Radiation Dose Use Cases
<<Edits to this page are no longer useful. This work has now been moved back into the Radiation Exposure Management Profile Draft.>>
The Radiation Dose Profile is intended to facilitate a variety of radiation dose related Use Cases.
Generally, the Profile will not directly mandate feature details (such as the contents or scope of summary reports a hospital dose information manager system might generate for policy monitoring), but it will attempt to mandate sufficient infrastructure (such as the transactions to get the necessary dose details from modalities to a dose information manager) to make it feasible for products to implement the features to meet the use cases.
This page is a discussion area to flesh out various use cases. They will be used as test cases for the specification in the Draft Profile.
Where possible, include details such as specific attributes, analysis methods and report contents.
- 1 Example of some of the problems
- 2 Broad Scenario from the Profile Proposal
- 3 Regulatory Use Case
- 4 Registry Use Case - French Ministry of Health
- 5 Registry Use Case - ACR
- 6 Hospital Use Case – Impact Evaluation of Specific Patient
- 7 Hospital Use Case - Policy Compliance Monitoring
- 8 Hospital Use Case – Prevention / Improvement
- 9 EMR Use Case - Patient History / Preferences
- 10 Clinical Trial Use Case
Example of some of the problems
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 divided 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.
Broad Scenario from the Profile Proposal
When Mr. Smith's order is placed, the physician can query the system about 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.
Regulatory Use Case
Several groups interested in regulatory issues (IEC, FDA, AAPM, etc) drove definition of the DICOM Dose SR Objects in the first place. How do they envision the objects being used?
ACR is not a regulator of dose. They will collect dose information at a national level, but not directly for regulatory purposes. (Analysis of the collected data might inform guidelines that could find their way into accreditation).
European regulation is based on European Directive Euratom 1997/43/EC. The application of the directive is mandatory in all EU countries (27 today) since 2000. All european national regulations have to be compatible with the directive. Euratom specifies a follow up for each patient and statistics at population level.
<The following national subsections should focus on national specifics/deviations/additions, and the common details should be moved up here to the European level.>
Today two regulations specify the dosimetric information that a user of ionising radiation must provide in France:
- One text, published in March 2004, concerns Dose Reference Levels (DRL) (Arrêté du 12 février 2004 relatif aux niveaux de référence diagnostiques en radiologie et en médecine nucléaire – Journal Officiel de la République Française du 16 mars 2004) (Since this seems to be in support of a registry, the details are listed below in the Registry section)
- Another text, published in September 2006, concerns dosimetric information that must be present in the medical record of the diagnostic examination (Arrêté du 22 septembre 2006 relatif aux informations dosimétriques devant figurer dans un compte rendu d'acte utilisant les rayonnements ionisants - Journal Officiel de la République Française du 29 septembre 2006).
For each medical procedure using ionising radiation, the medical record must include information allowing estimation of the dose received by the patient.
In Diagnostic or Interventional Radiology for procedures concerning the head, neck, thorax, abdomen or pelvis, it must record:
- type of examination (radiography or fluoroscopy)
- dose area product (DAP)
- If DAP is not known and the patient is either under 16yrs, a woman of childbearing potential getting a pelvis exam, or a pregnant woman getting a justified pelvis and/or abdomen exam, then it must record:
- high voltage (kV)
- available details of the following:
- intensity time product
- source-skin distance
- fluoroscopy time
- intensity for each field size
- number of exposures in graphy
In Mammography the requested information is:
- mean glandular dose
- If the device doesn’t allow direct measurement, an estimation could be provide by the product of number of exposures and dose on phantom measured each six month during mandatory external QA.
