Seeking care for cancer relevant symptoms.

Although it has been estimated that a fourth of the individuals who die of cancer each year might have survived if their cancer had been detected earlier, little is known about the characteristics of individuals who seek care for cancer relevant symptoms. This study reports data from a 1 yr panel study of 1210 individuals in metropolitan Los Angeles who were asked to report any symptoms they had experienced during each six week interval between interviews. These symptoms were classified according to the National Ambulatory Symptom Classification and then rated for their cancer relevance by three university-based oncologists working independently. There were 35 symptoms rated as cancer relevant. Inter-rated reliability was 0.73. Analysis is directed towards the distribution of these symptoms in the population, the distribution of physician visits for these symptoms and the structural and social psychological factors that predict the decision to seek medical care for these symptoms. The implications of the results for health education are twofold. First, because these symptoms occur homogeneously across subgroups, education programs targeted to specific subpopulations are likely to be inefficient. Second, even if successful education campaigns could be undertaken, it is unlikely that all physicians would agree that all of the increased demand for services would be appropriate.