Web Exclusive

The Speed of Thought

Combat Radiologist Burnout by Reducing Technology Hang-Ups

By Samir Mehta, MD

In today’s health care environment, radiologists are tasked with being the Medical Six Million Dollar Man. As providers, you’re expected to read images faster, be more accurate, and achieve quicker report turnaround times to satisfy referring physicians and meet critical patient care needs. In essence, you’re being asked to work as fast as you think. 

To reach this thought-speed goal, you need technological intervention. But technology is only truly valuable if it is designed around your needs. As the most tech-savvy and tech-dependent medical specialty, radiology depends on technology for advancement, but we in the field often feel it ignores our practical workflow needs. Previous technical advances—from PACS to speech recognition to the EHR—have fallen short of their promise of workflow simplification. They still leave you to pull the data you need for interpretation from disparate sources. Advanced software solutions could tie it all more closely together.

Consequently, radiology needs renovation—not a major overhaul of how the specialty functions but rather a handful of small changes that could profoundly improve your practice. What would it mean to focus solely on the images and be confident that your dictations always populate correctly in the report? Reinventing productivity this way requires unifying radiology’s siloed workflow components: the worklist, viewer, report, and AI. Optimal renovation flows from facilitating the needs of the radiologist first.

A collection of innovative micro-optimizations designed to lift the burden of technological impediments can help reach this goal. Redesigned workflow that addresses these hurdles could increase productivity and reduce your chances of burnout. Following are five small changes that can make a big impact on your practice.

1. Integrating Visual Navigation: For efficiency and accuracy, you must dedicate your brainpower and visual focus to images. Breaking that concentration creates delays and added work. Current radiology systems, however, undermine this need for more structured reporting to convey results to referring physicians.

Your practice, like many others, may have hundreds of available reporting templates that reside on a different screen from the viewer. That separation can be at odds with your search pattern. As a result, you’re forced to constantly have your head on a swivel, pulling your eyes away from the scan. You’re routinely double-checking that you’re looking at the right part of the template when dictating.

Integrating your visual navigation between the viewer and the reporter eliminates the continuous back-and-forth motion. Doing so will noticeably improve your ergonomic setup and can accelerate turnaround times.

2. Connecting Silos: Successful image interpretation is a multistep process that relies on several workflow components. However, these siloed pillars don’t communicate well. AI output isn’t always presented in the proper context, and the functionality of existing solutions is limited. As a result, these tools often don’t provide any true clinical value.

For example, an AI algorithm may detect a lung nodule, but its output is not overlaid on the originally acquired series; this disrupts workflow. The algorithm also won’t indicate whether the nodule was present on a prior study. Additionally, even after you’ve checked the algorithm’s finding on the acquired axial in this case, you still must dictate the entire finding into your report. Integrating these silos would make it easier to pinpoint previous images and reports that may contain valuable context about a patient’s condition. 

3. Defragmenting Data: Massive amounts of data live in EHRs and radiology systems. However, integration between the two repositories is lacking. This lack puts you in a precarious position, as you’re viewed as being solely responsible for any nuggets of patient health data that appear in images. 

When you detect findings unrelated to the current exam’s chief complaint, it’s difficult to determine whether they’re relevant without reviewing prior studies. If you dismiss the finding without a proper assessment, you run the risk of the patient going without needed intervention. Unfortunately, sifting through the ocean of information to find related prior images and details in patient histories is time-consuming. For example, a system that could automatically highlight previous reports of bony erosion throughout the body and hang respective prior studies in the viewer could provide valuable context to a current study that demonstrates a bony erosion in the hand. 

Currently, due to fragmented EHR/PACS integration, these types of analyses take too long. Radiology could benefit from an integrated system that allows you to comb through EHRs for prior diagnoses and old radiology reports that show possible disease progression efficiently.

4. Standardizing Hanging Protocols: Radiologists are creatures of habit, and no two are alike. That includes your preferences for hanging protocols. However, several hurdles currently exist to trip up your eyes and slow you down. Existing PACS systems can’t always effectively differentiate between prior exams to select the optimal ones for your review. It’s a problem that emerges when images are captured on different scanner models or brands, or if they’re gathered in a different institution. Variation in imaging series names is also a stumbling block to efficiency and productivity.  

As a result, you may perform mental gymnastics routinely as you repeatedly reorient images. Manually rearranging a protocol not only lengthens turnaround time but can also exacerbate provider burnout. 

In addition, hanging protocols are static. A dynamic hanging protocol that unifies and automates how you use viewports and change windows between report headers could further streamline the interpretation process. For example, when you’re reading a brain MRI, a system that allows you to say, “hang stroke” or tab into the stroke section in your report and proactively hang and maximize the image series (DWI/ADC) can be vital for stroke detection. Ultimately, standardizing hanging protocols could save you between 10 seconds and one minute per image, totaling up to 30 minutes or more every day.

5. Streamlining Log-Ins: If you’re like an increasing number of radiologists, you interpret studies for multiple institutions, requiring you to have a unique log-in for each PACS, EHR, and e-mail system. Even if these institutions rely on the same software, each installation is different, so you can’t save and transfer your settings from one system to another.

On the best of days, logging out of one system and into another is a minor hassle with minimal impact. It can be critical, however, in emergency situations. If you receive an urgent call in the middle of the day, the seconds or minutes lost to logging out of one system and accessing another has the potential to negatively impact patient outcomes, particularly with stroke care. Implementing an overlay system that provides access to all software tools via one log-in could alleviate this issue.  

To optimize radiology’s workflow for looming patient care needs, micro-optimization innovations such as these are needed. Each small step brings you closer to working as fast as you think.

— Samir Mehta, MD, is head of clinical operations for Sirona Medical.