An actionable metric for Radiology workflow that opens exam slots and reduces contrast waste

by Timothy P Szczykutowicz, Ph.D., DABR

You don’t have to be a domain expert to understand that when a CT technologist repeats a scan, the exam will take longer. If the exam administered iodinated contrast agent, a repeat may also cost the institution more money in a wasted contrast dose. Such inefficiency in radiology can easily be directly related to time/money loss for the hospital and radiologists. And in some cases, repeats represent a patient safety/quality issue [1].

The concept of reject/repeat analysis in digital radiography (DR) and mammography (MG) has been known for many years and it is being implemented for regulatory, radiation protection, and quality improvement purposes. However, our lab’s recent work has highlighted that the repeat rates in CT significantly contribute to time loss in radiology [2] and that they are a source of excess radiation [1]. The main reasons for repeat CT imaging are protocol selection error, issue with contrast administration, poor protocol instructions, protocol settings error and patient motion or noncooperation [1, 3, 4]. In short, the repeat metric shines a light on staff training issues or poorly set-up scanner protocols.

The first company to commercialize our CT repeat rate metric is Qaelum.  Qaelum implemented our repeat rate metric in a new focussed-on-quality platform called FOQAL (FOQAL-CT repeat, Qaelum, Belgium). I’m excited to work with a dose monitoring company that is offering more than traditional dose monitoring tools. When I speak with most radiology administrators and radiologists, they have a strong desire to keep their patients protected from higher than needed radiation doses, but see dose monitoring usually through a compliance/regulatory lens. What I see administrators really wanting are ways to extend their personnel and equipment to scan more patients. This is especially true in the current COVID19 pandemic. 

The new Qaelum software solution is based on our lab’s algorithm that recognizes standard protocols and alerts the user when there are scans in these protocols that are unnecessarily repeated (Fig 1). Some earlier attempts by vendors in this space failed to understand combo orders and optional delayed phases. Qaelum’s solution is robust to such things because my lab spent years developing the repeat analysis algorithm to be robust to real clinical workflows. Furthermore, Qaelum’s team has a wealth of experience and wisdom developing and incorporating quality solutions into a wide array of hospital settings. Besides the generic idea of calculating the repeat rate, Qaelum’s FOQAL-CT repeat product analyzes repeated exams and assesses the financial impact on the department (Fig 2). Thus, it becomes a valuable assistant to improve the efficiency in a tangible way. 

 Fig 1. Repeat rates per repeat type and per ‘base pattern’ (i.e. identified standard protocol) are visualized in an intuitive way. Improvement plans begin from the protocols with highest repeat rates.

Fig 2. Repeated scans are a source of extra cost for the CT department, in terms of time loss or contrast waste. Quantifying the financial liability assists in a tangible way. Here, Qaelum has summed the extra time spent on median repeated exams versus non repeated exams of the same type to demonstrate how many extra exam slots could be opened up if the site was able to reduce repeat imaging. Their FOQAL-CT repeat also allows the user to enter a rough cost estimate for lost exam revenue, letting site administrators gauge how much lost revenue they have from repeated scans.

A lot of metrics I see just show a needle for things like patient volume or radiologist interpretation volume. They are helpful for a big picture view, but do not provide us with actionable advice. Qaelum’s FOQAL-CT repeat will allow one to easily see scanners, technologists, protocols, and even specific phases of a protocol that are causing trouble. I know, because we performed a multi-center study at UW Madison on CT pulmonary artery exams for embolism detection and demonstrated we could do this [3]. We could detect specific technologists at specific locations who were underperforming compared to other sites, for the same protocol!

The software also identifies protocols with clinically relevant repeats (e.g., optional delayed phases in the setting of a slow bleed in a trauma patient). This means the metric appreciates when a radiologist actively adjusts the exam to the patient situation by not calling these events out as false positives. These cases are either identified by the solution automatically, or if performed rarely, the user can teach them to the software through a user specific interface (Fig 3).

Fig 3. Not all repeated scans are superfluous. Clinically relevant repeats can be excluded from the repeat rate analysis in an easy way, with possibility to also select the type of repeat that needs to be excluded. This allows for rare and clinically needed delayed or optional phases to not raise false alarms. For commonly performance optional phases, like delayed phase in the setting of a slow bleed, the algorithm will learn those behaviors on its own without user input needed.

If you want to know more about our results and conclusions on this topic, don’t miss our presentation “Automated analysis of CT repeated scans in a large European hospital” at ECR 2022. You can also visit Qaelum’s booth at ECMP 2022 in Dublin or find out more about the product under this link https://qaelum.com/solutions/ct-repeat

References:

• Szczykutowicz T, Rose S. “First Look: Repeat Rates in CT.” Radiology Management, 2019, 41(5).
• Rose S, Viggiano B, Bour R, Bartels C, Szczykutowicz T. “A Multiinstitutional Study on Wasted CT Scans for Over 60,000 Patients.” American Journal of Roentgenology, 2020, 215:5, 1123-1129. DOI: 10.1016/j.jacr.2021.02.014.
• Rose S, Viggiano B, Bour R, Bartels C, Kanne J, Szczykutowicz T. “Applying a New CT Quality Metric in Radiology: How CT Pulmonary Angiography Repeat Rates Compare Across Institutions.” Journal of the American College of Radiology, 2021, 18(7). DOI: 10.1016/j.jacr.2021.02.014.
• Szczykutowicz T, Rose S. “Repeat/Reject imaging in CT: An unnecessary source of patient dose and a tax to radiology resources.” European Congress of Radiology, ECR 2020, Vienna, Austria. DOI: 10.26044/esi2020/ESI-02989.

Timothy P Szczykutowicz, Ph.D., DABR Associate Professor, University of Wisconsin Madison Director of Clinical Operations, CT Protocol Project

Dr. Szczykutowicz “Stick” is an Associate Professor of Radiology at the University of Wisconsin. Protocols developed by his team have been shipped to 3,500 sites around the globe. Dr. Szczykutowicz is the author of 40 papers, 2 book chapters, the book “The CT Handbook”, 4 patents, and is a consultant to 5 companies.