body language, and read lips. An individual who averts his or her eyes from the examiner’s gaze may be demoralized, discouraged, or depressed. Slow, sighing respirations may indicate depression rather than respiratory insufficiency. The astute clinician can also watch vital-sign monitor screens, looking for changes that can signify intense affect.
Does the patient respond to the mention of a favorite hobby or a grandchild with a smile or with tears? Is the patient’s affect labile or consistent with the content of the discussion? Emotional lability is not usually an indicator of MDD; instead, it suggests frontal lobe dysfunction. One can probe for affect by joking and observing the patient’s reaction.
A patient who can move his or her arms can be asked to write, draw, or point to a letter or a picture board. One simple screening test that can be used is human figure drawing (i.e., having the patient draw a picture of a person and another of what the patient thinks is wrong with the person). Typically, drawings by depressed patients convey their sense of dejection or a disordered understanding of their dilemma.
Table 199.3 Methods of Assessing Depression in Sensorially Compromised Patients | |
|---|---|
|
Some tracheostomized patients may have the oxygenation status, control of respiratory muscles, and ability to manage secretions sufficient to use a Passy-Muir valve, which permits exhaled air to pass the larynx and thus allows the patient to speak. Alternatively, electronic voice-output communication aids may be used. These devices pair prerecorded messages or synthesized speech with labeled icons; patients communicate messages by touching buttons on display screens or on touch-sensitive keyboards. Speech pathologists have knowledge of and access to such technology. Methods of assessing depression in sensorially compromised patients are summarized in Table 199.3.
Differential Diagnosis of Depression
Causes Related to Medical Conditions
A variety of medical illnesses can cause affective disorders, contribute to their occurrence, and worsen their severity (Table 199.4). Clues that depression is due to a medical illness include
older age at onset of symptoms, lower incidence of a family history of depression, and changes in personality and cognition. A thorough history (including a review of systems), physical (including neurologic) examination, and laboratory testing can distinguish between primary (i.e., due to a psychiatric condition) and secondary (i.e., due to a medical condition) causes of depression. For secondary causes, treatment of the underlying illness is usually more effective than is the use of psychotropic medications.
older age at onset of symptoms, lower incidence of a family history of depression, and changes in personality and cognition. A thorough history (including a review of systems), physical (including neurologic) examination, and laboratory testing can distinguish between primary (i.e., due to a psychiatric condition) and secondary (i.e., due to a medical condition) causes of depression. For secondary causes, treatment of the underlying illness is usually more effective than is the use of psychotropic medications.
Table 199.4 Medical Conditions Associated with Depressive Symptoms | |
|---|---|
|
Perhaps the most important differential diagnosis to consider in a patient who appears to have MDD is hypoactive delirium. The key feature that distinguishes it from depression is inattention (i.e., an inability to focus and sustain alertness on a given stimulus and to resist distraction by other stimuli). Delirium is discussed in Chapter 197.
Causes Related to Medical Treatments
The pharmacologic agents most often responsible for depression in the ICU are antihypertensives, beta-blockers, antiarrhythmics, and steroids (Table 199.5). Some medications may cause depression only after several weeks or even months of continuous use. If a drug regimen or a dosage increase appears to be temporally related to the patient’s depression, the dose should be lowered or the medication eliminated entirely. If the medication cannot be stopped without serious risk to the patient, the depression should be treated.
Steroids
Depression, mania, psychosis, and delirium are frequent side effects of corticosteroid therapy. Mood symptoms are dose-dependent and usually occur within the first 2 weeks of therapy, although they can arise on the first day. A practical rule of thumb holds that neuropsychiatric adverse effects are common with prednisone ≥ 80 mg per day (or equivalent), uncommon ≤ 30 mg per day, and not uncommon in between. Although it has been suggested that women are more likely to develop steroid-induced adverse effects, the apparent increased frequency may be due to the higher prevalence of rheumatologic diseases in women. Corticosteroid-induced mood disorders are generally reversible with dosage reduction or discontinuation of the medication.
Laboratory Evaluation of Depression
Although the clinical interview and mental status examination are the most important components of psychiatric diagnosis, the use of laboratory tests is essential to exclude organic causes of depression. Although there is no consensus on the laboratory tests necessary in a patient with new-onset mood disorder, Table 199.6 lists those tests that should be considered. Thyroid-stimulating hormone is not on this list because many critically ill patients have abnormal thyroid biochemical profiles but do not have intrinsic thyroid disease. Syphilis and hypovitaminosis are rarely the sole causes of depression; tests for these conditions should be ordered only when there is a specific indication for them. Neuroimaging, electroencephalography, and cerebrospinal fluid analysis are relatively indicated in patients with new-onset psychiatric symptoms, altered cognition, new neurologic symptoms, seizures, and fever. The more of these features a patient has, the more important these additional tests become.
