Chapter 1. Differential Diagnosis Step by Step
© American Psychiatric Publishing
The process of DSM-5 differential diagnosis can be broken down into six basic steps: 1) ruling out Malingering and Factitious Disorder, 2) ruling out a substance etiology, 3) ruling out an etiological medical condition, 4) determining the specific primary disorder(s), 5) differentiating Adjustment Disorder from the residual Other Specified and Unspecified conditions, and 6) establishing the boundary with no mental disorder. A thorough review of this chapter provides a useful framework for understanding and applying the decision trees presented in the next chapter.
Step 1: Rule Out Malingering and Factitious Disorder
The first step is to rule out Malingering and Factitious Disorder because if the patient is not being honest regarding the nature or severity of his or her symptoms, all bets are off regarding the clinician’s ability to arrive at an accurate psychiatric diagnosis. Most psychiatric work depends on a good-faith collaborative effort between the clinician and the patient to uncover the nature and cause of the presenting symptoms. There are times, however, when everything may not be as it seems. Some patients may elect to deceive the clinician by producing or feigning the presenting symptoms. Two conditions in DSM-5 are characterized by feigning: Malingering and Factitious Disorder. These two conditions are differentiated based on the motivation for the deception. When the motivation is the achievement of a clearly recognizable goal (e.g., insurance compensation, avoiding legal or military responsibilities, obtaining drugs), the patient is considered to be Malingering. When the deceptive behavior is present even in the absence of obvious external rewards, the diagnosis is Factitious Disorder. Although the motivation for many individuals with Factitious Disorder is to assume the sick role, this criterion was dropped in DSM-5 because of the inherent difficulty in determining an individual’s underlying motivation for his or her observed behavior.
The intent is certainly not to advocate that every patient should be treated as a hostile witness and that every clinician should become a cynical district attorney. However, the clinician’s index of suspicion should be raised 1) when there are clear external incentives to the patient’s being diagnosed with a psychiatric condition (e.g., disability determinations, forensic evaluations in criminal or civil cases, prison settings), 2) when the patient presents with a cluster of psychiatric symptoms that conforms more to a lay perception of mental illness rather than to a recognized clinical entity, 3) when the nature of the symptoms shifts radically from one clinical encounter to another, 4) when the patient has a presentation that mimics that of a role model (e.g., another patient on the unit, a mentally ill close family member), and 5) when the patient is characteristically manipulative or suggestible. Finally, it is useful for clinicians to become mindful of tendencies they might have toward being either excessively skeptical or excessively gullible.
Step 2: Rule Out Substance Etiology (Including Drugs of Abuse, Medications)
The first question that should always be considered in the differential diagnosis is whether the presenting symptoms arise from a substance that is exerting a direct effect on the central nervous system (CNS). Virtually any presentation encountered in a mental health setting can be caused by substance use. Missing a substance etiology is probably the single most common diagnostic error made in clinical practice. This error is particularly unfortunate because making a correct diagnosis has immediate treatment implications. For example, if the clinician determines that psychotic symptoms are due to Cocaine Intoxication, it usually does not make sense for the patient to immediately start taking an antipsychotic medication unless the psychotic symptoms are putting the patient (or others) in immediate danger. The determination of whether psychopathology is due to substance use often can be difficult because although substance use is fairly ubiquitous and a wide variety of different symptoms can be caused by substances, the fact that substance use and psychopathology occur together does not necessarily imply a cause-and-effect relationship between them.
Obviously, the first task is to determine whether the person has been using a substance.This entails careful history taking and physical examination for signs of Substance Intoxication or Substance Withdrawal. Because substance-abusing individuals are notorious for underestimating their intake, it is usually wise to consult with family members and obtain laboratory analysis of body fluids to ascertain recent usage of particular substances. It should be remembered that patients who use or are exposed to any of a variety of substances (not only drugs of abuse) can and often do present with psychiatric symptoms. Medication-induced psychopathology is more and more common, and very often missed, especially as the population ages and many individuals are taking multiple medications. Although it is less common, toxin exposure should be considered, especially for people whose occupations bring them into contact with potential toxins.
Once substance use has been established, the next task is to determine whether there is an etiological relationship between it and the psychiatric symptomatology. This requires distinguishing among three possible relationships between the substance use and the psychopathology: 1) the psychiatric symptoms result from the direct effects of the substance on the CNS (resulting in diagnosis of Substance-Induced Disorders in DSM-5; e.g., Cocaine-Induced Psychotic Disorder, Reserpine-Induced Depressive Disorder); 2) the substance use is a consequence (or associated feature) of having a primary psychiatric disorder (e.g., self-medication); and 3) the psychiatric symptoms and the substance use are independent. Each of these relationships is discussed in turn.
1. In diagnosing a Substance-Induced Disorder, there are three considerations in determining whether there is a causal relationship between the substance use and the psychiatric symptomatology. First, you must determine whether there is a close temporal relationship between the substance or medication use and the psychiatric symptoms. Then, you must consider the likelihood that the particular pattern of substance/medication use can result in the observed psychiatric symptoms. Finally, you should consider whether there are better alternative explanations (i.e., a non-substance/medication-induced cause) for the clinical picture.
