DYSURIA

ESSENTIAL INQUIRIES

Fever.
Nausea or vomiting.
New back or flank pain.
Vaginal discharge.
Pregnancy risk.
Structural abnormalities.
Instrumentation of urethra or bladder.

General Considerations

Dysuria (painful urination) is a common reason for adolescents and adults to seek urgent medical attention. An inflammatory process (eg, infection; autoimmune disorder) underlies most causes of dysuria. In women, cystitis will be diagnosed in up to 50–60% of cases and has an incidence of 0.5–0.7% per year in sexually active young women. The key objective in evaluating women with dysuria is to exclude serious upper urinary tract disease, such as acute pyelonephritis, and sexually transmitted diseases. In elderly men, dysuriamay be a symptom of prostatitis. In contrast, in younger men, urethritis accounts for the vast majority of cases of dysuria.

Clinical Findings

A. Symptoms

Well-designed cohort studies have shown that some women can be reliably diagnosed with uncomplicated cystitis without a physical examination or urinalysis, and randomized controlled trials show that telephone management of uncomplicated cystitis is safe and effective. An increased likelihood of cystitis is present when women report multiple irritative voiding symptoms (dysuria, urgency, frequency), fever, or back pain (likelihood ratios = 1.6–2.0). Inquiring about symptoms of vulvovaginitis is imperative. When women report dysuria and urinary frequency, and deny vaginal discharge and irritation, the likelihood ratio or culture-confirmed cystitis is 24.5. In contrast, when vaginal discharge or irritation is present, as well as dysuria or urinary frequency, the likelihood ratio is 0.7. Gross hematuria in women with voiding symptoms usually represents hemorrhagic cystitis but can also be a sign of bladder cancer (particularly in older patients) or upper tract disease. Failure of hematuria to resolve with antibiotic treatment should prompt further evaluation of the bladder and kidneys. Chlamydial infection should be strongly considered among women age 25 years or younger who are sexually active and seeking medical attention for a suspected urinary tract infection for the first time or have a new partner. 

Because fever and back pain, as well as nausea and vomiting, are considered harbingers of (or clinical criteria for) acute pyelonephritis, women with these symptoms shouldusually be examined by a clinician prior to treatment in order to exclude coexistent urosepsis, hydronephrosis, or nephrolithiasis. Other major risk factors for acute pyelonephritis (among women 18–49 years of age) relate to sexual behaviors (frequency of sexual intercourse three or more times per week, new sexual partner in previous year, recent spermicide use), as well as diabetes mellitus and recent urinary tract infection or incontinence. Finally, pregnancy, underlying structural factors (polycystic kidney disease,
nephrolithiasis, neurogenic bladder), immunosuppression, diabetes, and a history of recent bladder or urethral instrumentation usually alter the treatment regimen (antibiotic choice or duration of treatment, or both) of uncomplicated cystitis. In men, lower urinary tract symptoms and depressive symptoms are strongly associated and reciprocally related. Erectile dysfunction increases the odds of both disorders.

B. Physical Examination

The presence of fever, tachycardia, or hypotension should alert the clinician to the possibility of urosepsis and the potential need for hospitalization. A focused examination in women, in uncomplicated circumstances, could be limited to ascertainment of costovertebral angle tenderness and to a lower abdominal and pelvic examination, if the history suggests vulvovaginitis or cervicitis.

C. Diagnostic Studies

1. Urinalysis—Urinalysis is probably overutilized in the evaluation of dysuria. The probability of culture-confirmed urinary tract infection among women with a history and physical examination compatible with uncomplicated cystitis is about 70–90%. Urinalysis is most helpful when the woman with dysuria does not have other typical features of cystitis. Dipstick detection (> trace) of leukocytes, nitrites, or blood supports a diagnosis of cystitis. When both leukocyte and nitrite tests are positive, the likelihood ratio is 4.2, and when both are negative, the likelihood ratio is 0.3. The negative predictive value of urinalysis is not sufficient to exclude culture-confirmed urinary tract infection in women with multiple and typical symptoms; and randomized trial evidence shows that antibiotic treatment is beneficial to women with typical symptoms and negative urinalysis dipstick tests.

2. Urine culture—Urine culture should be considered for all women with upper tract symptoms (prior to initiating antibiotic therapy), as well as those with dysuria and a negative urine dipstick test. In symptomatic women, a clean-catch urine culture is considered positive when 102–103 colony-forming units/mL of a uropathogenic organism is detected.

3. Renal imaging—When severe flank or back pain is present, the possibility of complicated kidney infection (perinephric abscess, nephrolithiasis) or of hydronephrosis should be considered. Depending on local availability, acceptable imaging options to assess for hydronephrosis include abdominal radiographs, renal ultrasound, or CT scanning. To exclude nephrolithiasis, noncontrast helical CT scanning is more accurate than intravenous urography and is rapidly becoming the diagnostic test of choice for this purpose. In a meta-analysis, the positive and negative likelihood ratios of helical CT scanning for diagnosis of nephrolithiasis were 23.2 and 0.05, respectively.

Differential Diagnosis

The differential diagnosis of dysuria in women includes acute cystitis, acute pyelonephritis, vaginitis (Candida, bacterial vaginosis, Trichomonas, herpes simplex), urethritis/ cervicitis (Chlamydia, gonorrhea), and interstitial cystitis/ painful bladder syndrome. Nucleic acid amplification tests from first-void urine or vaginal swab specimens are highly sensitive for detecting chlamydial infection. Other infectious pathogens associated with dysuria and urethritis in men include Mycoplasma genitalium and Enterobacteriaceae.

Treatment

Definitive treatment is directed to the underlying cause of the dysuria. An evidence-informed algorithm for managing suspected urinary tract infection in women is shown in Figure 2–3. This algorithm supports antibiotic treatment of most women with multiple and typical symptoms of urinary tract infection without performing urinalysis or urine culture.

1.In women who have risk factors for sexually transmitted diseases, consider testing for Chlamydia. The US Preventive Services Task Force recommends screening for Chlamydia for all women 25 years or younger and women of any age with more than one sexual partner, a history of sexually transmitted disease, or inconsistent use of condoms.

2.A complicated UTI is one in an individual with a functional or anatomic abnormality of the urinary tract, including a history of polycystic renal disease, nephrolithiasis, neurogenic bladder, diabetes mellitus, immunosuppression, pregnancy, indwelling urinary catheter, or recent urinary tract instrumentation.

3.The only physical examination finding that increases the likelihood of UTI is costovertebral angle tenderness, and clinicians may consider not performing this test in patients with typical symptoms of acute uncomplicated UTI (as in telephone management).

Figure 2–3. Proposed algorithm for evaluating women with symptoms of acute urinary tract infection (UTI). (Modified and reproduced, with permission, from Bent S et al. Does this woman have an acute uncomplicated urinary tract infection? JAMA. 2002 May 22–29;287(20):2701–10.)

Symptomatic relief can be provided with phenazopyridine, a urinary analgesic that is available over-the-counter; it is used in combination with antibiotic therapy (when a urinary tract infection has been confirmed) but for no more than 2 days. Patients should be informed that phenazopyridine will cause orange/red discoloration of their urine and other bodily fluids (eg, some contact lens wearers have reported discoloration of their lenses). Rare cases of methemoglobinemia and hemolytic anemia have been reported, usually with overdoses or underlying renal dysfunction.

In cases of interstitial cystitis/painful bladder syndrome (see Chapter 23), patients will often respond to a multimodal approach that may include urethral/vesicular dilation, biofeedback, cognitive behavioral therapy, antidepressants, dietary changes, vaginal emollients, and other supportive measures.

When to Refer

• Anatomic abnormalities leading to repeated urinary infections.
• Infections associated with nephrolithiasis.
• Persistent interstitial cystitis/painful bladder syndrome.

When to Admit

• Severe pain requiring parenteral medication or impairing ambulation or urination (such as severe primary herpes simplex genitalis).
• Dysuria associated with urinary retention or obstruction.

ACUTE HEADACHE

ESSENTIAL INQUIRIES

Age > 50 years.
Rapid onset and severe intensity (ie, “thunderclap” headache).
Fever.
Trauma.
Vision changes.
HIV infection.
Current or past history of hypertension.
Neurologic findings (mental status changes, motor or sensory deficits).

General Considerations

Headache is a common reason that adults seek medical care, accounting for approximately 13 million visits each year in the United States to physicians’ offices, urgent care clinics, and emergency departments. A broad range of disorders can cause headache. This chapter will deal only with the approach to the acute nontraumatic headache in adolescents and adults. The challenge in the initial evaluation of acute headache is to identify which patients are presenting with an uncommon but life-threatening condition. In the emergency department setting, approximately 1% of patients seeking medical attention for acute headache will have a life-threatening condition, whereas the  prevalence of life-threatening conditions in the office practice setting is considerably lower.

Regardless of the underlying cause, headache is currently believed to occur as a result of the release of neuropeptides from trigeminal nerve endings that encapsulate the blood vessels of the pia mater and dura mater, resulting in neurogenic inflammation. Because this represents a final common pathway for many etiologies, diminution of headache in response to typical migraine therapies (such  as serotonin receptor antagonists or ketorolac) does not rule out critical conditions such as subarachnoid hemorrhage or meningitis as the underlying cause.

