The Incidence and Risk Factors for Emergency Department Imaging in Acute Renal Colic

Introduction

Controversy exists over the best choice of imaging for renal colic. Noncontrast CT provides the greatest accuracy, but is associated with radiation exposure, while ultrasound (U/S) provides a reasonably accurate, but less sensitive, alternative without the risk of radiation. ^{1 –4}

Current urologic guidelines are equivocal on the use of U/S and CT ^5,6 and the American College of Emergency Medicine's “Choosing Wisely” recommends avoiding CT in young healthy emergency department (ED) patients (age <50) with a known history of kidney stones who are presenting with symptoms consistent with uncomplicated renal colic, ⁷ but recommendations for de novo renal colic patients are absent.

Considering the increasing incidence of stone disease, ^8,9 rising concerns around radiation exposure from diagnostic imaging, and paucity of population-based data on the utilization of CT scans for the diagnosis of renal colic, it is important to evaluate current trends. This information will be important for informing clinical practice guidelines and targeting knowledge translation efforts to influence practice patterns. Accordingly, the objective of our study was to examine the incidence of use of CT scans and U/S in renal colic ED patients and measure the burden of repeat imaging within 30 days, as well as to assess the influence of patient and hospital factors on diagnostic imaging modality.

Subject/Patients and Methods

Results

During the study time frame, there were 60,269,378 ED visits in the province of Ontario, of which 585,356 (<1%) had a primary diagnosis of renal colic. After applying the study exclusion criteria and the 180-day washout period to identify index events, there were 397,491 renal colic ED visits in our final cohort occurring in 342,739 patients (Table 1).

The baseline characteristics of the study cohort including patient demographics and institutional characteristics are reported in Table 2. The median age was 49 (37–60), 59.4% were male and 16.8% had a history of kidney stones.

Imaging was performed in 268,055 (67.4%) renal colic events with an ED visit, with CT being performed for 68.2% of events, U/S for 26.7% of events, and both modalities on the same day for 5.2% of events.

Within 30 days of initial imaging, 21.2% (n = 56,780) of patients underwent repeat imaging at a median of 10 days (4–19) from initial imaging. U/S (58.9%, n = 33,421) was used more commonly for repeat imaging, compared with CT scan (39.8%, n = 22,597), although with a greater median time to imaging [12 (5–20) vs 7 (2–17) days, respectively] (Table 3). Overall, repeat imaging was more common in those who underwent an initial U/S (28%) compared with those who had an initial CT scan (18.5%). When stratified by initial imaging modality, those who had an initial U/S were more likely to have a CT (62.8%) than a repeat U/S (35.5%) within 30 days (Table 4). Conversely, those who initially had a CT were much more likely to have U/S (71.7%), when repeat imaging was performed (Table 4).

A total of 58,448 (22%) patients who underwent imaging at their initial ED visit avoided a CT scan during presentation and through 30 days.

Risk factor analysis demonstrated a higher risk ratio for undergoing initial imaging (either U/S or CT) at presentation to the ED with increasing age, male sex, later year of cohort entry, history for DM or IBD, and presentation to a higher volume ED. Conversely, those with a history for kidney stones, higher comorbidity (aggregated diagnosis group [ADG]) score, rural residence, and those presenting to an academic hospital had a reduced risk ratio of undergoing imaging at ED presentation (Table 5).

Risk factor analysis also demonstrated that a higher risk ratio for undergoing a CT scan was associated with increasing age, male sex, later year of cohort entry, history for DM or IBD, and presentation to a hospital with a larger number of beds or a higher volume of ED visits. Alternatively, those with a history for kidney stones, a higher comorbidity (ADG) score, rural residence, and those presenting to an academic hospital had a reduced risk ratio of undergoing a CT scan (Table 6).

Interestingly, the use of CT in the initial ED visit and compared with 2010 increased over time, with the risk ratio for having a CT scan rising annually to its peak (1.32, 95% CI 1.29–1.34) in 2020, the final year of our study (Table 6).

Lastly, risk factor analysis demonstrated that a higher risk ratio for undergoing U/S was associated with female sex, younger age, later year of cohort entry, history for kidney stones, and presentation to an academic hospital, a hospital with ≥800 beds, or a hospital with higher volume of ED visits (Table 6).

