“Medically unexplained” symptoms and symptom disorders in primary care: prognosis-based recognition and classification

Many patients consult their GP because they experience bodily symptoms. In a substantial proportion of cases, the clinical picture does not meet the existing diagnostic criteria for diseases or disorders. This may be because symptoms are recent and evolving or because symptoms are persistent but, either by their character or the negative results of clinical investigation cannot be attributed to disease: so-called “medically unexplained symptoms” (MUS).

MUS are inconsistently recognised, diagnosed and managed in primary care. The specialist classification systems for MUS pose several problems in a primary care setting. The systems generally require great certainty about presence or absence of physical disease, they tend to be mind-body dualistic, and they view symptoms from a narrow specialty determined perspective. We need a new classification of MUS in primary care; a classification that better supports clinical decision-making, creates clearer communication and provides scientific underpinning of research to ensure effective interventions.

We propose a classification of symptoms that places greater emphasis on prognostic factors. Prognosis-based classification aims to categorise the patient’s risk of ongoing symptoms, complications, increased healthcare use or disability because of the symptoms. Current evidence suggests several factors which may be used: symptom characteristics such as: number, multi-system pattern, frequency, severity. Other factors are: concurrent mental disorders, psychological features and demographic data. We discuss how these characteristics may be used to classify symptoms into three groups: self-limiting symptoms, recurrent and persistent symptoms, and symptom disorders. The middle group is especially relevant in primary care; as these patients generally have reduced quality of life but often go unrecognised and are at risk of iatrogenic harm. The presented characteristics do not contain immediately obvious cut-points, and the assessment of prognosis depends on a combination of several factors.

Three criteria (multiple symptoms, multiple systems, multiple times) may support the classification into good, intermediate and poor prognosis when dealing with symptoms in primary care. The proposed new classification specifically targets the patient population in primary care and may provide a rational framework for decision-making in clinical practice and for epidemiologic and clinical research of symptoms.