Diagnostic Reference Levels in Medical Imaging

Draft document: Diagnostic Reference Levels in Medical Imaging
Submitted by Anja Almén , Skanes University Hospital
Commenting as an individual

The recommendation is a valuable document concerning a variety of aspects of diagnostic reference levels. In the optimisation process especially in diagnostic radiology the use of diagnostic reference levels could have actual impact on the radiation dose to the patients and should be considered important in the medical sector.

If the length of the document is considered a problem, some sections dealing with optimisation could perhaps be shortened. For example in chapter 5 a number of optimisation possibilities is mentioned, e.g. in paragraph 235 p. 87 even some specific dose reduction actions is mentioned. This could be considered the second step after establishing a local DRL.

The credibility of the system of DRL relies on up-to-date comparable results, on all levels – local, regional and national. It should be no doubt that derived numbers are comparable and no systematic “errors” occur and effect the DRL (especially local) obtained. As stated in the report, the procedures – imaging technique, technical parameters, number of images etc. - should be altered due to the medical purpose of the procedure. An example is of course performing one or several scans for different medical indications in CT. Many examples in the report also refer to the anatomical region exposed. In section 2.4 the issue is addressed but it is helpful to further give advice. In this section chest radiography is mentioned leaving out e.g. the much more challenging field of CT with a number of protocols used for different medical conditions. The issue of medical purpose and the connection to the radiation dose could also be further explained. Part of target audience could be less familiar with the clinical environment. The report needs a somewhat clearer recommendation on how to deal with this and to what extent and how this should influence the methodology when deriving DRL.

How to deal with the use of different imaging technology for the same medical task is also important to address. The recommendations give a rather vague advice on this. The difference between radiology and nuclear medicine is evident. In nuclear medicine the type of radiopharmaceutical is often connected to a DRL. Some studies suggest that it could be useful to derive DRL when technology (e.g. detector technology) in radiology fundamentally differ. Further advice on this could be of great value. (The mentioned difference in technology in paragraph 95 p. 44 could actually be influenced by different medical indication not only technology.)

It could be desirable to get a clearer view on how to deal with image-guided procedures. In some countries DRL:s exist for the most standardised treatments such as PCI. Some countries report on merging different image-guided procedures to a common DRL due to similarities in the treatment. Some additional advice could be useful as this field is challenging.

The use different methods to adjust the derived quantities to the patient characteristics such as using SSDE, could be even more stressed in the recommendations. The derived DRL:s on the local level could be used extensively and serve as the assessment of radiation dose and is used also for example to derive collective effective dose even on a national level. Increasing precision in the assessment of absorbed dose should be encouraged.  

It could also be of value to further explain and report in uncertainties in derived dose quantities as well as uncertainties in estimated local DRL. Also limitations of the dose quantities in the assessment of dose to the patients could be further addressed. These issues influence the credibility of the system.

Chapter 7 includes many technology specific issues – and could be outdated very fast and some technical parameters mentioned is perhaps not used extensively today. 

Annex A. is great for the reader to get traceability.