When you evaluate a person in mental distress, your goal should be to create a therapeutic alliance, but the tangible result of an evaluation is a diagnosis. This diagnosis is the foundation of a treatment plan.

In earlier versions of DSM, diagnoses were described in a multiaxial, or five-axis, system. Practitioners broke a diagnosis into five components: mental illnesses, personality disorders, general medical conditions, psychosocial problems, and global functioning. At its best, the multiaxial system encouraged practitioners to understand a person’s distress from multiple perspectives: a biological account of mental illness, a psychological account of personality, a mechanistic account of physical illness, a subjective list of psychosocial factors, and a standardized assessment of functioning. At its worst, the multiaxial system reinforced divisions between mind and body; allowed personality disorders to be used as pejorative slurs; included inconsistent accounts of psychosocial functioning; and jumbled together categories, lists, and assessments. It turned out to be a messy recipe.

The authors of DSM-5 (American Psychiatric Association 2013) reorganized the multiaxial system into a problem list. For physicians, the problem list is familiar because it is already in use throughout medicine. Nonphysicians, however, may benefit from a brief introduction to the problem list. Simply put, a problem list is a comprehensive and hierarchical catalog of the problems addressed during a current encounter.

To benefit communication, the items on the list should be standardized. There are many ways to account for mental distress and mental illness. Individual practitioners may focus on dysfunctional neural circuits, traumatic experiences, or maladaptive personality traits. When these practitioners wish to speak with each other, they need a standard list. The standard list we favor is DSM-5 because it is the consensus diagnostic system of contemporary psychiatry, our way for mental health practitioners to work together while we await a diagnostic system with greater validity.

One reminder that we are awaiting a diagnostic system with improved validity is that the diagnoses generated by a DSM-5 interview are called disorders rather than diseases or illnesses. Physicians usually think in terms of diseases, which can be described as pathological abnormalities in the structure and function of body organs and systems. Patients usually present with illnesses, their experience of pathological abnormalities or of being sick. From a distance, diseases and illnesses may seem like the same experience viewed from the different perspectives of patient and physician. However, diseases and illnesses are often divergent experiences, not just different perspectives, as anthropologists have repeatedly documented (Estroff and Henderson 2005).

Disorders are a kind of middle path between disease and illness. The term disorder acknowledges the complex interplay of biological, social, cultural, and psychological factors in mental distress. Broadly speaking, a disorder simply indicates a disturbance in physical or psychological functioning. Use of the disorder label to describe mental distress draws attention to how mental distress impairs a person’s functioning, suggests the complex interplay of events that result in mental distress, and implicitly acknowledges the limits of our knowledge about the causes of mental distress (Kendler 2012). Practitioners in the field do not yet know enough to be more precise. We accept the ongoing use of disorder in our diagnostic systems as an opportunity for humility and a spur to further study but primarily as a way to communicate together.

For DSM-5 to work as a common language, practitioners need to agree on the features of a specific diagnosis. Standardization does not work without specificity. Imagine a recipe that lists “a serving of fat” as an ingredient. Someone following the recipe would be confused. Did the author of the recipe mean a spoonful of bacon drippings or a half cup of coconut oil? Each is possible, but each results in a different dish; the recipe becomes more of a personal inspiration than a communal instruction. Similarly, practitioners should recognize that characterizing a person as having “an unspecified mental disorder” inadequately communicates the precise nature of the patient’s illness to other practitioners.

We encourage practitioners to select the most specific diagnosis for which a patient qualifies. If you believe a grandmother is depressed, determine not only whether her depression constitutes a major depressive episode but also whether it is a single or recurrent episode, with or without psychotic features, and whether it is mild, moderate, or severe. This level of specificity enables communication with other practitioners and informs their treatment. For example, although we recognize that an adult with a mild first depressive episode would be treated differently from an adult with a severe recurrent depressive episode with psychotic features, we would have trouble deciding how to proceed with an adult diagnosed with a nonspecific mental disorder. Identifying a specific disorder improves communication with a patient’s other practitioners while also communicating to the patient and his caregivers your diagnostic ability and understanding of his particular illness. Diagnosis is, itself, a response to a patient’s suffering, because giving a specific name to the seemingly unnameable is salutary. (Diagnosis also improves communication with regulators and third-party payers, who frequently reimburse better for more specific diagnoses.)

Sometimes, however, a specific diagnosis is inappropriate. When you are uncertain or need additional information, a provisional diagnosis is always preferable to a specific but inaccurate one. The goal is to eventually arrive at the most specific diagnosis possible. It is discouraging to review medical records in which a person’s diagnosis has remained poorly characterized for years.

Even if your diagnoses lack specificity, you can make them comprehensive. They should include all problems—mental disorders, general medical conditions, and psychosocial problems—that are currently diminishing a person’s ability to function. As you know by now, we use DSM-5 to describe mental disorders, including the adverse effects from psychiatric treatment, as described in Section II of DSM-5. You also need to describe general medical conditions that currently affect a person’s function. You do not need to list well-healed injuries. To describe psychosocial problems that influence a person’s health, we favor using the standardized list of ICD-10 codes. Some of the most relevant Z codes are found in Chapter 12, “Rating Scales and Alternative Diagnostic Systems,” but the complete list of Z codes, numbered Z00–Z99, is found in the ICD chapter called “Factors Influencing Health Status and Contact With Health Services,” which can be found online at http://www.cms.gov/Medicare/coding/ICD10/2016-ICD-10-CM-and-GEMs.html.

Finally, the patient’s mental disorders, general medical conditions, and psychosocial problems should be ordered hierarchically. Those problems that are the focus of your treatment at a specific time should lead the list. For example, an older adult may have hypertension, but if you are treating him for an episode of major depressive disorder following an intentional overdose, then his first two problems are his major depressive disorder and his suicide attempt. If you evaluate him again 2 months later and he has recovered from his ingestion and his depressive symptoms are decreased, then his depressive episode and suicide attempt would be lower on the problem list. A well-ordered problem list communicates the focus of your treatment to everyone who reviews your record.