Treatment of Depression
Patients who meet the criteria for MDD are usually treated with a somatic therapy (including pharmacotherapy and electroconvulsive therapy [ECT]), alone or in combination with psychotherapy (Table 199.7). In critical care units, somatic therapies are the most widely used treatments for depression. Pharmacotherapy may be used in critical care units also for patients who have an adjustment disorder with depressed mood, particularly when these patients have several neurovegetative symptoms. A patient who is neither eating nor sleeping and who lacks the energy to participate in his or her rehabilitation may be helped considerably by antidepressants, especially psychostimulants.
Table 199.5 Drugs Associated with Depressive Symptoms | |
|---|---|
|
Each type of pharmacotherapy has its own indications and contraindications, but general rules are available when
selecting an antidepressant [31,32]. The most common rule is to choose a medication with a side-effect profile that best fits a patient’s needs. For instance, a patient who is having trouble sleeping will benefit from a sedating antidepressant. Conversely, a patient who has severe psychomotor retardation may benefit from a more stimulating antidepressant. With the exception of the psychostimulants, all antidepressants require approximately 4 to 6 weeks until full antidepressant effects are noted, although some response can occur in 1 to 2 weeks. Obviously, in critical care units, quicker effects are generally needed. Stimulants and ECT work more quickly, usually within several days. Patients with depression may also manifest considerable anxiety and may be helped by the use of an anxiolytic while awaiting response to an antidepressant. Psychotically depressed patients (with delusions or hallucinations) may need antipsychotics for control of symptoms.
selecting an antidepressant [31,32]. The most common rule is to choose a medication with a side-effect profile that best fits a patient’s needs. For instance, a patient who is having trouble sleeping will benefit from a sedating antidepressant. Conversely, a patient who has severe psychomotor retardation may benefit from a more stimulating antidepressant. With the exception of the psychostimulants, all antidepressants require approximately 4 to 6 weeks until full antidepressant effects are noted, although some response can occur in 1 to 2 weeks. Obviously, in critical care units, quicker effects are generally needed. Stimulants and ECT work more quickly, usually within several days. Patients with depression may also manifest considerable anxiety and may be helped by the use of an anxiolytic while awaiting response to an antidepressant. Psychotically depressed patients (with delusions or hallucinations) may need antipsychotics for control of symptoms.
Table 199.6 Laboratory Evaluation of Depression | |
|---|---|
|
Table 199.7 Comparative Properties of Some Antidepressants | |
|---|---|
|
Psychostimulants
Psychostimulants have been used to treat depressive symptoms since their development in the 1930s, but they fell into disrepute when they became known as drugs of abuse in the 1950s and 1960s. Since then, there have been numerous reports on the use of stimulants in the treatment of depressed patients, particularly apathetic and geriatric patients; recently, there has been a renewed interest in the use of psychostimulants in depressed, medically ill patients who are intolerant of other medications
[33]. Thought to be particularly effective in patients with cancer and stroke, their rapid onset is of great use in any setting, including the ICU, where speed of recovery is crucial. For example, they are valuable in patients who are difficult to wean from mechanical ventilation [34].
[33]. Thought to be particularly effective in patients with cancer and stroke, their rapid onset is of great use in any setting, including the ICU, where speed of recovery is crucial. For example, they are valuable in patients who are difficult to wean from mechanical ventilation [34].
The psychostimulants most commonly used are dextroamphetamine (Dexedrine) and methylphenidate (Ritalin). Both appear to work through the direct neuronal release of dopamine and norepinephrine; dextroamphetamine blocks catecholamine reuptake and weakly inhibits monoamine oxidase. Both of these psychostimulants are predominantly excreted by the kidneys, although dextroamphetamine also undergoes a complex biotransformation.
The usual effects of stimulants are to increase motor behavior, increase arousal, and decrease appetite; however, in patients who are anorexic on the basis of depression, appetite is paradoxically increased, likely through dopaminergic stimulation of the nucleus accumbens. Their antidepressant effect is usually evident in the first 2 days of treatment, if not earlier. In a review of 66 patients hospitalized on medical-surgical wards at Massachusetts General Hospital, 93% achieved maximum benefit within 2 days of use [35,36]. Stimulants do not show anticholinergic effects or cause orthostatic hypotension. They can increase heart rate and blood pressure and can cause coronary spasm and cardiac arrhythmias; however, these effects are rare (even with preexisting cardiac abnormalities) at the low doses (5 to 20 mg/day) usually used for the treatment of depression [35]. In fact, stimulants have been used safely and effectively in a broad spectrum of patients, including those with critical illness, and have shown little potential for abuse or dependence. Contraindications to stimulant use include the concurrent use of α-methyldopa (which becomes a sympathoamine when metabolized), monoamine oxidase inhibitors (MAOIs), and bronchodilators; and pregnancy, seizures, delirium, psychosis, significant hypertension, and active angina [37].
