“John is an unemployed veteran who is back again for his hypertension. His doctor said that he needs a different medicine. He had something like this once before, but he isn’t certain whether it was when he was a little boy or not.”

“Mr. Smith is a 42-year-old black man who enters our hospital for the second time with a chief complaint of a ‘nosebleed’ of about 6 hours duration. The history of the present illness begins 5 years ago when a blood pressure of 180/120 was recorded by a plant physician at the time of a routine preemployment examination.

“He sought no medical advice for this and remained asymptomatic until 4 years ago when he noted the gradual onset of headaches, usually but not always pounding, occurring on an average twice a week. There were no aggravating or alleviating factors (other than aspirin), nor any associated symptoms, until 2 years ago, when the headaches became more severe and blurred vision supervened. For this latter reason, he visited our emergency room (ER) and was admitted. At that time, he was also told for the first time that he had blood in his urine, although he himself recalls no change in its color … (etc.).”

Time may actually be of several dimensions, as we shall shortly see. At first glance, it might seem that the dimension of time has already been presented for the symptom of the chief complaint because the duration is a statement in this dimension. And for some symptoms, this is true (e.g., “cyanosis since birth” in cases of congenital heart disease with right-to-left shunting). But in other cases, the issue is much more complicated than a simple issue of duration.

For instance, consider this account of crescendo angina:

Here the patient is describing the duration of the individual attack, the duration of the relief of symptoms, and the frequency of the attacks, all of which are expressed in units of time and each of
which must also be described in the record within the chronologic dimension. If one had only the duration of the symptom, one could not make the diagnosis of crescendo angina.

With further questioning, one could add still another time measure: How long the patient could walk up that hill before the “squeezing-aching” would appear.

It will become obvious that there are many other important symptoms existing in time that we have not discussed here. To give one additional example, students are taught to ask whether there is an “aura” or warning in evaluating the complaint of syncope. But even more useful is the time dimension of the aura because patients with vasovagal syncope tend to have a long aura (about 2.5 minutes), while those with cardiac syncope usually have a very brief aura (<3 seconds) rather than none at all (Martin et al., 1983).

When the onset of the illness is vague or the description seems somewhat sparse, a useful question to ask the patient is this: “When is the last time you can remember feeling perfectly healthy?” Use exactly these words.

The purpose of this question is to get the patient talking. It is not intended to determine when the history of the present illness began, although sometimes it will do that. The idea is to increase the number of historic facts at your disposal. After you get all the facts, you will be able to determine the beginning of the present illness. In addition to sharpening up the early history of the present illness by giving the patient a chance to reflect and recall, this question often stimulates patients to talk about other symptoms that they had momentarily pushed out of consciousness or that the listener had not attended to at first.

It is important to understand that this question is not the same as, “When did you first get sick?” or even, “When is the last time you did not have (the given symptom)?” Those are not really open-ended questions. The question as stated in the first paragraph is open-ended, provided that you listen to the answer without interrupting.

A feature of timing that is pertinent to a wide range of specific symptoms is the relationship of the onset to any drugs that the patient may be taking. The probability of various side effects may be found in the Physicians Desk Reference. However, this author frequently relearns the dictum taught at Parkland Memorial Hospital: “Any drug can do anything.” Adverse reactions may be paradoxical or difficult to distinguish from the condition being treated, as with suicidality or violent behavior associated with certain antidepressants (Kauffman, 2009). They may be rare, unanticipated from the drug’s purported mechanism of action, or simply underreported, for example, irritability and aggressive behavior on cholesterol-lowering drugs (Golomb et al., 2004; Tatley and Savage, 2007). Adverse effects do not necessarily have an immediate onset. Some, such as cognitive impairment due to diazepam, occur as a long-lived drug accumulates. Allergic or idiosyncratic reactions can occur unpredictably in patients who have previously tolerated a drug very well. The post hoc ergo proper hoc fallacy is always a risk when attributing a finding to a drug (see Chapter 27), yet continued use of a drug for fear of the fallacy can do continued harm if the drug truly is the culprit.