· You should consider whether a temporal relationship exists between the substance/medication use and the onset or maintenance of the psychopathology. The determination of whether there was a period of time when the psychiatric symptoms were present outside the context of substance/medication use is probably the best (although still fallible) method for evaluating the etiological relationship between substance/medication use and psychiatric symptoms. At the extremes, this is relatively straightforward. If the onset of the psychopathology clearly precedes the onset of the substance/medication use, then it is likely that a non-substance/medication-induced psychiatric condition is primary and the substance/medication use is secondary (e.g., as a form of self-medication) or is unrelated. Conversely, if the onset of the substance/medication use clearly and closely precedes the psychopathology, it lends greater credence to the likelihood of a Substance-Induced Disorder. Unfortunately, in practice this seemingly simple determination can be quite difficult because the onsets of the substance/medication use and the psychopathology may be more or less simultaneous or impossible to reconstruct retrospectively. In such situations, you will have to rely more on what happens to the psychiatric symptoms when the person is no longer taking the substance or medication. Psychiatric symptoms that occur in the context of Substance Intoxication, Substance Withdrawal, and medication use result from the effects of the substance or medication on neurotransmitter systems. Once these effects have been removed (by a period of abstinence after the withdrawal phase), the symptoms should spontaneously resolve. Persistence of the psychiatric symptomatology for a significant period of time beyond periods of intoxication or withdrawal or medication use suggests that the psychopathology is primary and not due to substance/medication use. The exceptions to this are Substance/Medication-Induced Major or Mild Neurocognitive Disorder, in which by definition the cognitive symptoms must persist after the cessation of acute intoxication or withdrawal or medication use, and Hallucinogen Persisting Perception Disorder, in which following cessation of use of a hallucinogen, one or more of the perceptual symptoms that the individual experienced while intoxicated with the hallucinogen (e.g., geometric hallucinations, flashes of color, trails of images of moving objects, halos around objects) are reexperienced. The DSM-5 criteria for substance/medication-induced presentations suggest that psychiatric symptoms be attributed to substance use if they remit within 1 month of the cessation of acute intoxication, withdrawal, or medication use. It should be noted, however, that the need to wait 1 full month before making a diagnosis of a primary psychiatric disorder is only a guideline that must be applied with clinical judgment; depending on the setting, it might make sense to use a more extended duration or a shorter duration depending on your concern for avoiding false positives versus false negatives with respect to detecting a substance/medication-induced presentation. Some clinicians, particularly those who work in substance use treatment settings, are most concerned about the possibility of misdiagnosing a substance/medication-induced presentation as a primary mental disorder that is not caused by substance use and might prefer allowing 6–8 weeks of abstinence before considering the diagnosis to be a primary mental disorder. On the other hand, clinicians who work primarily in psychiatric settings may be more concerned that given the wide use of substances among patients seen in clinical settings, such a long waiting period is impractical and might result in an overdiagnosis of Substance-Induced Disorders and an underdiagnosis of primary mental disorders. Moreover, it must be recognized that the one-size-fits-all 1-month time frame applies to a wide variety of substances and medications with very different pharmacokinetic properties and a wide variety of possible consequent psychopathologies. Therefore, the time frame must be applied flexibly, considering the extent, duration, and nature of the substance/medication use.
Sometimes, it is simply not possible to determine whether there was a period of time when the psychiatric symptoms occurred outside of periods of substance/medication use. This may occur in the often-encountered situation in which the patient is too poor a historian to allow a careful determination of past temporal relationships. In addition, substance use and psychiatric symptoms can have their onset around the same time (often in adolescence), and both can be more or less chronic and continuous. In these situations, it may be necessary to assess the patient during a current period of abstinence from substance use or to stop the medication suspected of causing the psychiatric symptoms. If the psychiatric symptoms persist in the absence of substance/medication use, then the psychiatric disorder can be considered to be primary. If the symptoms remit during periods of abstinence, then the substance use is probably primary. It is important to realize that this judgment can only be made after waiting for enough time to elapse so as to be confident that the psychiatric symptoms are not a consequence of withdrawal. Ideally, the best setting for making this determination is in a facility where the patient’s access to substances can be controlled and the patient’s psychiatric symptomatology can be serially assessed. Of course, it is often impossible to observe a patient for as long as 4 weeks in a tightly controlled setting. Consequently, these judgments must be based on less controlled observation, and the clinician’s confidence in the accuracy of the diagnosis should be more guarded.
· In determining the likelihood that the pattern of substance/medication use can account for the symptoms, you must also consider whether the nature, amount, and duration of substance/medication use are consistent with the development of the observed psychiatric symptoms. Only certain substances and medications are known to be causally related to particular psychiatric symptoms. Moreover, the amount of substance or medication taken and the duration of its use must be above a certain threshold for it to reasonably be considered the cause of the psychiatric symptomatology. For example, a severe and persisting depressed mood following the isolated use of a small amount of cocaine should probably not be considered to be attributed to the cocaine use, even though depressed mood is sometimes associated with Cocaine Withdrawal. Similarly, cannabis smoked in typical moderate doses rarely causes prominent psychotic symptoms. For individuals who are regular substance users, a significant change in the amount used (either a large increase or a decrease in amount sufficient to trigger withdrawal symptoms) may in some cases cause the development of psychiatric symptoms.
