Family physicians frequently diagnose and treat psychiatric disorders, particularly in patients enrolled in managed care plans. One study revealed that 25 to 30 percent of patients presenting to primary care physicians have psychiatric disorders.¹ Although estimates of missed psychiatric diagnoses are probably inflated and overemphasized,² studies indicate that 30 to 80 percent of these cases are undetected by primary care physicians.¹

To avoid missing a psychiatric diagnosis, family physicians should use a systematic approach when assessing a patient for psychiatric symptoms, much like the approach used in the medical review of systems. The psychiatric review of symptoms (PROS) is one such approach. Adapted from clinical psychiatric practice, the PROS is a series of questions designed to rapidly screen for the major psychiatric disorders. A similar screening tool, called the PRIME-MD, has been shown to be useful and to have a diagnostic accuracy rate of 88 percent.³ However, the PRIME-MD has several disadvantages: it is time-consuming, it requires specialized training and it does not provide a mechanism for screening for dementia, psychosis and personality disorders. For these reasons and others, the PRIME-MD has not enjoyed widespread use among primary care physicians.⁴

Within the context of the history and physical examination, the PROS should generally follow the medical history, since some initial rapport-building helps set the stage for more personal psychiatric questions. While the medical review of systems is easily remembered by keeping in mind the major organ systems from head to toe, the PROS, of course, elicits information about more abstract symptoms and may be more difficult to remember. The mnemonic “Depressed Patients Seem Anxious, So Claim Psychiatrists” may be useful, as follows:

1. Depression and other mood disorders (major depression, bipolar disorder, dysthymia).

2. Personality disorders (primarily borderline personality disorder).

3. Substance abuse disorders.

4. Anxiety disorders (panic disorder with agoraphobia, obssessive-compulsive disorder).

5. Somatization disorder, eating disorders (these two disorders are combined because both involve disorders of bodily perception).

6. Cognitive disorders (dementia, delirium).

7. Psychotic disorders (schizophrenia, delusional disorder and psychosis accompanying depression, substance abuse or dementia).

General Interviewing Approaches to Psychiatric Symptoms

Psychiatric symptoms are often difficult for patients to discuss. Topics such as suicidality, substance abuse and obsessive-compulsive rituals can arouse feelings of shame, embarrassment or despair. Several interview methods can be used to approach sensitive topics in a nonthreatening manner.

The interview technique of normalization involves introducing a behavioral topic by first making a statement to let the patient know that you consider the behavior in question to be a normal, or at least an understandable, response to a mood or situation. For example, the topic of alcohol abuse can be approached with a question like “With all the stress you've been under, I wonder if you've been drinking more lately?”

Symptom assumption⁵ is a similar technique in which a question is phrased in a way that implies you already assume the patient has engaged in a particular behavior. This technique communicates to the patient that you will not be surprised or offended by a positive response. For example, a patient who has indicated suicidal ideation might be asked, “What kinds of ways have you thought about to hurt yourself?” A patient who abuses alcohol and is suspected of having polysubstance abuse as well might be asked, “What sort of drugs do you usually use when you drink?”

Transitioning techniques are often used in psychiatric interviewing to facilitate a rapid series of questions on sensitive topics. Rather than abruptly switching from topic to topic (as is appropriate in the medical review of systems), a previous topic or a previous response is used as a jumping-off point for the next question. Thus, the topic of suicidality might be approached with the statement,“Earlier you mentioned that you didn't know how much more of this you could take. Have you had thoughts of wanting to escape it by dying?”

As in the medical review of systems, the best approach for the PROS is to begin with broad screening questions and proceed to specific symptoms if the patient's response to the screening question is positive. While a negative response to a given screening question decreases the likelihood of a disorder, the sensitivity of such screening is never perfect, and answers should be interpreted within the context of the patient's entire history and physical examination.

Mood Disorders

For depression, the simple question, “Are you depressed?” is effective. A study of terminally ill patients revealed that this straightforward approach had a 100 percent sensitivity and specificity in diagnosing major depression, outperforming elaborate screening instruments such as the Beck Depression Inventory.⁶ When depression screening is positive, the next step is to determine the presence of neurovegetative symptoms of depression. This information is helpful not only in confirming the diagnosis but also in identifying specific target symptoms to monitor after antidepressant therapy is initiated.