Clinical Findings

A. Symptoms

A careful history and physical examination should aim to identify causes of acute headache that require immediate treatment. These causes can be broadly classified as imminent or completed vascular events (intracranial hemorrhage, thrombosis, vasculitis, malignant hypertension, arterial dissection, or aneurysm), infections (abscess, encephalitis, meningitis), intracranial masses causing intracranial hypertension, preeclampsia, and carbon monoxide poisoning. Having the patient carefully describe the onset of headache can be helpful in diagnosing a serious cause. Report of a sudden-onset headache that reaches maximal and severe intensity within seconds or a few minutes is the classic description of a “thunderclap headache” and should precipitate workup for subarachnoid hemorrhage, since the estimated prevalence of subarachnoid hemorrhage in the setting of a thunderclap headache is 43%.

The general medical history can also guide the need for additional work-up. A new headache in a patient over the age of 50 or with a history of HIV disease under most circumstances (including a normal neurologic examination) warrants neuroimaging immediately (Table 2–8). When the patient has a medical history of hypertension—particularly uncontrolled hypertension—a complete search for criteria satisfying a diagnosis of “malignant hypertension” is appropriate to determine the correct urgency level of hypertension management (see Chapter 11). Headache and hypertension associated with pregnancy may be due to preeclampsia. Episodic headache associated with the triad of hypertension, heart palpitations, and sweats is suggestive of pheochromocytoma. In the absence of thunderclap headache, advanced age, and HIV disease, a careful physical examination and detailed neurologic examination will usually determine acuity of the work-up and need for further diagnostic testing.

Patient symptoms can also be useful for diagnosing migraine headache in the absence of the “classic” migraine pattern involving scintillating scotoma followed by unilateral headache, photophobia, and nausea and vomiting (Table 2–9). The presence of three or more of these features can establish the diagnosis of migraine (in the absence of other clinical features that warrant neuroimaging studies), and the presence of none or one of these features (provided it is not nausea) can help to rule out migraine.

B. Physical Examination

Critical components of the physical examination of the patient with acute headache include vital sign measurements, neurologic examination, and vision testing with funduscopic examination. The finding of fever with acute headache warrants additional maneuvers to elicit evidence of meningeal inflammation, such as Kernig and Brudzinski signs. Besides malignant hypertension, significant hypertension can also be a sign of intracranial hemorrhage, preeclampsia, and pheochromocytoma. Patients over 60 years of age should be examined for scalp or temporal artery tenderness.

Careful assessment of visual acuity, ocular gaze, visual fields, pupillary defects, the optic disk, and retinal vein pulsations is crucial. Diminished visual acuity is suggestive of glaucoma, temporal arteritis, or optic neuritis. Ophthalmoplegia or visual field defects may be signs of venous sinus thrombosis, tumor, or aneurysm. Afferent pupillary defects can be due to intracranial masses or optic
neuritis. Ipsilateral ptosis and miosis suggest Horner syndrome and in conjunction with acute headache may signify carotid artery dissection. Finally, papilledema or absent retinal venous pulsations are signs of elevated intracranial pressure—findings that should be followed by neuroimaging prior to performing lumbar puncture (Table 2–8).

Mental status and complete neurologic evaluations are also critical and should include assessment of motor and sensory systems, reflexes, gait, cerebellar function, and pronator drift. Any abnormality on mental status or neurologic evaluation warrants emergent neuroimaging (Table 2–8).

C. Diagnostic Studies

Guidance on when to perform neuroimaging is summarized in Table 2–8. Under most circumstances, a noncontrast head CT is sufficient to exclude intracranial hypertension with impending herniation, intracranial hemorrhage, and many types of intracranial masses (notable exceptions include lymphoma and toxoplasmosis in HIVpositive patients, herpes simplex encephalitis, and brain abscess). When appropriate, a contrast study can often be ordered to follow a normal noncontrast study. A normal neuroimaging study does not sufficiently exclude subarachnoid hemorrhage and should be followed by lumbar puncture. In patients for whom there is a high level of suspicion for subarachnoid hemorrhage or aneurysm, a normal CT and lumbar puncture should be followed by angiography within the next few days (provided the patient is medically stable). Lumbar puncture is also indicated to exclude infectious causes of acute headache, particularly in patients with fever or meningeal signs. Cerebrospinal fluid tests should routinely include Gram stain, white blood cell count with differential, red blood cell count, glucose, total protein, and bacterial culture. In appropriate patients, also consider testing cerebrospinal fluid for VDRL (syphilis), cryptococcal antigen (HIV-positive patients), acid-fast bacillus stain and culture, and complement fixation and culture for coccidioidomycosis. Storage of an extra tube with 5 mL of cerebrospinal fluid is also prudent for conducting unanticipated tests in the immediate future. Consultation with infectious disease experts regarding local availability of polymerase chain reaction tests for specific infectious pathogens (eg, herpes simplex 2) should also be considered in patients with evidence of centralnervous system infection but no identifiable pathogen. 

In addition to neuroimaging and lumbar puncture, additional diagnostic tests for exclusion of life-threatening causes of acute headache include erythrocyte sedimentation rate (temporal arteritis; endocarditis), urinalysis (malignant hypertension; preeclampsia), and sinus CT or radiograph (bacterial sinusitis, independently or as a cause of venous sinus thrombosis).

Treatment

Treatment should be directed at the cause of acute headache. In patients in whom migraine or migraine-like headache has been diagnosed, early treatment with NSAIDs or triptans can often abort or provide significant relief of symptoms. The effectiveness of NSAIDs and triptans appears to be equivalent, although combination sumatriptan-naproxen appears to provide a greater sustained response (2–24 hours) compared with monotherapy with either agent alone. Nonoral medication administration is often necessary for patients with significant nausea and vomiting, in which case nasal and subcutaneous triptans and intramuscular ketorolac are therapeutic options. Other causes of acute headache, such as subarachnoid hemorrhage, intracranial mass, or meningitis, usually require emergent treatment in the hospital.

When to Refer

• Frequent migraines not responsive to standard therapy.
• Migraines with atypical features.
• Chronic daily headaches due to medication overuse.

When to Admit

• Need for repeated doses of parenteral pain medication.
• To facilitate an expedited work-up requiring a sequence of procedures and neuroimaging.
• Monitoring for progression of symptoms and neurologic consultation when the initial emergency department work-up is inconclusive.
• Pain severe enough to impair activities of daily living or limit partic

FATIGUE & CHRONIC FATIGUE SYNDROME

ESSENTIAL INQUIRIES

Weight loss.
Fever.
Sleep-disordered breathing.
Medications.
Substance use.

General Consideration

Fatigue, as an isolated symptom, accounts for 1–3% of visits to generalists. The symptom of fatigue is often poorly described and less well defined by patients than symptoms associated with specific dysfunction of organ systems. Fatigue or lassitude and the closely related complaints of weakness, tiredness, and lethargy are often attributed to overexertion, poor physical conditioning, sleep disturbance, obesity, undernutrition, and emotional problems. A history of the patient’s daily living and working habits may obviate the need for extensive and unproductive diagnostic studies.

Chronic fatigue syndrome may be due to retrovirus infection or an immune dysregulation mechanism, or both. Recent studies, however, have failed to show any differences in levels of xenotropic murine leukemia-virusrelated virus in US patients with and without chronic  fatigue syndrome. The diagnosis of chronic fatigue syndrome remains hotly debated because of the lack of a gold standard. Persons with chronic fatigue syndrome meeting specific criteria (such as those from the CDC) report a greater frequency of childhood trauma and psychopathology and demonstrate higher levels of emotional instability and self-reported stress than persons who do not have chronic fatigue. Neuropsychological, neuroendocrine, and brain imaging studies reveal neurobiologic abnormalities in most patients, but none with a consistent pattern. Sleep disorders have been reported in 40–80% of patients with hronic fatigue syndrome, but their treatment has provided only modest benefit, suggesting that it is an effect rather than a cause of the fatigue. Veterans of the Gulf War show a tenfold greater incidence of chronic fatigue syndrome compared with nondeployed military personnel.

Clinical Findings

A. Fatigue

Clinically relevant fatigue is composed of three major components: generalized weakness (difficulty in initiating activities); easy fatigability (difficulty in completing activities); and mental fatigue (difficulty with concentration and memory). Important diseases that can cause fatigue include hyperthyroidism and hypothyroidism, CHF, infections (endocarditis, hepatitis), COPD, sleep apnea, anemia, autoimmune disorders, irritable bowel syndrome, and cancer. Alcoholism, side effects from such drugs as sedatives, and b-blockers may be the cause. Psychological conditions, such as insomnia, depression, anxiety, panic attacks, dysthmia, and somatization disorder, may cause fatigue. Common outpatient infectious causes include mononucleosis and sinusitis. These conditions are usually associated with other characteristic signs, but patients may emphasize fatigue and not reveal their other symptoms unless directly asked. The lifetime prevalence of significant fatigue (present for at least 2 weeks) is about 25%. Fatigue of unknown cause or related to psychiatric illness exceeds that due to physical illness, injury, medications, drugs, or alcohol.

B. Chronic Fatigue Syndrome

A working case definition of chronic fatigue syndrome indicates that it is not a homogeneous abnormality, and here is no single pathogenic mechanism (Figure 2–2). No physical finding or laboratory test can be used to confirm the diagnosis of this disorder. 

The evaluation of chronic fatigue syndrome includes a history and physical examination as well as complete blood count, erythrocyte sedimentation rate, serum chemistries blood urea nitrogen (BUN), electrolytes, glucose, creatinine, and calcium; liver and thyroid function tests antinuclear antibody, urinalysis, and tuberculin skin test; and screening questionnaires for psychiatric disorders. Other tests to be performed as clinically indicated are serum cortisol, rheumatoid factor, immunoglobulin levels, Lyme serology in endemic areas, and tests for HIV antibody. More extensive testing is usually unhelpful, including antibody to Epstein-Barr virus. There may be an abnormally high rate of postural hypotension. MRI scans may show brain abnormalities on T2-weighted images—chiefly small, punctate, subcortical white matter hyperintensities, predominantly in the frontal lobes; however, brain MRI is not recommended in the routine evaluation of chronic fatigue syndrome.

