Register for Updates | Search | Contacts | Site Map | Member Login

ICRP: Free the Annals!

View Comment

Submitted by Sue Edyvean, Public Health England
   Commenting on behalf of the organisation
Document Diagnostic Reference Levels in Medical Imaging
 

General comments


1.     The ICRP committee has worked hard to produce a comprehensive draft document. We recognise the considerable amount of work involved in producing such a document.


2.     We are encouraged that the concept of considering image quality has been introduced.


3.     There seems to be an extensive amount of repetition from chapter to chapter. Sometimes key aspects are presented in different ways, leading to apparent different interpretations which can lead to confusion. The report has potential to be more concise, consistent and cohesive.


4.     If the aim of repetition is for each chapter to be a ‘stand-alone-chapter’, then additional work is needed to ensure that key points are fully (and consistently) presented.


5.     It would be helpful to have a brief summary (could be in point form, or identified when discussed in the executive summary) of the key topics that are different between the previous DRL report, and this one. Topics such as: clinical indication, local DRLs (if they are different), dealing with survey distributions where technology differences indicate significantly different dose levels, median versus mean, the shift from a DRL being used primarily for outliers (if that is the case).


6.     Structure of chapters 3, 4 and 5: The rationale for the allocation of the various imaging modalites into the given chapters is not always clear to the reader, and may not be helpful to the end user. More detail is given in the specific comments. It might be helpful to review this.


7.     It would be helpful to have a full discussion on the change from mean to median, and the implications. One aspect is mentioned in chapter 2 (para 80), but further discussion would be helpful. See also specific comments.


8.     Local DRLs could be described more clearly and fully. The descriptions (or implied interpetation) do not appear to be consistent (see comments in Glossary and in chapter 2).


9.     The terminology heirarchy in the report is ‘local’, ‘national’, ‘regional’ (e.g. European). However in the UK a ‘region’ is usually sub-national. This is perhaps the case in some other countries. A suggestion is to change ‘regional’ to ‘supra-national’.


10.  Whilst we can see the term ‘DRL quantity’ is needed, the report becomes very full of phrases holding the abbreviation ‘DRL’ – whether ‘DRL quanity’ or ‘DRL value’. Whilst it is helpful to have such a term as ‘DRL quantity’ it can potentially be confusing initially, especially to the casual or uninitiated reader. A suggestion is to sometimes use another term (that does not include ‘DRL’) in the text. For example ‘appropriate dose quantity’, ‘appropriate dose indicator’, ‘appropriate dose metric’. This could be indicated in the Glossary – for example: ‘DRL quantity, also referred in the text as the appropriate dose indicator’.


11.  The report has stressed the need for consideration of image quality – which is commendable. However there is an assumption that values that are less than the 50% value of the DRL curve represent image quality that is not good enough, and must be investigated. Combine this with an investigation of doses above 75% value, it implies that only centres that have DRL quanties that are within 50% and 75% of the percentile range are ‘good enough’. However if the interpretation is not as simplistic as it has seemed in our reading of the document, then a suggestion is for the text to be clarified to ensure that the intended interpretation is imparted.


12.  A practical point from the perspective of those using the document might be that there may be a danger of getting too immersed in quartiles/means/medians etc (although it is important to discuss these), and thereby losing sight of what a DRL is actually supposed to be – a level which in standard practice, for standard patients etc, wouldn’t be expected to be routinely exceeded – effectively the start of the high dose ‘tail’ commonly seen on dose data distributions. This is how DRLs were first intended (see ICRP 73, page 23) and was the fundamental starting point. It was then observed (at that time) that the relevant point corresponded to the third quartile value which was a convenient point to use. This probably will always be the case where there are a significant number of non-optimised departments/rooms included in the data collection. However, post-optimisation – particularly in a country or region with a long history of attention to patient dose – the dose data distribution may be very different – more of a normal distribution - and use of a 3rd quartile may place perfectly ‘OK’ departments above the DRL (ie by definition 25% of rooms will be above the national DRL set at 75%, but their value may be reasonably ‘close’ to the mean). Therefore in UK we have had additional guidance on what is a significant difference in relation to DRL comparison.


