CT imaging history for patients presenting to the ED with renal colic--evidence from a multi-hospital database

Numerous investigators have raised concerns regarding the potential relationship between ionizing radiation dose from medical imaging and risk of human malignancy [1‐5]. Depending on the area imaged, a single CT scan is estimated to increase the lifetime risk of cancer by as much as 1 in 2000 [3]. Despite such risk, the use of CT has grown significantly over the past 20 years to evaluate for pulmonary embolism and a variety of intra-abdominal conditions. The average growth rate of CT utilization for chest pain complaints alone was 28.1% per year between 2001 and 2007 [6].

Several publications in the mid-1990s touted the use of CT to accurately identify the presence of nephrolithiasis and alternative conditions in ED patients with suspected renal colic. Within just a few years, unenhanced CT of the abdomen and pelvis supplanted intravenous pyelography (IVP) as the diagnostic study of choice for evaluating patients with suspected renal colic. [7] Similar to the observed trends for chest imaging, CT utilization for the evaluation of patients with flank pain grew exponentially [7, 8]. Westphalen et al. found that from 1996 to 2007 there was a 10-fold increase in CT scan use for patients with suspected kidney stone, but the authors found no associated change in the proportion of diagnosis of kidney stone, diagnosis of significant alternate diagnoses, or admission to the hospital [7]. A similar study conducted from 2000 to 2008 revealed a significant increase in computerized tomography use, a significant decrease in x-ray use, historically the preferred imaging study for this condition, while ultrasound use remained stable. They also found that despite the increased use of CT scan, the proportion of patients diagnosed with kidney stone did not increase [8].

The growing use of CT scan as the diagnostic modality of choice in renal colic patients is particularly concerning because of the recurrent nature of the condition. The average patient with renal colic has a 7% chance of recurrence per year and a 50% chance of a recurrence within 10 years [9]. A single center, retrospective chart review including 356 patient encounters over a 10-month period by Broder et al. found that 79% of patients presenting to emergency departments with recurrent symptoms suggesting renal colic underwent at least 2 CT scans [10]. Katz et al. estimated that a typical single detector unenhanced CT and single multi-detector unenhanced CT of the abdomen and pelvis exposes a patient to 6.5 mSv and 8.5 mSv respectively [5]. Unfortunately, within their single-center cohort of 4562 patients undergoing repeat CT scan for suspected renal colic, they found patients with three or more unenhanced CT scans of the abdomen and pelvis were exposed to ionizing radiation doses ranging from 20 to 154 mSv. Such exposure falls within a range that Brenner et al. found suggestive of significantly increased cancer risks (10–50 mSv for an acute exposure and 50–100 mSv for a protracted exposure) [4].

Prior studies related to CT scan use in renal colic have been mostly small, single center studies or have relied on national, administrative databases [5, 7‐15]. We are unaware of prospective emergency department-based studies, to date, that evaluated repeated patient exposure for renal colic imaging within a multi-center database. The objective of our study was to evaluate whether our patients had been exposed to multiple CTs for the evaluation of renal colic within a regional hospital system and to identify patient characteristics that were associated with CT use during the current ED visit.

Patient characteristics are shown in Table 1. Of the 213 patients included in the final study group, 59% were male and 67% were Hispanic. The mean age was 38 +/− 11 years. Sixty-two percent of patients reported a prior history of kidney stones. The most common presenting symptom was flank pain and hematuria was detected by urinalysis in 76% of patients (Table 2).

Table 3 reveals that 60% (95% CI = 53–66%) of patients received a CT scan at the ED current visit and 30% of patients underwent US. Only 4% of patients underwent both a CT and an US and 14% of patients had no imaging performed at their current visit. Hydronephrosis was present on 55% (95% CI = 46–63%) of CTs and 28% (95% CI = 19–40%) of ultrasounds performed during the current visit. A renal stone was identified on 90% (95% CI = 83–94%) of CTs and 38% (95% CI = 45–74%) of ultrasounds performed in those cases.

No significant differences were observed in the proportion of patients for the current ED visit who received CT with respect to: female vs. male (62% vs. 56%; p = 0.4), mean age (37+/− 9 years vs. 39+/− 11 years; p = 0.2), and Hispanic vs. non-Hispanic white (63% vs.63%; p = 0.96). Patients with a prior stone history were more likely than those with no history to receive an EDCV CT (88% vs. 16%; p < 0.001). Patients with a urinalysis positive for blood were no more likely to undergo CT than patients with no blood on urinalysis (61% vs. 60%; p = 0.4).