Some examples of quantity are “three tablespoons of sputum each morning,” “two-pillow orthopnea,” “about one-half cup of bloody emesis,” or “claudication that had formerly appeared after four or five level blocks, now occurring regularly at one block.”

Some symptoms, such as pain, have no international units but can still be expressed on an analog scale from 0 to 10, where 0 is the absence of the symptom and 10 is an extreme, such as “so bad you would have killed yourself.”

Other symptoms cannot be described in cardinal numbers but still require a salient, concrete description. Adjectives (such as “terrible”) are much less useful than the answer to the question, “What could you do in the past that you cannot do now because of (this symptom)?” The patient might respond that he can no longer walk across the room to the toilet because of shortness of breath.

Although location might well have been covered in the chief complaint, if the chief complaint is the leading symptom of the present illness, the use of the patient’s own words might preclude a precise description at that point. The pain in the “tummy” might be in the epigastrium, hypochondrium, periumbilical area, suprapubic area, or even the colon and rectum. A misinterpretation of the term “tummy” by the physician will, at best, delay the diagnosis. Similarly, the “hip” can be the buttock, the rectum, the actual hip joint, the iliac crest, the skin on the lateral surface of the buttock, or nodes in the inguinal or femoral areas. Sometimes it is the doctor who is imprecise; he might change the patient’s “thigh,” “knee,” “calf,” “ankle,” or “instep” into the “left lower extremity,” thereby homogenizing, instead of refining, these terms.

Some symptoms, such as weakness, do not always have a location. This is also important to establish and to note in the record. The differential diagnosis of weakness in the right hand is completely different from that of generalized weakness that is not focal.

For certain symptoms, particularly pain, the dimension of location also includes radiation, which is where the symptoms move (see the sections on the “Durga and pseudo-Durga syndromes”).

The patient should be asked about aggravating factors in an open-ended manner: “What kinds of things might make it worse?” Although aggravating factors are usually analyzed in conjunction with alleviating factors, a separate question should be used for the latter. If you ask the patient, “What kinds of things make this better or worse?” most patients will answer only the second part of the question and will never return to the alleviating factors. The physician who asks this question may then remember “no data” as the alleviating factor. (This, of course, is one form of the dreaded double question [see Chapter 2]. If you audit your tape recordings carefully, you will be amazed by how often you use double questions. Avoiding them is a learned skill.)

To illustrate the utility of the aggravating and alleviating factors, consider two patients with left-sided anterior chest pain. One patient’s pain is induced by exercise and strong emotions but consistently relieved by rest and sublingual nitroglycerin; this is characteristic of angina pectoris. The other patient’s pain is aggravated by sneezing, coughing, and respiration but alleviated by shallow breathing and splinting of the left side of the chest; this patient has pleurisy. Although the patient may not always have such a classic history, the physician will never learn of these diagnostic
clues, even when readily available, if he does not ask the proper questions.

The quality of the symptom is most important for the very symptoms that at first glance seem to be least susceptible to a qualitative description. For instance, Samuel Levine used to tell his students that if the patient’s chest pain was glibly and clearly communicated it was probably not myocardial in origin. The difficulty that the patient experienced in describing the pain was in itself a diagnostic clue.

When a patient complains of “spells,” “falling out,” “risings” (or “kernels”), weakness, back pain, or fatigue, a good opening question is, “Could you tell me more about what that was like?”

Not all the information obtained in response to that question will necessarily be in the dimension of quality. In fact, some patients will proceed to describe almost everything else except the quality: where he was when the symptom occurred, the time of day, statements of other persons, opinions of other physicians, and accounts of the symptoms of other family members. Ask such patients, “Could you tell me more about what it felt like to you?” If the patient truly seems to display alexithymia (the inability to describe mood and feelings), try the question, “What might I have had that would feel most like what you are describing?”