13.  On the basis of the above comment, the philosophy of DRLs should be emphasized, and separated from the mechanism of their use. Also it is important to state that uncertainties must always be considered when comparing doses. (For information: At least one dose management system company is putting national DRLs into their software. Therefore it is essential that a clear message, and understanding as to how national DRLs are meant to be used, is conveyed).


14.  Guidance on the mechanism of specific use is however important, especially for comparisons with studies from other countries or with historical reviews. It may therefore be helpful to have a table that clearly defines all of this terminology. This could be a ‘go to’ table, for readers wishing to undertake dose surveys, and would also ensure that the whole document is self consistent. (It could go into chapter 2). An example table is given – (as text for ICRP web site) with our interpretations of terminology. (It is presented as an example to how this could work, not as a definitive solution). Worked through examples in an appendix may also be of use. Elsewhere we consider mean versus median – for the purpose of this table, and as proposed in the report, median is given.


15.  All comments and suggestions are made with the intention of clarifying interpretation. If our comments indicate a mis-interpretation of the intended meaning of any statement, then a re-wording of the statement is suggested.


16.  ‘Go to’ table example


Key: Column (1)Term, (2) Description of area, (3) Description of value, (4) Suggested notation (for clarity and consistency), (5) Comment, (6) Alternative (if no acquired data available)


(1)   Median DRL quantity


(2)   A single radiology unit (‘x-ray room’)


(3)   Median value of the DRL quantities for the given clinical indication, and examination


(4)   Room median


(5)    -


(6)   For submission and comparison to National DRLs


 


(1)   Local DRL


(2)   A healthcare facility consisting of a number of ‘x-ray rooms’ ideally under the same governance


(3)   Median of the median value of those x-ray rooms within the ‘local’ area


(4)   LDRLmedian  , or LDRL (median)


(5)   For local use for clinical staff


(6)   Adopt from regional, national or supra-national


 


(1)   National DRL (third quartile)


(2)   Country


(3)   Third quartile of the median value of those x-ray rooms within the national survey


(4)   NDRL75 or NDRL(75)


(5)   National median also presented in survey results


(6)   Adopt from supra-national


 


(1)   Supra-national DRL (e.g. European)


(2)   A number of countries – this could be within a contintent (eg within Europe), a continent (Europe), or international. E.g. European DRLs, International DRLs


(3)   Third quartile of the median value of those x-ray rooms within the supra-national survey, or for ease median value of the national values


(4)   SN DRL (median of national), or SN DRL (median of national), e.g. European DRL (median of national)l


(5)   A pragmatic approach to combining values from national surveys


(6)   -


 


An additional level, between Local and National, could be considered, in order to separate the various interpretations of ‘local’. However though this may probably introduce too many levels. E.g.:


(1)   sub-national DRL (third quartile)


(2)   A larger group of x-ray facilities, not necessarily under the same governance


(3)   Third quartile of the median value of those x-ray rooms within the sub-national region


(4)   Sub-national (or UK regional) DRL75


(5)   For larger spread of data


(6)   Adopt from national or supra-national


 


Specific comments


Preface


Page 7, -,Line 11 ff


Would be good to have references. It is difficult to date actual beginnings from reports now currently published on the internet. Cf NEXT first report on internet is given as 1984, NRPB first report is also in the 1980s. Both organisations it seems were doing work prior to that, presumably in internally published reports, or as external reports but not scanned and published on the internet.


For interest, and background to the UK:


Following the report of Lord Adrian in the mid 1950’s, he then requested physicists in hospitals to undertake a survey of gonad doses from radiological investigations (1957) “Committee on Radiological Hazards to Patients, Lord Adrian, The British Journal of Radiology 1957 30:354, 285-285 “ http://www.birpublications.org/doi/abs/10.1259/0007-1285-30-354-285


Also see II. Comments on the Diagnostic Aspects of the Adrian Committee Reports, G. M. Ardran


The British Journal of Radiology 36 (424) pp: 235-237 (1963)


S Rae, A survey of the radiation doses to patients from diagnostic radiology NRPB Radiol. Prot. Bull. No 19 (1977).


