In Reply We thank Bernheim for his thoughtful comments on our recently published Original Investigation. 1 The American College of Radiology (ACR) and Centers for Medicare & Medicaid Services recommend a maximum volumetric computed tomography dose index (CTDIvol) of 3 mGy for lung cancer screening. 2,3 As a measure of the average dose per slice, CTDIvol will reflect many of the scanning decisions made by the radiologists and technologists, and reliably measure average doses that patients will receive.However, the CTDIvol will not capture all decisions made by radiologists and technologists. The total scan length should be minimized to avoid irradiating the thyroid and liver, and the examination should include only a single scan-not multiple scans. These recommendations are not always followed, and neither of these decisions will be reflected in the CTDIvol. Therefore, we also used effective dose (ED) as a measure to describe dose and to reflect the consistency of lung cancer screening computed tomography across hospitals. The ED threshold we used comes from the American Association of Physicists in Medicine (AAPM) statement, which specifies that lung cancer screening protocols should result in an ED of 1 mSv or less, and further indicates in tables that this corresponds to 3.0 mGy CTDIvol. 4 This characterization of the ED threshold of 1 mSv or less as an analog to the ACR CTDIvol criterion of 3.0 mGy led us to use it as a cutoff in our study 1 and to attribute the cutoff to the ACR. However, given that this analog of the ED threshold was developed by the AAPM, it was an error to attribute it to ACR directly.Our study 1 highlights that while lung cancer screening computed tomography scans should be single-slice scans, multiple scans are sometimes performed, and this poor quality practice is only reflected in the ED measure. We found greater variation in the ED measure compared with the CTDIvol measure because of this. We believe that the 1 mSv ED threshold in the AAPM protocol is the best analog to the ACR and Centers for Medicare & Medicaid Services guidelines, and is the ideal benchmark to assess performance. However, we could have also used as a benchmark the observed 1.5 mSv average