Article, Urology

Medical expulsive therapy use in emergency department patients diagnosed with ureteral stones

a b s t r a c t

Objective: Recent studies have clarified the role of alpha-blockers, such as tamsulosin, for patients diagnosed with Ureteral stones b 10 mm not requiring an urgent intervention. Prior studies have reported low rates of use of MET by emergency physicians. We sought to describe patterns of alpha-blocker use and to determine factors associ- ated with utilization in patients diagnosed with ureterolithiasis in the ED. Methods: We used data from a randomized trial of CT scan vs. ultrasound in participants with suspected urolith- iasis enrolled at 15 EDs between October 2011 and February 2013. The use of medical expulsive therapy was identified by the prescription of an alpha-blocker, calcium channel blocker, or steroid at the ED visit. The preva- lence of alpha-blocker use in participants with ureteral stones on imaging was calculated, and multivariable models were used to examine risk factors for utilization.

Results: Of the 524 participants who were identified with a ureteral stone on CT scan and discharged from the ED, 375 (71.4%) received an alpha-blocker, and 2 (b 1%) received a steroid. There was no significant difference in alpha-blocker use for participants based on Stone size or location. However, there was a 3.6-fold difference in alpha-blocker use between the lowest and highest use ED sites. In the multivariable analysis, ED site was inde- pendently associated with utilization of alpha-blockers.

Conclusions: Alpha-blockers were prescribed in more than two-thirds of patients with a Distal ureteral stone on imaging, a much higher prevalence than previously reported. There was substantial variability in alpha-blocker use based on ED site.

Introduction

Background

urinary stone disease is a common condition that causes acute, se- vere pain when stones lodge in the ureter, frequently resulting in an emergency department (ED) visit [1-3]. Medical expulsive therapy (MET), including alpha-blockers, steroids, and calcium channel blockers, has been extensively studied for improving the rate of stone passage in patients who do not require immediate urologic intervention [4-6,22]. Both the American Urologic Association and European Urologic Association have broadly recommended that patients with a new

* Corresponding author at: Department of Emergency Medicine, 505 Parnassus Avenue, San Francisco, CA 94143, USA.

E-mail address: [email protected] (R.C. Wang).

diagnosis of ureteral stone b 10 mm and do not require and urgent uro- logic intervention “should be offered an appropriate medical therapy to facilitate stone passage” [7]. This recommendation has been refuted by a large multicenter trial of participants with ureteral stones of all loca- tions and sizes up to 10 mm, which did not show a benefit of tamsulosin or nifedipine [8]. Recently, two meta-analyses have clarified the role of MET for patients diagnosed with ureteral stones. Both meta-analyses have found strong evidence that tamsulosin and other alpha-blockers are efficacious and should be prescribed for patients with ureteral stones 5-10 mm in size [9,10].

Prior studies have reported low rates of MET use by emergency phy- sicians, ranging from 1.1% to 14% of patients diagnosed with urolithiasis [3,11-13]. These prior studies have concluded that MET is “underutilized by emergency physicians”, citing a “problem in knowledge translation”, which has implications for quality of patient care and the economic bur- den of urolithiasis [12-14]. These studies used the National Hospital

http://dx.doi.org/10.1016/j.ajem.2017.02.040

Ambulatory Medical Care Survey (NHAMCS), or the MarketScan Com- mercial Claims and Encounters Database, a medical and drug insurance Claims database, which defined eligibility for MET based on ICD-9 codes for urolithiasis. However, randomized trials of MET determined eligibil- ity based on the presence of a ureteral stone on CT scan [6]. Thus, the prior studies that reported MET “underutilization” are limited in their ability to measure appropriate MET use.

Importance

One of every 11 Americans suffer from urolithiasis [15]. From 1992 to 2009, ED visits for urinary stone disease nearly doubled, and it is es- timated there are now more than a million annual ED visits for suspected USD in the U.S. [3,16]. The use of Health care resources for these patients has also increased, rising from $2.1 billion in 2000 to over $5 billion in 2006 [17,18]. MET is an ED intervention that could help with morbidity from Kidney stones in certain situations. However, there is a lack of studies to examine patterns of MET use by emergency physicians that include imaging data to allow for the identification of el- igible patients.

