A randomized, double-blind, sham-controlled trial comparing two screening devices for radiation contamination.

Abstract

This exploratory study compared the screening ability of a newly introduced radiation detection portal with a traditional Geiger counter for detection of radiation contamination in the setting of a mass casualty training exercise. Following a pretrial evaluation of interobserver reliability for Geiger counter use, 30 volunteers were randomly assigned to don gowns containing three disks, each of which was either a sham resembling the radioactive samples or an actual cesium-137 sample; each subject participated a minimum of four times with different gowns each time. Each subject underwent standard radioactivity screening with the Geiger counter and the portal. Interobserver reliability was excellent between the two Geiger counter screeners in the pretrial exercise, correctly identifying 101 of 102 sham and radioactive samples (κ = 0.98; 95% confidence interval [CI] = 0.94 to 1.00). For radioactively labeled subjects across all bodily locations, the portal (43/61, or 70.5%; 95% CI = 58.1% to 80.5%) was less sensitive than the Geiger counter screening (61/61, or 100%; 95% CI = 92.9% to 100%), which resulted in a portal false-negative rate of 29.5%. For radiation detection in the posterior thorax, the portal radiation screening (4/19, or 21.1%; 95% CI = 8% to 43.9%) was less accurate than the Geiger counter (19/19, or 100%; 95% CI 80.2% to 100%). In contrast, there were no major differences between the portal and the Geiger counter for radiation detection at the left shoulder, right shoulder, or sham (nonradiation) detection. There were no false-positive detections of the sham-labeled subjects for either device, yielding a specificity of 100% for both screening modalities. Geiger counter screening was more sensitive than, and equally specific to, radiation detection portal screening in detecting radioactively labeled subjects during a radiation mass casualty drill.