How one radiology department leveraged a semi-automated workflow to solve its ‘plain film problem’

After one Texas academic tertiary center expanded its community presence, its imaging department was saddled with 22% more radiographs and radiologists too overworked to read them.

The conundrum became widely known as the “plain film problem,” researchers explained in a case study shared Thursday in JACR. Attempts to mandate or incentivize rads to interpret these excess exams fell flat: the “carrot-and-stick” approaches proved ineffective, they noted.

That’s when Texas Children’s Hospital and Baylor College of Medicine providers turned to the idea of “nudge theory.” The concept posits that changing the environment is more effective at influencing human behavior rather than implementing top-down rules.

“We saw an opportunity to change our reading environment to make interpreting radiographs the default, not something to be incentivized,” Marla B.K. Sammer, MD, with both Houston-based organizations, and colleagues explained.

Solving the “plain film” problem

In March 2019, the organization deployed its new software program which adds a subset of unread radiographs to all of its five modality- and location-based worklists.

A filter automatically selects where to send exams based on many variables, including priority, exam code, demographics, location, and more. Further rules dictate how exams are assigned, which can be round-robin, least busy RVU and fair distribution (based on RVUs and other exams read by radiologists).

At the outset, about 20% of daytime radiographs were auto-assigned, with each worklist receiving 20 exams at most. Two distribution windows were set: one in the morning and another one hour before the end of the day, with 10 minute breaks to allow for triaging.

The program proved a success and was expanded in January 2020 to include subspecialty exams, including inpatient and urgent imaging. 

Keys to success

Evaluating the software’s success, Sammer et al. reported the mean and median radiograph turnaround times dropped from 43 to 41 minutes and 25 to 23 minutes, respectively.

Additionally, complaints about the plain film problem stopped showing up in emails to the division chief, messages posted to the department and during monthly meetings. The authors regarded this as primary proof of their achievement.

Sammer and co-authors also identified a few keys to their project.

Semi-automating the process helped ensure only rads with the capacity to read extra exams did so. Consistency in distribution hours and exam assignments and transparency were also crucial, they explained.

And the providers made sure to emphasize fairness, autonomy, consistency and transparency to mitigate known factors of radiologist burnout, Sammer and colleagues concluded.

Read more about the project in the Journal of the American College of Radiology here.