· You should also consider other factors in the presentation that suggest that the presentation is not caused by a substance or medication. These include a history of many similar episodes not related to substance/medication use, a strong family history of the particular primary disorder, or the presence of physical examination or laboratory findings suggesting that a medical condition might be involved. Considering factors other than substance/medication use as a cause for the presentation of psychiatric symptoms requires fine clinical judgment (and often waiting and seeing) to weigh the relative probabilities in these situations. For example, an individual may have heavy family loading for Anxiety Disorders and still have a cocaine-induced panic attack that does not necessarily presage the development of primary Panic Disorder.
2. In some cases, the substance use can be the consequence or an associated feature (rather than the cause) of psychiatric symptomatology. Not uncommonly, the substance-taking behavior can be considered a form of self-medication for the psychiatric condition. For example, an individual with a primary Anxiety Disorder might use alcohol excessively for its sedative and antianxiety effects. One interesting implication of using a substance to self-medicate is that individuals with particular psychiatric disorders often preferentially choose certain classes of substances. For example, patients with negative symptoms of Schizophrenia often prefer stimulants, whereas patients with Anxiety Disorders often prefer CNS depressants. The hallmark of a primary psychiatric disorder with secondary substance use is that the primary psychiatric disorder occurs first and/or exists at times during the person’s lifetime when he or she is not using any substance. In the most classic situation, the period of comorbid psychiatric symptomatology and substance use is immediately preceded by a period of time when the person had the psychiatric symptomatology but was abstinent from the substance. For example, an individual currently with 5 months of heavy alcohol use and depressive symptomatology might report that the alcohol use started in the midst of a Major Depressive Episode, perhaps as a way of counteracting insomnia. Clearly the validity of this judgment depends on the accuracy of the patient’s retrospective reporting. Because such information is sometimes suspect, it may be useful to confer with other informants (e.g., family members) or review past records to document the presence of psychiatric symptoms occurring in the absence of substance use.
3. In other cases, both the psychiatric disorder and the substance use can be initially unrelated and relatively independent of each other. The high prevalence rates of both psychiatric disorders and Substance Use Disorders mean that by chance alone, some patients would be expected to have two apparently independent illnesses (although there may be some common underlying factor predisposing to the development of both the Substance Use Disorder and the psychiatric disorder). Of course, even if initially independent, the two disorders may interact to exacerbate each other and complicate the overall treatment. This independent relationship is essentially a diagnosis made by exclusion. When confronted with a patient having both psychiatric symptomatology and substance use, you should first rule out that one is causing the other. A lack of a causal relationship in either direction is more likely if there are periods when the psychiatric symptoms occur in the absence of substance use and if the substance use occurs at times unrelated to the psychiatric symptomatology.
After deciding that a presentation is due to the direct effects of a substance or medication, you must then determine which DSM-5 Substance-Induced Disorder best describes the presentation. DSM-5 includes a number of specific Substance/Medication-Induced Mental Disorders, along with Substance Intoxication and Substance Withdrawal. Please refer to Decision Tree for Excessive Substance Use Decision Tree for Excessive Substance Use in Chapter 2, “Differential Diagnosis by the Trees,” for a presentation of the steps involved in making this determination.
Step 3: Rule Out a Disorder Due to a General Medical Condition
After ruling out a substance/medication-induced etiology, the next step is to determine whether the psychiatric symptoms are due to the direct effects of a general medical condition. This and the previous step of the differential diagnosis make up what was traditionally considered the “organic rule-outs” in psychiatry, in which the clinician is asked to first consider and rule out “physical” causes of the psychiatric symptomatology. Although DSM no longer uses words such as organic, physical,and functional, to avoid the anachronistic mind-body dualism implicit in such terms, the need to first rule out substances and general medical conditions as specific causes of the psychiatric symptomatology remains crucial. For similar reasons, the phrase “due to a medical condition” is avoided in DSM because of the potential implication that psychiatric symptomatology and mental disorders are separate and distinct from the concept of “medical conditions.” In fact, from a disease classification perspective, psychiatric disorders are but one chapter of the International Classification of Diseases (ICD), as are infectious diseases, neurological conditions, and so forth. When the phrase “due to a medical condition” is used, what is really meant is that the symptoms are due to a medical condition that is classified outside the ICD mental disorders chapter—that is, a nonpsychiatric medical condition. In DSM-5 and this handbook, therefore, the phrase “medical condition” is modified with adjectives such as another, other, or general to clarify that the etiological condition, like a mental disorder, is a medical condition but that it is differentiated from psychiatric medical conditions by virtue of being nonpsychiatric.
From a differential diagnostic perspective, ruling out a general medical etiology is one of the most important and difficult distinctions in psychiatric diagnosis. It is important because many individuals with general medical conditions have resulting psychiatric symptoms as a complication of the general medical condition and because many individuals with psychiatric symptoms have an underlying general medical condition. The treatment implications of this differential diagnostic step are also profound. Appropriate identification and treatment of the underlying general medical condition can be crucial in both avoiding medical complications and reducing the psychiatric symptomatology.
