Outcome Ascertainment
This cohort was followed for occurrence of cancer and other chronic diseases by home visits biennially with all surviving cohort members or next of kin (typically their spouse or children) if the cohort members were deceased. The biennial active follow-up for the cohort was virtually complete, with a response rate of 99.8% for the first follow-up survey between 2000 and 2002 and 98.7% for the second follow-up survey between 2002 and 2004. In addition to the in-person active follow-up, we searched the records from the population-based Shanghai Cancer Registry on a monthly basis to assure a timely and complete ascertainment of new cancer cases in this study cohort. The death certificate data from the Shanghai Vital Statistics were used to update vital status of the cohort members and identify causes of death. Medical charts from hospitals were reviewed and the pathologic characteristics of the tumor were recorded. The majority of the cases (n = 246; 96.1%) were pathologically confirmed, with the remainder (n = 10; 3.9%) diagnosed with endoscopy, radiography (barium-air double contrast radiography and/or computed tomographic scans), or ultrasound.