Dose Watch is a multi-modality and multivendor data collection system that can easily be connected to CT,
angiographic systems,
conventional diagnostic mammography and conventional flat panel X-ray.
The software adapts the collection of dosimetric indicators adjusted for the characteristics of the equipment and is able to handle different types of linking modality: Radiation Dose Structured Report (RDSR),
Modality Performed Procedure Step (MPPS),
image DICOM header (RAW),
and File Transfer Protocol (FTP) connection.
DoseWatch performs automated analysis but permits also off-line analysis,
generating two Excel files,
"General Data Export" and "Detailed Data Export".
Depending on the user needs,
patient data can be exported as fully anonymized or partially anonymized.
For our study the software was linked to a CT scanner (Brilliance iCT 256,
Philips,
Best,
The Netherlands) and to a digital angiography equipment (Philips Allura FD20).
The reliability of the MPPS connection system was assessed for both modalities.
Data analysis was performed on exported data.
Dose Watch identified the CT dose through the "Local Study description",
which was derived from the request in the RIS system,
and "Protocol study name" from the CT scan.
We avoided the use of "Local Study Description" because was not necessarily related to the required exam.
The angiographic exams were identified merely by the "Local Study description".
Results
There were not connection issues and adequate recording of data for dose exposure evaluation has been verified.
It should be remembered that the modality in use affects the amount of information that are stored for each examination.
For connection with the CT it has been necessary to use the MPPS protocol,
which guarantees the transfer of a rather limited number of data.
In angiography it was possible to adopt the RDSR mode (Radiation Dose Structured Report),
which is the easier solution for file management,
even if the amount of information transferred is not standardized but depends on the equipment brand,
age and software version.
To use efficiently a dose tracking program is essential to check out how many investigation protocols are used on a single equipment.
In fact,
an excessive number of protocols denotes a standardized way of working and could affect the validity of statistical analysis.
85 different protocols show up in our CT investigation database,
but 87% of the activity is carried out using only 14 different investigation protocols (Figure 1).
In addition,
39 protocols are used less than 10 times per year.
In the evaluation of dose exposure aspects,
were taken into consideration the more frequently performed examination types:
- brain
- chest
- abdomen
- chest – abdomen
The 98.5 % of head examinations are investigated with the same type of protocol,
with 7430 tests performed in the period under analysis.
86 % of these tests was carried out with a single acquisition.
The graph (Figure 2) shows the distribution of DLP values (mGy · cm) for investigations carried out with a single series,
the mean,
the median,
75th percentile value and diagnostic reference level ( LDR ) value provided in the current legislation.
Notice that it would require a re-evaluation of the execution modalities of examinations,
to reduce further the dose,
paying particular attention in the selection of acquisition range.
However,
this is not a particularly critical situation.
730 examinations of the chest are available for analysis.
Five different imaging protocols are used ,
two of them are chest-specific (HRCT) and three others are generic ,
difficult to differentiate on dose-indicators analysis.
It is a survey example where there is the need to check if selection of protocols is correct.
About 2000 examinations of abdominal area were recorded in the database,
with acquisition mode analysis quite more complicated than for head and thorax.
Table 1 shows DLP average values (in mGy · cm) as a function of acquisition series number,
for principally available protocols (those used less than 10 times during observation period have been removed).
The third place for execution frequency is occupied by one of the acquisition protocols from the district "Chest - Abdomen" (indicated as Thorax / TORACE_ADD_MDC) ,
for which there are about 300 investigations available to be analyzed; them were carried out with a variable number of series,
according to diagnostic question.
Approximately 100 surveys of the same district were acquired with an alternative protocol (Thorax / TOR - ADD_2_ELICHE).
The comparison between the average DLP values,
on the basis of serial number acquired,
does not show significant differences from a dose exposure angle of view (Figure 3,
values obtained on less than 10 examinations are not shown).
For this type of examination LDR is not available.
The same investigation on the database related to interventional procedures has instead highlighted a real issue.
For these investigations,
the procedures are identified only by the "Local Study description" that matches a description taken from nomenclature used by the RIS; in case of surveys carried out in emergency setting,
it is show only one of the protocols selected on the angiogram, which are very general (head,
abdomen,
peripheral); so there is no information on procedure performed because Dose Watch does not export this information.
Crossing the informations from dose tracking program database and another archive "manually" managed by the operators with clinical indications related to each procedure,
the correlation between "Local Study Description" and clinical procedures performed was analyzed.
Table 2 lists the "Local Study description" associated with interventional procedures in the abdominal area and "clinic" classification of clinical procedures performed in the same district.
Dose exposure indicators available are the DAP (Dose Area Product) and the air accumulated KERMA,
in the interventional reference point .
As seen in the top chart ,
the results obtained from calculating the DAP average value for each "Local Study description" are absolutely not related to those deduced by averaging the dose exposure indicator for the different clinical procedures (Figure 4).
To effectively use DoseWatch it is fundamental the univocal association between the clinical procedure and from x-ray exposition related data,
otherwise the analysis should become lesser solid and,
in extreme cases,
unreliable.