This differential diagnosis can be difficult for four reasons: 1) symptoms of some psychiatric disorders and of many general medical conditions can be identical (e.g., symptoms of weight loss and fatigue can be attributable to a Depressive or Anxiety Disorder or to a general medical condition); 2) sometimes the first presenting symptoms of a general medical condition are psychiatric (e.g., depression preceding other symptoms in pancreatic cancer or a brain tumor); 3) the relationship between the general medical condition and the psychiatric symptoms may be complicated (e.g., depression or anxiety as a psychological reaction to having the general medical condition vs. the medical condition being a cause of the depression or anxiety via its direct physiological effect on the CNS); and 4) patients are often seen in settings primarily geared for the identification and treatment of mental disorders in which there may be a lower expectation for, and familiarity with, the diagnosis of medical conditions.
Virtually any psychiatric presentation can be caused by the direct physiological effects of a general medical condition, and these are diagnosed in DSM-5 as one of the Mental Disorders Due to Another Medical Condition (e.g., Depressive Disorder Due to Hypothyroidism). It is no great trick to suspect the possible etiological role of a general medical condition if the patient is encountered in a general hospital or primary care outpatient setting. The real diagnostic challenge occurs in mental health settings in which the base rate of general medical conditions is much lower but nonetheless consequential. It is not feasible (nor cost-effective) to order every conceivable screening test on every patient. You should direct the history, physical examination, and laboratory tests toward the diagnosis of those general medical conditions that are most commonly encountered and most likely to account for the presenting psychiatric symptoms (e.g., thyroid function tests for depression, brain imaging for late-onset psychotic symptoms).
Once a general medical condition is established, the next task is to determine its etiological relationship, if any, to the psychiatric symptoms. There are five possible relationships: 1) the general medical condition causes the psychiatric symptoms through a direct physiological effect on the brain; 2) the general medical condition causes the psychiatric symptoms through a psychological mechanism (e.g., depressive symptoms in response to being diagnosed with cancer—diagnosed as Major Depressive Disorder or Adjustment Disorder); 3) medication taken for the general medical condition causes the psychiatric symptoms, in which case the diagnosis is a Medication-Induced Mental Disorder (see “Step 2: Rule Out Substance Etiology” in this chapter); 4) the psychiatric symptoms cause or adversely affect the general medical condition (e.g., in which case Psychological Factors Affecting Other Medical Condition may be indicated); and 5) the psychiatric symptoms and the general medical condition are coincidental (e.g., hypertension and Schizophrenia). In the real clinical world, however, several of these relationships may occur simultaneously with a multifactorial etiology (e.g., a patient treated with an antihypertensive medication who has a stroke may develop depression due to a combination of the direct effects of the stroke on the brain, the psychological reaction to the resultant paralysis, and a side effect of the antihypertensive medication).
There are two clues suggesting that psychopathology is caused by the direct physiological effect of a general medical condition. Unfortunately, neither of these is infallible, and clinical judgment is always necessary.
· The first clue involves the nature of the temporal relationship and requires consideration of whether the psychiatric symptoms begin following the onset of the general medical condition, vary in severity with the severity of the general medical condition, and disappear when the general medical condition resolves. When all of these relationships can be demonstrated, a fairly compelling case can be made that the general medical condition has caused the psychiatric symptoms; however, such a clue does not establish that the relationship is physiological (the temporal covariation could also be due to a psychological reaction to the general medical condition). Also, sometimes the temporal relationship is not a good indicator of underlying etiology. For instance, psychiatric symptoms may be the first harbinger of the general medical condition and may precede by months or years any other manifestations. Conversely, psychiatric symptoms may be a relatively late manifestation occurring months or years after the general medical condition has been well established (e.g., depression in Parkinson’s disease).
· The second clue that a general medical condition should be considered in the differential diagnosis is if the psychiatric presentation is atypical in symptom pattern, age at onset, or course. For example, the presentation cries out for a medical workup when severe memory or weight loss accompanies a relatively mild depression or when severe disorientation accompanies psychotic symptoms. Similarly, the first onset of a manic episode in an elderly patient may suggest that a general medical condition is involved in the etiology. However, atypicality does not in and of itself indicate a general medical etiology because the heterogeneity of primary psychiatric disorders leads to many “atypical” presentations.
Nonetheless, the most important bottom line with regard to this task in the differential diagnosis is not to miss possibly important underlying general medical conditions. Establishing the nature of the causal relationship often requires careful evaluation, longitudinal follow-up, and trials of treatment.
Finally, if you have determined that a general medical condition is responsible for the psychiatric symptoms, you must determine which of the DSM-5 Mental Disorders Due to Another Medical Condition best describes the presentation. DSM-5 includes a number of such disorders, each differentiated by the predominant symptom presentation. Please refer to 2.29 Decision Tree for Etiological Medical Conditions in Chapter 2, “Differential Diagnosis by the Trees,” for a presentation of the steps involved in making this determination.
Step 4: Determine the Specific Primary Disorder(s)
Once substance use and general medical conditions have been ruled out as etiologies, the next step is to determine which among the primary DSM-5 mental disorders best accounts for the presenting symptomatology. Many of the diagnostic groupings in DSM-5 (e.g., Schizophrenia Spectrum and Other Psychotic Disorders, Anxiety Disorders, Dissociative Disorders) are organized around common presenting symptoms precisely to facilitate this differential diagnosis. The decision trees in Chapter 2 provide the decision points needed for choosing among the primary mental disorders that might account for each presenting symptom. Once you have selected what appears to be the most likely disorder, you may wish to review the pertinent differential diagnosis table in Chapter 3, “Differential Diagnosis by the Tables,” to ensure that all other likely contenders in the differential diagnosis have been considered and ruled out.
