Criteria-based diagnostics generate excessive and incorrect diagnoses

Pål Gjerden About the author

Psychiatric diagnoses are for the most part based on standardised forms as diagnostic instruments. The use of forms oversimplifies a complex reality and leads to overdiagnosis.

Psychiatry strives to defend its position as a medical specialty. Instead of accepting that psychiatry and clinical medicine are essentially different, there is an insistence that psychiatry is scientifically based in the same way as clinical medicine. A pivotal premise for this is that the basis for all therapy must be a so-called correct diagnosis.

Psychiatric diagnoses are a litany of symptoms which experts consider reasonable to merge. Historically, psychiatric diagnoses have been a combination of speculation, theories and prejudices. Any attempt to define what is valid in psychiatric diagnostics is fraught with uncertainty.

The main argument for our current system of diagnoses is the desire to create a reliable system. In order to achieve this, criteria-based diagnostics are recommended – as evident from the last three national professional guidelines for psychiatry from the Norwegian Directorate of Health (1 – 3). The therapist must tick «yes» or «no» in boxes on a form, or grade symptoms on a numerical scale. The diagnosis is made when a predetermined score is achieved.

Simplified description

A plethora of such forms exist, and common to them all is the simplistic description of symptoms. The fact that mental illness involves complicated, diffuse symptoms that may be invisible or difficult to verbalise is ignored. The patient’s personal history is considered to be of little relevance.

Auditory hallucinations may be a cardinal symptom of psychosis. For obvious reasons, hallucinations cannot be verified and are difficult to describe. Are they voices or thoughts, or something in between? Are the voices inside the head or do they arise from outside? Are the people who are speaking real? Is the patient certain of the voices, or in doubt? It is not a diagnostic requirement that the patient must be convinced that the hallucinations represent reality.

In practice, specialists will differ in their assessments of this (4). Many psychiatrists – but certainly not all – would say that if the person is aware that thoughts/voices are generated inside his head, he has not lost contact with reality and is presumably not psychotic. For a psychotic patient, the question of whether he hears voices may be completely meaningless. Of course he hears voices – if he is not deaf. The most basic diagnostic concepts may be that diffuse.

Symptoms are not diagnoses

Despite fulfilment of all diagnostic criteria, the diagnosis may be incorrect. Identical symptoms may arise in different contexts and define different diagnoses. The diagnoses change, new diagnoses are added and others fall out of vogue – although we must assume that the reality remains the same. It is nevertheless appropriate to retain some categories; it is significant to know whether the person is psychotic or not, even though the benefit of more refined diagnostics may be a matter for discussion. It is not of decisive importance whether the diagnosis is paranoid psychosis or paranoid schizophrenia – in principle the treatment will be the same. However, it is important to distinguish a primary psychotic disorder from an obsessive compulsive or dissociative disorder. These are patients who require different approaches, both in terms of psychotherapy and drug therapy.

The consequences of misdiagnosis may be grave, particularly with regard to diagnoses of psychosis. To be misdiagnosed with psychosis early in life may be devastating, over the years I have had the misfortune to witness several such instances. I cannot remember having read a single guideline that mentions the possibility of false positive diagnoses as unfortunate consequences of relying on criteria-based diagnostic forms.

Ticking boxes rather than performing an assessment

Certain diagnostic criteria may have value in some contexts, for example in research. All research implies some kind of simplification. The problem is that these simple checkbox forms are increasingly perceived as representing a kind of truth. In psychiatry today, assessing a patient has increasingly become synonymous with crossing off items on a pile of forms, all equally simple and all with a minimum space for personal reflection. The professional and personal evaluation has almost become absent in psychiatry – the result after filling out a form or two is defined as the answer.

All of these forms provide less information than a clinical interview. In my opinion they give an oversimplified and inadequate picture of the patient. Relying on them as purveyors of truth may mislead the therapist and be potentially harmful for the patient.

The argument for using these diagnostic forms is primarily to avoid overlooking a diagnosis that would not be made on the basis of a clinical assessment. This leads to an increased number of diagnoses, in other words, to overdiagnosis. The most obvious example is the Mini International Neuropsychiatric Interview, MINI (5), which was developed as a triage instrument for research purposes. This has apparently been forgotten.

The MINI form is used extensively as a diagnostic tool and is recommended by the Norwegian Directorate of Health, even though the Norwegian Knowledge Centre for the Health Services points out the risk of overdiagnosis and finds no evidence for either its validity or reliability (6, 7). You would have a very carefree life if you did not conform to at least one diagnosis after having been assessed by the MINI form.

It is my opinion that psychiatric diagnosis according to criteria-based forms results in overdiagnosis. Overdiagnosis equals misdiagnosis.

Anbefalte artikler