Executive Summary


Page 9,Para. (b),


Line 8


Line 8: suggest change to “one of the principles …”


The ICRP uses the radiological protection terms: Justification, Optimisation and Dose limits in general, and for medical exposure there are no dose limits.  In a hospital,  the terms Acceptance test, Periodical test and examination optimisation are used. For examination optimisation the organ doses, image quality and diagnostic information are assessed in order for there to be sufficent information for the diagnostic task with the least organ doses.


The use of optimisation in this report refers sometimes to the ICRP approach and sometimes to the hospital practice and they are not necessarily the same.


Executive Summary


Page 9


Para. (b)


Line 11-Line 12


DRLs are not intended for use in radiotherapy (radiation therapy). However it might be helpful to note that recently we are seeing an increasing number of UK radiotherapy departments adopting local reference levels for CT scans undertaken as part of the radiotherapy planning process, and verification imaging undertaken as part of the radiotherapy treatment process. Can consideration be given to recommending the introduction of reference levels for imaging undertaken as part of the radiotherapy planning and verification process?


Executive Summary


Page 10


Para. (g)


Line 4-Line 5


‘They are derived from an arbitrary threshold’


Replace with


‘They are given as an arbitrary threshold (usually 75%) from a distribution’


Some guidance is helpful for comparison of data between institutes and facilities.


Executive Summary


Page 10


Para. (i)


Line 18


‘Compliance’


Is this an appropriate word to use for DRLs? What does ‘compliance with DRLs’ mean? Suggest ‘A dose level below the DRL value does not, by itself...’


Executive Summary


Page 10


Para. (i)


Line 18-Line 24


The statement that - below the 50th percentile, image quality should be taken as the priority, implies that half of users should investigate IQ instead of dose. If that is not the intention then suggest re-wording.


Given that data that is submitted for the publication of national DRLs should come from diagnostic data, it is difficult to believe that 50% of users have an image quality problem. For example the document mentions iterative reconstruction in CT as a reason for significantly lower dose. In that circumstance with survey data of FBP and IR scanners pooled together, a user at (for example) the 60th percentile with FBP may have optimised better than a user at the 40th percentile with IR.


Suggest: “Image quality is always to be considered in optimisation. The median may be considered as a balance point of image quality and dose in the general view, however users who have values significantly lower than the median of the national or regional distribution may need to look at image quality as a priority”.


Executive Summary


Page 11


Para. (n)


Line 5,6


‘...tasks of collating data on DRL quantities and setting DRLs’


Suggest replace with: ‘..tasks of collating appropriate data for setting DRLs’


This is in order to limit the number of times DRLs appears in a sentence. Too many times and the reader goes word blind and cannot read the message. Try as far as possible to not use ‘DRL quantities in the text’ (as the reader’s brain doesn’t always get to the ‘quantities’ word in time to realise that it is not about ‘DRL values’, but ‘DRL quantities’).


(If this change is not accepted, then the sentence needs a comma after ‘setting DRLs’ (line 6))


Executive Summary


Page 11


Para. (o)


Line 13


Delete the second ‘from’


Change “are collected from at several different health facilities.”  to “are collected from several different health facilities.”


Executive Summary


Page 11


Para. (o)


Line 14-15


This is an appropriate definition.


Executive Summary


Page 11


Para. (p)


Line 16


A region in the UK is at the sub-national level. Suggest a new term ‘supranational’ (see general comments)


Executive Summary


Page 11


Para. (p)


Line 17-18


To use the median value of national DRLs for a regional (supranational) DRL must be very carefully considered, and may not be of benefit for any individual country.


Executive Summary


Page 11


Para. (q)


Line 19-21


This frequency is rather high. Are many countries able to apply this frequency? This may be too high given the burden of work with current processes.