Treatment

A. Fatigue

Management of fatigue involves identification and treatment of conditions that contribute to fatigue, such as cancer, pain, depression, disordered sleep, weight loss, and anemia. Resistance training and aerobic exercise lessens fatigue and improves performance for a number of chronic conditions associated with a high prevalence of fatigue, including CHF, COPD, arthritis, and cancer. Continuous positive airway pressure is an effective treatment for obstructive sleep apnea. Psychostimulants such as methylphenidate have shown inconsistent results in randomized trials of treatment of cancer-related fatigue.

B. Chronic Fatigue Syndrome

A variety of agents and modalities have been tried for the treatment of chronic fatigue syndrome. Acyclovir, intravenous immunoglobulin, nystatin, and low-dose hydrocortisone/ fludrocortisone do not improve symptoms except in patients with postural hypotension, some of whom report response to increases in dietary sodium as well as fludrocortisone, 0.1 mg orally daily. There is a greater prevalence of past and current psychiatric diagnoses in patients with this syndrome. Affective disorders are especially common. Patients with chronic fatigue syndrome have benefited from a comprehensive multidisciplinary intervention, including optimal medical management, treating any ongoing affective or anxiety disorder pharmacologically, and implementing a comprehensive cognitive-behavioral treatment program. Cognitive-behavioral therapy, a form of nonpharmacologic treatment emphasizing self-help and aiming to change perceptions and behaviors that may perpetuate
symptoms and disability, is helpful. Although few patients are cured, the treatment effect is substantial. Response to cognitive-behavioral therapy is not predictable on the basis of severity or duration of chronic fatigue syndrome, although patients with low interest in psychotherapy rarely benefit. Graded exercise has also been shown to improve functional work capacity and physical function. At present, cognitive-behavioral therapy and graded exercise are the treatments of choice for patients with chronic fatigue syndrome. A 2011 randomized trial (PACE trial) has confirmed the independent benefits of cognitive behavioral therapy and graded exercise, and found no benefit of adaptive pacing therapy.

In addition, the clinician’s sympathetic listening and explanatory responses can help overcome the patient’s frustrations and debilitation by this still mysterious illness. All patients should be encouraged to engage in normal activities to the extent possible and should be reassured that full recovery is eventually possible in most cases.

When to Refer

• Infections not responsive to standard treatment.
• Difficult to control hyperthyroidism or hypothyroidism.
• Severe psychological disease.
• Malignancy.

When to Admit

• Failure to thrive.
• Fatigue severe enough to impair activities of daily living.

INVOLUNTARY WEIGHT LOSS

ESSENTIAL INQUIRIES

Age.
Caloric intake.
Fever.
Change in bowel habits.
Secondary confirmation (eg, changes in clothingsize).
Substance use.
Age-appropriate cancer screening history.

General Considerations

Body weight is determined by a person’s caloric intake, absorptive capacity, metabolic rate, and energy losses. The metabolic rate can be affected by a multitude of medical conditions through the release of various cytokines such as cachectin and interleukins. Body weight normally peaks by the fifth or sixth decade and then gradually declines at a rate of 1–2 kg per decade. In NHANES II, a national survey of community-dwelling elders (age 50–80 years), recent involuntary weight loss (> 5% usual body weight) was reported by 7% of respondents, and this was associated with a 24% higher mortality.

Etiology

Involuntary weight loss is regarded as clinically significant when it exceeds 5% or more of usual body weight over a 6- to 12-month period and often indicates serious physical or psychological illness. Physical causes are usually evident during the initial evaluation. Cancer (about 30%), gastrointestinal disorders (about 15%), and dementia or depression (about 15%) are the most common causes. When an adequately nourished-appearing patient complains of weight loss, inquiry should be made about exact weight changes (with approximate dates) and about changes in clothing size. Family members can provide confirmation of weight loss, as can old documents such as driver’s licenses. A mild, gradual weight loss occurs in some older individuals. It is due to changes in body composition, including loss of height and lean body mass and lower basal metabolic rate, leading to decreased energy requirements. However, rapid involuntary weight loss is predictive of morbidity and mortality in any population. In addition to various disease states, causes in older individuals include loss of teeth and consequent difficulty with chewing, alcoholism, and social isolation.

Clinical Findings

Once the weight loss is established, the history, medication profile, physical examination, and conventional laboratory and radiologic investigations (such as complete blood count, serologic tests including HIV, thyroid-stimulating hormone [TSH] level, urinalysis, fecal occult blood test, chest radiography, and upper gastrointestinal series) usually reveal the cause. When these tests are normal, the second phase of evaluation should focus on more definitive gastrointestinal investigation (eg, tests for malabsorption; endoscopy) and cancer screening (eg, Papanicolaou smear, mammography, prostate specific antigen [PSA]). 

If the initial evaluation is unrevealing, follow-up is preferable to further diagnostic testing. Death at 2-year followup was not nearly as high in patients with unexplained involuntary weight loss (8%) as in those with weight loss due to malignant (79%) and established nonmalignant diseases (19%). Psychiatric consultation should be considered when there is evidence of depression, dementia, anorexia nervosa, or other emotional problems. Ultimately, in approximately 15–25% of cases, no cause for the weight loss can be found.

Differential Diagnosis

Malignancy, gastrointestinal disorders (eg, malabsorption, pancreatic insufficiency), dementia, depression, anorexia nervosa, hyperthyroidism, alcoholism, and social isolation are all established causes. “Meals on Wheels” is a useful mnemonic for remembering the common treatable causes of involuntary weight loss in the elderly (see box below).

Meals on Wheels: A Mnemonic
for Common Treatable
Causes of Unintentional
Weight Loss in the Elderly.

Medication effects
Emotional problems, especially depression
Anorexia tardive (nervosa), alcoholism
Late-life paranoia
Swallowing disorders
Oral factors (eg, poorly fitting dentures, cavities)
No money
Wandering and other dementia-related behaviors
Hyperthyroidism, hypothyroidism, hyperparathyroidism, hypoadrenalism
Enteric problems (eg, malabsorption)
Eating problems (eg, inability to feed self)
Low-salt, low-cholesterol diets
Social problems (eg, isolation, inability to obtain preferred foods), gallstones

Treatment

Weight stabilization occurs in most surviving patients with both established and unknown causes of weight loss through treatment of the underlying disorder and caloric supplementation. Nutrient intake goals are established in relation to the severity of weight loss, in general ranging from 30 to 40 kcal/kg/d. In order of preference, route of administration options include oral, temporary nasojejunal tube, or percutaneous gastric or jejunal tube. Parenteral nutrition is reserved for patients with serious associated problems. A variety of pharmacologic agents have been proposed for the treatment of weight loss. These can be categorized into appetite stimulants (corticosteroids, progestational agents, dronabinol, and serotonin antagonists); anabolic agents (growth hormone and testosterone derivatives); and anticatabolic agents (omega-3 fatty acids, pentoxifylline, hydrazine sulfate,
and thalidomide).

When to Refer

• Weight loss caused by malabsorption.
• Persistent nutritional deficiencies despite adequate supplementation.
• Weight loss as a result of anorexia or bulimia.

When to Admit

• Severe protein-energy malnutrition, including the syndromes of kwashiorkor and marasmus.
• Vitamin deficiency syndromes.
• Cachexia with anticipated progressive weight loss secondary to unmanageable psychiatric disease.
• To carefully manage electrolyte and fluid replacement in protein-energy malnutrition and avoid “re-feeding syndrome.”

FEVER & HYPERTHERMIA

ESSENTIAL INQUIRIES

Age.
Localizing symptoms.
Weight loss.
Joint pain.
Injection substance use.

Immunosuppression or neutropenia.
History of cancer.
Medications.
Travel.

General Considerations

The average normal oral body temperature taken in mid-morning is 36.7°C (range 36–37.4°C). This spectrum includes a mean and 2 standard deviations, thus encompassing 95% of a normal population, measured in mid-morning (normal diurnal temperature variation is 0.5–1°C). The normal rectal or vaginal temperature is 0.5°C higher than the oral temperature, and the axillary temperature is correspondingly lower. Rectal temperature is more reliable than oral temperature, particularly in patients who breathe through their mouth or in tachypneic states.

Fever is a regulated rise to a new “set point” of body temperature. When stimuli act on monocyte-macrophages, these cells elaborate pyrogenic cytokines, causing elevation of the set point through effects in the hypothalamus. These cytokines include interleukin-1 (IL-1), tumor necrosis factor (TNF), interferon-gamma, and interleukin-6 (IL-6). The elevation in temperature results from either increased heat production (eg, shivering) or decreased loss (eg, peripheral vasoconstriction). Body temperature in cytokine-induced fever seldom exceeds 41.1°C unless there is structural damage to hypothalamic regulatory centers. 

1. Fever

Fever as a symptom provides important information about the presence of illness—particularly infections—and about changes in the clinical status of the patient. The fever pattern, however, is of marginal value for most specific diagnoses except for the relapsing fever of malaria, borreliosis, and occasional cases of lymphoma, especially Hodgkin disease. Furthermore, the degree of temperature elevation does not necessarily correspond to the severity of the illness. In general, the febrile response tends to be greater in children than in adults. In older persons, neonates, and in persons receiving certain medications (eg, NSAIDs or corticosteroids), a normal temperature or even hypothermia may be observed. Markedly elevated body temperature may result in profound metabolic disturbances. High temperature during the first trimester of pregnancy may cause birth defects, such as anencephaly. Fever increases insulin requirements and alters the metabolism and disposition of drugs used for the treatment of the diverse diseases associated with fever.

