Abstract

The prevalence of symptoms in the community is sometimes used as an index of untreated morbidity. However, such an index can be very misleading unless it makes allowance for differences in symptom severity between declared and undeclared patients. Recent attenders at one health centre were compared with controls who had not seen their GPs for at least three months. Comparisons of symptom severity were made between attenders who had reported one or more of seven selected symptoms and non-attenders who said they were troubled by the same symptoms. For the symptoms selected, it was found that a high proportion of sufferers in both groups were of at least two months' chronicity. Prevalence rates of up to 33% for backache and tiredness in older women were found in non-attending controls, but symptom severity was significantly less than in patients who had recently consulted their doctors with the same symptom. Symptoms were both more prevalent and more severe among women than among men. It seemed unlikely, however, that this difference could explain the higher consultation rates for women, because the same excess of women over men persisted among consulters. It seems that for these symptoms increasing severity is associated with an increasing probability of attending the surgery but that the symptoms functions more often as a background factor than as a precipitant.