Want to increase your reporting efficiency? Check your PACS/RIS. That's the message of researchers from a top London facility who found that, depending on the PACS/RIS, between 33% and 51% of reporting time can be spent on tasks not related directly to image interpretation.
The group evaluated the time it took to report chest x-rays on two PACS/RIS, identifying the amount of "useful" and "wasted" time for each. Led by fourth-year medical student Abhinaya Chandrashekar from Imperial College, the researchers did not specify which system performed better, merely noting stark differences between the two.
"Optimizing PACS/RIS technology with an emphasis on efficient radiology reporting has the potential to help deal with plain-film reporting backlogs that are well-recognized in the U.K.," she and her colleagues wrote in an e-poster from the 2018 European Society of Thoracic Imaging (ESTI) and European Society of Cardiovascular Radiology (ESCR) annual meeting in Geneva, Switzerland.
Going head to head
PACS and RIS are integral parts of the reporting process and help radiologists with productivity. However, despite efforts to improve overall efficiency in radiology, little is known about the impact of PACS/RIS on radiologists' reporting efficiency, explained the authors, who sought to change that by evaluating the efficiency of two different PACS/RIS from a user perspective.
In the study, five radiologists cumulatively reported 200 chest x-rays in five batches. They used an Agfa Healthcare PACS/RIS and a Carestream PACS/Soliton RIS at two different institutions. Three radiologists reported solely on the Agfa PACS/RIS, one radiologist reported solely on the Carestream PACS/Soliton RIS, and one radiologist reported on both systems. The researchers timed each radiologist using an electronic stopwatch with split-timer functionality so as to evaluate "wasted" reporting time and "useful" reporting time.
Directly looking at a patient's x-ray and dictating the report was deemed "useful" time. All other time was "wasted" and included the following:
- Clicking on the x-ray image
- Waiting for the system to process and open the patient's x-ray
- Retrieving the patient's clinical information details
- Retrieving the patient's previous relevant radiographs
- Resolving error messages
- Correcting dictation spelling errors due to inaccurate voice recognition.
To minimize bias, each radiologist was timed by the same individual. Once the radiologist reported five chest x-rays, the timer was stopped immediately, and the total, useful, and wasted time values were recorded. The radiologist was then provided with a break of five minutes, after which the reporting of another batch of five x-rays commenced. No interruptions were allowed during reporting.
"To account for the inevitable variation in total speed of reporting amongst the radiologists, given varying levels of chest x-ray complexity and reporting expertise, the relative proportion of 'useful' and 'wasted' time in comparison to total time was measured, rather than the raw data values," Chandrashekar and colleagues wrote.
They found average wasted time per five chest x-rays was approximately 51% and 33% for the two systems. For the radiologist reporting on both systems, a substantial difference in average time taken to report five chest x-rays was noted between the two systems: 9 minutes versus 2 minutes. Average overall time taken to report 100 chest x-rays, excluding breaks, was more than double using one system (127.64 minutes) compared with the other (55.05 minutes).
The results are summarized in the table below.
| Results of reporting on 2 PACS/RIS | ||
| System 1 | System 2 | |
| Average time taken to read 5 x-rays | 6.38 min | 2.75 min |
| Median time taken to read 5 x-rays | 5.22 min | 2.80 min |
| Total time taken to read 100 x-rays excluding breaks | 127.64 min | 55.05 min |
| Average "useful" time per consultation | 49.16% | 67.08% |
| Average "wasted" time per consultation | 50.84% | 32.92% |
"The choice of PACS/RIS utilized can substantially influence radiologists' reporting times," they concluded.
