Radiation Dose Profile - Work Page
- Kevin O'Donnell (Toshiba)
- 1 Summary
- 2 2. The Problem
- 3 3. Key Use Case
- 4 4. Standards & Systems
- 5 5. Technical Approach
- 6 6. Support & Resources
- 7 7. Risks
- 8 8. Open Issues
- 9 Deployment Issues
It is difficult to monitor radiation dose for individual patients and almost impossible to assemble and compare such statistics for a site or a population or a type of exam.
DICOM has an SR format for radiation dose events and a protocol for exchanging them.
A Radiation Dose profile could require compliant radiating devices to produce such reports and could define transactions to actors that collect, analyze and present such information.
Euratom and ACR have published guidelines requiring/encouraging dose tracking. Individuals from SFR are willing to participate in Profile development.
The main challenges are dealing with the chicken-and-egg problem and avoiding inconsistent implementations.
2. The Problem
Tracking radiation dose delivered to patients for medical purposes is gathering increasing attention from professional societies and regulatory groups. Publications include: European directive Euratom 97/43, ACR Dose Whitepaper and some Japanese regulations.
Many such groups have expressed a desire to establish standards of practice based first on a quantitative understanding of current practice, however they have found it prohibitively difficult to collect such data. Needs have been expressed to track both dose to individual patients and to assemble population data.
Physicians are asked to follow guidelines, like constraining their patients dose to As Low As Reasonably Achievable (ALARA), but it can be difficult for them to see what their patient dose actually was, and almost impossible to assemble comparison values as to what is reasonably achievable.
In most of the world, although it should be, it is not possible to do things like:
- view the dose a patient (or particular organs) received for a certain exam, hospital stay or course of treatment
- view a patients cumulative dose history
- determine if a given patient dose exceeds maximum guidelines or is otherwise an "outlier" requiring investigation
- compute the population "dose profile" for a certain hospital or region
- compute the population "dose profile" for a certain pathology
- compare "dose profiles" against other sites/regions, against local policy targets or against standards of practice
In the vast majority of medical procedures involving radiation, the potential benefit to the patients health far outweighs the potential risk, but the trade-off should not be forgotten, and without the support of technological mechanisms, a conscious evaluation if the trade-off is seldom done.
The problem is that although standards exist (see below) such dose tracking has not been widely deployed due to:
- difficulty coordinating the roles of the different equipment involved (which actors should do what)
- difficulty coordinating tracking across multiple departments and multiple institutions
- the need to converge on one of the available standard approaches
- a chicken-egg situation between the dose record producers, the dose record receivers and the dose record consumers
These are, of course, many of the typical problems that IHE is used to addressing.
3. Key Use Case
A more detailed discussion of use cases can be found on the Radiation Dose Use Cases page. If you have additional input, please add it there.
The following use cases were a couple quick examples for the Profile Proposal phase.
First use case scenario
Frank Smith is hospitalized following an automobile accident.
The ER doctor sends him for an abdominal CT. Later, his attending physician sends him to Radiology once for a chest X-ray and twice for a right arm and left leg X-ray. A referral to Cardiology results in a coronarography (fluoroscopy and radiographies were done).
The hospital would like to believe its doctors are following site policy and its practices are in line with national norms, but doesn't know if that is actually true and has little or no data to back it up. It has little or no tools for implementing a dose management or reduction plan and no tools to measure progress towards its targets.
Some of the modalities included a few dose details in the headers of the images. Those images are distributed across the Radiology and Cardiology PACS systems. A few of the Reporting workstations display dose values some of the time when the details in the image header match what it expects to find. Other systems put dose details in some of the MPPS messages sent to the RIS and the CIS. All the HIS knows is the names of the radiology procedures ordered. No system has a complete view and due to inconsistency of data format and distribution, none of them attempt to compile a dose record.
Correlation with Mr. Smith's radiotherapy earlier this month and again next month at the clinic across town is completely out of the question.
Second use case scenario
When Mr. Smith's order is placed, the system can provide guidance to the physician, taking into account the dose of considered procedures and current patient dose.
At the end of the stay, the HIS knows all ionizing radiation events that occurred during the stay of Mr. Smith, keeps track of all data and calculates and archives the total dose delivered during the stay. Mr. Smith had several visits to this hospital before. The HIS consolidates the total radiation dose delivered through all visits. Data is sent to the regional EHR system to be consolidated with other enterprises Mr. Smith visits during his life. Data is also transmitted, after anonymization, to a national radiation dose registry. The registry provides periodic reports back to the facility with regional and national benchmarks of radiation dose. Using this report, Mr. Smith’s hospital can compare its dose profile by modality, exam type, and pathology to facilities of the same type, in the same region, and to the nation as a whole.