In Computed Tomography for examinations concerning the head, neck, thorax, abdomen or pelvis, the requested information is:
- dose length product (DLP) must be indicated with distinction between sequences concerning head and neck on one side and trunk on the other side
- If the DLP is not available, the following information can be provided:
- examination length
- CTDIw or CTDIvol
- For women with childbearing potential or pregnant women (justified examination) it is mandatory to indicate:
In Nuclear Medicine
- radiopharmaceutical name
- administered activity
- way of administration
German & Dutch Regulatory
Germany and Netherlands dose information does not go above the Hospital level. Dose information is required and must be auditable.
Germany is working on additions to DICOM Basic Diagnostic Imaging Report (SupXXXX in WG-20) to include Radiation Regulation related details such as:
- Pregnancy Status
- Indications for Procedure
- Physician Responsible for Indication
- Performing Person (who administered)
- Radioactive Substance Administered
- Radiation Exposure (text description of "the exposure")
- Performing Person’s Organization Name
Spanish dose information does not go above the Hospital level. Dose information is required in order to audit the use of diagnostic equipment (comparing with Dose Reference Levels) and must be auditable by the responsible health authority. Also the responsible health authority and the Nuclear Security Council will guarantee that the distribution of the estimations of resulting individual doses is determined, for the population and the significant groups of reference of the population, whose results will be sent to the Ministry of Health and Consumption.
Registry Use Case - French Ministry of Health
French regulations mandate the submission of Dose Reference Levels data to a federal registry.
Every year each specialist must submit dosimetric information (at least 20 patients) for each of at least two examination types (generally the most frequent or most irradiating examinations).
In Conventional Radiology the requested dosimetric information is the entrance surface dose (in mGy) for one exposure or the dose area product (DAP in mGy.cm²) for one examination or for the overall procedure. Examinations of interest are:
- Lung (PA)
- Lung (profile)
- Lumbar spine (AP)
- Lumbar spine (profile)
- Skull (AP)
- Skull (profile)
- Lung (PA) – 0-1 year
- Lung (PA) – 5 years
- Lung (profile) – 5 years
- Skull (PA or AP) – 5 years
- Skull (profile) – 5 years
- Pelvis (AP) – 0-1 year
- Pelvis (AP)- 5 years
- Abdomen – 5 years
In Computed Tomography the requested dosimetric information is the weighted CT dose index (CTDIw in mGy) for each sequence, and the dose length product (DLP in mGy.cm) for each sequence or for the overall procedure. Nevertheless the volume CT dose index (CTDIvol in mGy) is also accepted. Examinations of interest are:
In Nuclear Medicine (out of scope of the REM Profile this year) the requested dosimetric information is the injected activity (in MBq) and the type of radiopharmaceutical. Examinations of interest are:
- Bone scintigraphy
- Lung scintigraphy
- Thyroid scintigraphy
- Cardiac SPECT
- Cardiac scintigraphy
- Kidney scintigraphy
- Dynamic kidney scintigraphy
- Brain SPECT
- Scintigraphy of somatostatin receptors.
- PET FDG
Registry Use Case - ACR
ACR is forming a national registry for Radiation Dose. The initial focus is CR, DR and CT. They have not yet thought through the implications for diagnostic or interventional XA.
The primary function is to benchmark dose protocoling with the goal of minimizing dose to the population.
ACR is interested in collecting a cumulative log:
- “Protocol used”
- Reason for exam?
- Patient gender, age – to permit age/sex specific guidelines (different risks)
- Patient weight – affects dose, affects imaging, used in guidelines
- Patient ID (pseudonymized) – to evaluate individual dose histories
- Technologist ID (pseudonymized) – to compare inter-operator variability
- Time of acquisition? – to compare shift variability?
- Equipment details
- Date of Last Calibration – affects dose? Want to track compliance?
ACR has no plan to become a national personal dose record but are very interested in collecting data about “frequently scanned patients” and the cumulative dose.
Records should be at the encounter level.
The ACR service would be to review the performance of a given hospital against the benchmarks and perhaps provide performance summaries for “comparable” facilities.