Step 5: Differentiate Adjustment Disorders From the Residual Other Specified or Unspecified Disorders
Many clinical presentations (particularly in outpatient and primary care settings) do not conform to the particular symptom patterns, or they fall below the established severity or duration thresholds to qualify for one of the specific DSM-5 diagnoses. In such situations, if the symptomatic presentation is severe enough to cause clinically significant impairment or distress and represents a biological or psychological dysfunction in the individual, a diagnosis of a mental disorder is still warranted and the differential comes down to either an Adjustment Disorder or one of the residual Other Specified or Unspecified categories. If the clinical judgment is made that the symptoms have developed as a maladaptive response to a psychosocial stressor, the diagnosis would be an Adjustment Disorder. If it is judged that a stressor is not responsible for the development of the clinically significant symptoms, then the relevant Other Specified or Unspecified category may be diagnosed, with the choice of the appropriate residual category depending on which DSM-5 diagnostic grouping best covers the symptomatic presentation. For example, if the patient’s presentation is characterized by depressive symptoms that do not meet the criteria for any of the disorders included in the DSM-5 chapter “Depressive Disorders,” then Other Specified Depressive Disorder or Unspecified Depressive Disorder is diagnosed (rules regarding which of these two categories to use are provided in the next paragraph). Because stressful situations are a daily feature of most people’s lives, the judgment in this step is centered more on whether a stressor is etiological rather than on whether a stressor is present.
DSM-5 offers two versions of residual categories: Other Specified Disorder and Unspecified Disorder. As the names suggest, the differentiation between the two depends on whether the clinician chooses to specify the reason that the symptomatic presentation does not meet the criteria for any specific category in that diagnostic grouping. If the clinician wants to indicate the specific reason, the name of the disorder (“Other Specified Disorder”) is followed by the reason why the presentation does not conform to any of the specific disorder definitions. For example, if a patient has a clinically significant symptomatic presentation characterized by 4 weeks of depressed mood, most of the day nearly every day, which is accompanied by only two additional depressive symptoms (e.g., insomnia and fatigue), the clinician would record Other Specified Depressive Disorder, Depressive Episode With Insufficient Symptoms. If the clinician chooses not to indicate the specific reason why the presentation does not conform to any of the specific disorder definitions, the Unspecified Disorder designation is used. For example, if the clinician declines to indicate the reason why the depressive presentation does not fit any of the specified categories, the diagnosis Unspecified Depressive Disorder is made instead. The clinician might choose the unspecified option if there is insufficient information to make a more specific diagnosis and the clinician expects that additional information may be forthcoming, or if the clinician decides it is in the patient’s best interest not to be specific about the reason (e.g., to avoid offering potentially stigmatizing information about the patient).
Step 6: Establish the Boundary With No Mental Disorder
Generally, the last step in each of the decision trees is to establish the boundary between a disorder and no mental disorder. This decision is by no means the least important or easiest to make. Taken individually, many of the symptoms included in DSM-5 are fairly ubiquitous and are not by themselves indicative of the presence of a mental disorder. During the course of their lives, most people may experience periods of anxiety, depression, sleeplessness, or sexual dysfunction that may be considered as no more than an expected part of the human condition. To be explicit that not every such individual qualifies for a diagnosis of a mental disorder, DSM-5 includes with most criteria sets a criterion that is usually worded more or less as follows: “The disturbance causes clinically significant distress or impairment in social, occupational, or other important areas of functioning.” This criterion requires that any psychopathology must lead to clinically significant problems in order to warrant a mental disorder diagnosis. For example, a diagnosis of Male Hypoactive Sexual Desire Disorder, which includes the requirement that the low sexual desire causes clinically significant distress in the individual, would not be made in a man with low sexual desire who is not currently in a relationship and who is not particularly bothered by the low desire.
Unfortunately, but necessarily, DSM-5 makes no attempt to define the term clinically significant. The boundary between disorder and normality can be set only by clinical judgment and not by any hard-and-fast rules. What may seem clinically significant is undoubtedly influenced by the cultural context, the setting in which the individual is seen, clinician bias, patient bias, and the availability of resources. “Minor” depression may seem much more clinically significant in a primary care setting than in a psychiatric emergency room or state hospital where the emphasis is on the identification and treatment of far more impairing conditions.
In clinical mental health settings, the judgment regarding whether a presentation is clinically significant is often a nonissue; the fact that the individual has sought help automatically makes it “clinically significant.” More challenging are situations in which the symptomatic picture is discovered in the course of treating another mental disorder or a medical condition, which, given the high comorbidity among mental disorders and between mental disorders and medical conditions, is not an uncommon occurrence. Generally, as a rule of thumb, if the comorbid psychiatric presentation warrants clinical attention and treatment, it is considered to be clinically significant.