Determination of the setting in which a symptom began is useful in direct proportion to the breadth of the interviewer’s view of the word “setting.” A narrow interviewer may only inquire, and episodically at that, as to the patient’s position (as in syncope), time of day (as for ulcer pain), recent ingestion of alcoholic beverages (for determining the etiology of pancreatitis), and so forth. This restricted view leads to a number of problems.

Narrow inquiries into “setting” are not open-ended information-gathering devices but branch-point signs at best, or, more usually, a means of confirming diagnoses already suggested to the interviewer. The answers to “setting” questions generated in this fashion cannot be used for research purposes because they are not asked of all patients. They cannot be used for learning purposes for a related reason (i.e., their unknown specificity). They are not useful in patient care because they are post facto to diagnoses considered for some other reason, or are simply window dressing, as in the following example.

One morning, I was presented a 28-year-old black man who had been admitted with crushing chest pain, precordial Q waves, huge ST-segment elevations, T wave inversions, and increased levels of the cardiac enzymes. The diagnosis of myocardial infarction was evident to everyone.

Most of the history of the present illness was a recitation of the following facts:

When I inquired why I was being given this information, I was told that the patient had been diagnosed (correctly) as suffering from a myocardial infarction and that these were the risk factors for atherosclerotic coronary artery disease. Although the patient did not have any of these risk factors, the cardiology consultant had nonetheless suggested atherosclerosis as the etiology of the disease. Because of the narrowness of focus in contemporary medicine (engendered, in part, by such post facto lists), unusual diagnoses, such as coronary artery arteritis, a coronary embolus, or a congenital anomaly of the coronary circulation, were not considered. Once the patient was labeled as having atherosclerosis, despite the absence of risk factors, all investigations were stopped cold.

Open-ended setting questions will often provide important clues to a correct diagnosis that will be missed by the Procrusteans (i.e., those who insist that the patient’s story must fit their preconceived notions, just as travelers had to fit the mythical innkeeper’s bed). The patient might start talking about something he had not told anyone before, for example, recent recurring spells of despondency, leading to further questions and possibly to a diagnosis of anxiety or depression on a positive basis, rather than the always treacherous exclusionary one. Also, open-ended questions help the interviewer understand the patient as a person.

Another error of the Procrusteans is to insist that the setting must be the one expected on a post facto basis, regardless of what the patient says. If the patient has told you that his symptoms do not occur in a setting of strong emotional arousal, you should not insist that they do. I have seen this error made in more than one patient suffering with (undiagnosed) acute intermittent porphyria, whose symptoms had been attributed to “nerves” by several physicians.

A broad interpretation of setting may lead to the discovery of new diseases, as in the story of Soma Weiss’ streetcar conductor (see Chapter 18). An interest in the setting is also important to the mental hygiene of the physician. If you do not have this basic curiosity about your patients and their illnesses, you will soon be practicing applesorter medicine. After a few years of seeing the apples only in the single dimension of size and ignoring their colors, tastes, smells, bumps, bruises, and curious travels, your medical practice will become onedimensional, causing boredom and burnout in your life’s work.

Finally, a broad understanding of the onset setting may be useful in management as well as diagnosis. The physician should be particularly sensitive to absences and partings. If the onset of the illness was in the setting of the loss of a significant other, then a similar loss of the physician (who may be quite significant to the patient) may be expected to produce perturbations both in the patient and in the patient-physician relationship. Loss can be permanent (death or abandonment), temporary (vacation or illness on the part of the physician), or anticipated (either based in reality or the fantasy of the patient).

For the Junior Student. If you do attempt to elicit the history in the manner described here, you may find yourself in conflict with some of the medical residents with whom you must work. They will want you to “get in and get out” and “just get the facts” as they perceive them. While their approach may yield certain short-term benefits, it will deprive you of the opportunity to learn certain enduring skills, which are not easily acquired later. You must decide
whether you want to learn a difficult but valuable technique from the patients assigned to you or whether you would prefer to be a donut-and-cup bearer to the tin gods of the buffed chart.