4. Standards & Systems
Some existing systems that might be involved include:
Systems that schedule studies (RIS), administer radiation (modalities, and therapy systems), and store and report the results (PACS?, HIS?, EHR?)
Some standards which might be relevant to the solution include:
- DICOM SR Dose objects (PS3.3 A.35.8 X-Ray Radiation Dose, Supp 127 CT Radiation Dose)
- DICOM IOD’s attributes (CT IOD, XR IOD, etc.)
- DICOM MPPS attributes
- Other standards in radiotherapy and nuclear medicine? (PS 3.3 A.18 RT Dose?)
- Dose measurement protocols and guidelines from professional and/or regulatory bodies
5. Technical Approach
Existing actors that could be used or might be affected by the profile:
- Acquisition Modality - Store dose information to Archive; might be required to support sending to multiple destinations
- (Image) Archive - Accept & store dose information; support query & retrieve of it
Possible new actors:
- Dose Information Manager - Retrieve (or Receive) Dose Information from Archive, process/compile as necessary, submit reports to national dose registries, generate local reports
- Dose Information Consumer - Might retrieve or receive dose from archive or modality, do fancy 3D displays, etc.
- (Dose) Display - retrieve and present dose information to a user
New transactions (standards used)
- Store Dose Information - Similar to Images Stored. Use existing DICOM SR Specs.
- Query Dose Information - ditto
- Retrieve Dose Information - ditto
- Submit Dose Report - Sends information to a repository or policy system. Probably CDA or XML for content. Probably XDS or some Web Service for transport.
- Retrieve Dose Benchmarks - Get dose benchmarks (from a registry or other source of policy or comparative data) for comparison. (By demographics, by procedure type, by pathology, etc.)
Impact on existing integration profiles
New integration profiles needed
Consider a Radiation Dose Management Profile.
Breakdown of tasks that need to be accomplished
- Deal with open issues
- Document the above Transactions and Actors
- Solicit input from National groups
- Solicit input from Mammo, Cardio and Rad-Onc (mostly out of scope) Domains
6. Support & Resources
Members of SFR have indicated their support for this work and Joël Chabriais & Philippe Puech have agreed to participate in development. GMSIH and Karima Bourquard are in favour of the work and may support the development. Bernard Aubert representative of IRSN will participate to provide expertise on radiation dose follow-up issues. IRSN is considering a National Radiation Dose Registry in France in respect of Euratom art. 12.
Members of ACR have indicated their support for this work and Laura Coombs & David Channin have agreed to help write the profile. They are particularly interested in the role of a Dose Registry since Laura at ACR is currently working to establish such a registry. Paul Nagy from SIIM/UofMaryland is also involved in the ACR Dose Registry project and would like to contribute and review the work.
Members of SEFM (Spanish Medical Physicist Society) have been "waiting a long time" for such work and would like to support the work. Julio Almansa has agreed to participate.
Dr. Wallis of SNM has offered to provide input on Nuclear Medicine.
Alain Gauvin from the McGill University Health Center has offered to contribute his experience as a diagnostic medical physicist and radiation safety officer.
VA is interested in Dose Reporting and is considering setting up a Dose Registry. They would like to be involved in Public Comment.
Some additional background
Final report by The UK Clinical Advisory Committee on diagnostic imaging on the safety and efficacy of CT screening - July 2007 expresses a need for an integrated framework for policing and compliance of national and international protocols for radiation exposure.
The French Ministry of Health (Roselyne Bachelot), decided in July that Radiation Dose Follow-up is now a major work item for the ministry. The ASN (Autorité de Sûreté Nucléaire - Nuclear Safety Authority) is now in charge to develop the necessary infrastructure. On September 10, ASN and IRSN (Institut de Radioprotection et de Sûreté Nucléaire - Radioprotection and Nuclear Safety Institute) met at SFR (Société Française de Radiologie - French Radiological Society) and decided to investigate how to build a National Radiation Dose Registry, SFR will organize information for RIS vendors in France, ESR (European Society of Radiology) supports the profile and coordinate action at a European level for RIS and Modalities Vendors. ASN, IRSN and SFR will continue to work together on the topic with GMSIH. ASN, IRSN and SFR strongly support the profile proposal. SFR is willing to organize a highly visible event during JFR'2008 (French Congress of Radiology) to inform attendees of the work and a demo of the profile at JFR'2009 with the support of ASN and IRSN.