Executive Summary


Page 11


Para. (r)


Line 22-28


The establishment, or setting, of local DRLs is a legal requirement in the UK, regardless of the availability of national/international DRLs.


Executive Summary


Page 11


Para. (r)


Line 25


Delete ‘within that limited area’. This may be mis-interpreted as a geographic area.


Local DRLs needs a clear, consistent, description elsewhere in the document.


Executive Summary


Page 11


Para. (s)


Line 30


‘DRL Quantity’. See general comments. This is an example.


Executive Summary


Page 11


Para. (t)


Line 34-36


Need to understand what is meant by the local level, where the data comes from, how it works in practice.


Executive Summary


Page 11


Para. (u)


Line 38


Corrective action is not always required when a local DRL value exceeds the national DRL. The words “,if shown to be required,” should be inserted in line 38 between “and” and “a corrective action plan”. 


Executive Summary


Page 12


Para. (w)


Line 4


‘administered radiation’


Can see this is intended to be a general term to cover both external x-rays, and internal radiation from administered activity in nuclear medicine. However the term ‘administered’ is usually used in nuclear medicine. It would help for this phrase to be expanded, or changed.


If kept, or an alternative used, it should be in the Glossary


Executive Summary


Page 12


Para. (x)


Line 9


10 kg banding is not necessarily required for all examinations (eg. paediatric heads?)


Glossary


Page 13


 -


Line 2


If ‘achievable dose’ is being proposed by ICRP – then it should be included in the Glossary.


However even if it is not being proposed, it does appear in the text elsewhere, and therefore should be in the Glossary. Also in this instance there is a precedent for inclusion – cf Notification value which is a value that another organisation uses, appears elsewhere in the text and also appears in the Glossary


Glossary


Page 13


 -


Line 9


‘Ka,r’


Should ‘a,r’ be subscripts? This gets confusing later in the text when there is a list of parameters separated by commas, so suggest subscripts are used.


Glossary


Page 14


 -


Line 7


A DRL is really only to give an indication for dose values that are too high. Suggest to delete ‘low’.


The distribution of doses can be used to make some judgement, but not the DRL (DRL value)..


Glossary


Page 14


 -


Line 19


‘DRL quantity’


This becomes too many ‘DRL’ words when ‘DRL quantity’ is put into the text, and does not make for an easy read. (For example the paragraph on lines 31-35 for ‘DRL value’)


Suggest that it is either changed, or if kept, then when it is referred to elsewhere in the text, that every effort is made to not use it Suitable alternatives could be ‘appropriate dose metric’, ‘relevant quantity’, ‘appropriate dose indicator’, ‘dose indicators used for DRLs’  etc etc


See also comments on ‘DRL value’


Glossary


Page 14


 -


Line 21


Use the subscripts after the K symbol. This is not easy to read.


Glossary


Page 14


 -


Line 31


‘selected’. Elsewhere this is an ‘arbitrary’ value. Need to be more consistent. Suggest use ‘arbitrary’ throughout


Glossary


Page 14


 -


Line 31


‘numerical’ – suggest change to: ‘notational’


Glossary


Page 14


 -


Line 31-35


Too many ‘DRL quantities’ in this paragraph


Glossary


Page 14


 -


Line 31-35


This implies that a local DRL is at the 75% percentile. Elsewhere in the document it is implied or given as the median.


If it is meant to be at the 75% level:


It is seen that it would definitely be helpful for clarity of terminology that a ‘DRL’ is usually the value at the third quartile of a distribution. However locally one might expect values to be closer to the median of the ‘local’ distribution, and therefore this may not be a helpful parameter.


Also, practically this will cause confusion in the UK. The local DRL , as used in the UK, is not the third quartile – it usually the average of room means. See IPEM report 88 for a good description.


As a general point for this report - ‘Local’ perhaps needs a clearer and consistent definition, with clear examples. Alternatively it should be allowed to be interpreted according to need (but with suitable annotation to describe how it is reached in order to allow comparison).  (See also p15 lines 14-16).