The eight neurovegetative symptoms of depression can be easily remembered with the mnemonic “SIGECAPS” (Table 1). Used by psychiatry residents at Massachusetts General Hospital (where it was devised by Dr. Carey Gross), the mnemonic refers to a prescription one might write for a depressed, anergic patient—SIG: Energy CAPSules. Each letter refers to one of the major diagnostic criteria for major depressive disorder, as listed in Table 1.

Rather than asking about each of the neurovegetative symptoms separately, a more efficient approach is to ask,“How has your depression affected your life over the past couple of weeks? For example, how has it affected your sleep? Your appetite?” and so forth. For patients who seem reluctant to admit to a depressed mood (or who have poor insight), beginning with the question, “Do you have any problems sleeping?” provides a nonthreatening introduction to a discussion of depressive symptoms.

Suicidal Ideation

The most effective approach for assessing suicidal ideation is to ask first about passive suicidal ideation. This sensitive area may be introduced with the question, “With all the depression you've been dealing with, have you ever had the thought that you'd be better off dead?” The most common response is the reassuring, “Oh sure, the thought has crossed my mind, but I'd never do anything to hurt myself.” However, if active suicidal ideation is present, one should determine if the patient has a suicide plan in place and estimate how realistic and imminent it is. A patient with a specific suicide plan should undergo urgent psychiatric evaluation.

Dysthymia

Dysthymia is a chronic depression lasting at least two years that does not meet symptomatic criteria for major depression. The diagnosis is usually easily made with the question, “When was the last time you remember not feeling depressed?” Typically, the patient with dysthymia answers “many years”; indeed, the average duration of the disorder is 16 years.⁷

Bipolar Disorder

A common oversight in psychiatric screening is to neglect to ask questions aimed specifically at diagnosing bipolar disorder. A failure to recognize the presence of bipolar disorder can lead to serious problems, since antidepressant therapy can precipitate manic episodes in such patients.⁸ Rapid screening for a history of mania is made challenging by the high potential for false-positive responses. Many patients report periods of euphoria and high energy that represent normal variations in mood rather than mania.

The following screening question is helpful: “Have you had periods of feeling so happy or energetic that your friends told you were talking too fast or that you were too ‘hyper’?” If the screen is positive, the mnemonic “DIG-FAST”can be used to recall the cardinal symptoms of mania (Table 2).

Borderline Personality Disorder

An important personality disorder for the family physician to recognize is borderline personality disorder. Patients with this disorder have high comorbidity with other psychiatric disorders and high rates of suicidal ideation, and they cause particular treatment difficulties, including hostility toward care-givers and low rates of treatment compliance.⁹

The central feature of patients with borderline personality disorder is a morbid fear of abandonment with consequential pathologic responses to perceived rejection.⁹ Such patients may demand inappropriate amounts of time or support from a primary care physician, and they may become hostile and demanding or suicidal if these needs are not met. While screening for borderline personality disorder is not always practical during the initial visit, the mnemonic “I DESPAIRR” (Table 3) is useful for recalling the criteria of the disorder, and some of these questions can be posed during follow-up visits as time and circumstances allow.

Asking about each trait in turn is not advised, both because this approach interrupts the flow of the interview and because it is time-consuming. Instead, these traits are best uncovered within the context of a brief survey of a patient's relationship history. Patients with borderline personality disorder often have a lifelong trail of ruptured and stormy relationships within several spheres, including the family, school, work, romance and marriage. While a lengthy social history is generally not possible during an initial medical visit, the family practitioner should be alert to the following “red flags”:

1. A history of doctor shopping.

2. A history of legal suits against physicians or other professionals.

3. A history of suicide attempts.

4. A history of several brief marriages or intimate relationships.

5. An immediate idealization of you as a “wonderful doctor,” especially if the patient compares you with disappointing caregivers of the past.

6. Excessive interest in your personal life, eventually leading to invitations to socialize with you. Behavior of this type implies boundary violations, and its purpose is to cement a relationship with the physician, allaying the patient's ever-present fear of abandonment.

