By Dr. Stuart M. Lane
Posted: Friday, January 1, 1999
ARTICLES
Publication Date: January 1, 1999
Adam Stern, a 48-year-old teacher, went to the emergency room of Community Hospital after he'd awakened in the middle of the night in a pool of sweat and with severe pain in the center of his chest. Over the previous few months, usually after eating his evening meal, he had experienced brief episodes of similar chest pressure and pain, and liquid antacids had not relieved them. Mr. Stern had been smoking two packs of cigarettes daily for 10 years and had been first diagnosed with essential hypertension (an elevation in arterial blood pressure whose cause is unknown) two years before. Sedentary, he had gained 10 pounds during the previous year.
On examination in the emergency room, Mr. Stern's pain subsided, and neither his chest nor his abdomen was tender. But he remained anxious and sweaty. His blood pressure was 170/100 mm Hg (millimeters mercury), and his pulse was 110 beats per minute. According to Mr. Stern's history, his father had died at age 64 of a heart attack, and an uncle and brother of his suffered from angina pectoris—severe, spasmodic chest pain associated with an undersupply of blood to the heart.
Mr. Stern's history and findings from his physical prompted a resident to order electrocardiography and a blood test for cardiac enzymes. The resultant findings did not suggest that Mr. Stern had had a heart attack. He was hospitalized for observation. Serial electrocardiograms were unremarkable, and his cardiac enzyme values remained normal. Thus, a heart attack—clinically referred to as an acute myocardial infarction (AMI)—was ruled out. The cause of Mr. Stern's chest pain, however, was not determined. He was discharged the next day.
Chest Pain and Heart Attacks
Chest pain is a common complaint among adults and may be a symptom of a lethal illness. About 5 million persons a year visit American ERs because of acute chest pain. AMI is ruled out in more than half of patients with chest pain who are admitted to coronary care units for evaluation. Among the many possible causes of chest pain are anxiety disorders, chest-wall muscle strain, cholecystitis (inflammation of the gallbladder), cocaine use, gastroesophageal reflux disease (GERD) [see "Heartburn: Reflections on an 'Old Flame'" Priorities, Vol. 9, No. 1, 1997, page 39], ischemic heart disease (characterized by narrowing of the lumen of coronary arteries), lung cancer, peptic ulcers, pneumonia, and various musculoskeletal disorders.
Every year in the United States, approximately 650,000 patients are diagnosed with AMI, and the condition may go undiagnosed in some 25,000 ER patients. Because many disorders can cause chest pain, deciding whether to hospitalize a patient whose symptoms and history suggest an AMI is often difficult. Reportedly, AMI in ER patients goes undiagnosed 1—8 percent of the time, and undiagnosed AMI accounts for the largest percentage of malpractice claims in emergency medicine. According to findings from some studies, approximately 2 percent of patients suffering a myocardial infarction or unstable angina are mistakenly discharged. On the other hand, hospitalization in the U.S. to rule out AMI in patients who prove not to have recently had an AMI costs roughly $1.5—3.5 billion annually.
Science, Art, and "Managed Care"
Other fairly common disorders that are diagnostically challenging include colorectal cancer, multiple sclerosis, osteoarthritis, psychosis, and temporomandibular joint (TMJ) disease (a jaw condition). Such disorders are not inherently elusive—each has distinctive clinical characteristics. Diagnosing them tends to be difficult because of fundamental imperfections in American healthcare.
As high-tech as American healthcare has become, with such advances as in biophysical imaging techniques, medicine continues to be both a science—a system by which biomedical knowledge is brought to bear—and an art—an activity wherein whether the practitioner succeeds or fails may well depend on how experienced, observant, heedful, discerning, and creative he or she is. For example, how therapeutically useful data from high-tech diagnostic tests are depends on whether the physician ordered the appropriate tests; whether the physician orders the appropriate tests may depend on whether he or she has obtained adequate information from the patient; and obtaining adequate information from the patient may depend on whether the physician has spent enough time with the patient or on whether the patient trusts the physician. By personally seeking information from patients about such areas as family, schooling, hopes, and fears, the physician may not only obtain particulars crucial to diagnosis—he or she may also establish a rapport conducive to patient candor, which can vastly facilitate diagnosis and can optimize treatment.
