The Most Common Diagnostic Errors in Psychiatry and How to Avoid Them

Why psychiatric diagnosis is difficult

Diagnosing a mental disorder differs from many diagnoses in the other fields of medicine. There is no single laboratory or imaging test that would confirm or rule out, for example, depression or an anxiety disorder. The clinician (a psychiatrist or a clinical psychologist) relies mainly on the interview, on observation, and on what the patient and their environment report. A diagnosis is therefore a reconstruction of the clinical picture, not the reading of a test result.

An additional difficulty is introduced by the non-specificity of symptoms. Low mood, anxiety, sleep problems, difficulties with concentration or irritability appear in many different disorders, and also in states that are not a disorder at all. The same symptom may be part of a depressive episode, a reaction to stress, a physical illness (an illness of the body), or the effect of a substance. A symptom on its own does not point to a diagnosis; it only points to a direction for further thinking.

Added to this are practical factors: limited consultation time, incomplete information, the variability of the picture over time, and the fact that a patient may be unable or unwilling to provide everything that matters. ICD-11 diagnostics provides an organised map of disorders and a shared language, but it does not remove these difficulties. An awareness of the typical errors is part of the clinical craft and reduces the risk of mistake.

Error one: a hasty diagnosis without differential diagnosis

The most common and most serious error is making a diagnosis on the basis of a first impression, without carrying out a differential diagnosis. Differential diagnosis is the systematic consideration of alternative explanations for the same picture of symptoms and the checking of which of them fits best. Skipping this step means the clinician stops at the first hypothesis that came to mind.

The psychological mechanism behind this error is sometimes called premature closure: the moment at which a clinician treats a matter as settled before having actually considered it. It is often accompanied by what is known as anchoring, that is, an excessive attachment to the first piece of information or the first label heard, for example from earlier documentation or from the patient themselves.

The consequences can be serious. A classic example is the diagnosis of recurrent depressive disorder (code 6A71) in a person in whom it was never checked whether, in the past, there had been an episode of elevated mood, increased energy and reduced need for sleep. If such an episode occurred, the correct diagnosis may be bipolar disorder (code 6A60 or 6A61), and the management looks different. Skipping differential diagnosis here leads to an error with real consequences.

The protection against this error is discipline of thinking. The ICD-11 guidelines, collected in the CDDR document (Clinical Descriptions and Diagnostic Requirements), include, for each disorder, separate notes on the boundaries with other entities. Deliberately asking oneself what else could explain this picture is a simple yet effective barrier against premature closure.

Error two: ignoring the duration and course of symptoms

Many mental disorders differ from transient states not in the type of symptoms, but in how long the symptoms persist and how they develop over time. For this reason the ICD-11 guidelines often indicate time thresholds: a minimum period of symptoms that allows a given diagnosis to be considered at all. Skipping this requirement is a frequent source of errors.

An example is diagnosing generalised anxiety disorder (code 6B00) in a person who, for two weeks, has been intensely worried about a specific, difficult matter. This disorder requires that excessive anxiety and worry persist for many months and concern many areas of life. A brief, situational reaction, even an intense one, does not meet this description and may instead be an understandable reaction to stress rather than a disorder.

The error also works the other way round. Acute and transient psychotic disorder has, by definition, a short course, and the persistence of psychotic symptoms over a longer period directs the diagnosis towards other entities, such as schizophrenia (code 6A20). A clinician who does not carefully establish the onset of symptoms and their dynamics risks attributing the picture to the wrong entity.

Assessing the course also includes distinguishing a single episode from a recurrent state, and a symptomatic phase from remission (a period in which the symptoms have resolved). ICD-11 treats this information as part of the diagnosis, not as a secondary detail. A diagnosis made without a time axis is incomplete and prone to error.

Error three: overlooking physical causes and the effect of substances

Mental health symptoms do not always result from a primary mental disorder. They may be a direct consequence of a physical illness, of the action of medications, or of the use of psychoactive substances, including alcohol. Overlooking this possibility is one of the most serious errors, because it leads to treating the symptom rather than its actual cause.

