No Left Turns in Radiology
What do UPS trucks have to do with data and radiology?
Have you ever heard the urban legend that UPS drivers don’t make left turns? As it turns out, it’s not actually an urban legend, a fact Woojin Kim, MD, assistant professor of radiology at the
Hospital of the University of Pennsylvania, used to illustrate the incredible power big data might hold for radiology.
During his presentation, as part of the ACR 2015 session titled “New Payment Models Meet Big Data,” Kim noted that many years ago UPS began looking for a way to streamline its business. Around the same time, the company noticed that an inordinate number of UPS trucks were crashing while turning against traffic (making left-hand turns). Company leaders knew they had to make a change.
That’s when company engineers hit on a novel idea: by analyzing delivery route data, they formulated “route-optimization plans to increase efficiency, reduce fuel consumption and get drivers back to their centers earlier,” all of which involved mandating that their drivers no longer make left-hand turns. The engineers realized that saving one minute per driver per day along delivery routes would translate into a savings of $14.5 million per year. Meanwhile, saving one mile per driver per day would, over the course of a single year, save the company an additional $30 million. What began as a safety analysis eventually yielded a method for saving the company significant amounts of money.
Kim’s point with respect to radiology is that when radiologists mine meaningful data — not just data for data’s sake, but data that can translate into value — they can improve patient care and save health care systems money. Using the example of Syed F. Zaidi, MD, president and CEO of Radiology Associates of Canton, Inc. (RAC), in Canton, Ohio, Kim showed how one radiologist can make a difference to both quality patient care and the bottom line. By examining patient data, Zaidi and his associates at RAC have managed to reduce inpatient stays by three days by streamlining the care they provide. Since the typical inpatient stay amounts to $2,000, this translates into $6,000 in savings per patient per stay. Multiplied over many patient stays, analyzing big data has helped Zaidi save his hospital a significant amount of money.
Building upon Kim’s theme of big data’s potentially transformative effect on radiology, Richard Duszak Jr., MD, FACR, who also presented during the session, went on to herald the coming era of payment system change. Although the new generation of payment models is still being formulated, big data will play a role in how physicians — including radiologists — are compensated in the near future, he noted. In an environment where payment models are evolving from a fee-for-service format to one focused on value, radiologists will thrive only if they help determine their place within the new system.
According to Duszak, big data is the key element of a project undertaken by the Harvey L. Neiman Health Policy Institute (HPI), of which he is the chief medical officer and senior research fellow. The HPI is working to determine how radiologists can lobby effectively for a seat at the compensation table. One of the most likely forms a “fee-for-value” model will take, argues Duszak, will involve bundled payments. However, since there are over 700 Diagnosis Related Group (DRG) categories, upon first glance it looks like a daunting challenge to negotiate compensation based on over 700 different varieties of bundled payments.
Collecting big data, however, can help in this effort. The HPI has launched the “Inpatient Bundled Payment Model Initiative” to develop an evidence-based prioritized strategic framework for identifying encounters in which initial bundled payment modeling might be most impactful for radiology. Mining data from CMS, Duszak and Danny Hughes, PhD, senior director, health policy research and senior research fellow at the HPI, have secured all fee-for-service claims data for a 5 percent random sample of Medicare beneficiaries from 2009-2011. The institute hopes to use this information to carve out a place for radiologists in this new, prospective bundled payment model. In addition, the researchers have used Part A (hospital) claims data along with using separate Part B (physician services) claims data in their research.
What Duszak and Hughes found when they ran the numbers may move the needle on fair radiologist compensation within a future bundled payment system. The researchers discovered that only five percent of DRGs contained 50 percent of all imaging costs, even though imaging was present in 99.6 percent of uniquely identifiable DRGs. Imaging volumes followed a similar pattern. These kinds of analytics allow radiologists to quickly determine which DRGs may be more appropriate for their practices to consider for bundled payment models.
In the same way that UPS made a seemingly counter-intuitive rule against its drivers making left-hand turns, radiologists must convince insurers to compensate them as integral members of the health care team. This task may seem counter-intuitive, given that health care reform passed in recent years does not seem geared toward radiologists. One of the most compelling ways radiologists will make a convincing case for fair compensation, however, is by presenting evidence in the form of value-based metrics, many of which will be derived by mining big data.
By Chris Hobson, Imaging 3.0 Content Manager