For Students of All Levels. If you want to excel at history taking, you must read about the problems that you are seeing. You will often want to return to the patient to obtain more information about certain dimensions of a symptom based on your reading.

For the Attending. The patient’s culture and language are extremely important parts of the setting, both for the interviewing process and the composition of the history. Unfortunately, this is very difficult to teach to the beginning student who is going through an acculturation process himself. By the time the young physician can appreciate these issues fully, he no longer needs to be reminded of them by the written word. The only effective method of instruction is at the bedside with a live patient.

Inconstant dimensions are those that do not apply to all symptoms but that are, nevertheless, very important for some, such as color, clarity, consistency, and so forth.

Pain is the body’s signal that something is wrong. The perception of pain consists of a somatic neuronal input and a psychologic interpretation. Pain as a manifestation of various specific ailments is discussed below. Frequently, however, pain persists beyond the time when an injury is healed or the neuronal source is apparent. Chronic, intractable “pain without lesion” is increasingly recognized as a medical problem in its own right. Nearly 15% of the population in Western societies is believed to suffer daily chronic pain; in more than 10%, the pain is severe enough to interfere with daily living (Mäntyselkä et al., 2003).

Symptomatic treatment with high-dose opioids has become increasingly accepted, though fraught with hazard because of the abuse potential of these medications, the incentives inherent in a black market for a highly desired substance, and the threat to physicians from overzealous regulators and prosecutors (see Chapter 2). Physicians who undertake the management of chronic patients with controlled substances need to stay abreast of the everchanging regulatory and legal climate. Particular care must be taken in monitoring and documenting the effect of pain and of treatment on the patient’s life and ability to work or carry on activities of daily living.

Pain is unavoidably a subjective experience. Attempts to quantitate it usually rely on a visual analog scale or ranking the severity of pain on a scale such as “0” for “pain free” to “10” for “the worst pain you can imagine” or “pain so bad you would kill yourself.” One clinical pain diagram is shown in Fig. 25-5. These days, students probably cannot avoid seeing the “smiley face” posters in their hospitals, provided by a manufacturer of analgesics.

Despite proclamations from the Joint Commission on Accreditation of Health Care Organizations, pain is not a “fifth vital sign.” It is a symptom reported by the patient and not an objective measurement of a physiologic process necessary to sustain life.

One attempt to measure pain quantitatively is described below.

Patients who report chills should be asked whether the bed actually shook or moved on the floor. A true shaking chill also tends to last a definite period, on the order of 30 minutes. Shaking chills, as opposed to a sensation of chilliness, have diagnostic implications and are always high on the list of findings that must be explained by the primary diagnosis. They also imply that the fever was high, even though the patient may not have measured it.

Night sweats should be described in quantitative, practical terms. For example, “Were the sweats so bad that you had to change your pajamas?” Or, “Did you have to change the pillowcase to get back to sleep?” Or, “Did your wife change the sheets?” Note that patients who sleep in warm weather in non-air-conditioned rooms do not soak the sheets as highly febrile patients do. They simply cast the sheets off and roll away from any dampness, and of course, they are never awakened by a chill.

The old teaching that viral diseases did not cause bed-shaking chills is probably incorrect; patients with dengue fever, for instance, shake the bed.

Itching is usually a manifestation of cutaneous disease but may also result from a systemic condition (Bernhard, 1987). A drug history is most important. Phenothiazines, tolbutamide, erythromycin, anabolic hormones, estrogen, progesterone, and testosterone may cause itching by inducing cholestasis, narcotics through histamine release, and aspirin through prostaglandin effects. Any drug may
cause pruritus through an idiosyncratic effect, even if the patient has been taking it for years.

Inquire about exposures to fiberglass, dusts, and chemicals. In addition, in many patients with unexplained pruritus, a veterinary examination may identify exposure to ectoparasites from pets.

Itching severe enough to waken the patient at night should arouse suspicion of scabies or dermatitis herpetiformis. If these conditions are ruled out in a patient with severe itching, consider a systemic cause.