2. Hyperthermia

Hyperthermia—not mediated by cytokines—occurs when body metabolic heat production or environmental heat load exceeds normal heat loss capacity or when there is impaired heat loss; heat stroke is an example. Body temperature may rise to levels (> 41.1°C) capable of producing irreversible protein denaturation and resultant brain damage; no diurnal variation is observed.

Neuroleptic malignant syndrome is a rare and potentially lethal idiosyncratic reaction to major tranquilizers, particularly haloperidol and fluphenazine; however, it has also been reported with the atypical neuroleptics (such as olanzapine or risperidone). Serotonin syndrome resembles neuroleptic malignant syndrome but occurs within hours of ingestion of agents that increase levels of serotonin in the central nervous system, including serotonin reuptake inhibitors, monoamine oxidase inhibitors, tricyclic antidepressants, meperidine, dextromethorphan, bromocriptine, tramadol, lithium, and psychostimulants (such as cocaine, methamphetamine, and MDMA). Clonus and hyperreflexia are more common in serotonin syndrome whereas “lead pipe” rigidity is more common in neuroleptic malignant syndrome. Neuroleptic malignant and serotonin syndromes share common clinical and pathophysiologic features to malignant hyperthermia of anesthesia (see Chapters 25 and 38).

3. Fever of Undetermined Origin

(See Fever of Unknown Origin, Chapter 30)

Most febrile illnesses are due to common infections, are short-lived, and are relatively easy to diagnose. In certain instances, however, the origin of the fever may remain obscure (“fever of undetermined origin,” FUO) even after protracted diagnostic examination. In upper respiratory tract infections, fever typically lasts no more than 3–5 days, beyond which additional evaluation should be considered. he term “FUO” has traditionally been reserved for unexplained cases of fever exceeding 38.3°C on several occasions for at least 3 weeks in patients without neutropenia or immunosuppression (see Chapter 30). Some authors have advocated for changing the definition of FUO in frail elderly nursing home patients to a persistent oral or tympanic membrane temperature of 37.2°C (99°F) or greater, or persistent rectal temperature of 37.5°C (99.5°F) or greater.

Differential Diagnosis

See Table 2–7. The causes of FUO among elderly patients differ from those in younger patients, and a definitive diagnosis can be achieved in a larger proportion of elderly patients (up to 70%). Although the general diagnostic approach and distribution of infectious and noninfectious causes are roughly similar between age groups, tuberculosis and temporal arteritis are particularly more common causes of FUO in the elderly.

Treatment

Most fever is well tolerated. When the temperature is > 40°C, symptomatic treatment may be required. A temperature > 41°C is likely to be hyperthermia and thus not cytokine mediated, and emergent management is indicated. (See Heat Stroke, Chapter 37.)

A. General Measures for Removal of Heat

Regardless of the cause of the fever, alcohol sponges, cold sponges, ice bags, ice-water enemas, and ice baths will lower body temperature. They are more useful in hyperthermia, since patients with cytokine-related fever will attempt to override these therapies.

B. Pharmacologic Treatment of Fever

1. Antipyretic drugs—Antipyretic therapy is not needed except for patients with marginal hemodynamic status. Aspirin or acetaminophen, 325–650 mg every 4 hours, is effective in reducing fever. These drugs are best administered continuously rather than as needed, since “prn” dosing results in periodic chills and sweats due to fluctuations in temperature caused by varying levels of drug.

2. Antimicrobial therapy—In most febrile patients, empiric antibiotic therapy should be deferred pending further evaluation. However, empiric antibiotic therapy is sometimes warranted. Prompt broad-spectrum antimicrobials are indicated for febrile patients who are clinically unstable, even before infection can be documented. These include patients with hemodynamic instability, those with neutropenia (neutrophils < 500/mcL), others who are asplenic (surgically or secondary to sickle cell disease) or immunosuppressed (including individuals taking systemic corticosteroids, azathioprine, cyclosporine, or other immunosuppressive medications) (Tables 30–4 and 30–5), and those who are HIV infected (see Chapter 31). For treatment of fever during neutropenia following chemotherapy, outpatient parenteral antimicrobial therapy with an agent such as ceftriaxone can be provided effectively and safely. If a fungal infection is suspected in patients with prolonged fever and neutropenia, fluconazole is an equally effective but less toxic alternative to amphotericin B.

C. Treatment of Hyperthermia

Discontinuation of the offending agent is mandatory. Treatment of neuroleptic malignant syndrome includes dantrolene in combination with bromocriptine or levodopa (see Chapter 25). Treatment of serotonin syndrome includes administration of a central serotonin receptor antagonist cyproheptadine or chlorpromazine alone or in combination with a benzodiazepine (see Chapter 25). In patients for whom it is difficult to distinguish which syndrome is present, treatment with a benzodiazepine may be the safest therapeutic option.

When to Refer

Once the diagnosis of FUO is made, referral to infectious disease specialist or rheumatologist may be appropriate to guide specific additional tests.

When to Admit

• Malignant hyperthermia.
• Heat stroke.
• For measures to control temperature when it is > 41°C or when associated with seizure or other mental status changes.

LOWER EXTREMITY EDEMA

ESSENTIAL INQUIRIES

History of venous thromboembolism.
Symmetry.
Pain.
Dependence.
Skin findings.

General Considerations

Acute and chronic lower extremity edema present important diagnostic and treatment challenges. Lower extremities can swell in response to increased venous or lymphatic pressures, decreased intravascular oncotic pressure, increased apillary leak, and local injury or infection. Chronic venous insufficiency is by far the most common cause, affecting up to 2% of the population, and the incidence of venous insufficiency has not changed during the past 25 years. Venous insufficiency is a common complication of DVT; however, only a small number of patients with chronic venous insufficiency report a history of this disorder. Venous ulcer formation commonly affects patients with chronic venous insufficiency, and management of venous ulceration is labor-intensive and expensive. Other causes of lower extremity edema include cellulitis, musculoskeletal disorders (Baker cyst rupture, gastrocnemius tear or rupture), lymphedema, CHF, cirrhosis, and nephrotic syndrome as well as side effects from calcium channel blockers, minoxidil, or thioglitazones.

Clinical Findings

A. Symptoms and Signs

Normal lower extremity venous pressure (in the erect position: 80 mm Hg in deep veins, 20–30 mm Hg in superficial veins) and cephalad venous blood flow require competent bicuspid venous valves, effective muscle contractions, and normal respirations. When one or more of these components fail, venous hypertension may result. Chronic exposure to elevated venous pressure by the postcapillary venules in the legs leads to leakage of fibrinogen and growth factors into the interstitial space, leukocyte aggregation and activation, and obliteration of the cutaneous lymphatic network. These changes account for the brawny, fibrotic skin changes observed in patients with chronic venous insufficiency, and the predisposition toward skin ulceration, particularly in the medial malleolar area.

Among common causes of lower extremity swelling, DVT is the most life-threatening. Clues suggesting DVT include a history of cancer, recent limb immobilization, or confinement to bed for at least 3 days following major surgery within the past month (Table 2–6). A search for alternative explanations is equally important in excluding DVT. Bilateral involvement and significant improvement upon awakening favor systemic causes (eg, venous insufficiency, CHF, and cirrhosis). “Heavy legs” are the most frequent symptom among patients with chronic venous insufficiency, followed by itching. Pain, particularly if severe, is uncommon in uncomplicated venous insufficiency. Lower extremity swelling and inflammation in a limb recently affected by DVT could represent anticoagulation failure and thrombus recurrence but more often are caused by postphlebitic syndrome with valvular incompetence. Other causes of a painful, swollen calf include ruptured popliteal cyst, calf strain or trauma, and cellulitis.

Lower extremity swelling is a familiar complication of therapy with calcium channel blockers (particularly felodipine and amlodipine), thioglitazones, and minoxidil. Bilateral lower extremity edema can be a presenting symptom of nephrotic syndrome or volume overload caused by renal failure. Prolonged airline flights (>10 hours) are associated with increased risk of edema. In those with low to medium risk of thromboembolism (eg, women taking oral contraceptives), long flights are associated with a 2% incidence of asymptomatic popliteal DVT.

B. Physical Examination

Physical examination should include assessment of the heart, lungs, and abdomen for evidence of pulmonary hypertension (primary, or secondary to chronic lung disease), CHF, or cirrhosis. Some patients with cirrhosis have pulmonary hypertension without lung disease. There is a spectrum of skin findings related to chronic venous insufficiency that depends on the severity and chronicity of the disease, ranging from hyperpigmentation and stasis dermatitis to abnormalities highly specific for chronic venous insufficiency: lipodermatosclerosis (thick brawny skin; in advanced cases, the lower leg resembles an inverted champagne bottle) and atrophie blanche (small depigmented macules within areas of heavy pigmentation). The size of both calves should be measured 10 cm below the  tibial tuberosity and elicitation of pitting and tenderness performed. Swelling of the entire leg or swelling of one leg 3 cm more than the other suggests deep venous obstruction. In normal persons, the left calf is slightly larger than the right as a result of the left common iliac vein coursing under the
aorta.

An ulcer located over the medial malleolus is a hallmark of chronic venous insufficiency but can be due to other causes. Shallow, large, modestly painful ulcers arecharacteristic of venous insufficiency, whereas small, deep, and more painful ulcers are more apt to be due to arterial insufficiency, vasculitis, or infection (including cutaneous diphtheria). Diabetic vascular ulcers, however, may be painless. When an ulcer is on the foot or above the mid calf, causes other than venous insufficiency should be considered.