Dose DataMed, a project granted by European Union wished such a work for a long time and is happy about this proposal and will possibly provide some support.
The ACR Dose Index Registry (DIR) will conduct a fall 2007 pilot program involving seven facilities and launch nationwide in 2008. This registry will initially collect dose estimates from CR and DR.
There are some thoughts in Canada about a patient carrying a lifetime dose record which would get updated by a facility when the patient visits for a scan (sort of like immunization records are done in some places).
<List technical or political risks that will need to be considered to successfully field the profile.>
Vendors might not implement the profile. During JFR'07, some discussion with vendors employees show that they don't understand why the profile use the Radiation Dose SR instead of old MPPS attributes. It seems that there is a need for the vendors participating (or not) in the profile development to organize a strong education among their employees worlwide to explain why it is not relevant to continue to use MPPS for Radiation Dose Follow up purpose but it is necessary to move to the Radiation Dose SR.
- Need to keep vendors in the loop (e.g. those who meet at NEMA MII)
It might be hard to get consensus on dose attributes between different national groups.
- Consider specifying a superset that each country can focus on their part of. If there are points of disagreement, consider dealing with them in national extensions or alternate templates. The technical committee will ensure that all types of dose measurements maybe entered. Setting or recommending dose policies is clearly outside the scope of the profile. Since various International Regulatory Groups may analyze/interpret things differently, there will be no attempt to dictate the use.
Controversies exist over what are the most relevant/useful measurements to make and how those should be be interpreted.
- An IHE Profile will need to limit requirements to common denominator measurements and to empower analysis but not to dictate interpretation/policy.
8. Open Issues
How should manual entry of dose information be handled: In CR, often the X-Ray device, dosimeter and CR reader are from different vendors and not connected. Manual entry could also handle legacy systems. Could a CR reader submit an SR dose report? Is there enough data CR IODs?
- A solution for manual entry should be supported to handle legacy and primitive systems. A PC supporting manual entry might be indistinguishable from a Modality actor in this profile so, maybe they just claim to be modalities. Will need to investigate the CR question. Note for some modalities, manual entry could be painful. If manual input is allowed, then calculations should be clearly marked as potentially erroneous.
- It may be necessary to choose a simple, but standardized dose exposure unit that could help simplifying exposure calculation for exposures to modalities that do not allow manual input and then, provide objective data.
- In theory, the Dose SR plays right into CR. A CR X-Ray device could send the dose SR to the PACS, while the CR reader sends the image. Apparently some CRs support a standard called x-con to transfer exposure details from the X-Ray device to the Reader.
The initial focus is on X-ray based diagnostic devices. Do we want to address SPECT and PET imaging now? Do we also want to address therapeutic devices in this profile? Bedside devices? Substances? (seeds, radionuclides)
- Focus on X-ray based diagnostic devices. NM and therapeutic devices could be added later as additional actors (assuming the clinical and DICOM work has been done). Bedside x-rays are similar to mobile ultrasound; the same protocols will work, they just need to support caching and dumping reports.
What types of dose information need to travel into HL7 space? Who should take care of translating them from DICOM? What should they be translated into? CDA?
- Some information likely will need to travel into HL7 space. It will be translated by whatever actor lives on the border. It will likely be translated into CDA or XML since DICOM is poorly supported outside the imaging dept. XDS may be a good transport for outside the enterprise. If CDA is chosen consider working with DICOM WG-20 for help/advice publishing in HL7.
Should we place any timing requirements on the submissions? I.e. how quickly they must, or how late they're allowed to, or some relation to the study?
- It would be nice to avoid such requirements. Will need to consider if they are necessary.
First priority is to begin capturing the data using standards. Later it may be useful to tie this into the National Registry Submission work being done in IHE Cardiology and IHE Quality.
Are there additional complexities related to Dose Modulation that might need addressing?
- It was pointed out by one of the CT vendors that for some dual kV systems it's not just the mA that gets modulated.
AAPM Mtg in Houston in July 2008 might be a good place to introduce the Profile.
John Boone (Chair of the AAPM Science Council) might be able to help set this up, and is also interested in involving the AAPM in the Public Comment phase.