Predicting medically unexplained physical symptoms and health care utilization - A symptom-perception approach

Objectives: The present study investigated the contribution of demographic characteristics (age, gender, socioeconomic status [SES]) and symptom-perception variables to unexplained physical symptoms and health care utilization. In addition, the consequences of the use of four frequently applied symptom-detection methods for relations among study variables were examined. Method: A group of 101 men and women were administered a standardized interview and several questionnaires. Their general practitioners (GPs) rated (un)explained symptoms and consultations over the previous year. Results: Path analyses showed that direct and indirect effects on symptoms and GP consultations depend on method of symptom detection, the largest difference being between self-reported symptoms and registered symptoms. The model including self reported common symptoms demonstrated the direct and indirect effects of the symptom-perception variables: chronic disease, negative affectivity, selective attention to bodily sensations, and somatic attribution. In the model including registered symptoms, only chronic disease and SES showed effects on symptoms and GP consultations. Conclusion: This study demonstrates the usefulness of a symptom-perception approach to the experience of unexplained symptoms, the importance of selection of a symptom-detection method, and the need for different models for the explanation of daily experienced symptoms and their presentation in health care. (C) 2002 Elsevier Science Inc. All rights reserved.