Systemic diseases associated with itching include renal failure (due to secondary hyperparathyroidism); endocrine conditions (thyroid dysfunction, diabetes mellitus, and carcinoid syndrome); hematologic conditions (such as hemochromatosis and polycythemia vera); and hepatobiliary disease, especially with biliary obstruction (as in primary biliary cirrhosis or carcinoma causing extrahepatic obstruction—note that itching due to the retention of bile salts may precede the appearance of jaundice). Itching produced by bathing (aquagenic pruritus) may be a symptom of polycythemia vera, Hodgkin disease, mastocytosis, or aging (Phillips, 1992). Nearly 30% of patients with Hodgkin disease have severe burning pruritus (Bernhard, 1987). Other malignant causes include visceral cancers, mycosis fungoides, multiple myeloma, and central nervous system tumors.

Hemoptysis literally means “to spit up blood,” and hematemesis means “to vomit up blood.” As actually used, hemoptysis refers to the experience of coughing up blood, and by extension, it refers to any blood produced through the mouth but originating in the pulmonary system.

The four main causes of hemoptysis used to be bronchogenic carcinoma, bronchiectasis, rheumatic mitral valve disease (especially mitral stenosis), and tuberculosis. While this remains a useful mnemonic for the previously asymptomatic patient who presents with a sudden hemoptysis, it is less accurate today. The prevalence of bronchiectasis, tuberculosis, and rheumatic heart disease has decreased substantially. In a study conducted in Jerusalem between 1990 and 1995, the most common causes of hemoptysis were bronchiectasis (20%), lung cancer (19%), bronchitis (18%), and pneumonia (16%) (Hirshberg et al., 1997). In a study in Kansas City, Missouri, the most common cause was bronchitis (26%) (Reisz et al., 1997).

A serious but treatable cause that may be overlooked in both children and adults is an inhaled foreign body. This may occur in persons who have no risk factors such as memory loss, seizures, or alcoholism. The hemoptysis may be massive in the case of inhaled vegetable matter, which may induce florid bronchiectasis and vascular changes (Dore et al., 1997). The patient is likely to have a chronic cough and may have a history of sudden onset of cough followed by wheezing or dyspnea (Al-Majed et al., 1997). This condition illustrates that to make a diagnosis, preferably by means other than surgical pathology, one must think of it.

In the areas of the world where tuberculosis is still a problem, more specific etiologies of hemoptysis need to be considered, including Rasmussen aneurysm or secondary aspergillosis. (Taking one item out of a differential diagnosis and performing a more sophisticated differential of that condition, as was just done, is called “going up to the next level of complexity.”)

In about 20% of cases, no cause for hemoptysis is found despite extensive investigation (and a lengthy list of uncommon possibilities). The risk of recurrence is low (Dore et al., 1997). Knowledge of facts like this will help a physician reassure a patient who has experienced a frightening symptom, but should not serve as a pretext to forgo a thorough investigation, particularly if the symptom is recurrent.

Of course, hematemesis is caused by an entirely different set of diagnostic possibilities. It usually implies a source of blood proximal to the ligament of Treitz and therefore includes gastric ulcer, gastric carcinoma, Osler-Weber-Rendu syndrome, esophageal varices, and so forth. Conceivably, it could be caused by epistaxis, assuming that the patient swallowed the blood and vomited it later.

The physician is responsible for making the distinction if the patient reports, “I spit up blood.” To qualify it as hematemesis, the patient should have actually gagged and regurgitated the bolus of blood. On the other hand, patients with hemoptysis usually can recall a clear episode of coughing. Often, they can point to the side of the thorax from which the blood came. Unfortunately, even excellent interviewers may not be able to make the distinction 100% of the time. However, one becomes more skilled by interviewing many patients with this complaint prior to knowing the anatomically determined source of the bleeding.

