Limitations of interview-based risk assessment of RF exposure from appliances.

This study was conducted to assess the quality of interview-based exposure estimates obtained in a large epidemiologic case-control study: The Northern Germany Leukemia and Lymphoma Study (1997-2002) (NLL). The NLL used standardized, face-to-face, computer-assisted interviews to record subjects' lifetime use of radiofrequency (RF)-emitting appliances such as cellular telephones, cordless telephones, baby monitors, and television headphones. Exposure assessment comprised 3 levels of precision: ever use, gross vs. net appliance-years, and lifetime cumulative exposure hours. In the current study, the authors analyzed data from 3041 interviews of NLL controls, representing an age-stratified random sample of the general populations of 6 counties in Northern Germany. Weighted kappa coefficients for gross vs. net appliance-years for men were 0.59 (95% confidence interval [CI] = 0.46, 0.71) for baby monitors and 0.98 (95% CI = 0.97, 0.99) for cordless phones; for women, the coefficients were 0.68 (95% CI = 0.56, 0.79) and 0.97 (95% CI = 0.94, 0.98), respectively. Weighted kappa values were considerably lower when net appliance-years and lifetime cumulative exposure hours were compared. Study results demonstrated that interview information on use of RF-emitting appliances, when measured at different levels of precision, can result in misclassification and biased risk estimates.