A longtime leader in healthcare improvement, we’re developing new ways to revolutionize the industry.
Key Takeaways:
PINC AI™ data shows that up to 10 percent of hospital inpatients who should have received an advanced image as part of their recommended care did not due to a nationwide shortage of contrast media.
Contrast media went into shortage[i] in mid-April after COVID-19 lock downs in China halted the Shanghai-based production center responsible for nearly 80 percent of product for one of the two primary suppliers serving U.S providers[ii]. Premier fill rate data shows that the shutdown triggered an almost immediate dip in contrast media fill rates that persists today.
Premier leverages manufacturer data, member reporting and fill rate trends as mechanisms to help determine the health of the supply chain. For drugs, Premier considers a healthy fill rate to be above 90%, and anything that falls below 80 percent is an early indication that demand is outpacing supply and that shortages may be imminent.
During the shortage, providers were forced to conserve supplies, which often meant limiting imaging services with contrast media to medical emergencies only.
However historically, less has been known about exactly how many procedures are deferred due to a particular shortage, or the effects on patient care.
To determine the impact, PINC AI™ data scientists examined five clinical conditions typically reliant on advanced imaging with contrast media (COVID patients, mothers and babies, hip/knee and spinal surgical patients). The data showed deep drop offs in imaging utilization starting in April and continuing through the summer, when the contrast media shortage was most acute.
Looking just at computerized tomography (CT) scans, for instance, data show that as many of 10 percent of patients across clinical focus areas who were expected to receive an advanced image did not. Nationwide, the average number of patients who did not receive imaging services that would be typically expected hovered around 5 percent during April and May, and around 6 percent in July before finally recovering at the end of this summer.
Contrast media is commonly used to enhance the quality of advanced images such as CT scans and magnetic resonance imaging (MRI). More advanced than a typical X-ray, these images can depict bone as well as blood vessels, organs and muscle. Producing a highly detailed image, these technologies are often used to diagnose common conditions, detect tumors or other abnormalities and help physicians prepare for surgeries. An estimated 50 million imaging exams[iii] are conducted with contrast media each year.
Premier is committed to helping protect members from shortages like contrast media, and providing access to the vital products they need to ensure continuity of care.
For more:
Data
De-identified patient level data from the PINC AI™ Healthcare Database was utilized for this analysis. The dataset was limited to inpatient stays related to one of five clinical conditions, as defined by Centers for Medicare & Medicaid Services: COVID-19, mothers, babies, hip and knee arthroplasty and spinal surgery. Imaging modalities were the most prominent contrast-eligible imaging techniques across the evaluated clinical cohorts. The response variable was defined as the binary occurrence of contrast medium in the presence of a contrast-eligible image. Extracted data elements included a set of patient clinical and demographic characteristics comprised of length of stay, age, gender, race, point of origin, discharge status, MS-DRG, cost type and payor type. For evaluation purposes, CMS-defined geographic region was also extracted. The data was separated into baseline and evaluation datasets consisting of 190,854 and 523,044 inpatient stays respectively, across 801 distinct facilities. Baseline data was limited to discharges occurring in 2019, while the evaluation dataset was limited to discharges occurring between January 2020 through August 2022. Given the recency of the COVID-19 cohort, the baseline period was narrowed to discharges occurring between January through December 2020, while the evaluation period was narrowed to discharges from January 2021 to August 2022.
Methods
The baseline dataset was used to establish expected contrast utilization. Expected utilization, defined as the probability of a contrast-based imaging given the administration of a contrast-eligible image, is computed at the patient level using multiple logistic regression, thereby controlling for patient-level clinical and demographic characteristics. Models were stratified by clinical condition and geographic region for both imaging techniques. Observed and expected utilization was summed to the facility and month level and expressed as a ratio of total observed to expected utilization of contrast media. A Welch’s two-tailed t-test evaluated significant deviations between observed and expected utilization.
[i] https://www.aha.org/advisory/2022-05-12-shortage-contrast-media-ct-imaging-affecting-hospitals-and-health-systems.
[ii] https://www.reuters.com/business/healthcare-pharmaceuticals/ge-unit-boosts-medical-dye-output-china-covid-lockdown-cuts-supplies-2022-05-10.
[iii] https://www.nytimes.com/2022/05/26/health/dye-contrast-scan-shortage.html.