Table 4 shows patients’ prior CT histories. Out of the 213 patients, 118 (55%; 95% CI = 49–62%) had at least one prior CT, 46 (22%; 95% CI = 16–28%) had ≥3 prior CTs, and 29 patients (14%; 95% CI = 10–19%) had ≥10 prior CTs. Patients who did not receive a CT at the current visit had a significantly higher mean prior number of CTs than those who had current visit CT (5.1+/− 7.7 vs 2.2+/− 4.9; p < 0.001). Patients who received an US at their current visit also had a significantly higher mean number of prior CTs (4.9+/− 7.6 vs 2.8+/− 5.6; p < 0.001). Characteristics associated with positive prior CT history included: female gender (42% vs. 28%; p = 0.02), flank pain (40% vs. 13%; p < 0.001), and prior renal stone by history (48% vs. 10%; p < 0.001).

Patients with a reported prior stone were more likely to receive a renal ultrasound as the only imaging study during the EDCV (33% vs. 15%; p = 0.003). Patients with a history of one or more CT scans were also more likely to undergo only an ultrasound compared to those with no prior CT (35% vs. 22%; p = 0.046). For patients who received only an US at their current visit, there were no significant differences observed with respect to female vs. male (26% vs. 26%; p = 1.0), mean age (37+/− 9 vs. 38+/− 11), or Hispanic vs. non-Hispanic white (23% vs. 25%;p = 0.4).

Unenhanced CT is a rapid and highly accurate imaging modality with reported specificity and sensitivity between 91 and 100% for diagnosing nephrolithiasis in patients presenting with flank pain [12, 13]. In theory, physicians and their patients benefit from CT use as it may also identify alternative significant diagnoses. However, somewhat surprisingly, the exponential increase in the use of CT scan for evaluation of flank pain has not resulted in significant improvements such as finding alternate diagnoses, better diagnosis of kidney stone, changing rates of admissions to the hospital, and/or identifying those in need of immediate urologic intervention [7, 14].

Several recent studies propose that renal ultrasound should be used as the first or even only diagnostic study when evaluating renal colic. The advantages of ultrasound include a low cost and widespread availability [16]. Investigators have reported that utilizing ultrasound as the initial diagnostic study of choice for renal colic is associated with lower cumulative radiation exposure without significant differences in complications, serious adverse events, pain scores, return emergency department visits, or hospitalizations [15]. In addition, studies comparing CT and US have shown a specificity of 95–100% and sensitivity between 61 and 93% for diagnosing ureterolithisis with US. These results suggest that US may be the ideal screening test, while CT is be reserved for situations when US is unavailable or non-diagnostic [12, 13].

Another proposed benefit to using CT is the ability to accurately measure stone size and degree of urinary obstruction. CT can also provide information about the need for further urologic intervention such as stenting [9]. However, authors of a recent study observed that the vast majority of patients receiving a CT scan as part of their work-up for renal colic did not require admission or immediate intervention for their kidney stone. Thus, there is a large population of low risk patients with suspected renal stones that could safely be managed with either ultrasound alone, expectant management or delayed CT [14]. Unfortunately, despite evidence that ultrasound is a cost-effective and useful alternative without inherent radiation risks, CT utilization continues to be the prevalent modality for diagnosis of ureterolithiasis and renal colic in the ED [7, 8, 10, 14].

We believe that our study is novel for utilizing a multi-hospital database across a geographic region to determine the prior imaging exposure of our ED patients. While surely not inclusive of all prior studies that might have been obtained at free-standing and non-system hospitals, our hospital system captures the majority of ED visits in a 12-county region. Further, we suggest that our study adds to the current understanding of imaging patterns for ED patients as we examined patient characteristics and clinical history that was associated with CT use both at the current visits and past visits respectively.

Our study has several limitations that warrant discussion. Our study population was predominantly Hispanic and from lower socioeconomic groups. Thus, our findings may not be generalizable to other settings. As we noted previously, our methodology likely underestimates the prior imaging exposure of our patients as we cannot account for CTs performed outside of our hospital system. Regardless of such a limitation, our records identified a large percentage of patients with a history of recurrent imaging, so the findings are, nonetheless concerning. Finally, although our data suggest some reason for hope that the practice of emergency physicians may be evolving toward primary use of ultrasound for work-up of renal colic, our practice setting may not be generalizable to others as we have a teaching environment with residents and clinical faculty providing care to patients. Future investigators should examine the evolution of imaging practice for ED renal colic evaluation in other types of institutions and for other patient populations.

Within our ED cohort of renal colic patients, 55% had at least one prior CT. The mean number of prior CTs was lower for patients receiving CT on EDCV, and Ultrasound (US) alone was used more often in patients with prior stone history vs. those with no prior history.