Diagnostic Reference Levels in Medical Imaging

Draft document: Diagnostic Reference Levels in Medical Imaging
Submitted by Dietmar Nosske, Federal Office for Radiation Protection (BfS) Germany
Commenting on behalf of the organisation

General remarks:

This document is well structured and good to read. It is appreciated that many details for the compilation and application of DRL are clearly specified. It is helpful to have summarising statements at the beginning of all chapters. On the other hand, the document includes many repetitions. Moreover, in many general statements only X-ray and no nuclear medicine diagnostics seem to be considered. This is true especially for chapter 7.


Special remarks:

More relevant remarks:

Page 78, lines 11 ff, page 82, par. (204): It is recommended to set DRLs for CT examinations per sequence instead for the whole examination because the DLP of a sequence is a matter of optimisation for which DRLs are an effective tool while the number of sequences is a matter of medical justification.

Page 78, lines 18ff, page 84, par. (219): It is recommended not to use weight-based administered activities (MBq/kg) as DRLs for (most) nuclear medicine investigations because i) this concept is in contrast to the concept for diagnostic radiology where DRLs are for standard patients and ii) it is not clear for which investigations the DRLs should be weight-based. The criterion "all types of nuclear medicine investigations except for those where the radiopharmaceutical is concentrated predominantly in a single organ" is not totally clear. For example, is this valid for bone scintigraphy (99mTc phosphates are predominantly concentrated in bone)? On the other hand, 99mTc-pertechnetate for thyroid scintigraphy is not only concentrated in thyroid but also in salivary glands and in some walls of the alimentary tract.

Page 92, lines 22ff: Weight bands with 10 kg intervals may be too rough for the first year of life. For this, weights representing infants older than 3 months, newborns and even premature infants may be needed. In nuclear medicine, fractions of administered activities given by the EANM with 2 kg intervals have been proven to be a good tool.

Page 101, lines 23 and page 106, par. (281): A DRL value is (only) "exceeded" if in an audit the median value in a particular facility exceeds the DRL; it is "consistently exceeded" if this is true in more than one audit. According to this, a step concept should be implemented with different actions if the DRL is exceeded or consistently exceeded.


Minor remarks:

Page 11, line 13: The second "from" should be deleted.

Page 14, lines 19ff: "administered activity" should also be mentioned as a DRL quantity. The same is valid for par. (44) on page 33.

Page 20, line 3: Hesse et al., 2005 is missing in the references.

Page 34, par. (48): It is not clear how the practices involved in the use of DRLs in nuclear medicine are different from those in diagnostic radiology.

Page 41, line 5f: Exclusion of the highest and lowest 5% of the data (elimination of outliers) is not necessary if percentiles are determined instead of mean values. The same is valid for par. (102) on page 45.

Page 43, last line of Table 2.3: Later on, the administered activity per body weight is recommended as the DRL quantity in nuclear medicine. Then the unit would be MBq/kg.

Page 51, line 8: It should be "Administered radiation activities" rather than "Administered radiation doses".

Page 67, line 32: It should be "patients" instead of "patints".

Page 86, lines 15ff: If DRLs in nuclear medicine are given as administered activities per body weight then DRLs should not only be based on administered activities for average-sized patients.

Page 105, line 6: Fig. 7.1 instead of Fig. 8.1.

Page 105, line 18: SD should also be explained.

Page 119, line35: The DRL is not only dependent on the radionuclide but also on the radiopharmaceutical.