But in this turbulent era of attenuated patient—physician relationships, physicians may tend to view patients less as persons than as cases—conglomerates of diseased organs, malfunctions, symptoms, and emotional problems. Moreover, because various specialists and other health professionals, at sites remote from one another, may compile information on a patient, the patient may lack a sense of who is in charge, tire of providing personal information repeatedly, and become uncooperative. Physicians in "managed care" settings may work with restraints on (a) the time they spend with patients, (b) diagnostic-test options, and (c) referrals to specialists.
An Overview of Diagnosing
The process of diagnosing typically begins with the asking of key questions concerning the patient's chief complaint and its history. A physical examination follows. In light of the clinical findings—the findings from the interview and the physical—the physician often orders several lab tests, with the goal of ranking possible diagnoses by integrating clinical and lab data. Then the physician considers the risk/benefit ratios of diagnostic and therapeutic options and, to plan treatment and its monitoring, discusses such options with the patient.
In general, patients describe their health problems and those of their relatives very subjectively. Often, information about a symptom that particularly troubles a patient contributes little to the process of diagnosing, while a complaint the patient considers minor or not worth mentioning is clinically important. Moreover, sometimes the patient's replies are inaccurate because of a language and/or social barrier; forgetfulness; disorientation; or a fear of disability, death, or familial repercussions of his or her illness.
Because physical signs of disease are objective and may be the only real evidence of disease on hand, findings from physical exams should be recorded immediately. Many inaccuracies stem from delays in recording such findings. In some cases the frequent repeating of physicals may be appropriate but will not accord with the guidelines of the patient's health care plan.
Numerous lab tests are easily available, and lab testing is often crucial to diagnosis. In medicine, "lab tests" usually refers to blood tests. Diagnostic tests in widespread use to which the expression usually does not refer include electrocardiography, magnetic resonance imaging (MRI), ultrasonography, and various radiologic exams (e.g., CAT scanning). But no diagnostic test is foolproof, and correctly interpreting the data from such tests is often difficult. Data from tests that are not specific enough—i.e., that can register reactions to too many different entities indiscriminately and thus often yield false positive data—or that are not "sensitive" enough—i.e., that do not adequately measure minuscule changes or scarce entities—may mislead a physician. After the physician has confirmed any abnormality suggested by the test data, he or she must decide whether the abnormality is serious and whether to order other tests. Even when neither clinical findings nor lab findings are very accurate, consolidating them can lead to a more reliable narrowing of diagnostic possibilities than can using either group of findings alone.
"Differential diagnosis" refers to diagnosis by ruling out at least one physical or mental disorder that might account for a patient's symptom(s). The more a patient's symptoms correspond to those of a recognized syndrome—one of many textbook groups of symptoms that together distinguish a disorder—the fewer the disorders the physician must rule out. Matching overall findings with a recognized syndrome usually does not exactly reveal the cause of an illness, but it considerably shortens the list of possible causes and may suggest specific additional tests.
Once More—From the Heart?
When Adam Stern was in Community Hospital's ER reporting chest pain, his history and findings from his physical exam made it appear likely that he had very recently had a heart attack (an AMI). Thus he was admitted as an in-patient. Because serial electrocardiograms and cardiac enzyme values did not substantiate the diagnosis of AMI, he was discharged the next day. But electrocardiography is not a very "sensitive," or delicate, test. Indeed, if particular ECG findings were prerequisite to hospitalizing putative AMI patients, about half of all patients who come to the ER because of a heart attack would leave without having been properly diagnosed and treated.