Finally, some conditions that can impair functioning, such as Uncomplicated Bereavement, may still not qualify for the use of an Other Specified or Unspecified Disorder category because they do not represent an internal psychological or biological dysfunction in the individual, as is required in the DSM-5 definition of a mental disorder. Such “normal” but impairing symptomatic presentations may be worthy of clinical attention, but they do not qualify as a mental disorder and should be diagnosed with a category (usually a V or Z code, corresponding to ICD-9-CM or ICD-10-CM, respectively) from the DSM-5 Section II chapter “Other Conditions That May Be a Focus of Clinical Attention,” which is included after the mental disorders chapters.
Differential Diagnosis and Comorbidity
Differential diagnosis is generally based on the notion that the clinician is choosing a single diagnosis from among a group of competing, mutually exclusive diagnoses to best explain a given symptom presentation. For example, in a patient who presents with delusions, hallucinations, and manic symptoms, the question is whether the best diagnosis is Schizophrenia, Schizoaffective Disorder, or Bipolar Disorder With Psychotic Features; only one of these can be given to describe the current presentation. Very often, however, DSM-5 diagnoses are not mutually exclusive, and the assignment of more than one DSM-5 diagnosis to a given patient is both allowed and necessary to adequately describe the presenting symptoms. Thus, multiple decision trees may need to be consulted to adequately cover all of the important clinically significant aspects of the patient’s presentation. For example, a patient who presents with multiple unexpected panic attacks, significant depression, binge eating, and excessive substance use would require a consideration of the following decision trees: panic attacks (Decision Tree for Panic Attacks), depressed mood (Decision Tree for Depressed Mood), appetite changes or unusual eating behavior (Decision Tree for Appetite Changes or Unusual Eating Behavior), and excessive substance use (Decision Tree for Excessive Substance Use). Moreover, because of comorbidity within diagnostic groupings, multiple passes through a particular decision tree may be required to cover all possible diagnoses. For example, it is well recognized that if a patient has one Anxiety Disorder (e.g., Social Anxiety Disorder [Social Phobia]), he or she is more likely to have other comorbid Anxiety Disorders (e.g., Separation Anxiety Disorder, Panic Disorder). The anxiety decision tree (Decision Tree for Anxiety), however, helps to differentiate among the various Anxiety Disorders, and therefore a pass through the tree will result in the diagnosis of only one of the Anxiety Disorders. Multiple passes through the anxiety tree, answering the key questions differently each time depending on which anxiety symptom is the current focus, are needed to capture the comorbidity.
The use of multiple diagnoses is in itself neither good nor bad as long as the implications are understood. A naïve and mistaken view of comorbidity might assume that a patient assigned more than one descriptive diagnosis actually has multiple independent conditions. This is certainly not the only possible relationship. In fact, there are six different ways in which two so-called comorbid conditions may be related to one another: 1) condition A may cause or predispose to condition B; 2) condition B may cause or predispose to condition A; 3) an underlying condition C may cause or predispose to both conditions A and B; 4) conditions A and B may, in fact, be part of a more complex unified syndrome that has been artificially split in the diagnostic system; 5) the relationship between conditions A and B may be artifactually enhanced by definitional overlap; and 6) the comorbidity is the result of a chance co-occurrence that may be particularly likely for those conditions that have high base rates. The particular nature of the relationships is often very difficult to determine. The major point to keep in mind is that “having” more than one DSM-5 diagnosis does not mean that there is more than one underlying pathophysiological process. Instead, DSM-5 diagnoses should be considered descriptive building blocks that are useful for communicating diagnostic information.
How to Use the Handbook: Case Example
To demonstrate how to use the diagnostic tools provided in this handbook to determine a differential diagnosis, consider the following case, adapted from DSM-5 Clinical Cases, edited by John W. Barnhill, M.D. (pp. 32–34).
Adapted with permission from Heckers S: “Sad and Psychotic,” in DSM-5 Clinical Cases. Edited by Barnhill JW. Washington, DC, American Psychiatric Publishing, 2014, pp. 32–34. Copyright © 2014 American Psychiatric Association.
John is a 25-year-old single, unemployed white man who has been seeing a psychiatrist for several years for management of psychosis, depression, anxiety, and abuse of marijuana and alcohol.
After an apparently normal childhood, John began to show dysphoric mood, anhedonia, low energy, and social isolation by age 15. At about the same time, John began to drink alcohol and smoke marijuana every day. In addition, he developed recurrent panic attacks, marked by a sudden onset of palpitations, diaphoresis, and thoughts that he was going to die. When he was at his most depressed and panicky, he twice received a combination of sertraline 100 mg/day and psychotherapy. In both cases, his most intense depressive symptoms lifted within a few weeks, and he discontinued the sertraline after a few months. Between episodes of severe depression, he was generally seen as sad, irritable, and amotivated. His school performance declined around tenth grade and remained marginal through the rest of high school. He did not attend college, which had been his parents’ expectation, but instead lived at home and did odd jobs in the neighborhood.