Chronic cough is one of the top five reasons for consulting family physicians. The diagnosis may be missed for years, especially if the physician focuses solely on the chest. There are receptors that can provoke a reflex cough in the ear, the esophagus, and the pericardium, as well as the tracheobronchial tree (Bellanti et al., 2000).

Important dimensions include exacerbating and relieving factors. Cough due to asthma may be provoked by exposures or exercise. Psychogenic cough, which may be quite severe, is relieved by sleep or distraction. Ask about response to past therapies. Relief from steroids suggests an inflammatory condition, from antibiotics, an infection, most often sinusitis. Proton pump inhibitors may relieve cough resulting from gastroesophageal reflux, whether due to microaspiration or receptors in the esophagus.

Always take a drug history: Up to 20% of patients taking ACE (angiotensin-converting enzyme) inhibitors have a cough. Important associated symptoms or signs include sinus congestion and postnasal drip, cardiac arrhythmias (which can cause a cough relieved by correcting the arrhythmia), and symptoms related to acid reflux. The presence or absence of purulent sputum is usually of little diagnostic help (Stulbarg, 2003).

An unusual cause of cough, generally associated with headache, toothache, fever, or upper limb ischemic symptoms, is temporal arteritis (Hellmann, 2002). A high index of suspicion in patients older than about 50 may lead to early diagnosis and prevention of blindness or other devastating complications.

Think of risk factors for possible infectious etiologies. Ask about fevers, night sweats, past skin tests, and constitutional symptoms (weight loss, anorexia, fatigue, etc.). A recent viral illness may cause a postviral sensitivity with cough that lasts for months. In patients who might have been exposed, perhaps without being aware of it, pertussis is surprisingly common, even in adults who have been vaccinated. Pertussis is diagnosed in about 20% of patients who present to an
ER with a cough lasting from 3 to 12 weeks. Consider Mycobacterium avium complex (MAC) in patients who may be immunosuppressed.

Although physicians may gain some clues from listening to the cough, the patient’s description is generally not diagnostically helpful (Stulbarg, 2003).

A history of a negative chest radiograph is not helpful, as standard films are quite insensitive. Computerized tomography may reveal an endobronchial tumor, tuberculosis, interstitial disease, bronchiectasis, or the nodular infiltrate of MAC in a patient with a normal or nearly normal radiograph.

The three most common causes of cough in patients with a normal chest X-ray who are nonsmokers and not taking an ACE inhibitor are gastroesophageal reflux disease, postnasal drip syndrome, and asthma, with the majority of patients having more than one cause (Mello et al., 1996).

In children, consider congenital malformations, cystic fibrosis, and aspiration of foreign bodies. In adolescents, repetitive cough may be the presenting symptom of Tourette syndrome (Hogan and Wilson, 1999).

One 30-year cough was instantly cured by removing a hair touching the patient’s eardrum (Stulbarg, 2003).

The diagnosis of difficulty swallowing can generally be made on the basis of the history. Dr Stuart Danovitch of Washington, DC, writes that progressive, short-duration dysphagia, chiefly for solids and ultimately for liquids, is exceedingly characteristic of cancer. Weight loss is generally associated. Progressive dysphagia for solids over years characterizes peptic stricture and would be supported by a history of heartburn and antacid ingestion. Nonprogressive, intermittent dysphagia for solids, especially occurring early in a meal, is typical for a lower esophageal (Schatzki) ring. Slowly progressive dysphagia, over a period of years, for both liquids and solids suggests achalasia. Intermittent, minimally progressive dysphagia and odynophagia characterize esophageal spasm, particularly when accompanied by chest pain. Reviewing previous medical records helps obtain a good picture of the longitudinal progression of the disease (S. Danovitch, personal communication, 2004).

The differential diagnosis of abdominal pain in various locations is discussed in Chapter 20. A general observation is that lower abdominal disorders may present with upper and midabdominal complaints, whereas upper abdominal disease such as cholecystitis or pancreatitis are far less likely to present with lower abdominal complaints (S. Danovitch, personal communication, 2004).