Psychostimulants should be the first consideration in treating depression in critically ill patients. Patients are started on 5 mg of methylphenidate or 2.5 to 5 mg of dextroamphetamine in the morning. The dose is increased by 5 mg per day (for methylphenidate) or 2.5 to 5 mg per day (for dextroamphetamine) until a therapeutic effect is detected or until a maximum dose of 20 mg has been reached. Heart rate and blood pressure should be monitored as closely as necessary. Stimulants are usually given for at least 1 to 2 weeks after depressive symptoms have fully remitted. In most cases, after stimulants are stopped, depression does not recur.
Stimulants taken in overdose may cause seizures, coma, hallucinations, paranoia, hyperthermia, hypertension, cardiac arrhythmias, angina, and circulatory collapse. The major treatment for overdose is to acidify the urine (which enhances renal excretion) and to use supportive measures for all other abnormalities.
Modafinil (Provigil)—a wakefulness-promoting medication approved for narcolepsy, shift work sleep disorder, and obstructive sleep apnea/hypopnea syndrome—may be a beneficial alternative to the psychostimulants.
Selective Serotonin Reuptake Inhibitors
The SSRIs are a class of antidepressants that causes a potent and selective blockade of serotonin reuptake. Since the introduction of fluoxetine (Prozac) in 1987, SSRIs have become the most widely prescribed class of antidepressants. Other SSRIs include sertraline (Zoloft), paroxetine (Paxil), fluvoxamine (Luvox), citalopram (Celexa), and escitalopram (Lexapro). They are far less anticholinergic, antihistaminergic, and anti-α1-adrenergic than the older tricyclic antidepressants (TCAs) and, therefore, are associated with far fewer side effects. They also have fewer cardiovascular effects and do not commonly cause orthostatic hypotension.
Pharmacokinetics
SSRIs are well absorbed from the gastrointestinal tract, and absorption is generally unaffected by food and antacids. They have a large volume of distribution and are highly protein-bound. They are extensively metabolized in the liver, where they are oxidized, methylated, and conjugated. The elimination half-lives of sertraline, paroxetine, fluvoxamine, and citalopram are approximately 1 day (although sertraline has a mildly active metabolite with a half-life of 66 hours); this allows once-a-day dosing. Fluoxetine has a half-life of 2 to 3 days and a highly active metabolite (norfluoxetine) with a mean half-life of 6.1 days. Fluoxetine takes a much longer time to reach steady state and, more importantly for drug overdoses, can take weeks to months to be fully cleared. Elimination half-lives are dose-dependent (i.e., higher doses and lengthier usage are associated with higher plasma levels and longer half-lives). SSRIs show wide interindividual variation in pharmacokinetics and do not yet have a clearly established dose-response curve.
Metabolic Impairment
Fluoxetine, sertraline, fluvoxamine, and citalopram are unaffected by renal dysfunction [38,39]. Paroxetine, although minimally excreted in the urine (like other SSRIs), shows increased plasma concentrations in the setting of renal disease [38]. Fluoxetine, sertraline, paroxetine, and citalopram doses should be reduced by at least half in patients with liver disease [38]. Fluvoxamine has been used in patients with cirrhosis and hepatic encephalopathy without adverse effects [40]. The hepatic clearance, not the plasma concentration, of fluvoxamine is affected by cirrhosis. Therefore, the dosage frequency, rather than the total dosage, should be altered [40]. In elderly individuals, fluoxetine does not have altered pharmacokinetics; in contrast, sertraline and paroxetine have increased plasma levels and slower clearance. Although citalopram has a 30% longer half-life in the elderly, the frequency and severity of side effects are not higher in this group [41].
Side Effects
SSRIs can cause tremulousness, agitation, irritability, insomnia, anorexia, nausea, vomiting, diarrhea, excess sweating, and sexual dysfunction (i.e., decreased libido, erectile and orgasmic dysfunction). The syndrome of inappropriate antidiuretic hormone is an uncommon adverse effect reported with all of the SSRIs; especially in critically ill patients, other causes of hyponatremia should be sought before attributing the metabolic derangement to the SSRI. The SSRIs do not typically cause clinically significant changes in heart rate, blood pressure, or the electrocardiogram (ECG). Overdoses of SSRIs are discussed in Chapter 124.
Theoretically, SSRIs can cause angina or myocardial infarction (MI) due to the direct vasoconstrictive effects of serotonin on damaged myocardium. When fluoxetine therapy is initiated, serum serotonin levels rise for the first 2 weeks and then return to baseline. This mechanism has been implicated in 3 cardiac deaths that occurred 10 days after initiation of fluoxetine [42]. This theoretical concern should extend to other SSRIs as well.
Related posts:
Full access? Get Clinical Tree
