Radiology Volumes Are Different From What Finance Is Reporting. Why Does That Keep Happening?

Many of us who have worked in hospitals have been there; you are in a meeting with Finance, and their handouts have what is supposed to be Radiology volumes, but they are different from what Radiology’s internal reports show. So, whose numbers are right?

Radiology is one of those areas where you can have different volumes for the same thing, and they can all be correct. I hope this article can explain what perhaps is being counted and help prepare you with the right questions to ask the next time you are presented with inconsistent data.

Consider the following:

One patient can have multiple outpatient exams over time. In this scenario, it is important to distinguish among counting unique patients, unique exams, and unique cases.

One patient can have multiple exams in one session. Although it is a single time slot, there may be multiple exams in the Radiology Information System (RIS) associated with that single patient. Again, there is a need to distinguish between counting patients and counting exams.

One exam can produce multiple CPT codes. Although exam codes are usually one-to-one with billing codes, there may be occasions where an exam can produce multiple billing codes. Again, what is being reported?

In a single case, you can have multiple exams from different modalities. A patient may have an x-ray and an MRI, or an Ultrasound and a CT. It is important to find out how the data are being reported. Is only one of them being counted, or is the case being duplicated and counted once in each column?

Now, just to add a little more confusion, beware that there are different terms used to describe the same things. Here are some:

  • A Case is sometimes called an Encounter. In the financial systems, charges and revenues, and direct and indirect costs are calculated and reported at the Encounter level. To avoid double counting revenues and costs, Finance may assign a case with multiple modality exams to only one modality based on some hierarchy (e.g., MRI trumps x-ray).
  • An Exam is sometimes called an Accession. In RIS each exam has a unique accession number. As a side note, for the purposes of “big data” analysis (my specialty!) accessions are extremely important when it comes to process/operations analyses since timestamps, exam codes, along with demographic information can be found at that level.
  • A CPT Code is sometimes called billing code, fee code, charge code, or simply “procedure”. In my experience, the latter is what I have seen used most often.

Now, in case you are starting to feel confident that you understand everything above, allow me to introduce you to Nuclear Medicine!

If there is an area where the discrepancy among reported volumes is greatest, it is most likely Nuclear Medicine. The reason is, in one word, “injections”.

In Nuclear Medicine, a patient is injected with a radio-pharmaceutical and then scanned. Sometimes, as in the case of a stress test, that process is done twice.

Injections take technologist time and are often scheduled in the RIS on a particular resource (i.e., room). Nuclear Medicine sections often choose to count injections as procedures, which helps explain why volume discrepancies can get big. In the case of a stress test, one case could be counted four times (2 injections + 2 scans) in Radiology’s internal reports.

How these procedures are counted also needs to be understood. Some Nuclear Medicine sections create a no-charge statistical fee codes for the purposes of counting injections, while other sections base their count on the number of times the injection room is utilized. Neither is wrong, as long as the counting process is consistent.

Finally, I will reiterate what I mentioned previously: all the volumes reported can be correct, they are simply counting different things.

Hopefully, this article helps explain the discrepancies, and helps you ask the right questions when presented with volumes (or asked to produce volume reports).

Reality Isn’t Always Pretty: “See” Your Opportunities

In a previous blog, I showed how you can look at a year’s worth of data to look at the average utilization of an MRI machine. The average number of minutes of “patients-on-table” was 550 /day, and the average number of patients was around 14. In this blog, I take it a step further and look at what a typical 550 minute /day looks like.

I selected a day with 557 minutes of patients-on-table with 14 cases, and plotted each case, to the minute.

A screenshot of a cell phone

Description automatically generated

The visual itself is revealing. In theory, each yellow block is an opportunity. Although I argued previously that you cannot expect 100% utilization, the amount of yellow in this graphic should raise some questions.

Managers should sit with frontline supervisors and staff to better understand the causes for these downtimes. This graphic can serve to facilitate the discussion.

Whatever the causes are, including “no-shows”, failed safety questionnaires, or claustrophobia, they should be tracked and addressed.

The process of assembling the graphic also revealed that there may be an opportunity to standardize data entry among the technologists. There were a couple of examples of “overlap” between patients. One may be due to continuing to process the images after the patient left and the next patient walked in, but in one case, both patients were begun the same time!

My advice to all radiology administrators is don’t wait for a process improvement project to begin thinking about your data’s integrity–start standardizing the process of entering timestamps now, and perform frequent audits.

“Big data” analysis is extremely helpful, but it requires a certain level of trust in the data. Even if it is difficult to capture the exact time of when a patient goes into or out of a room, it is still possible to make assumptions that reasonably estimate those times, but that requires consistency and adherence to the process by all staff.

Although experienced managers often have a good grasp of how their areas are performing, sometimes it takes a carefully crafted image to shine a light on the reality of a process.

The same analytics process can be used for other modalities. Also, beyond Radiology, other entities with available data (e.g., physician practices) can benefit from such an analysis.

If you would like your utilization data analyzed in a similar fashion, please reach out to me via my email at ELYanalytics@Outlook.com

Find the Nuggets in your River: MRI Utilization.

If you are an administrator in a hospital or a diagnostic imaging center, then chances are you are already  looking at your MRI utilization. MRI exams have a relatively large contribution margin and being able to “squeeze” an extra exam here or there can have a huge financial long-term impact.

You cannot aim for 100 % utilization—that would be unachievable and certainly inadvisable. There will always be claustrophobic patients, late patients, and “no shows”, among other legitimate reasons for lower than 100% utilization. Also, a schedule should have “gaps” to accommodate emergency add-on cases and allow for exam overruns.

A good starting utilization goal is 85%. In this example (Table 1), I am using a year’s worth of data to look at room utilization, defined as number of minutes patients are on the table divided by the number of minutes a room is staffed. This particular machine had 65% utilization. The addition of one 45-minute exam every weekday would bring the utilization to 71% and generate between $75,000 and $200,000 annually (assumed a contribution margin range between $300 to $800 /case). This is only from one machine reaching 71% utilization only.

Table 1

The next challenge is where to find those opportunities. One potential area is to look at your first case of the day—is it starting on time? A quick visual (Graph 1) of the average starting time for a whole year reveals that the “begin” time for the first case is all over the place. If the first exam is scheduled for 7 AM, then there may be opportunities for improvement.

Graph 1

The example presented here uses one year’s-worth of data. It is high level and meant to show potential opportunities. A lot can happen in one year. A second iteration of the analysis should probably look at the latest quarter.

As I have recommended in a previous blog, it is important to sit with the client and those who know the operation best, before coming up with any recommendations. Only they can tell you if the schedule has changed within the year, and exactly at what moment a technologist “begins” an exam (i.e., is it when the patient enters the room or when the patient is on the table?).

This type of detailed analysis is difficult to achieve using canned reports. Those may be quick and easy, but they do not give you the flexibility you need when looking for improvement opportunities.

Raw data, on the other hand, is like a raging river—it can be overwhelming, but it can hide some golden opportunities in its depths. In the hands of someone who can navigate it, those opportunities can be surfaced.