C. Diagnostic Studies

Most causes of lower extremity swelling can be demonstrated with color duplex ultrasonography. Patients without an obvious cause of acute lower extremity swelling (eg, calf strain) should have an ultrasound performed, since DVT is difficult to exclude on clinical grounds. A predictive rule allows a clinician to exclude a lower extremity DVT in patients without an ultrasound if the patient has low pretest probability for DVT and has a negative sensitive D-dimer test (the “Wells rule”). Assessment of the ankle-brachial pressure index (ABPI) is important in the management of chronic venous insufficiency, since peripheral arterial disease may be exacerbated by compression therapy. This can be performed at the same time as ultrasound. Caution is required in interpreting the results of ABPI in older patients and diabetics due to decreased compressibility of their arteries. A dipstick urine test that is strongly positive for protein can suggest nephrotic syndrome, and a serum creatinine can help
estimate kidney function.

Treatment

Treatment of lower extremity edema should be guided by the underlying etiology. See relevant chapters for treatment of edema in patients with CHF (Chapter 10), nephrosis (Chapter 22), cirrhosis (Chapter 16), and lymphedema (Chapter 12). Edema resulting from calcium channel blocker therapy responds to concomitant therapy with ACE inhibitors or angiotensin receptor blockers.

In patients with chronic venous insufficiency without a comorbid volume overload state (eg, CHF), it is best to avoid diuretic therapy. These patients have relatively decreased intravascular volume, and administration of diuretics may result in acute renal insufficiency and oliguria. The most effective treatment involves (1) leg elevation, above the level of the heart, for 30 minutes three to four times daily, and during sleep; (2) compression therapy; and (3) ambulatory exercise to increase venous return through calf muscle contractions. A wide variety of stockings and devices are effective in decreasing swelling and preventing ulcer formation. They should be put on with awakening, before hydrostatic forces result in edema. To control simple edema, 20-30 mm Hg is usually sufficient; whereas, > 30-40 mm Hg is usually required to control moderate to severe edema associated with ulcer formation. Escin, 50 mg orally twice daily (or as horse chestnut seed extract, 300 mg orally twice daily) has been shown in several randomized trials to be equivalent to compression stockings and can be quite useful in nonambulatory patients; however, this medication is not approved by the US Food and Drug Administration. Patients with decreased ABPI should be managed in concert with a vascular surgeon. Compression stockings (12–18 mm Hg at the ankle) are effective in preventing edema and asymptomatic thrombosis associated with long airline flights in low- to medium-risk persons. See Chapter 12 for treatment of venous stasis ulcers.

When to Refer

• Chronic lower extremity ulcerations requiring specialist wound care.
• Nephrotic syndrome should be managed with nephrology consultation.
• When there is coexisting severe arterial insufficiency (claudication) complicating treatment with compression stockings.

When to Admit

• Pending definitive diagnosis in patients at high risk for DVT with normal lower extremity ultrasound.

• Concern for impending compartment syndrome.
• Severe edema that impairs ability to ambulate or perform activities of daily living.

PALPITATIONS

ESSENTIAL INQUIRIES

Forceful, rapid, or irregular beating of the heart.
Rate, duration, and degree of regularity of heart beat.
Age at first episode.
Factors that precipitate or terminate episodes.
Light-headedness or syncope.
Neck pounding.
Chest pain.

General Considerations

Palpitations are defined as an unpleasant awareness of the forceful, rapid, or irregular beating of the heart. They are a common presenting complaint and are usually benign; however, they are occasionally the symptom of a life-threatening arrhythmia. To avoid missing a dangerous cause of the patient’s symptom, clinicians sometimes pursue expensive and invasive testing when a conservative diagnostic evaluation is sufficient. The converse is also true; in one study, 54% of patients with supraventricular tachycardia were initially wrongly diagnosed with panic, stress, or anxiety disorder. A disproportionate number of these misdiagnosed patients are women.

Clinical Findings

A. Symptoms

Although described by patients in a myriad of ways, guiding the patient through a careful description of their palpitations may indicate a mechanism and narrow the differential diagnosis. Pertinent questions include the age at first episode; precipitants; and the rate, duration, and degree of regularity of the heart beat during the subjective palpitations. The examiner can ask the patient to “tap out” the rhythm with their fingers. The circumstances associated with onset and termination can also be helpful in determining the cause. Palpitations that start and stop abruptly suggest supraventricular or ventricular tachycardias. Patient-terminated palpitations using vagal maneuvers (such as the Valsalva maneuver) suggests supraventricular tachycardia.

Three common descriptions of palpitations are (1) “flip-flopping” (or “stop and start”), often caused by premature contraction of the atrium or ventricle, with the perceived “stop” from the pause following the contraction, and the “start” from the subsequent forceful contraction; (2) rapid “fluttering in the chest,” with regular “fluttering” suggesting supraventricular or ventricular arrhythmias (including sinus tachycardia) and irregular “fluttering” suggesting atrial fibrillation, atrial flutter, or tachycardia with variable block; and (3) “pounding in the neck” or neck pulsations, often due to “cannon” A waves in the jugular venous pulsations that occur when the right atrium contracts against a closed tricuspid valve.

Palpitations associated with chest pain suggests ischemic heart disease, or if the chest pain is relieved by leaning forward, pericardial disease is suspected. Palpitations associated with light-headedness, presyncope, or syncope suggests hypotension and may signify a life-threatening cardiac arrhythmia. Palpitations that occur regularly with exertion suggests a rate-dependent bypass tract or hypertrophic
cardiomyopathy. If a benign etiology for these concerning symptoms cannot be ascertained at the index visit, then ambulatory monitoring or prolonged cardiac monitoring in the hospital might be warranted.

Noncardiac symptoms should also be elicited since the palpitations may be caused by a normal heart responding to a metabolic or inflammatory condition. Weight loss may suggest hyperthyroidism. Palpitations can be precipitated by vomiting or diarrhea that leads to electrolyte disorders and hypovolemia. Palpitations associated with hyperventilation, hand tingling, and nervousness are  common when anxiety or panic disorder is the root cause.

B. Physical Examination

It is uncommon for the clinician to have the opportunity to examine a patient during an episode of palpitations.

However, careful cardiovascular examination can find abnormalities that can increase the likelihood of specific cardiac arrhythmias. The midsystolic click of mitral valve prolapse can suggest the diagnosis of a supraventricular arrhythmia as the cause for the palpitations. The harsh holosystolic murmur of hypertrophic cardiomyopathy, which occurs along the left sternal border and increases with the Valsalva maneuver, suggests atrial fibrillation or ventricular tachycardia. The presence of dilated cardiomyopathy, suggested on examination by a displaced and enlarged cardiac point-of-maximal impulse, increases the likelihood of ventricular tachycardia and atrial fibrillation. In patients with chronic atrial fibrillation, in-office exercise (eg, a brisk walk in the hallway) may reveal an intermittent accelerated ventricular response as the cause of the palpitations. The clinician should also look for signs of hyperthyroidism, such as tremulousness, brisk deep tendon reflexes, fine hand tremor, or signs of stimulant drug use (such as dilated pupils or skin or nasal septal lesions). The presence of visible neck pulsations (LR, 2.68; 95% CI, 1.25–5.78) inassociation with palpitations increases the likelihood of  atrioventricular nodal reentry tachycardia.

C. Diagnostic Studies

The two cardiac studies that are commonly used in the initial evaluation of a patient with palpitations are the 12-lead ECG and ambulatory monitoring devices, such as the Holter monitor or the event recorder.

A 12-lead ECG should be performed on all patients reporting palpitations because it can provide evidence for a wide variety of causes. Although in most instances a specific arrhythmia will not be detected on the tracing, a careful evaluation of the ECG can help the clinician deduce a likely etiology in certain circumstances.

For instance, bradyarrhythmias and heart block can be associated with ventricular ectopy or escape beats that may be experienced as palpitations by the patient. Evidence of prior myocardial infarction by history or on ECG (eg, Q waves) increases the patient’s risk for nonsustained or sustained ventricular tachycardia. Ventricular preexcitation (Wolff-Parkinson-White syndrome) is suggested by a short PR interval (< 0.20 ms) and delta waves (upsloping PR segments). Left ventricular hypertrophy with deep septal Q waves in I, VL, and V4 through V6 is seen in patients with  hypertrophic obstructive cardiomyopathy. The presence of left atrial enlargement as suggested by a terminal p-wave force in V1 more negative than 0.04 msec and notched in lead II reflects a patient at increased risk for atrial fibrillation. A prolonged QT interval and abnormal T-wave morphology suggests the long-QT syndrome, which puts patients at increased risk for ventricular tachycardia.

For high-risk patients (Table 2–5), further diagnostic studies are warranted. A step-wise approach has been suggested—starting with ambulatory monitoring devices (Holter monitoring if the palpitations are expected tooccur within the subsequent 72-hour period, event monitoring if less frequent), followed by invasive electrophysiologic testing if the ambulatory monitor records a worrisome arrhythmia or if serious arrhythmias are strongly suspected despite normal findings on an appropriate ambulatory monitor.

In patients with a prior myocardial infarction, ambulatorycardiac monitoring or signal-averaged-ECG are appropriate next steps to assess ventricular tachycardia. ECG exercise testing is appropriate in patients who have palpitations with physical exertion and patients with suspected coronary artery disease. Echocardiography is useful when physical examination or ECG suggests structural abnormalities or decreased ventricular function.

