STONE Score for Uncomplicated Ureteral Stone

Why did you develop the STONE Score? Was there a particular clinical experience or patient encounter that inspired you to create this tool for clinicians?

We found that over the years, and in the United States in particular, it had become a knee-jerk reaction to perform a CT scan on anyone with suspected kidney stone. However, it hasn’t been shown to alter management that much, and sometimes it’s fairly obvious that patients have a kidney stone. A CT scan has ionizing radiation that can cause cancer later, it’s expensive and time-consuming, and if it’s not needed we wanted to find a way that we could clinically predict who would have a kidney stone to try to help avoid unnecessary imaging.

What pearls, pitfalls and/or tips do you have for users of the STONE Score? Do you know of cases where it has been applied, interpreted, or used inappropriately?

There are some caveats that are listed at the top of the score page, which are to be careful about not using it in anyone who has a potentially complicated kidney stone—anyone who has evidence of an infection, or pre-existing renal disease. In those cases, you would generally want to get imaging because you wouldn’t want to miss a stone. But for uncomplicated kidney stone, about 80% of those will pass on their own, so imaging doesn’t necessarily help you much in those cases. So that’s the main caveat: make sure you’re not concerned about an infected stone or other pre-existing renal disease.

What recommendations do you have for doctors once they have applied the STONE Score? Are there any adjustments or updates you would make to the score based on new data or practice changes?

A reason to get a CT scan in a patient is really two-fold when you suspect a kidney stone. One is when you think maybe it’s a kidney stone, but there might be something else that’s bad. And we found that having a high STONE Score inversely correlates with the likelihood of having a significant alternate diagnosis. So it can help you to know that they’re likely to have a stone and unlikely to have something else. The other reason you might need a CT scan is that the patient might need an intervention. There’s a minority of those, maybe 20%, that would require an intervention. But it could expedite appropriate intervention if you have imaging. I think it’s helpful to be used in a shared decision making model, where you could tell the patient they are 90% likely to have a kidney stone, and 80% of those are likely to pass on their own, without anything other than pain control. However there is a small proportion that do need an intervention, and we could do a CT scan here and now but we don’t have to.

The other thing I would say is that it can help you to choose the right kind of CT scan. Kidney stones in particular are very amenable to reduced radiation dose CT, and if you are going to do a CT scan, it helps to pick those patients appropriately.

How do you use the STONE Score in your own clinical practice? Can you give an example of a scenario in which you use it? You mention shared decision making; how many patients are actually okay with not getting a CT scan after you have that conversation with them?

It’s a good question, and I don’t have an exact percentage for you, but it’s certainly a substantial number of patients that, given the choice (and especially if their pain is controlled in the emergency department) are perfectly happy to go home without getting a CT scan. Particularly when they understand that there’s some radiation involved, as well as the time and expense, even if it’s covered by insurance. I’ve been generally pleasantly surprised by that. There are certainly some people who just want to know and just want to get a CT scan, and I’m okay with that, but I think it’s a particularly good situation as there are different ways to approach it. Frankly it’s not dangerous for an uncomplicated stone, because if it doesn’t pass, it tends to cause ongoing symptoms, and they can follow up and get a CT scan at a later time. But it would be a minority of patients who would actually require intervention.

Any other research in the pipeline that you’re particularly excited about?

We’ve tried to put this in place to implement either avoidance for CT imaging entirely or appropriate use of reduced-dose CT. We have recently completed an expert consensus on imaging in these patients that involves emergency medicine, radiology, and urology, and are in the process of publishing that work. We are also hoping to implement elements of the STONE Score into clinical decision support (CDS) that can be part of the electronic health record (EHR) to help guide appropriate imaging on a consistent basis.

Any other thoughts on managing kidney stones?

I think it’s worth emphasizing is that there seems to be this mantra or cultural norm in the United States that all first-time stones need a CT. But our research and my own clinical experience doesn’t necessarily support that. If you have someone who comes in at 3 am and clearly has a kidney stone by the STONE Score and clinical experience, and you can get their pain relieved, they don’t necessarily require a CT scan. And in fact, the history of kidney stone is not part of the STONE Score. I have some theories on why that didn’t come out in either the derivation or validation phase, but even if you have a prior stone or not, the STONE Score still performs well. I want to emphasize that in my opinion, not all first-time stone requires a CT. And that’s certainly a practice in other countries - they don’t necessarily do a CT just because it’s a first-time stone.

We also talked about the incorporation of ultrasound, the STONE PLUS (Point-of-care Limited Ultrasound), and in particular, for patients who are in the moderate STONE Score zone, that’s where ultrasound is most helpful. If their score is high, they’re likely to have a stone whether or not they have hydronephrosis, because not all stones have hydro. But if you’re not sure, ultrasound performed as needed can be very helpful.

What’s your theory on why history of stone doesn’t show up as being significant in the data?

I think it was patients who may have been looking for narcotic pain medication and felt that if they said they had a kidney stone, they would get it, because they’d had a prior kidney stone. But in the derivation and validation phase, we actually didn’t see that. It didn’t necessarily pan out that a CT would show a stone. So it was not one of the variables that reached significance.

What prompted you to perform a secondary analysis in 2024 of the data used in the initial derivation and validation of the STONE score?

In response to the call to action to examine clinical calculators that include race adjustments that may create or perpetuate health care inequities, we revised the original STONE score for predicting a symptomatic kidney stone by replacing “Origin or race” (“O”) with “Obvious or gross hematuria” as a predictor. The updated STONE score showed high discrimination similar to that of the original STONE score.

Christopher L. Moore, MD, is a professor of emergency medicine at Yale University. He is also Chief of the Emergency Ultrasound Section in the Department of Emergency Medicine. Dr. Moore’s research focuses primarily on diagnostic testing and diagnostic imaging in the emergency setting.