Glossary


Page 14


 -


Line 31-35


Lines 34 and 35 state that ‘Regional DRLs can also be based on the median values of the available national DRLs’. This is understandably a pragmatic solution. Does the commission have a view as to how different the resultant values would be; ie obtaining a Regional DRL from the 75% percentile of all source data (room medians), or the median from national DRLs (75% percentile of room medians)?


Glossary


Page 14


 -


Line 30-35


‘DRL value’


Suggest terminology which includes the percentage point e.g.  ‘DRL75’ or DRL(75).


At the very minimum the ICRP commission should encourage that wherever DRL is stated that it is also accompanied by the region (local, national) and the basis on which it is quoted (75%, median etc).


It will make no sense for future and retrospective comparisons unless this is made clear whenever a ‘DRL’ is quoted (and there is already seemingly confusion in the literature).


This is particularly more so after this report is published, where room median values are being suggested (as opposed to the mean) to create a national distribution, as well as (in some places in this report) the local DRL is being suggested to be third quartile rather than mean or median.


Glossary


Page 15


 -


Line 11


Please use kilogram instead of kilogramme.


Glossary


Page 15


 -


Line 14 -16


Local diagnostic reference level


It would be helpful to have a clearer defintion of local DRL vs national DRL (cf line 33). It is currently not clear nor expanded upon sufficiently. How many centres make up a ‘local’ area?


This used to be the mean value for the patient sample. This needs making very clear if the definition has changed. There are seemingly two proposed changes – median, and third quartile.


See comment above for page 14, line 31-35


Glossary


Page 15


 -


Line 14


‘Local diagnostic reference level ‘. Change to ‘Local DRL’ or ‘Local diagnostic reference level (Local DRL)’ for consistency with ‘DRL value’ Glossary entry


Glossary


Page 15


 -


Line 28


‘Exposure’.  Change to: ‘Radiation exposure’


Glossary


Page 15


 -


Line 33


It would be helpful to have a clearer defintion of local DRL vs national DRL (cf lines 14-16).


Glossary


Page 15


 -


Line 33


Change to ‘National DRL’ or ‘National diagnostic reference level (National DRL)’ for consistency with ‘DRL value’ Glossary entry


Chapter 1


Page 19


 -


Line 15


‘median of the distribution of a DRL quanity ..’


Either this is suggesting the median value to be a local DRL (as per in the UK – where the mean usually – as opposed to the median – is used) but the document is not calling it a local DRL.


Or this is referring to a national DRL value, in which case change to: ‘median of the distribution of medians of a DRL quanity ..’


Chapter 1


Page 19


 -


Line 19


‘measured’ change to: ‘measured or determined’


Chapter 1


Page 19


 -


Line 25


‘is not a good indicator’ . Change to: ‘is not necessarily a good indicator’


Chapter 1


Page 19


 -


Line 31


“Values DRL quantities” change to : “Values of DRL quantities …”


This is a key example where excessive use of ‘DRL’ in the terms ’DRL quantities’, ‘DRL values’ and ‘DRLs’ can lead the reader to be word blind and not understand what is being said. Suggest rephrase – see suggestions above in comment on Glossary - page 14, line 19.


Chapter 1, Section 1.1


Page 20


Para. (1)


Line 3


The national reference values (and adopted as national DRL values) for general radiology, given in Hart et al. 2009 are now out of date, having been replaced by those in Hart et al. 2012.


The current UK CT national reference values (and adopted as national DRL values):-  Shrimpton, P. C., Hillier, M. C., Meeson, S., Golding, S. J. 2014  Doses from Computed Tomography (CT) Examinations in the UK -2001 review. PHE-CRCE-013  Public Health England, UK


In addition, in the UK these adopted values are now collated to one page, on the following link, which may be a useful additional reference.  https://www.gov.uk/government/publications/diagnostic-radiology-national-diagnostic-reference-levels-ndrls


Chapter 1, Section 1.3


Page 20


Para. (5)


Line 30-32


The information given here does not seem to match what is said in the Executive Summary.