Differential Diagnosis

When assessing a patient with palpitations in an urgent care setting, the clinician must ascertain whether the symptoms represent (1) an arrhythmia that is minor and transient, (2) significant cardiovascular disease, (3) a cardiac manifestation of a systemic disease such as thyrotoxicosis, or (4)
a benign somatic symptom that is amplified by underlying psychosocial characteristics of the patient.

Palpitations in patients with a known history of cardiac disease or palpitations that occur during sleep increase the ikelihood of a cardiac arrhythmia. A history of panic disorder  or palpitations that last < 5 minutes make a cardiac arrhythmia slightly less likely. Patients who seek medical attention in the emergency department instead of a medical clinic are more likely to have a cardiac etiology (47% versus 21%), while psychiatric causes are more common among patients with palpitations who seek medical attention in office practices (45% versus 27%). In a study of patients who went to a university medical clinic with the chief complaint of palpitations, etiologies were cardiac in 43%, psychiatric in 31%, and miscellaneous in 10% (including illicit drugs, medications, anemia, thyrotoxicosis, and mastocytosis).

Cardiac arrhythmias that can result in symptoms of palpitations include sinus bradycardia; sinus, supraventricular, and ventricular tachycardia; premature ventricular and atrial contractions; sick sinus syndrome; and advanced atrioventricular block.

Nonarrhythmic cardiac causes of palpitations include valvular heart diseases, such as aortic insufficiency or stenosis, atrial or ventricular septal defect, cardiomyopathy, congenital heart disease, and pericarditis.

Noncardiac causes of palpitations include fever, dehydration, hypoglycemia, anemia, thyrotoxicosis, and pheochromocytoma. Drugs such as cocaine, alcohol, caffeine, and pseudoephedrine can precipitate palpitations, as can prescription medications, including digoxin, phenothiazines, theophylline, and b-agonists as well as ephedra-containing herbal remedies or supplements, which are  banned by the US Food and Drug Administration.

The most common psychiatric causes of palpitations are anxiety and panic disorder. The release of catecholamines during a panic attack or significant stress can trigger an arrhythmia. Asking a single question, “Have you experienced brief periods, for seconds or minutes, of an overwhelming panic or terror that was accompanied by racing heartbeats, shortness of breath, or dizziness?” can help identify patients with panic disorder.

Treatment

After ambulatory monitoring, most patients with palpitations are found to have benign atrial or ventricular ectopy and nonsustained ventricular tachycardia. In patients with structurally normal hearts, these arrhythmias are not associated with adverse outcomes. Abstention from caffeine and tobacco may help. Often, reassurance suffices. If not, or in very symptomatic patients, a trial of a b-blocker may be prescribed. For treatment of specific atrial or ventricular arrhythmias, see Chapter 10.

When to Refer

• For electrophysiologic studies.
• For advice regarding treatment of atrial or ventricular arrhythmias.

When to Admit

• Palpitations associated with syncope or near-syncope, particularly when the patient is aged 75 years or older and has an abnormal ECG, hematocrit < 30%, shortness of breath, respiratory rate > 24/min, or a history of CHF.
• Patients with risk factors for a serious arrhythmia.

CHEST PAIN

Chest pain onset, character, location/size, duration,
periodicity, and exacerbators.
Shortness of breath.
Vital signs.
Chest and cardiac examination.
Electrocardiography.
Biomarkers of myocardial necrosis.

General Considerations

Chest pain (or chest discomfort) is a common symptom that can occur as a result of cardiovascular, pulmonary, pleural, or musculoskeletal disease, esophageal or othergastrointestinal disorders, or anxiety states. The frequency and distribution of life-threatening causes of chest pain, such as acute coronary syndrome (ACS), pericarditis, aortic dissection, pulmonary embolism, pneumonia, and esophageal perforation, vary substantially between clinical settings. Systemic lupus erythematosus, rheumatoid arthritis, and HIV are conditions that confer a strong risk for coronary artery disease. Because pulmonary embolism can present with a wide variety of symptoms, consideration of the diagnosis and rigorous risk factor assessment for venous thromboembolism (VTE) is critical. Classic VTE risk factors include cancer, trauma, recent surgery, prolonged immobilization, pregnancy, oral contraceptives, and family history and prior history of VTE. Other conditions associated with increased risk of pulmonary embolism include CHF and COPD. Sickle cell anemia can cause acute chest syndrome. Patients with this syndrome often have chest pain, fever, and cough.

Although uncommon in the office setting, delays in diagnosing life-threatening causes of chest pain can result in serious morbidity and mortality.

Clinical Findings

A. Symptoms

Myocardial ischemia is usually described as dull, aching sensation of “pressure,” “tightness,” “squeezing,” or “gas,” rather than as sharp or spasmodic. Ischemic symptoms usually subside within 5–20 minutes but may last longer. Progressive symptoms or symptoms at rest may represent unstable angina. Prolonged chest pain episodes might represent myocardial infarction, although up to  one-third of patients with acute myocardial infarction do not report chest pain. When present, pain due to myocardial ischemia is commonly accompanied by a sense of anxiety or uneasiness. The location is usually retrosternal or left precordial. Because the heart lacks somatic innervation, precise localization of pain due to cardiac ischemia is difficult; the pain is commonly referred to the throat, lower jaw, shoulders, inner arms, upper abdomen, or back. Ischemic pain may be precipitated or exacerbated by exertion, cold temperature, meals, stress, or combinations of these factors and is usually relieved by rest. However, many episodes do not conform to these patterns; and atypical presentations of ACS are more common in the elderly, women, and persons with diabetes. Other symptoms that are associated with ACS include shortness of breath; dizziness; a feeling of impending doom; and vagal symptoms, such as nausea and diaphoresis. In the elderly, fatigue is a common presenting complaint of ACS. Likelihood ratios for cardinal symptoms considered in the evaluation of acute myocardial infarction are summarized in Table 2–4.

Hypertrophy of either ventricle or aortic stenosis may also give rise to chest pain with less typical features. Pericarditis may produce pain that is greater when supine than upright and may increase with respiration, coughing, or swallowing. Pleuritic chest pain is usually not ischemic, and pain on palpation may indicate a musculoskeletal cause. Aortic dissection classically produces an abrupt onset of tearing pain of great intensity that often radiates to the back; however, this classic presentation occurs in a small proportion of cases. Anterior aortic dissection can also lead to myocardial or cerebrovascular ischemia.

Pulmonary embolism has a wide range of clinical presentations, with chest pain present in only 75% of cases. The chief objective in evaluating patients with suspected pulmonary embolism is to assess the patient’s clinical risk for VTE  zased on medical history and associated signs and symptoms (see above and Chapter 9). Esophageal perforation of the thoracic region is another cause of chest pain, with most cases resulting from medical procedures of the esophagus.

B. Physical Examination

Findings on physical examination can occasionally yield important clues to the underlying cause of chest pain; however, a normal physical examination should never be used as the sole basis for ruling-out most diagnoses, particularly ACS and aortic dissection. Vital sign measurement, including pulse oximetry, is always the first step for assessing the urgency and tempo of the subsequent examination and diagnostic work-up.

Findings that increase the likelihood of ACS include diaphoresis, hypotension, S3 or S4 gallop, pulmonary crackles, or elevated jugular venous pressure (see Table 2–4). Although chest pain that is reproducible or worsened with palpation strongly suggests a musculoskeletal cause, up to 15% of patients with ACS will have reproducible chest wall tenderness. Pointing to the location of the pain with one finger has been shown to be highly correlated with nonischemic chest pain. Aortic dissection can result in differential blood pressures (> 20 mm Hg), pulse amplitude deficits, and new diastolic murmurs. Although hypertension is considered the rule in patients with aortic dissection, systolic blood pressure < 100 mm Hg is present in up to 25% of patients.

A cardiac friction rub represents pericarditis until proven otherwise. It can best be heard with the patient sitting forward at end-expiration. Tamponade should be excluded in all patients with a clinical diagnosis of pericarditis by assessing pulsus paradoxus (a decrease in systolic blood pressure during inspiration > 10 mm Hg) and inspection of jugular venous pulsations. Subcutaneous emphysema is common following cervical esophageal perforation but present in only about one-third of thoracic perforations (ie, those most commonly presenting with chest pain).

The absence of physical examination findings in patients with suspected pulmonary embolism usually serves to ncrease the likelihood of pulmonary embolism, although a normal physical examination is also compatible with the much more common conditions of panic/anxiety disorder and musculoskeletal disease.

C. Diagnostic Studies

Unless a competing diagnosis can be confirmed, an ECG is warranted in the initial evaluation of most patients with acute chest pain to help exclude ACS. ST segment elevation is the ECG finding that is the strongest predictor of acute myocardial infarction (see Table 2–4); however, up to 20% of patients with ACS can have a normal ECG. A 2007 study concluded that patients with suspected ACS can be safely removed from cardiac monitoring in the emergency department if they are pain-free at initial physician assessment and have a normal or nonspecific ECG. This decision rule had 100% sensitivity for serious arrhythmia (95% confidence interval, 80–100%), but deserves further validation. Clinically stable patients with cardiovascular disease risk factors, normal ECG, normal cardiac biomarkers, and no alternative diagnoses (such as typical GERD or costochondritis) should be followed-up with a timely exercise stress test that includes perfusion imaging. The ECG can also provide evidence for alternative diagnoses, such as pericarditis and pulmonary embolism. Chest radiography is often useful in the evaluation of chest pain, and is always indicated when cough or shortness of breath accompanies chest pain. Findings of pneumomediastinum or new pleural effusion are consistent with esophageal perforation. Sixty-four–slice CT coronary angiography has been studied for diagnosing acute coronary syndromes and predicting outcomes in the emergency department. One study of 58 patients found a high positive predictive value for diagnosing acute coronary syndromes, and a negative predictive value of 97%. Another study found that it compared favorably with stress nuclear imaging in the detection and exclusion of ACS in low-risk chest pain patients. CT is the study of choice at most centers for the diagnosis of esophageal perforation as well as for aortic dissection (helical CT). Current generation highly-sensitive assays for cardiac troponin I have been studied in an effort to more rapidly prognosticate outcomes in patients presenting with suspected ACS. Patients presenting to the emergency department with chest pain of intermediate or high probability for ACS without electrocardiographic or biomarker evidence of a myocardial infarction can be safely discharged from an observation unit after stress cardiac magnetic imaging.