Months later, Mr. Stern had chest pain as he jogged. He underwent an exercise stress test, but findings from it were normal. After another nocturnal episode of severe chest pain, he was admitted to Community Hospital for testing, which resulted in the ruling out of coronary spasm as the cause of his chest pain. A cardiologist then referred him to a gastroenterologist.
That gastrointestinal (GI) disorders are a major source of noncardiac chest pain has long been recognized. Angina pectoris (also known as angina) and noncardiac chest pain share neural pathways. Each year in the U.S., some 25,000 to 75,000 patients are diagnosed with chest pain of esophageal origin, which often mimics angina. About 10—20 percent of all patients admitted to hospitals to determine whether they have had a heart attack have pain whose etiology is esophageal. Esophageal disorders are the cause of chest pain in approximately 50 percent of patients whose chest pain is noncardiac. Often, patients with esophageal chest pain are not evaluated for esophageal disorders during their hospital stay and are thus discharged without a definitive diagnosis.
The gastroenterologist examined the upper part of Mr. Stern's gastrointestinal tract with an endoscope, which usually consists of a flexible optical tube attached to a visual monitor. Mr. Stern also underwent an X-ray procedure called an "upper GI series." The doctor provisionally diagnosed him with GERD (see sidebar), which is marked by heartburn; instructed him in a regimen for minimizing reflux (the upward movement of stomach contents into the esophagus); and prescribed an antacid. The regimen included dietary restrictions, measures for weight reduction, and sleeping on a wedge to keep the chest above the legs. Mr. Stern shortly began following the doctor's advice and has not had chest pain since.
The Bottom Line
Preventing clinical situations such as that which Mr. Stern experienced in Community Hospital's ER is not simple, but state-of-the-art diagnostic software and reliable official guidelines for diagnosing and triaging can help enormously. Hospitals can prepare such guidelines by identifying, establishing, and formalizing key historical questions and key physical and diagnostic-test findings.
Stuart M. Lane, Ph.D., is on the faculties of the Joseph L. Mailman School of Public Health of Columbia University, New School University, Pace University, and Saint Joseph's College (Maine). His doctorate is in chemistry. His M.S. degree, from Columbia University, is in health administration.
A 'Burning' Issue: GERD Revisited
Diagnosing gastroesophageal reflux disease, or GERD, can be challenging even for skilled physicians. Although various diagnostic tests can clarify a diagnosis of GERD, the most useful diagnostic tool is the patient's own story. Because GERD can mimic disparate conditions, some of which are serious, a physician evaluating anyone whose symptoms suggest GERD must consider other diagnostic possibilities.
The most threatening condition GERD can mimic is a heart attack. Even a cardiologist may be unable to tell whether a patient's chest pain is due to heart disease or to GERD. The consequences of misinterpreting heart attack symptoms as GERD may be devastating.
Other conditions that can mimic GERD include chest-wall muscle strain, esophageal spasms, gallbladder attacks, and peptic ulcer disease.
Not every case of heartburn requires medical evaluation. If one's heartburn is occasional and responds well to antacids or to anti-reflux medicines called "H2 blockers," seeing a physician about the symptom may not be necessary. But even symptoms suggestive of GERD that are only occasional may not be consequences of GERD. Thus, any patient unsure about the cause of such symptoms should consult a physician.
Even patients with a confirmed diagnosis of GERD who are strictly following their treatment plan may develop GERD complications. GERD patients should notify their physicians if:
* heartburn they consider otherwise unremarkable does not respond to their usual anti—GERD measures;
* their stool is black;
* food seems to stick to their esophagus after they have swallowed;
* breathing difficulty dizziness, or sweating accompanies what appears to be heartburn;
* they vomit blood (which may range in color from bright red to black);
* they have chest pain of a quality or duration that is different from their GERD chest pain;
* their use of over-the-counter GERD medicines (e.g., antacids) has increased;
* symptoms attributed to GERD have become more frequent; or
* their body weight has decreased and the cause of the decrease is not evident.
—Michael Kirsch, M.D.
(From Priorities, Vol. 11, No. 1)