Hypothyroidism, some neurological disorders, deficiencies, adverse effects of medications, or conditions involving pain can produce a picture resembling depression, anxiety or a decline in cognitive function. For such situations ICD-11 provides separate categories of secondary mental syndromes, for example secondary mood syndrome (code 6E62), which states explicitly that the source of the symptoms is another illness.

Similarly, mental health symptoms may accompany substance use or withdrawal syndromes. Disorders due to alcohol use (code 6C40) can imitate or overlap with mood and anxiety disorders. If the picture is better explained by the action of a substance, a diagnosis of an independent, primary mental disorder is not appropriate.

The practical protection is to treat the ruling out of physical causes and substances as a mandatory step rather than an option. An interview covering chronic illnesses, medications taken, substance use and, where needed, additional investigations is not a formality. It is part of sound ICD-11 diagnostics, without which a diagnosis of a mental disorder cannot be regarded as certain.

Error four: confusing a disorder with a normal reaction

Sadness, anxiety, anger, tension and grief are natural parts of human life, not symptoms of illness. Their mere presence does not indicate a disorder. Confusing an understandable reaction with pathology, that is, the over-medicalisation of ordinary experiences, is an error just as serious as overlooking a genuine disorder.

ICD-11 builds the boundary with normality directly into the descriptions of disorders. The CDDR guidelines include separate notes indicating which reactions are expected and proportionate to the situation. Most diagnoses additionally require that the symptoms cause significant distress or significant impairment of functioning in important areas of life, such as relationships, work or education. This requirement serves as a cut-off threshold.

A good example is the reaction to loss. Grief after the death of a close person is a painful but natural process and is not, in itself, a disorder. A separate diagnosis, prolonged grief disorder (code 6B42), is considered only when the reaction is exceptionally long and impairing, clearly going beyond what is understandable in the given cultural and personal context.

The boundary between a disorder and normality is always set by the same dimensions: the type and severity of symptoms, their duration, and their impact on functioning and on the level of distress. A reaction proportionate to the situation, resolving within a reasonable time and not causing lasting disruption, lies within the range of normality, however unpleasant it may be.

Error five: counting symptoms instead of assessing essential features

A common misconception is treating a diagnosis like a simple survey: if you gather the right number of symptoms from a list, the diagnosis appears automatically. ICD-11 deliberately moves away from this kind of thinking. Instead of rigidly counting points, the guidelines use the concept of essential features: elements of the clinical picture whose presence is required for a given diagnosis to be considered at all.

The difference is fundamental. A person may recognise many accompanying symptoms in themselves and yet still not meet the description of a disorder if its core, defining feature is not present. In generalised anxiety disorder (6B00), the essential feature is persistent, difficult-to-control excessive anxiety or worry. Muscle tension, sleep problems or irritability are accompanying features that reinforce the picture but do not replace the essential feature.

The mechanical summing of symptoms leads to errors in both directions. A disorder can be over-diagnosed in a person with many non-specific complaints in whom the core of the picture is missing. A disorder can also be overlooked in a person with few but defining symptoms. The ICD-11 guidelines require an assessment of whether an essential pattern is present, not a mere tally.

This approach is more flexible but also more demanding. Screening tools and questionnaires are helpful and indicate that it is worth looking at a problem more closely, but a questionnaire result is not a diagnosis. A diagnosis arises only from the clinical interpretation of the whole picture.

Error six: overlooking the co-occurrence of disorders

In clinical practice an ideally single picture is rarely encountered. Co-occurrence is common, that is, the presence of more than one disorder in the same person. The error consists of stopping at the first accurate diagnosis and assuming it explains the whole, while part of the picture remains uncovered.

A classic example is focusing solely on a visible depressive episode and overlooking a co-existing anxiety disorder, a disorder due to substance use, or difficulties of a personality nature. The first diagnosis may be correct, but incomplete, and the overlooked part of the picture may significantly affect the patient's course and needs.