The ACR benchmark report is your report card. Benchmarking will be challenging because technology moves very quickly. Higher levels is not necessarily bad. Various factors can impact.
Hospital Use Case – Impact Evaluation of Specific Patient
A few days after a CT exam is carried out for a young female patient, that patient is identified as pregnant (this was not identified before the scans because tech didn’t ask, patient misunderstood, patient didn’t know, etc.). The referring physician is at that point aware that the newly discovered pregnancy of the patient is problematic in this context, and requests from the radiologist who had read the exam an evaluation on the risk to the fetus. The radiologist requests a hospital medical physicist to provide her with an estimate of the radiation dose received by the uterus in the course of the CT exam.
The medical physicist retrieves the images of the study in question, and determines with the help of the radiologist which series encompass the uterus. Knowing which series are of interest, the medical physicist is then able to leverage the dose indicators and weight of the patient contained in the image headers to sum up the total dose to the uterus. The dose value is then provided to the clinical coordinator of the department, who enters the dose estimate in the RIS. This causes a statement to be automatically appended to the report (which had already been signed off), and the status of the study is automatically demoted to "pending signature" as is the case when an addendum is added to an existing report. Before signing off the report, the radiologist completes the addendum with her estimate of the risk to the fetus given the dose measurement, and communicates this result by phone to the referring physician, who also receives the written addendum electronically signed by the radiologist a few days latter.
Hospital Use Case - Policy Compliance Monitoring
Hospitals generally have policies relating to patient radiation dose. They would like to have tools to help them monitor whether those policies are being followed.
Might want to look at average patient dose for each tech, for each shift, for each room.
Breakdown by patient age group, weight group, gender, procedure,
In theory, this is one way to counter dose creep. Image quality sets a low side limit on dose (too low and the images are unacceptable to the radiologists). Policy can be used to set a high side limit then monitor and encourage people to stay on the right side of the limit.
Hospital Use Case – Prevention / Improvement
As part of the departmental quality improvement program, the radiation safety officer (RSO) of the hospital accesses the RIS to carry out his bi-monthly assessment of radiation dose use. For a selected set of procedures, the dose-area product of a same procedure is evaluated between the few rooms where it may take place, and no significant variation is found on the average over the last 6 months. Another report of dose per procedure is then obtained, but this time the comparison is made between performing radiologists for interventional procedures, and performing technologists for CT and radiographic exams.
It becomes clear that a newly arrived radiologist tends to generate for a certain procedure 2 to 3 times the dose-area product of his colleagues, whose averages are met in a narrow cluster well below the value created by the newcomer. While the dose-area product sub-total in fluoroscopy is much more similar to the one of other radiologists (which is consistent with the average fluoroscopy time of the report), there is a significant discrepancy in dose-area product for the acquisition sub-total. The number of acquired images also corroborates this, as it is also relatively remote (higher) from the departmental average. The RSO writes a memo to the department chair about this situation, who raises the issue at the weekly radiologist meeting. Upon discussion, it becomes clear that the radiologist uses a supplementary acquisition which his colleagues do not utilize. After still more discussion, radiologists agree that the acquisition, although moderately useful, probably does not bring any information which would not be picked up in the rest of the exam, and it is agreed that it should not be done. The RSO controls the situation a month latter, and is reassured by the results which show that all radiologists now make use of a similar radiation dose for that study.
EMR Use Case - Patient History / Preferences
- patient talks to specialist and doesn't want a new scan
- the details should be in the patient history
- docs often don't think about it, we need to raise the flag, then give them details
- important to note that this has to go to the doc
- Techs will alert the rad when the order comes in. Rads are responsible for the tech behavior.
- tech may have some opportunity to bias the study to lower dose based on internal policy
Clinical Trial Use Case
A site proposes a low dose CT lung screening procedure but currently it's difficult to quantitatively demonstrate/validate the dose benefit, they can only report on the detection rates.
This profile could enable lots of such studies.