J Am Coll Radiol 2007;4:604-611. Copyright © 2007 American College of Radiology
Nancy A. Obuchowski, PhD, Darlene Holden, MD, Michael T. Modic, MD, Grace Cheah, RN, MSN, Alex Zhenghong Fu, PhD, Michael Brant-Zawadzki, MD, Raul Seballos, MDC, Tan-Lucien Mohammed, MD
Purpose: The authors performed a pilot randomized controlled trial of total-body screening to assess the feasibility of a full-scale study.
Materials and Methods: After informed consent, 50 asymptomatic people were randomized to either the intervention arm (total-body screening with multidetector computed tomography) or the control arm (no screening for 3 years). The study was approved by our institutional review board and was compliant with the Health Insurance Portability and Accountability Act. Images were interpreted independently by 6 radiologists from 2 institutions. Subjects in both study arms completed periodic health questionnaires and medical utilization forms over 2 years. Key outcome variables were the incidence of symptomatic disease, medical costs, and patient-reported health.
Results: Sixteen screened subjects (64%) had abnormal findings on screening. A second interpretation of the images yielded a similar overall rate but with considerable variability at the subject level. No cancers were detected. Ninety percent of subjects were compliant at 2 years. Medical costs were twice as high for screened subjects, with considerable between-subject variability. Screened subjects reported fewer physical limitations than unscreened subjects.
Conclusion: A full-scale randomized controlled trial of total-body screening will need to account for the large interreader variability in interpreting the images, the high rate of incidental findings, and the low prevalence of cancers. A full-scale study using mortality as the endpoint does not seem feasible at this time.
Key Words: Total-body screening, computed tomography, reader variability, randomized controlled trial