Around age 20, John developed a psychotic episode in which he had the conviction that he had murdered people when he was 6 years old. Although he could not remember who these people were or the circumstances, he was absolutely convinced that it had happened, something that was confirmed by continuous voices accusing him of being a murderer. He also became convinced that other people would punish him for what happened when he was 6 years old and thus he also feared for his life. Over the next 2 or 3 weeks, he became guilt ridden and preoccupied with the idea that he should kill himself by slashing his wrists, culminating in his being psychiatrically hospitalized because of his parents’ concerns that he would act on these delusions. Although his affect on admission was anxious, within a couple of days he also became very depressed with accompanying symptoms of dysphoria, prominent anhedonia, poor sleep, and decreased appetite and concentration. With the combined use of antipsychotic and antidepressant medications, both the depression and psychotic symptoms remitted after an additional 4 weeks. Thus, the total duration of the psychotic episode was approximately 7 weeks, 4 of which were also characterized by the depressive episode. He was hospitalized with the same pattern of symptoms two additional times before age 22, starting out with a couple of weeks of delusions and hallucinations related to his conviction that he had murdered someone when he was a child, followed by severe depression lasting an additional month. Both of those relapses occurred while he was apparently adhering to reasonable dosages of antipsychotic medication. For the past 3 years, John has been adherent to clozapine and has been without any further episodes of hallucinations, delusions, or depression.
John began to abuse marijuana and alcohol at age 15. Before the onset of psychosis at age 20, he smoked several joints of marijuana almost daily and binge drank on weekends with occasional blackouts. After the onset of psychosis, his marijuana use decreased significantly, yet he continued to have two more psychotic episodes through age 22 (as described above). He started attending Alcoholics Anonymous and Narcotics Anonymous groups, achieved sobriety from marijuana and alcohol at age 23, and has since remained sober.
This case presents with both prominent psychotic symptoms (delusions and hallucinations) and mood symptoms (depression). Thus, the clinician can start the differential diagnosis process with any of the following decision trees: delusions (Decision Tree for Delusions), hallucinations (Decision Tree for Hallucinations), or depressed mood (Decision Tree for Depressed Mood). Given the especially prominent nature of the delusions, we first start with the delusions decision tree (Decision Tree for Delusions). The first question, whether the beliefs are a manifestation of a culturally or religiously sanctioned belief system, can be answered “no” because John’s fixed belief that he murdered people when he was age 6 is not a manifestation of any sanctioned belief system and is thus appropriately considered to be a delusion. The next question, regarding whether his delusions are due to the physiological effects of a substance, must be seriously considered given the fact that his delusions first emerged at age 20 during a time when he was smoking several joints of marijuana almost daily. To answer this question, we need to consider Step 2 of the six differential diagnosis steps presented earlier in this chapter, which provides guidance on how to rule out a substance etiology. In determining whether there is a causal relationship between the marijuana use and the delusions, we need to determine whether all three of the following conditions are true: 1) that there is a close temporal relationship between marijuana use and the onset and maintenance of the delusions, 2) that the pattern of marijuana use is consistent (in terms of dosage and duration) with the development of delusions, and 3) that there is no alternative (i.e., non-substance/medication-induced) explanation for the delusions. Although it is not common for marijuana to cause florid delusions, heavy marijuana use in some vulnerable individuals can result in delusions during Marijuana Intoxication, so the second condition (i.e., substance use is heavy and/or prolonged enough to induce the symptom) is met. In evaluating the first condition, however, although the delusions emerged during heavy marijuana use, the fact that the delusions persisted in the hospital when John was abstinent from marijuana and then subsequently reoccurred when his marijuana use was minimal indicates that the delusions cannot be explained as a manifestation of his marijuana use. Thus, the answer to the second question in the delusions decision tree, regarding whether there is a cannabis etiology for the delusions, is “no.” The absence of any reported general medical conditions in John also rules out a medical etiology, and therefore the answer to the following question is also “no.”
After ruling out cultural and religious, substance/medication-induced, and general medical etiologies for John’s delusions, we then must differentiate among the primary psychotic and mood disorders as possible explanations for the delusions. The next question, which asks whether the delusions have occurred only in the context of an episode of elevated, expansive, or irritable mood, is answered “no” because of the absence of a history of manic or hypomanic symptoms. The subsequent question, about whether the delusions have occurred only in the context of an episode of depressed mood, is also answered “no” because the delusions also occurred at times when John was not experiencing a depressive episode (i.e., each psychotic episode is characterized by a several-week period of delusions before the development of the severe depressive symptoms).
The next block of questions in the delusions tree provides the differential diagnosis of non-mood-restricted delusions. The question inquiring whether the delusions last for 1 month or more is answered “yes” (i.e., each time the delusions have occurred, they lasted for several weeks), moving us for the first time to the right in the decision tree to consider the differential between Schizophrenia, Schizophreniform Disorder, Schizoaffective Disorder, Delusional Disorder, and Bipolar or Major Depressive Disorder With Psychotic Features. The subsequent question about whether the delusions are accompanied by other psychotic symptoms characteristic of Schizophrenia (i.e., hallucinations, disorganized speech, grossly disorganized or catatonic behavior, or negative symptoms) is also answered “yes” given that in John’s case the delusions of having murdered a person when he was a child are accompanied by accusatory auditory hallucinations. The next question (i.e., whether there is a history of Major Depressive or Manic Episodes) is answered “yes” given the history of recurrent Major Depressive Episodes, as is the following question (i.e., whether during an uninterrupted period of illness the psychotic symptoms occur concurrently with the mood episodes) because the delusions and hallucinations continued to persist after the Major Depressive Episodes emerged, thus indicating a period of overlap.