Substance Abuse Disorders

The best and quickest screen for alcoholism remains the tried-and-true CAGE questionnaire¹⁰ (Table 4), in which a positive response to two or more of the items implies a 95 percent chance of alcohol abuse or dependence. However, a recent study¹¹ suggests that the way in which physicians create the transition to the CAGE questions profoundly affects the questionnaire's sensitivity. In this study, 43 confirmed alcoholics were divided into two groups. In one group, the CAGE was introduced with an open-ended question, such as “Do you have a drink now and then?” In the second group, patients were first asked to quantitate their alcohol intake with the question,“How much do you drink?” The sensitivity of the CAGE questions was dramatically higher in the first group (95 percent) than in the second group (32 percent), demonstrating the importance of beginning the CAGE questions in a nonjudgmental way.

Many physicians have been taught that the two-question drinking test (“Have you ever had a drinking problem?” and “When was your last drink?”) is an effective rapid screen for alcoholism. The screen is considered positive if the patient answers “yes” to the first question and “within the past 24 hours” to the second question.¹² However, a study failed to confirm the results of the initial study, and the sensitivity of this screening tool is probably no better than 50 percent.¹³

Abuse of other substances is common in alcohol abusers, and a positive CAGE screen should be followed by a question about drug use. The symptom assumption technique is useful in this setting. The possibility of drug use can be introduced with the question, “Aside from drinking, what sorts of recreational drugs do you use regularly? Cocaine? Marijuana? Speed? Heroin?” Delivered in a matter-of-fact manner, this question communicates a nonjudgmental attitude and tends to decrease the patient's shame about admitting to drug use.

Anxiety Disorders

A good general screen for all anxiety disorders is the question, “Do you tend to be an anxious or nervous person?” A positive response should prompt the physician to screen for panic disorder, agoraphobia and obsessive-compulsive disorder. Time constraints during the history and physical examination preclude screening for all of the anxiety disorders; panic disorder with agoraphobia is selected because it is common, and obsessive-compulsive disorder because patients rarely divulge this information unless they are specifically asked.

Panic Disorder with Agoraphobia

For panic disorder, the straightforward question, “Do you have anxiety or panic attacks?” is useful. If the patient is confused about what is meant by the term “panic attacks,” the following explanation is usually sufficient: “A panic attack is a sudden rush of fear and nervousness that makes your heart pound and makes you afraid you're going to die or go crazy.”

Once a diagnosis of panic disorder is confirmed, questions about symptoms of agoraphobia should be asked because agoraphobia accompanies panic in the majority of cases.¹⁴ The question, “Have you had to limit where you can go because of your anxiety?” captures the gist of the disorder. Patients with panic disorder often have been gradually limiting their activities, especially those that require driving, without realizing that anxiety is the underlying cause of their self-imposed restrictions.

Obsessive-Compulsive Disorder

A good screening question for diagnosing obsessive-compulsive disorder is,“Do you have symptoms of obsessive-compulsive disorder, such as checking things repeatedly or washing your hands over and over?”A potential pitfall is a false-positive response from healthy patients who are adaptively compulsive or perfectionistic but do not meet the criteria for obsessive-compusive disorder. A question such as “Do your compulsions significantly interfere with your ability to live your life?” helps identify a clinically significant disorder.

Somatization Disorder

Since the criteria for somatization disorder are somewhat arbitrary, the following mnemonic may help in remembering them: “Recipe 4 Pain: Convert 2 Stomachs to 1 Sex.” The translation of this mnemonic is as follows: the presence of four pain symptoms (“Recipe 4 Pain”), one conversion symptom (“Convert”), two gastrointestinal symptoms (“2 Stomachs”) and one sexual symptom (“1 Sex”).

Eating Disorders

To elicit information about behavior related to an eating disorder, the first question can be, “Have you ever felt like you are overweight?” An unequivocal negative response probably excludes the diagnosis of an eating disorder. A positive response should be followed by more focused questions about the methods used to lose weight, asking the questions in a matter-of-fact way to help defuse the patient's embarrassment. For example, the patient may be asked, “Have you dieted? Used laxatives? Made yourself throw up?”A binging history can be elicited by asking the following question: “Do you go on eating binges in which you eat an unusually large amount of food within a two-hour period and feel that you can't control your eating?”

