Referral bias in thyroid cancer surgery: direction and magnitude.

AIM: This study was conducted to clarify the impact of referral bias in thyroid cancer surgery.
METHODS: Analysis of 1419 consecutive patients with papillary (n=653), follicular (n=248), and medullary thyroid cancer (n=518) referred to a specialist center for initial surgery or reoperation.
RESULTS: With increasing travel distance (successive postal code areas), mean age decreased among patients referred for initial surgery (from 53 to 35 years for papillary cancer, 65 to 49 years for sporadic medullary cancer, and 40 to 23 years for hereditary medullary cancer, all p< or =0.001; and from 65 to 54 years for follicular cancer, p=0.26). The significant decline in mean age continued among patients reoperated on for papillary cancer (from 53 to 43 years, p<0.001), but was lost among patients reoperated on for medullary cancers. For patients with differentiated, but not medullary cancers, greater travel distance was associated with higher frequencies of extrathyroidal extension at initial surgery (from 17% to 63% for follicular cancer, p=0.003) and reoperation (from 18% to 47% for papillary, and 5% to 44% for follicular cancer, all p<0.001), and higher frequencies of lymph node metastasis at initial surgery (from 23% to 58% for papillary, and 17% to 50% for follicular cancer, p< or =0.008) and reoperation (from 27% to 77% for papillary, and 0% to 35% for follicular cancer, all p<0.001).
CONCLUSION: Referral bias in thyroid cancer surgery can include two components working in opposite directions: age and extent of disease. Controlling for these components should reduce the impact of referral bias on thyroid cancer research.