The next question, which provides the crucial differential diagnostic distinction between Schizoaffective Disorder and Schizophrenia, asks whether, during an uninterrupted period of illness, the mood episodes have been present for a minority of the total duration of the active and residual phases of the illness. In John’s case, each of the psychotic episodes was present for approximately 7–8 weeks, with about 4 of those weeks characterized by the simultaneous occurrence of a severe Major Depressive Episode. Therefore, it is not the case that the mood episodes were present for only a minority of the time during an uninterrupted episode of illness (they were in fact present for a majority of the time), so the question is answered “no,” ruling out the diagnoses of both Schizophrenia and Schizophreniform Disorder. The next question, regarding whether delusions and hallucinations have occurred for at least 2 weeks in the absence of a Major Depressive Episode or Manic Episode, is answered “yes” (i.e., for the first 3 or 4 weeks of the psychotic episode, John was anxious but not suffering from significant depressed mood), bringing us to the terminal branch of the delusions decision tree (Decision Tree for Delusions) and the diagnosis of Schizoaffective Disorder. It should be noted that given the complete co-occurrence of the delusions and hallucinations during the psychotic episodes, had we started with the hallucinations tree (Decision Tree for Hallucinations) instead of the delusions tree, we would have gone through almost the exact same sequence of steps to arrive at the diagnosis of Schizoaffective Disorder, given the similarity of the branching structure of the delusions and hallucinations trees.
Alternatively, we could have approached this case from the perspective of John’s severe depressive symptoms and instead started with the depressed mood decision tree (Decision Tree for Depressed Mood). The first question in this tree inquires about a substance etiology for the depressive symptoms. Applying the same principles discussed above with regard to the relationship between John’s marijuana use and his delusions, this question can also be answered in the negative because although the marijuana use is sufficient to cause depressed mood, the fact that John continued to experience episodes of severe depression after he stopped his heavy use of marijuana indicates that, like the delusions, his depression cannot be considered to have been induced by the marijuana use. The next question asks whether the depression is due to the physiological effects of a general medical condition, and that question can also be answered “no” because of the absence of any history of medical problems. The next question asks whether the depressed mood was part of a Major Depressive Episode. The answer to that question is “yes” given that the depressive periods that developed after the onset of delusions and hallucinations were characterized by approximately 4 weeks of dysphoric mood, prominent anhedonia, poor sleep, decreased appetite, and reduced concentration, thus meeting syndromal criteria for a Major Depressive Episode. Note that the decision tree does not end at this point but that the diagnostic flow continues onward because Major Depressive Episode is not a codable diagnostic entity in DSM-5 but instead comprises one of the building blocks for the diagnoses of Bipolar I or Bipolar II Disorder, Major Depressive Disorder, and Schizoaffective Disorder. The next question, about the presence of clinically significant manic or hypomanic symptoms, is answered “no,” bringing us to a consideration of the relationship between the Major Depressive Episodes and the psychotic symptoms. The question about whether there is a history of delusions or hallucinations is answered “yes,” bringing us to the critical question as to whether the psychotic symptoms occur exclusively during Manic or Major Depressive Episodes. In John’s case, the psychotic symptoms have not occurred exclusively during the Major Depressive Episodes (i.e., the delusions and hallucinations occurred on their own for 3–4 weeks prior to the onset of the depressive episode), so the answer to this question is “no.” At this point in the depressed mood decision tree (Decision Tree for Depressed Mood), rather than being offered additional questions, we are told that a Schizophrenia Spectrum or Other Psychotic Disorder is present and are instructed to go to the delusions tree (Decision Tree for Delusions) or hallucinations tree (Decision Tree for Hallucinations) for the differential diagnosis, resulting in the diagnosis of Schizoaffective Disorder.
After arriving at the diagnosis of Schizoaffective Disorder through the use of the decision trees, we can refer to the DSM-5 classification in the Appendix to get the diagnostic code for Schizoaffective Disorder and/or we can review the differential diagnosis table for Schizoaffective Disorder in Chapter 3 (Table 3.2.2) to confirm that the key contenders to a diagnosis of Schizoaffective Disorder have been appropriately ruled out. The two main diagnostic contenders in this case are Schizophrenia and Major Depressive Disorder With Psychotic Features. Accordingly, the differential diagnosis table for Schizoaffective Disorder notes that Schizophrenia is differentiated from Schizoaffective Disorder by virtue of the fact that Schizophrenia is characterized by mood episodes that “have been present for a minority of the total duration of the active and residual periods of the illness.” In John’s case, each episode of the illness was characterized by a Major Depressive Episode being present for more than half of the time (i.e., about 4 weeks) of the total duration (i.e., 7–8 weeks), thus ruling out the diagnosis of Schizophrenia. Moreover, the table also notes that Schizoaffective Disorder is differentiated from Major Depressive Disorder With Psychotic Features by virtue of the fact that Major Depressive Disorder With Psychotic Features is characterized by psychotic symptoms that occur exclusively during Major Depressive Episodes. In John’s case, the psychotic symptoms were not confined exclusively to the depressive episodes, ruling out the diagnosis of Major Depressive Disorder With Psychotic Features.