Objectives

Using data from a recently completed randomized pragmatic trial, the Study of Ultrasonography versus Computed Tomography for Suspected Nephrolithiasis, we sought to examine patterns of MET pre- scription in fifteen academic US EDs [19]. Our objective was to deter- mine the current use of MET among patients with ureteral stones on CT, and identify factors associated with the failure to use alpha-blockers in eligible patients. We hypothesize that academic emergency physi- cians use MET at much higher rates in eligible patients than previously published, but that there is wide practice variation.

Methods

Study design and setting

This was an observational study using data from a recently conduct- ed randomized pragmatic trial, the Study of Ultrasonography versus Computed Tomography for Suspected Nephrolithiasis, (trial registration number: NCT01451931 at clinicaltrials.gov) [19]. This study was con- ducted at 15 academic EDs across the United States between October 2011 and February 2013. Details of the participating EDs have been re- ported [20]. Briefly, the participating sites were academic EDs with emergency medicine residencies and emergency ultrasound fellow- ships across the United States, with representation from a number of settings - urban, rural, university based and safety net hospitals. The sites varied by size, annual census, and patient population served. This randomized trial study was performed with institutional review board approval at each site and informed consent was obtained from all par- ticipants. This current study was performed with institutional review board approval at the University of California, San Francisco.

Participants

Adult participants with suspected kidney stones that required imag- ing (as determined by an attending emergency physician) who consented to study inclusion were randomly assigned to receive point-of-care ultrasound (POC ultrasound, performed by an ED physi- cian), radiology ultrasound, or CT as their initial imaging test. Patients were excluded from enrollment if they were pregnant, at high risk of a serious non-kidney stone diagnosis, had received a kidney transplant, required dialysis, had a known solitary kidney, or were N 285 lbs. if male or 250 lbs. if female. After participants were randomized to an ini- tial imaging test, the subsequent management of participants was up to the discretion of the emergency physicians, including decisions

regarding alpha-blocker prescription. We limited this analysis to study participants who were found to have a ureteral stone on CT. Participants who received an intervention at baseline were excluded as they re- ceived a urologic intervention prior to receipt of an alpha-blocker pre- scription in order to mirror eligibility criteria in previous randomized trials of MET.

Methods and measurements

Research coordinators used a standardized data collection form to collect detailed demographic, clinical, laboratory, and imaging data dur- ing the index ED. Prior to patient enrollment, research coordinators attended a two-day meeting to receive training study protocol, forms, and data collection. Additional weekly online meetings provided more in depth training regarding data collection. Detailed demographic, clin- ical, laboratory, and imaging data were collected during the index ED visit by trained research coordinators. Patients and physicians were di- rectly interviewed in real time during the index ED visits, including use of alpha-blockers, calcium channel blockers, and steroids. Specific med- ical expulsive therapy agents in each drug class were captured. These data were recorded on paper forms and faxed to a data-coordinating center, which provided immediate feedback for completeness. Research coordinators were blinded to the study hypothesis.

Outcome measures

The main outcome measure was the receipt of alpha-blocker pre- scription at the time of ED discharge. Each participant was identified as having received alpha-blockers (yes vs. no), and the type of alpha- blocker (tamsulosin vs. terazosin vs. doxazosin), calcium channel blockers (yes vs. no), and corticosteroids (yes vs. no) as adjuncts to ob- servation for stone passage. As alpha-blockers are known to have a class effect, we considered any alpha-blocker use as positive for the main outcome.

Exposures of interest

Based on prior studies, we identified variables associated with par- ticipant receipt of MET use, including gender, age, race, educational level, and insurance status, and ED site [3,11,12]. In addition to these variables, we evaluated stone size and location on CT. We combined race categories (Asian, Native American, Pacific Islander) as some cate- gories had b 5 observations.

