All feel the relentless pressure to improve efficiency and quality in their jobs. Diagnostic Radiology seems to be particularly targeted as the U.S. Centers for Medicare and Medicaid Services (CMS) announced 3% cuts to Radiology next year [link]. This article examines several features available in new ultrasound reporting software to significantly improve reporting efficiency.
Some have taken the first step in reporting automation – automatic measurement transfer.
Most do not realize there is an equally large opportunity in:
- Conditional Text Generation
- Automatic Fetching Prior Measurements and Calculations
- Integrated Clinical Decision Support
Embracing Reporting Automation
Ultrasound is a unique modality relying on sonographers for image acquisition, measurements, and preliminary impressions. Current workflows require duplicative efforts from sonographers and Radiologists to generate a report. Surprisingly, a minority of Radiology departments have implemented simple automation efforts, i.e., ultrasound DICOM SR data transfer into Powerscribe or other Radiology Reporting software [link]. I believe automation efforts lag for the same reasons that structured reporting initiatives are slow – the initial effort is large, and there is resistance to change.
Recent Imorgon Medical research found that 50% of Findings in a Radiology report come from qualitative data rather than measurement data [link]. The implication is that more than automating measurement transfer is required to improve reporting efficiency significantly. There is a significant opportunity with qualitative data. This article summarizes three features that maximize the return on any reporting automation effort.
1. Conditional Text Generation is next-level Ultrasound Reporting Software
Efficient ultrasound reporting necessitates swift translation of a sonographer’s findings, captured on a worksheet, into a finding or impression statement in a Radiologist’s report. Many Radiologists are familiar with structured reporting templates with default/normal findings prepopulated. However, when findings deviate significantly from normal, a Radiologist returns to more traditional dictation. Some Radiologists (a surprisingly small number) are familiar with measurement transfer from ultrasound. Conditional text generation similarly transfers sonographer observations directly into the reporting package.
The following lower extremity venous example is intentionally simple yet a powerful example in several ways: First, recreating this efficiency would be difficult without conditional text. Second, the sonographer is not likely to get over-ruled as they are the only ones who have scanned the patient. The Radiologist ensures clinical questions are addressed. The productivity improvement is beyond argument.
Lower Extremity Venous Report Template
Video 1 depicts an Imorgon Report Accelerator ultrasound worksheet with automated text generation. This example displays text to transfer. The following list highlights some of the productivity and quality improvement features:
- The green “findings” and “impressions” are similar to the default text in a Radiology template. A single selection adds the right-side findings and impressions.
- One selection of a non-occlusive femoral vein thrombus automatically updates both Findings and Impressions. The findings maintain the details of the exam and the impression notes a deviation and summarizes the remainder.
The image below illustrates mapping sonographer worksheet inputs into Powerscribe.
Conditional Text Generation goes beyond assigning work to another person.
It enables the person closest to the patient to generate report findings with a few selections.
2. Automatic Prior Measurements and Calculations
Illustration 2 showcases an ultrasound thyroid template by automatically retrieving prior DICOM SR measurements. There were four follow-ups over the past years. Ellipsoid volume and interval difference calculations are automatic. Judging by the measurement variation, previous measurements were only referenced after new measurements were taken. This simple example illustrates the opportunity of prefetching priors and automatic calculations to enhance report quality, reduce reporting time, and minimize diagnostic errors.
Thyroid Ultrasound Worksheet with Prior Measurements
Everyone knows to look up prior measurements,
yet everyone hates looking up priors.
3. Integrated Clinical Decision Support
Clinical decision support tools, like Thyroid Imaging Reporting and Data System (ACR TI-RADS), aim to standardize interpretation, reporting, and patient management. TI-RADS studies show a beneficial 17%-53% reduction in biopsies with a low false negative rate (NPV, 97.8%). (Grani et al.). However, existing tools have several shortcomings:
- Separate apps or web pages lack integration into the exam and sonographer worksheets
- Incomplete guidance implementation
- Results need manual transcription
Ideal clinical decision support includes worksheet integration, encompasses the entire clinical guideline, and automatically updates report documentation based on inputs.
Clinical Decision Support – TI-RADS
Video 2 illustrates the features of a modern ultrasound reporting system with an integrated clinical decision support system such as TI-RADS. Note the features:
- Automatically collapsing sections make unnecessarily long and complicated ultrasound worksheets accessible. The thyroid worksheet only shows measured nodules even though the worksheet template supports three nodules per lobe. Also, cystic and spongiform compositions hide unnecessary further characterizations to simplify the worksheet.
- TI-RADS guidelines are more fully implemented than other web implementations. A solid/mixed composition selection opens the remaining TI-RADS options. Additionally, echogenicity, taller-than-wide, and margins are exclusive radio buttons, while allowing multiple echogenic foci selections.
- Interval growth and small-size cases are integrated into the tool.
- No additional transcribing step is necessary as each selection updates impression and findings statements.
- TI-RADS grading, size, and text generation automate findings and boilerplate recommendations.
- All of this information is transferred directly into Powerscribe.
It is easy to imagine improved departmental workflow when all the integrated clinical decision support system features are combined.
It is easy to visualize productivity improvement with these features.
A key question is to ask is whether there will be more or less automation in the coming years?
Conclusion: Ultrasound Reporting Software Automates Radiology Reports
Reporting automation becomes necessary as Radiology faces continued reimbursement cuts and increased volumes. Ultrasound reporting software can now dramatically improve reporting efficiency in every Radiology department. Embracing this change will allow one to take advantage of the efficiency improvements that make your life better. The reporting automation features described are the beginning of automation features we will see in the coming years.
If this message resonates with you and you seek to explore its potential, please don’t hesitate to contact me at firstname.lastname@example.org. Lastly, we’d love to hear your thoughts – please share below.