The ACR DIR DR module is designed to monitor practice patterns and establish dose index benchmarks for common digital radiography (DR) examinations. The quality feedback reports provided by the module enable facilities to compare their dose data to national aggregate values to help determine if changes in their acquisition techniques or practice are needed to improve patient care.
Data Collected
The DIR DR module automatically de-identifies exam data sent to the registry and extracts DR-specific exposure and dose indices:
Exposure Index (EI)
Target Exposure Index (EIT)
Deviation Index (DI)
Kerma-Area Product (KAP)
Data elements describing patient age, body part, anatomical view, and station name are collected for data stratification. Other data elements that are collected, but not yet reported, can be found in the projection x-ray DIR data dictionary.
Exam descriptions are standardized using ACR Common radiology terminology and semantic structures to ensure data comparability across facilities and scanners. The large volume of data collected allows the DIR DR to establish dose index benchmarks for six distinct age groups (<1y, 1 to <5y, 5 to <10y, 10 to <15y, 15 to 18y, 18y and over).
Report Features
The DR module uses an interactive database that allows a registry participant to review their facility’s dose index statistics and compare facility results against that of all DIR DR participation sites. Various dashboards are available to benchmark facility performance using scatter plots, bar graphs, and box-and-whisker plots. Within an institution, comparisons can also be made across different DR manufacturers and across user-defined station groups. These visualization tools allow a participant to identify problematic DR exams, facilities, or stations to guide the effort of continuous quality improvement.
Compared to CT and fluoroscopy, a unique aspect of DR is that the dose indices are event-driven rather than examination-driven. A radiographic examination may consist of two or more anatomical views (e.g. posteroanterior and lateral chest views). Each anatomical view may have its own exposure technique. Therefore, it’s more meaningful and useful to report dose index statistics for each anatomical view separately.
Currently, however, only about 75% of the Radiation Dose Structured Reports (RDSRs) contain view information. Extracting view information from the RDSR is also a non-trivial task (DIR DR Anatomical Views). For these reasons, the DIR DR module reports dose index statistics at both the view level and the procedure level:
The DR Standardized Dose Index Report by View is limited to the most common (e.g. chest radiography) or higher kerma-area product (e.g. lumbar spine radiography) studies that are mapped to an ACR Common ID. Dose index statistics are reported by body part and anatomical view.
The DR Standardized Dose Index Report by Procedure includes all submitted studies mapped to an ACR Common ID. Dose index statistics are reported by ACR Common ID, Study Description, or Requested Procedure Description.
Use Case
Institutions can use the national dose index benchmarks (registry quartile values) as a quality check to look for opportunities to improve their techniques or practice.
In the example below, the DR Quality Committee at an institution would like to know how the DR dose indices at their institution compare to the national benchmarks. They start with the Bar Graph by View dashboard which displays the median EI values for various body parts and views at the institution as blue bars. As benchmarks, the registry 25th, median, and 75th percentile values are displayed as red lines. It’s clear to the committee that chest AP views acquired using fixed devices are a problematic area as the institutional median value greatly exceeds the registry 75th percentile.
