This summer, I started my residency in urology. My job for the first month was to see patients in the emergency department with urologic problems. Nephrolithiasis was one of the most frequently encountered diagnoses, and I soon learned the “drill” for working up kidney stones– check a white blood cell count, send off a urinalysis and urine culture, and, almost reflexively, order a CT scan.
Was the CT scan a necessary first imaging test? Or would many of those patients have been just as well off with an ultrasound instead, sparing them radiation while still leading us to the same diagnosis and treatment plan?
This week’s NEJM reports findings from a multi-center comparative effectiveness trial that examined how initial imaging affects outcomes for patients with suspected nephrolithiasis. The study randomly assigned over twenty seven hundred patients who’d presented to the emergency department with symptoms suggestive of kidney stones to one of three initial imaging modalities– point-of-care ultrasonography, performed by a provider in the emergency department; ultrasonography in the radiology department, performed by a radiologist; or an abdominal CT scan. Providers could order additional imaging subsequently, as they deemed appropriate. The main outcomes were the rate of high-risk diagnoses with complications that could be related to missed or delayed diagnoses (appendicitis with rupture, diverticulitis with abscess, bowel ischemia, pyelonephritis with urosepsis, and so on); cumulative radiation exposure over six months; and total cost.
There was no significant difference in the rate of high-risk diagnoses with complications across the three study arms (0.7% for point-of-care ultrasonography; 0.3% for radiology ultrasonography; and 0.2% for CT). There was also no difference across the study arms in the percentage of patients who experienced serious adverse events. But patients who underwent ultrasonography as the initial imaging modality were exposed to a lower cumulative dose of radiation than patients who underwent an initial CT scan (10.1mSv and 9.3 mSv for point-of-care and radiology ultrasonography, respectively, versus 17.2 mSv for CT scan; P<0.001).
“It should be emphasized…that ultrasonography when used alone is not very sensitive for detecting stones,” Dr. Gary Curhan of Brigham and Women’s Hospital in Boston writes in an accompanying editorial. He also points out that whether a patient had a known history of kidney stones may have influenced the interpretation of ultrasonographic findings. “It is possible that the characteristic shadowing or hydronephrosis would have been more likely to be reported in a patient with a history of stone disease, particularly if a recent imaging study had identified a stone. This latter possibility is supported by the study’s findings that among persons in the ultrasonography groups, those with a history of nephrolithiasis were less likely than those without such a history to undergo subsequent CT.” If a documented history of nephrolithiasis biased imaging interpretation, then the effectiveness of ultrasonography at making a correct diagnosis while “saving” patients from unnecessary radiation may be more limited than these study findings otherwise imply.
Time is another consideration. In this study, patients assigned to ultrasound by a radiologist ended up spending more time in the emergency department than patients in either of the other groups (5.1 hours for point-of-care ultrasonography; 6.4 hours for radiology ultrasonography; and 6.2 hours for CT scan; P<0.001). It is possible that the use of ultrasonography could result in delays in patient care. And while the point-of-care ultrasonography group spent less time in the emergency department, over 40% of those patients ended up getting a CT scan (versus 27% of patients who underwent radiology ultrasonography). In terms of cost, the average emergency room stay cost slightly less– $25 — for patients assigned to ultrasonography as compared to patients assigned to CT scan (the complete findings of the cost analysis were not reported in the paper).
After a diagnosis of stone disease has been made, management decisions may rely on information that can be obtained from CT but not from ultrasonography. If surgical management is being considered, a CT scan is valuable for localizing the stones and planning an operative approach. In addition, a CT scan may be more likely to identify additional stones that are not yet symptomatic, prompting a more aggressive regimen to prevent further stone formation.
As Curhan states, “Although we want to limit radiation exposure from all sources, the decision to use ultrasonography needs to be balanced against the additional information obtained by CT, which may influence subsequent clinical decisions.”
In your practice, do you routinely order CT scans for patients with suspected stone disease? For which patients do you first order an ultrasound? How will the findings of this study affect your approach to the diagnostic workup and management of nephrolithiasis?