The role of CT coronary calcium assessment in the evaluation of chest pain in the emergency department is still being defined. A study suggests that the coronary artery calcium score appears to be an effective initial tool for risk stratification of low- to intermediate-risk patients with possible acute coronary syndromes, on the basis of its high negative predictive value and additive diagnostic value.

In the evaluation of pulmonary embolism, diagnostic test decisions and results must be interpreted in the context f the clinical likelihood of VTE. A negative D-dimer test is helpful for excluding pulmonary embolism in patients with low clinical probability of VTE (3 month incidence = 0.5%); however, the 3-month risk of VTE among patients with intermediate and high risk of VTE is sufficiently high in the setting of a negative D-dimer test (3.5% and 21.4%, respectively) to warrant further imaging given the lifethreatening nature of this condition if left untreated. CT angiography (with helical or multidetector CT imaging) has replaced ventilation-perfusion scanning as the preferred diagnostic test for pulmonary embolism, having approximately 90–95% sensitivity and 95% specificity for detecting pulmonary embolism (compared with pulmonary angiography). However, according to guidelines published by the American Academy of Family Physicians
and the American College of Physicians (AAFP/ACP), the sensitivity of helical CT is probably not sufficiently high to exclude pulmonary embolism among patients with high clinical probability of VTE, in whom lower extremity ultrasound or pulmonary angiogram may be appropriate. When ventilation-perfusion scanning is performed, only results that are normal or high probability are useful for improving one’s clinical assessment of pulmonary embolism. Panic disorder is a common cause of chest pain, accounting for up to 25% of cases that present to emergency departments and a higher proportion of cases presenting in primary care office practices. Features that correlate with an increased likelihood of panic disorder include absence of coronary artery disease, atypical quality of chest pain, female sex, younger age, and a high level of self-reported anxiety.

Treatment

Treatment of chest pain should be guided by the underlying etiology. The term “noncardiac chest pain” is used to describe patients who evade diagnosis after receiving extensive work-up. Although under-studied, one small trial of patients with noncardiac chest pain found that about half fulfilled criteria for anxiety or depression and almost half reported symptom improvement with high-dose
proton-pump inhibitor therapy.

When to Refer

• Refer patients with poorly controlled, noncardiac chest pain to a pain specialist.
• Refer patients with sickle cell anemia to a hematologist.

When to Admit

• Failure to adequately exclude (to a sufficient degree) life-threatening causes of chest pain, particularly myocardial infarction, dissecting aortic aneurysm, pulmonary embolism, and esophageal rupture.
• Pain control for rib fracture that impairs gas exchange.

HEMOPTYSIS

 ESSENTIAL INQUIRIES

Smoking history.
Fever, cough, and other symptoms of lower respiratorytract infection.
Nasopharyngeal or gastrointestinal bleeding.
Chest radiography and complete blood count.

General Considerations

 

Hemoptysis is the expectoration of blood that originates below the vocal cords. It is commonly classified as trivial, mild, or massive—the latter defined as more than 200–600 mL (about 1–2 cups) in 24 hours. The dividing lines are arbitrary, since the amount of blood is rarely quantified with precision. Massive hemoptysis can be usefully defined as any amount that is hemodynamically significant or threatens ventilation, in which case the initial management goal is not diagnostic but therapeutic.

The lungs are supplied with a dual circulation. The pulmonary arteries arise from the right ventricle to supply the pulmonary parenchyma in a low-pressure circuit. The bronchial arteries arise from the aorta or intercostal arteries and carry blood under systemic pressure to the airways, blood vessels, hila, and visceral pleura. Although the bronchial circulation represents only 1–2% of total pulmonary
blood flow, it can increase dramatically under conditions of chronic inflammation—eg, chronic bronchiectasis— and is frequently the source of hemoptysis.

The causes of hemoptysis can be classified anatomically. Blood may arise from the airways in COPD, bronchiectasis, and bronchogenic carcinoma; from the pulmonary vasculature in left ventricular failure, mitral stenosis, pulmonary embolism, idiopathic pulmonary arterial hypertension, and arteriovenous malformations; or from the pulmonary parenchyma in pneumonia, inhalation of crack cocaine, or granulomatosis with polyangiitis (formerly Wegener granulomatosis). Diffuse alveolar hemorrhage is due to small vessel bleeding usually caused by autoimmune or hematologic disorders and results in alveolar infiltrates on chest radiography. Most cases of hemoptysis presenting in the outpatient setting are due to infection (eg, acute or chronic bronchitis, pneumonia, tuberculosis). Hemoptysis due to lung cancer increases with age, accounting for up to 20% of cases among the elderly. Less commonly (< 10% of cases), pulmonary venous hypertension (eg, mitral stenosis, pulmonary embolism) causes hemoptysis. Most cases of hemoptysis that have no visible cause on CT scan or bronchoscopy will resolve within 6 months without treatment, with the notable exception of patients at high risk for lung cancer (smokers older than 40 years). Iatrogenic hemorrhage may follow transbronchial lung biopsies, anticoagulation, or pulmonary artery rupture due to distal placement of a balloon-tipped catheter. No cause is identified in up to 15–30% of cases.

Clinical Findings

A. Symptoms

Blood-tinged sputum in the setting of an upper respiratory tract infection in an otherwise healthy, young (age < 40 years) nonsmoker does not warrant an extensive diagnostic evaluation if the hemoptysis subsides with resolution of the infection. However, hemoptysis is frequently a sign of serious disease, especially in patients with a high prior probability of underlying pulmonary pathology. One should not distinguish between blood-streaked sputum and cough productive of blood alone with regard to the evaluation plan. The goal of the history is to identify patients at risk for one of the disorders listed above. Pertinent features include past or current tobacco use, duration of symptoms, and the presence of respiratory infection. Nonpulmonary sources of hemorrhage from the nose or the gastrointestinal tract should also be excluded.

B. Physical Examination

Elevated pulse, hypotension, and decreased oxygen saturation suggest large volume hemorrhage that warrants emergent evaluation and stabilization. The nares and oropharynx should be carefully inspected to identify a potential upper airway source of bleeding. Chest and cardiac examination may reveal evidence of CHF or mitral stenosis.

C. Diagnostic Studies

Diagnostic evaluation should include a chest radiograph and complete blood count. Kidney function tests, urinalysis, and coagulation studies are appropriate in specific circumstances. Hematuria that accompanies hemoptysis may be a clue to Goodpasture syndrome or vasculitis. Flexible bronchoscopy reveals endobronchial cancer in 3–6% of patients with hemoptysis who have a normal (non-lateralizing) chest radiograph. Nearly all of these patients are smokers over the age of 40, and most will have had symptoms for more than 1 week. Bronchoscopy is indicated in such patients. High-resolution chest CT scan complements bronchoscopy and should be strongly considered in patients with normal chest radiograph and low risk for malignancy. It can visualize unsuspected bronchiectasis and arteriovenous malformations and will show central endobronchial lesions in many cases. High-resolution chest CT scanning is the test of choice for suspected small peripheral malignancies. Helical CT pulmonary angiography has become the initial test of choice for evaluating patients with suspected pulmonary embolism, although caution should be taken to avoid large contrast loads in patients with even mild chronic kidney disease (serum creatinine > 2.0 g/dL or rapidly rising creatinine in normal range). Helical CT scanning can be avoided in patients who are at “unlikely” risk for pulmonary embolism using the Wells clinical decision rule and who have a normal sensitive D-dimer test.

Treatment

The management of mild hemoptysis consists of identifying and treating the specific cause. Massive hemoptysis is life-threatening. The airway should be protected with endotracheal intubation, ventilation ensured, and effective circulation maintained. If the location of the bleeding site is known, the patient should be placed in the decubitus position with the involved lung dependent. Uncontrollable hemorrhage warrants rigid bronchoscopy and surgical consultation. In stable patients, flexible bronchoscopy may localize the site of bleeding, and angiography can embolize the involved bronchial arteries. Embolization is effective initially in 85% of cases, although rebleeding may occur in up to 20% of patients over the following year. The anterior spinal artery arises from the bronchial artery in up to 5% of people, and paraplegia may result if it is inadvertently cannulated.

When to Refer

• When bronchoscopic evaluation of lower respiratory tract is required, refer patients to a pulmonologist.
• Patients should be referred to an otolaryngologist for evaluation of upper respiratory tract bleeding source.
• Patients with severe coagulopathy complicating management should be referred to a hematologist.

When to Admit

• To stabilize bleeding process in patients at risk for or experiencing massive hemoptysis.
• To correct disordered coagulation (clotting factors or platelets, or both).
• To stabilize gas exchange.

DYSPNEA

ESSENTIAL INQUIRIES

Fever.
Cough.
Chest pain.
Vital sign measurements; pulse oximetry.
Cardiac and chest examination.
Chest radiography.
Arterial blood gas measurement.

