Primary care physicians surveyed in this study mistakenly interpreted improved survival and increased detection with screening as evidence that screening saves lives