Chapter 1, Section 1.3


Page 20


Para. (5)


Line 30-32


See comments on Executive summary with respect to this topic.


cf S Rae, A survey of the radiation doses to patients from diagnostic radiology Radiol. Prot. Bull. No 19 (1977). This is a paper archived NRPB publication.


Chapter 1


Page 21


Para. (5)


Line 1-2


Reference ‘Wall and Shrimpton, 1998’ should occur for UK not US.


Chapter 1, Section 1.3


Page 21


Para. (7)


Line 16-17


As with suggestion for Executive Summary section 5, page 11, line 38 suggest the words “,where shown to be required,” should be inserted between “and” and “a corrective action plan”. 


Chapter 1


Page 23


Para. (19)


Line 29-30


Effective dose has always been thought of as not appropriate for setting DRLs. If this statement (‘has been suggested’) is true it requires references to support it.


Chapter 1


Page 23


Para. (19)


Line 31


‘it does not assess the amount of ionising radiation used to perform a medical imaging task’


But it is a broad indicator.


Chapter 1


Page 23


Para. (19)


Line 29-31


Conversely it should be noted that comparison of dose indicators cannot always be taken as a comparison of effective dose or risk (for the same examination) as beam quality makes a considerable difference to the actual patient dose relative to the dose indicator. This is hard to account for within the DRL framework, however  it should certainly be considered when drawing conclusions from work on DRLs.


Suggest a comment to that effect.


Chapter 1


Page 23


Para. (19)


Line 31


‘also, effective dose is not readily available’


Some of the dose management systems are offering it, but calculated in many different ways. Suggest extending the text to include something like:


‘It is not always calculated in the same way, or quoted with the most current ICRP organ weighting factors or reccommended phantoms’.


Refs: ICRP 60, 103, ICRP adult phantom report 110, and Updated estimates of typical effective doses for common CT examinations in the UK following the 2011 national review Paul C Shrimpton, Jan T M Jansen, and John D Harrison, 2015, BJR, DOI: 10.1259/bjr.20150346


Chapter 1, Section 1.5.4


Pages 23 -24


Para. (20)


Line 36 ff


Suggest making a stronger point to emphasise this and to clarify the general point being made – e.g  ‘separate DRLs may be needed where technological advances or changes lead to significant consistent, identifable, differences in doses’


Chapter 1, Section


Page 24


Para. (22)


Line 11-16


This is a very important point and is not strongly emphasised enough in the document. Setting a DRL for a particular examination when the clinical indication is not considered can, for many examinations, be pointless.


Suggest emphasise more strongly in the document


Lists of examinations where the clinical indication is, or is not, important would be helpful. Though this may not perhaps be appropriate for this report.


Chapter 1


Page 24


Para. (21)


Line 6-10


‘This difficulty ... iterative reconstruction.’


These lines are effectively a repeat of information in para (20). Could delete.


Combine this para with para 20


Chapter 1


Page 24


Para. (25)


Line 31


‘with respect to single images’


Delete (is unnecessary, and confusing on first reading)


Chapter 1


Page 25


Para. (29)


Line 17


‘This report clarifies and defines some of these terms, such as local, national and regional DRLs. There has been some confusion regarding the proper use of local DRLs in certain situations’


This paragraph is specifically highlighting the need for adapting local DRLs where newer technology warrants a different value from national value (which may be based on older technology). Suggest a new separate paragraph starting with ‘There has been some confusion... since this is a specific comment.


However it seems that there is some inconsistency and lack of clarity throughout the report regarding local DRLs, at present it seems rather confusing. It would help to have more clarity on the ‘proper’ use of the local DRL. See comments wrt Glossary,p15, line33.


For example - Local could be described as an x-ray room, two or three x-ray rooms, a radiology department, a few radiology departments under the same governance etc etc.