ICD-11 explicitly allows and anticipates such situations. The clinician may make more than one diagnosis if the picture requires it, and additional codes and categories describing the course make it possible to reflect the situation more accurately. The guidelines do not force the whole picture into a single label.

The protection against this error is the deliberate question of whether the diagnosis made explains all the significant elements of the picture. If part of the symptoms goes beyond the description of the chosen entity, a second diagnosis should be considered, rather than stretching the first.

Error seven: overlooking cultural context, age and sex

Behaviours and experiences must be assessed in relation to a person's environment, culture, age and life situation. What is a symptom in one context may, in another, be the norm or an accepted way of expressing difficulties. An assessment detached from context leads both to over-diagnosis and to oversights.

ICD-11 and the CDDR guidelines draw attention to this explicitly: the disorders are accompanied by separate notes concerning cultural, age and sex differences. The same experiences may be named differently, shown differently and understood differently depending on the culture the patient comes from. A clinician unaware of these differences risks a mistaken interpretation.

Age is another important factor. The picture of many disorders differs in children, adults and older people. Some symptoms in children are easily confused with a stage of development, and in older people symptoms of a mood disorder are sometimes confused with natural ageing or with the onset of neurocognitive disorders. Difficulties with attention and activity, in turn, may suggest attention deficit hyperactivity disorder (code 6A05), the diagnosis of which, however, requires a developmental assessment and the presence of symptoms from an early period of life.

Sex affects both the epidemiology and the way symptoms are reported and recognised. Some disorders are sometimes over-diagnosed in one sex and overlooked in the other, if the clinician relies on a stereotyped picture. A deliberate consideration of context is part of a sound assessment, not an addition to it.

Error eight: over-reliance on the label and the labelling effect

Once made, a diagnosis can take on a life of its own. Subsequent clinicians, seeing it in the documentation, may accept it without re-verification, and interpret new symptoms through its lens. This phenomenon is sometimes called diagnostic overshadowing: an earlier label overshadows new information that could undermine it.

Particularly risky is attributing all new complaints to a known diagnosis. In a person with a diagnosed mental disorder, physical symptoms are sometimes too quickly regarded as an element of that disorder, while they may point to an independent illness of the body. It also works the other way round: in a person with a physical illness, a genuine mental disorder is sometimes overlooked.

Connected with this error is also the labelling effect. A diagnostic label affects how the patient is perceived and how they perceive themselves. A diagnosis that is inaccurate or made hastily is therefore not merely a technical mistake; it carries real social and personal consequences.

ICD-11 and good clinical practice treat a diagnosis as a working conclusion open to revision, not a closed verdict. A diagnosis should be reviewed as new information comes in. A readiness to revise one's own diagnosis is a sign of clinical maturity, not of weakness.

How to reduce the risk of error in practice

The first barrier is a structured process. Taking a full history, establishing the time axis of the symptoms, deliberately carrying out a differential diagnosis, and ruling out physical causes and substances takes more time than a first impression, but it is precisely this time that protects against most of the pitfalls described. The ICD-11 guidelines collected in the CDDR are a practical support here, because for each entity they indicate the boundaries with normality and with other disorders.

The second barrier is the deliberate slowing of the pace of thinking in situations typical of errors: when a diagnosis comes to mind too easily, when it rests mainly on someone else's earlier label, or when the picture does not fully fit any entity. The question of what else could explain this should be a routine, not an exception.

The third barrier is review over time and consultation. A diagnosis is not a one-off act; a patient's picture may change, and new information can be decisive. In difficult cases, a second opinion or consultation with another specialist reduces the risk of an error becoming entrenched. Organised tools supporting ICD-11 diagnostics, which make it easier to review related entities and their differential diagnosis, can also be helpful.

Finally, it is worth emphasising the role of the patient and their relatives. An organised account of when the difficulties appeared, how long they have lasted and how they affect daily life shortens the path to an accurate assessment. Knowledge of how diagnostic errors work also helps in asking the clinician good questions. The diagnosis, however, always remains the task of a qualified specialist, and descriptions of disorders are educational in character and do not replace a consultation.

Frequently asked questions