General Considerations

Dyspnea is a subjective experience or perception of uncomfortable breathing. However, the relationship between level of dyspnea and the severity of underlying disease varies widely across individuals. Dyspnea can result from conditions that increase the mechanical effort of breathing (eg, COPD, restrictive lung disease, respiratory muscle weakness), from conditions that produce compensatory tachypnea (eg, hypoxemia or acidosis), or from psychogenic origins. The following factors play a role in how and when dyspnea presents in patients: rate of onset, previous dyspnea, medications, comorbidities, psychological profile, and the severity of underlying disorder. In patients with established COPD, the patient-reported severity of dyspnea is superior to forced expiratory volume in 1 second (FEV1) in predicting quality of life and 5-year mortality.

Clinical Findings

A. Symptoms

The duration, severity, and periodicity of dyspnea influence the tempo of the clinical evaluation. Rapid onset, severe dyspnea in the absence of other clinical features should raise concern for pneumothorax, pulmonary embolism, or increased left ventricular end-diastolic pressure (LVEDP). Spontaneous pneumothorax is usually accompanied by chest pain and occurs most often in thin, young males, or in those with underlying lung disease. Pulmonary embolism should always be suspected when a patient with new dyspnea reports a recent history (previous 4 weeks) of prolonged immobilization, estrogen therapy, or other risk factors for deep venous thrombosis (DVT) (eg, previous history of thromboembolism, cancer, obesity, lower extremity trauma) and when the cause of dyspnea is not apparent. Silent myocardial infarction, which occurs more frequently in diabetic persons and women, can result in acute heart failure and dyspnea.

Accompanying symptoms provide important clues to various etiologies of dyspnea. When cough and fever are present, pulmonary disease (particularly infections) is the primary concern, although myocarditis, pericarditis, and septic emboli can also present in this manner. Chest pain should be further characterized as acute or chronic, pleuritic or exertional. Although acute pleuritic chest pain is
the rule in acute pericarditis and pneumothorax, must patients with pleuritic chest pain in the outpatient clinic have pleurisy due to acute viral respiratory tract infection. Periodic chest pain that precedes the onset of dyspnea is suspicious for myocardial ischemia as well as pulmonary embolism. When associated with wheezing, most cases of dyspnea are due to acute bronchitis; however, when acute bronchitis seems unlikely, the clinician should also consider new-onset asthma, foreign body, and vocal cord dysfunction.

When a patient reports prominent dyspnea with mild or no accompanying features, consider noncardiopulmonary causes of impaired oxygen delivery (anemia, methemoglobinemia, cyanide ingestion, carbon monoxide), metabolic acidosis due to a variety of conditions, panic disorder, and chronic pulmonary embolism.

B. Physical Examination

A focused physical examination should include evaluation of the head and neck, chest, heart, and lower extremities. Visual inspection of the patient’s respiratory pattern can suggest obstructive airway disease (pursed-lip breathing, use of extra respiratory muscles, barrel-shaped chest), pneumothorax (asymmetric excursion), or metabolic acidosis (Kussmaul respirations). Patients with impending upper airway obstruction (eg, epiglottitis, foreign body), or severe asthma exacerbation, sometimes assume a tripod position. Focal wheezing raises the suspicion for a foreign body or other bronchial obstruction. Maximum laryngeal height (the distance between the top of the thyroid cartilage and the suprasternal notch at end expiration) is a measure of hyperinflation. Obstructive airway disease is virtually nonexistent when a nonsmoking patient younger than 45 years has a maximum laryngeal height ≤ 4 cm (Table 2–2). Absent breath sounds suggests a pneumothorax. An accentuated pulmonic component of the second heart sound (loud P2) is a sign of pulmonary hypertension and pulmonary embolism.

A systematic review has identified several clinical predictors of increased LVEDP useful in the evaluation of dyspneic patients with no prior history of CHF (Table 2–3). When none is present, there is a very low probability (< 10%) of increased LVEDP, and when two or more are present, there is a very high probability (> 90%) of increased LVEDP.

C. Diagnostic Studies

Causes of dyspnea that can be managed without chest radiography are few: ingestions causing lactic acidosis, methemoglobinemia, and carbon monoxide poisoning. The diagnosis of pneumonia should be confirmed by chest radiography in most patients. When COPD exacerbation is severe enough to require hospitalization, results of chest radiography can influence management decisions in up to 20% of patients. Chest radiography is fairly sensitive and specific for new-onset CHF (represented by redistribution of pulmonary venous circulation) and can help guide treatment decisions in patients with dyspnea secondary to cardiac disease. End-expiratory chest radiography enhances detection of a small pneumothorax.

A normal chest radiograph has substantial diagnostic value. When there is no physical examination evidence of COPD or CHF and the chest radiograph is normal, the major remaining causes of dyspnea include pulmonary embolism, Pneumocystis jiroveci infection (initial radiograph may be normal in up to 25% ), upper airway obstruction, foreign body, anemia, and metabolic acidosis. If a patient has tachycardia and hypoxemia but a normal chest radiograph and electrocardiogram (ECG), then further tests to exclude pulmonary emboli are warranted (see Chapter 9), provided blood tests exclude significant anemia or metabolic acidosis. High-resolution chest CT is particularly useful in the evaluation of pulmonary embolism and has the added benefit of providing information about interstitial and alveolar lung disease.

Serum or whole blood brain natriuretic peptide (BNP or NT-proBNP) testing can be useful in distinguishing cardiac from noncardiac causes of dyspnea in the emergency department, since elevated BNP levels are both sensitive and specific for increased LVEDP in symptomatic persons. Cut points for ruling out acute heart failure increase with age, with one study recommending < 300  g/L to rule out heart failure, and three cut points for ruling-in heart failure (> 450 ng/L for age < 50 years; > 900 ng/L for age 50–75 years; and > 1800 ng/L for age > 75 years).

Persistent uncertainty following clinical examination and routine diagnostic testing warrants arterial blood gas measurement. With two notable exceptions (carbon monoxide poisoning and cyanide toxicity), arterial blood gas measurement distinguishes increased mechanical effort causes of dyspnea (respiratory acidosis with or without hypoxemia) from compensatory tachypnea (respiratory alkalosis with or without hypoxemia or metabolic acidosis) from psychogenic dyspnea (respiratory alkalosis). Carbon monoxide and cyanide impair oxygen delivery with minimal alterations in Po2; percent carboxyhemoglobin identifies carbon monoxide toxicity. Cyanide poisoning should be considered in a patient with profound lactic acidosis following exposure to burning vinyl (such as a theater fire or industrial accident). Suspected carbon monoxide poisoning or methemoglobinemia can also be confirmed with venous carboxyhemoglobin or methemoglobin levels.

Because arterial blood gas testing is impractical in most outpatient settings, pulse oximetry has assumed a central role in the office evaluation of dyspnea. Oxygen saturation values above 96% almost always correspond with a Po2 > 70 mm Hg, and values < 94% almost always represent clinically significant hypoxemia. Important exceptions to this rule include carbon monoxide toxicity, which leads to a normal oxygen saturation (due to the similar wavelengths of oxyhemoglobin and carboxyhemoglobin), and methemoglobinemia, which results in an oxygen saturation of about 85% that fails to increase with supplemental oxygen. A delirious or obtunded patient with obstructive lung disease warrants immediate measurement of arterial blood gases to exclude hypercapnia and the need for intubation, regardless of the oxygen saturation. When pulse oximetry yields equivocal results, assessment of desaturation with ambulation (eg, a brisk walk around the clinic) can be a useful finding (eg, when Pneumocystis jiroveci pneumonia is suspected) for confirming impaired gas exchange.

Episodic dyspnea can be challenging if an evaluation cannot be performed during symptoms. Life-threatening causes include recurrent pulmonary embolism, myocardial ischemia, and reactive airway disease. When associated with audible wheezing, vocal cord dysfunction should be considered, particularly in a young woman who does not respond to asthma therapy. Spirometry is very helpful in further classifying patients with obstructive airway disease but is rarely needed in the initial or emergent evaluation of patients with acute dyspnea.

Differential Diagnosis

Acute dyspnea, particularly as the chief complaint, demands urgent evaluation. Urgent and emergent conditions causing acute dyspnea include pneumonia, COPD, asthma, pneumothorax, pulmonary embolism, cardiac disease (eg, CHF, acute myocardial infarction, valvular dysfunction, arrhythmia, cardiac shunt), metabolic acidosis, cyanide toxicity, methemoglobinemia, and carbon monoxide poisoning. 

Treatment

The treatment of urgent or emergent causes of dyspnea should aim to relieve the underlying cause. Pending diagnosis, patients with hypoxemia should be immediately provided supplemental oxygen unless significant hypercapnia is present or strongly suspected pending arterial blood gas measurement. Dyspnea frequently occurs in patients nearing the end of life; whereas opioid therapy can provide substantial relief independent of the severity of hypoxemia, oxygen therapy appears to be most beneficial to patients with significant hypoxemia (see Chapter 5). There appears to be limited benefit of supplemental oxygen when patients with life-limiting illness have Pao2 > 55 mm Hg. In patients with severe COPD and hypoxemia, oxygen therapy improves mortality and exercise performance. Current clinical trials are attempting to determine benefits of supplemental oxygen in patients with COPD and moderate hypoxemia at rest or desaturation with exercise. Pulmonary rehabilitation programs are another therapeutic option for patients with moderate to severe COPD or interstitial pulmonary fibrosis.

When to Refer

• Patients with advanced COPD should be referred to a pulmonologist, and patients with CHF or valvular heart disease should be referred to a cardiologist following acute stabilization.
• Cyanide toxicity should be managed in conjunction with a toxicologis.

When to Admit

• Impaired gas exchange from any cause or high risk of pulmonary embolism pending definitive diagnosis.
• Suspected cyanide poisoning.