Then there is the use of median or third quartile – this may depend on the number of rooms coming under ‘local’, and how the value has been acquired.


The UK system – as outlined in IPEM report 88 – is a good reference; detailed but thorough.


Chapter 1


Page 26


Para. (34)


Line 13


The achievable dose value (50th percentile) seems fairly arbritrary. To immediately undergo image quality investigation under the achievable dose does not seem sensible. For example there are other confounding factors such as compatility of apparatus.


Chapter 1


Page 26


Para. (35)


Line 17-28


Dose should not be used as a quantity for image quality. Although it can be used as a guide to suggest that some radiologists, equipment, rooms, departments can operate at a lower dose and still have diagnostic images. These may have good practice, good equipment, good clinical staff – or they may have gone ‘too low’ . Maybe a better description (or references) as to how to evaluate / investigate image quality should be suggested.


Chapter 1, Section 1.6


Page 26


Para. (35)


Line 17-28


Is the ICRP suggesting that all participants who return a DRL value below the median (which is automatically 50% of participants who sent in a value or data for a given exposure) should investigate their image quality?


This seems to suggest that 50% of participants must automatically be thought to be potentially ‘getting it wrong’. This combined with the assessment of potential dose reduction for values above the 75 percentile means that for each examination only 25% of entered values will be considered adequate, from a perspective of either dose or image quality.


However assume the paragraph is saying that the median dose is the tipping point between being primarily interested in reducing dose (above it) to being primarily interested in ensuring adequate image quality, but that other factors still need to be taken into account, especially in cases of use of new technology and atypical patient cohorts.


It would help for this paragraph to be clarified. e.g. suggest that all dose quantity values above 75% , or below 25%, percentile values are investigated. This can be for image quality (too good or too poor), for technology factors, for patient factors etc etc.


Chapter 1, Section 1.7


Page 27


Para. (39)


Line 14-19


Suggest addition – ‘the ideal approach is to have a dose and image quality optimisation team consisting of a Radiologist, Radiographers, Medical Physics and other concerned staff’.


Chapter 2


Page 32


 -


This seems a relatively clear, precise, chapter


Chapter2


Page 32


 -


Line 16


Bullet point 4: Is this for a national survey? It would help to put it in the bullet point so that it can be read as a ‘stand alone’ point.


Chapter2


Page 32


 -


Line 16


Bullet point 4: Suggest, after ‘obtained’ :... ‘However smaller facilities should be not be ignored. They may represent expert specialised practice, or need to go through a survey process to gain dose awareness. Participating in a survey will naturally achieve this.’


Chapter2


Page 32


 -


Line 36-39


Bullet point 9: Mean versus median


This chapter has one paragraph (80), page 39, to support the change in recommendation from mean of the distribution to the median of the distribution of the DRL quantity. It might help to expand that.


Although the median is a more robust estimator than the mean, and may therefore be a better value for setting DRLs, the mean is sensitive to high dose outliers (as discussed in para 80). However it could be considered that those are the most important values for reduction of dose. If one is able to take dose reduction measures the mean value may show the results more clearly than the median value.


The transition period should be considered and addressed, due to having two bases for the contributing DRL quantity (mean or median) to a national survey. (See also next point)


Chapter2


Page 32


 -


Line 36-39


Bullet point 9:


Many national dose surveys currently use the mean value from hospitals, rather than the median, to set national DRLs. This change will affect comparisons.


Care must be taken in the transfer from using one to the other to minimise confusion and errors.


As median values tend to be lower than mean values there will be an apparent “drop in doses” if only the headline DRL values are looked at. This apparent drop in values will be largest for those examinations which have the most or highest value outliers.  These are the examinations that might most benefit from investigation or optimisation, with exceptions such as interventions where the complexity of some procedures may account for outliers. So an argument against the use of Local median results is that it may give a false sense of security. Most information would be gained by looking at both the median and the mean values and seeing how much they differ. 


