Numerous anecdotal articles and predictive models make claims that the COVID-19 pandemic claim has plummeted medical imaging volumes anywhere from 50%-70%. Confirming these suspicions, however, has been a challenge, due to the minimal number of sources with actual real-world data.

Researchers at University Hospitals Cleveland Medical Center/Case Western Reserve in Ohio set out to do just this, evaluating eight weeks of imaging within the COVID-19 healthcare crisis time frame.

"Imaging utilization during COVID-19 has been highly variable and will continue to be through the rest of 2020," Dr. Keval D. Parikh, a diagnostic radiology resident PGY-4/R-3 at University Hospitals Cleveland Medical Center/Case Western Reserve University, told HCB News. "Based on our patient population demographics and state policies, our experience with COVID-19 has thus far been different from the experiences of our counterparts in New York, Los Angeles, Texas and other parts of the country. Positive cases are continuing to peak in many parts of the country, including in Ohio, and different states are responding in different ways."

Parikh and his colleagues prepped for their assessment by creating an organizational structure with an Incident Command Center to ensure standardization of policies and procedures for addressing the pandemic across a health system of 18 hospitals. Its use enabled the researchers to assess system testing options and supply chain control of PPE, as well as repurpose staff that were not busy clinically to other areas of need.

A crisis time frame was set up and included imaging conducted between March 8 and April 30, with the first week of the date range and the 10 weeks prior to it used to determine the volume expected under ordinary conditions. A rolling seven-day total value was used to track and compare volume. Data was collected from a business intelligence and analytics platform that aggregates data primarily from the 3M M*Modal Scout System, the RIS and PACS system used to construct performance statistics by analyzing all radiologist-generated reports.

The researchers broke down total imaging utilization by patient setting (outpatient, ED, inpatient); imaging modality (X-ray, CT, mammography, MR, nuclear medicine/PET, US); and location. The locations in questions were the Central main campus of the healthcare system, as well as its smaller East and West Hospitals, all of which had the highest volumes. Revenue information was also collected from the hospital billing system.

A seven-day total value between April 7 and April 13 confirmed a drop of 55% in total imaging volume. Outpatient settings took the hardest hit at 68%, followed by EDs at 48%, and inpatient settings at 31%. Mammograms and nuclear medicine scans declined the most of all modalities, by 93% and 61%, respectively. Following them were declines in ultrasound (58%); MR (56%); plain films (53%); and CT (47%). "Outpatient studies tend to fall into this category of lower acuity studies that have decreased likelihood of altering clinical outcomes compared to the higher acuity ED and inpatient studies," said Parikh.

The main campus hospital experienced a 42% drop in volume but had less of a loss than the East Hospital at 60% and the West Hospital at 54%. Together, all three recorded a 48% total loss in imaging volume.

"Knowing that there may be a shift in volumes, decreasing in some areas and increasing in others, we were able to train and repurpose radiology technologists to reach out to patients with outstanding orders to schedule over 10,000 patients for future orders," said Parikh. "There were also staff repurposed to be screeners at doors into hospitals and ambulatory sites, to screen staff and patients for symptoms and temperatures. We also were able to shift technologists from free standing Emergency Rooms that were temporarily shut down due to volumes and use them at other sites where there was more need. This prevented furloughs and lost jobs."

The study also calculated revenue losses, with the Central hospital hitting its lowest point on April 14 with a 49% decrease from its technical component compared to normal practice prior to the pandemic. This coincided with the imaging volume decrease seen around April 13. The numbers are expected to help hospitals forecast demand in various patient settings, should stay-at-home or lockdown orders be declared in Ohio, as well as repurpose staff and change clinical policies before changes in patient volume take place.

Parikh says he and his colleagues have gained a number of insights for addressing volume loss and helping hospitals to recover as they open up once more. Among these tactics are extending patient scheduling calls to evenings and weekends to reschedule postponed exams; partnering with software companies to create virtual waiting rooms and maintain social distancing practices; universal masking mandates; and reducing or eliminating faculty incentive payments including overtime and internal moonlighting, while maintaining base compensation.

"We hope that our numbers and policies thus far will help health systems in different parts of the country prepare should their regional COVID-19 incidence rates and policies look similar to ours in this particular study over the remaining course of this pandemic," he said.

The researchers recommend that large and small institutions continue publicizing their actual data to show how the pandemic has affected different types of radiology practices, and have created a handbook of their own on how to keep patients and staff safe.

They are currently planning a similar study to evaluate the impact of COVID-19 on emergency settings, and hope to create another over the next three-to-six months on the recovery of volumes and return of patients for imaging studies in light of state policy changes expected throughout the fall and winter.

The findings were published in Academic Radiology.