A common pitfall is to screen for eating disorders only in women. Studies have pointed to a significant prevalence of these disorders in men, who constitute 10 to 15 percent of all anorexic and bulimic patients.¹⁵ Furthermore, 40 percent of men with bulimia are gay or bisexual, implying that screening for an eating disorder is particularly important in this population.¹⁶

Dementia

Early detection of dementia is increasingly important because currently available treatments, such as tacrine (Cognex) and donepezil (Aricept), are most useful in the early stages of the disease. Unfortunately, the utility of the most popular screening instrument, the Folstein Mini-Mental State Examination (MMSE),¹⁷ is hampered by high rates of false-positive diagnosis, especially among poorly educated patients.¹⁸ In a large-scale study (3,513 elderly patients) of the MMSE at the Mayo Clinic, unacceptably low positive predictive values led the researchers to conclude that the MMSE is ineffective when it is used to screen unselected patients in a general medical practice.¹⁹

For these reasons, a more reliable screening approach for the family physician is to conduct formal cognitive testing only in patients with a history that suggests a likelihood of cognitive impairment. Questions should directly address the possibility of short-term and long-term memory impairment. For short-term memory, studies have demonstrated the clinical utility of both a three-object recall and orientation to time and place (orientation to person is a measure of long-term memory).²⁰ For long-term memory, clinical studies have shown that general information questions and questions pertaining to remote personal information are most useful.²¹

Thus, a rapid screen for memory impairment proceeds in the following manner:

1. Ask the patient for the date and place.

2. Ask the patient to repeat three words—“ball, chair, purple”—and then to memorize them for later recall.

3. Ask about general information. This should be information that is uniformly taught in American society. Examples include the identity of famous persons, such as “Who was George Washington? Abraham Lincoln? John F. Kennedy? Martin Luther King, Jr.?” The patient can also be asked to roughly estimate the dates of famous wars (“When was the Civil War? World War II?”) and to name four major American cities. Recall of personal information can be tested if independent corroboration is possible. Examples include the patient's and spouse's birth dates, the name of the patient's high school and college, and the names and ages of children.

4. Ask the patient to repeat the three memorized words.

Poor performance on these screening tests, defined as significant disorientation to place and date, a greater than 50 percent failure rate on general and personal information items, and inability to remember at least two of the three words, suggests that more in-depth cognitive evaluation is needed, often including formal neuropsychologic testing.

Psychotic Disorders

Psychotic symptoms such as looseness of associations, bizarre delusions and hallucinations are generally easy to recognize, and patients manifesting these symptoms should be referred immediately for psychiatric treatment. However, many patients present with more subtle symptoms, which may not be discovered without some diagnostic “digging.” Because of the time constraints during an initial visit, screening for psychosis should be selective, focusing on those patients with a relatively high probability of harboring a psychotic thought process or content. Patients at high risk of psychosis include (1) patients with the diagnosis of major depression, substance abuse or dementia and (2) patients who appear guarded, suspicious or otherwise odd during the interview.

Screening questions for psychosis are often “piggy-backed” onto transitional questions referring to other symptoms previously described by the patient. For example, a depressed patient might be asked,“Depression sometimes causes people to have strange experiences, like hearing voices or feeling that others are trying to harm them. Has that happened to you?”A patient with dementia might be asked, “When you misplace things, do you sometimes think that they've been stolen?” or “Have you ever heard or seen people coming into your house?” For a substance-abusing patient, the approach might be to ask “Have drugs ever caused your mind to play tricks on you, like seeing things or having paranoid ideas?” Among substance abusers, psychotic ideation may result from acute intoxication (e.g., amphetamine or cocaine abuse), chronic use (e.g., alcoholic hallucinosis) or withdrawal (e.g., delirium tremens).

The best approach for ascertaining delusional ideation in someone suspected of active paranoia is to adopt the patient's viewpoint: “Have people been harrassing you or trying to harm you?” This communicates sympathy for the patient's perceptions and tends to defuse a guarded attitude.

This rapid screen is no substitute for a complete psychiatric evaluation in patients with more complicated or severe problems. Such patients should be referred for psychiatric consultation, but they can often return to their family physician for ongoing psychopharmacologic treatment.