Suggest implement a naming system e.g. NDRL75 (of medians). This  is perhaps too complex, however it serves as an example. We strongly suggest that the ICRP should recommend that, whenever a NDRL value is quoted, that the percentile level, and source of data (mean or median), are clearly stated. This will help in transitions and comparisons.


Chapter 2


Page 32


 -


Line 32-33


“…and should be traceable to a primary or secondary standard laboratory”. All secondary standard laboratories are traceable to a primary standard laboratory, so the ‘or secondary ‘should be removed, as it is potentially misleading.


Chapter 2


Page 33


Para. (42)


Line 17


“examiantions” typo. Change to  “examinations”


Chapter 2


Page 33


Para. (44)


Line 31


Occurs in many places: Ka,e should be written with sub-scripts a and e (Ka,e). It does not look right (and is confusing) when in a list where the comma is also used to separate parameters.


Chapter 2


Page 34


Para. (46)


Line 1-7


Suggest a change of wording in line 3 so that the second and third sentences reads “In some countries, hospitals or health authorities are legally required or are encouraged to set their own local DRLs and perform dose surveys to support them.  Local DRLs may be based on national values but it is preferable that they reflect local practises to encourage optimisation”.


Chapter 2


Page 36


Para. (Table 2.1)


mobile radiography as a low priority? Unsure of reason.


Chapter 2, Section 2.2


Page 36


Para. (57)


Line 3


Suggest add “and similar technologies” to the end of sentence 2 (line 4).


Suggest document should stress and recommend the setting of different DRL values for different technologies (including image reconstruction methods), as these may lead to very different doses being achievable for images of required diagnostic quality.


Chapter 2, Section 2.2


Page 36


Para. (58)


Consider adding as last sentence to paragraph 58:  “If in a country or region there exists two procedures for an examination with different DRLs then it is possible to set two DRL values and specify the examination and procedure. Especially, when new techniques emerge with an influence on the doses then this approach would allow for a transition period.”


Chapter 2


Page 37


Para. (60)


“However.....”


This is a different point, therefore should be put as a separate paragraph


Chapter 2


Page 38


Para. (68)


Line 4


Local DRLs in this context are mean values of local rooms. (IPEM report 88 (2004) is an excellent manual, giving clear precise scenarios, for collection and use of national and local DRLs).


However this is seemingly in contradiction to the Glossary where it seems to imply the 75th value of a local survey is to be used as the local DRL ( the local DRL meaning is cross-referenced to the DRL value, which states 75%).


This is confusing. If a local DRL is meant to be undefined, then that should be made clearer in the Glossary entry.


For reference: IPEM 88 (2004) proposes that attention needs to be given when a local room value is greater than 20% of the local DRL (which is set at the mean value of the distribution of dose indicator values). (This incidentally is equivalent in concept to using the 75% percentile of a distribution of dose indicator values i.e pay attention in a distribution when a value is 20-25% above the mean).


The mentality for a local DRL needs to be considered. Radiology professionals need to be able to either think of the local DRL as a target value (mean or median), or as a value they should perform under.  Since it is difficult to define by how much then we suggest that the local DRL remains as a mean or median.


If local values have been drawn from already optimised data, one hospital, or a few similar pieces of equipment, then the local DRL as the median makes sense. If ‘local’ values are drawn from a much larger group of facilities or wide ranging dose efficient equipment then the third quartile may be more helpful. However consideration must be given as to how the clinical professionals will view the values.


As per previous comments - suggest that to avoid confusion, and to help for consistency when comparing national values, or historical data, that this report recommends that that the DRL value has to be quoted accompanied by the basis of its derivation - this could be either a subscript, or a parenthesis which defines its origin. E.g. local DRL would LDRL(median), national DRL(75) or DRL75, European DRL (national median) etc. . In the literature, a DRL  - without its derivation, and regional coverage – has no meaning.


The text of the public area of this website is available for modification and reuse under the terms of the Creative Commons Attribution-Sharealike 3.0 Unported License and the GNU Free Documentation License (unversioned, with no invariant sections, front-cover texts, or back-cover texts)