Discussion

Our population-based study demonstrated that 67% of patients presenting with renal colic to the ED underwent cross-sectional imaging, with CT scan ‘being’ the most used modality (73%) followed by U/S (27%). Repeat imaging was performed in 21% of events and those undergoing an initial U/S had an increased likelihood of repeat imaging, which was more commonly a CT scan. Our study represents one of the largest population-based cohort studies examining imaging trends for renal colic and the first to specifically examine individual- and institution-level risk factors for CT scan and U/S.

CT scan utilization for suspected renal colic has risen dramatically in the United States, ^18,19 raising fears over radiation exposure, especially considering that patients may receive more than one CT scan in the course of their acute disease process. ²⁰ These concerns over radiation coupled with increasing awareness of costs have led to the controversy over the most appropriate diagnostic imaging modality for ED renal colic patients. ²¹ Further fueling the controversy are recent studies showing the accuracy of U/S with Doppler for diagnosing ureteral stones ^21,22 and a randomized study showing the lack of serious adverse outcomes and reduced radiation exposure when U/S is used first, compared with CT. ² Also complicating matters is that CT scans may better identify nonrenal pathology, do not require a specialized U/S technician, and they provide important prognostic information (i.e., accurate stone size, stone density, and skin-to-stone distance) not provided by U/S.

As a result, the American Urological Association guidelines on the surgical management of stones comment that “the use of U/S alone to direct SWL or URS treatment planning should be discouraged as U/S is inherently inaccurate in determination of stone size, and it provides no information on stone density.” ⁶

Despite the controversy over the best imaging modality for renal colic and the growing concerns over radiation exposure, our study confirms that CT remains the preferred modality. In fact, the use of CT increased over time with the risk ratio for having a CT scan rising annually to its peak in 2020, the last year of our study. In addition, those who did undergo U/S first on presentation to the ED had a greater likelihood of repeat imaging, in particular a higher likelihood of having a CT (63%) as opposed to repeat U/S (36%) within 30 days.

Our study demonstrated that undergoing a CT scan was associated with increasing age, male sex, later year of cohort entry, history for DM or IBD, and presentation to a hospital with a larger number of beds or higher volume of ED visits. This highlights the patient- and institution-level factors that need to be targeted with prevention strategies to reduce the utilization of CT scans, when possible, for cost reduction and to minimize patient exposure to ionizing radiation.

As might be expected, younger patients and females were less likely to get a CT scan, as one assumes ED physicians and urologists were cognizant of minimizing radiation exposure. Similarly, those with a history of kidney stones were less likely to get imaging, as it stands to reason that the diagnosis was more likely to be based on clinical grounds. Those living in rural communities were less likely to get imaging presumably because of more limited access to imaging, especially after hours. Lastly, patients presenting to academic hospitals were more likely to undergo U/S than a CT scan. This could be a result of better dissemination of the results of recent studies showing the safety and utility of U/S for evaluating renal colic ^2,22 in academic centers. Alternatively, there might be greater access to U/S in academic hospitals because of the higher complexity of care being provided, resulting in highly trained ultrasonographers being more readily available.

Our study has several limitations. First, given that we relied on administrative data for the determination of renal colic events, there is a risk for misclassification (i.e., classifying an event as renal colic when in fact it was a different diagnosis). Second, given it is an observational study, we can only hypothesize as to the underlying reasons for some of the trends and associations observed; in addition, there is the potential that unmeasured variables could have also contributed to the increased risk for imaging and type of imaging modality chosen in our study population. Lastly, the use of point-of-care U/S is not captured in our data set and might have impacted the use of formal imaging in those who did not undergo CT or U/S.

It will be important for future population-based studies to examine the impact of imaging modality at presentation for renal colic on patient outcomes.

Conclusions

Two-thirds of patients presenting with renal colic in Ontario underwent imaging, and despite growing concerns about radiation, CT was the most common modality utilized. Patients undergoing CT as their initial imaging did have a lower likelihood of repeat imaging within 30 days. The utilization of CT is increasing over time and was associated with increasing age, male sex, and presentation to a nonacademic hospital with a larger number of beds, and a higher volume of ED visits.