Statistical analysis

For all analyses we performed 2-sided significance testing and set a type I error rate at 0.05. We first calculated estimates of the prevalence of alpha-blocker use, defining use as a ratio of alpha-blockers prescribed/eligible patients. As the evidence for tamsulosin benefit is primarily in distal ureteral stones (vs. no benefit in proximal or mid- ureteral stones), we described alpha-blocker use according to stone lo- cation. Similarly, we examined alpha-blocker use according to stone size, as those with large stones (5-10 mm) receive the most benefit. We conducted univariate analyses to determine the strength of associa- tion between these potential predictors and the failure to prescribe alpha-blockers in eligible patients. We also constructed a Multivariable logistic regression models to determine the strength of association be- tween predictors of interest and the failure to prescribe alpha-blockers in eligible patients in those participants with distal ureteral stones. We included variables such as gender, age, race, educational level, insurance coverage, stone size, and ED site as potential predictors, as female gen- der, younger age, and those not employed full time were significantly less likely to receive medical expulsive therapy in prior studies. We con- sidered the ED site #5 as the reference as this site had the lowest prev- alence of non-use of alpha-blockers. Uncertainty of the estimated

prevalence and measures of associations was summarized using exact binomial 95% confidence intervals. Stata (StataCorp. 2013. Stata Statisti-

Table 2

Alpha-blocker use by stone size and location

cal Software: Release 13. College Station, TX) was used to perform the

statistical analysis.

Distal ureteral^b (N = 249)

Proximal and mid-ureteral^c

(N = 275) p-Value

3. Results

0-5 mm

5-10 mm

143/202 (70.8%)

34/44 (77.3%)

106/156 (68.2%)

85/110 (77.3%)

0.59

1.0

N 10 mm^a

2/3 (66.7%)

4/9 (44.4%)

0.1

Of 1623 subjects receiving CT (including 933 initially randomized to CT, 42 randomized to POCUS, 35 randomized to radiology US, and 613 who received a CT after initial randomization to an Ultrasound arm), 564 (34.8%) had a ureteral stone at the index ED visit. 524 were discharged home without a urologic intervention (See Appendix 1). The median age of participants diagnosed with ureterolithiasis was 40, and 68.4% were men (Table 1). Table 1 displays patient demographics, financial characteristics (educational level attained, insurance cover- age), and details of the stone location and stone size. 43/524 (8.2%) par- ticipants with a distal ureteral stone required urologic intervention within 30 days.

Main results

Missing data was identified in 5 participants (b 1%) who refused to answer the question regarding whether they had insurance coverage. Of the 524 participants, 374 (71.4%) were prescribed an alpha-blocker, and of these 366 (97.9%) were given tamsulosin, with the remainder re- ceiving terazosin or doxazosin. Steroids were prescribed in 2 patients, and calcium channel blockers were prescribed in 0 patients (overall 0.4% of 524 participants). Table 2 displays patterns of alpha-blocker use by stone size. The most common location of ureteral stone was dis- tal - 249/524 (47.5)%. Of the distal ureteral stones, 202/249 (81.1%) measured 0-5 mm, and 44/249 (28.6%) measured from 5 to 10 mm. Alpha-blockers were prescribed in approximately 70% of those with dis- tal ureteral stones. There was no difference between alpha-blocker use

Table 1

Characteristics of participants with ureteral stone on CT scan.

Overall (N = 524)

Male 358 (68.3)

Age (IQR) 40 (30-51)

Race

^a Medical expulsive therapy not indicated for ureteral stones N 10 mm.

^b Tamsulosin effective in distal ureteral stones.

^c Tamsulosin not effective in proximal or mid-ureteral stones.

in those with distal vs. non-distal stones. Similarly, N 70% of those with smaller stones as well as larger stones received alpha-blockers. 6 of 12 participants with stones N 10 mm received alpha-blockers; of the 6 par- ticipants with stones N 10 mm who received an alpha-blocker, 4 re- ceived a urologic consultation in the ED.

Table 3 displays the results of a multivariable analysis to identify predictors of alpha-blocker underutilization in patients with distal ure- teral stones. ED site was the strongest predictor of underutilization of alpha-blockers, with ED sites numbered 14, 7 and 15 having odds ratios of 11.8, 20.3, and 57.2 as compared to the reference site (5). Utilization was also less likely in those who reported a high school education com- pared to those who had attained a graduate school education (OR 3.7, 95% CI 1.1-12.9). This association was significantly associated with un- derutilization despite adjustment for gender, race, age, insurance cover- age, stone size, and ED site. However, we do not believe that physicians fail to use alpha-blockers in those with lower educational levels attained, but rather due to the correlation between educational levels and ED site (chi-square test = 137, p b 0.001).

We displayed alpha-blocker use in eligible patients stratified by stone size and ED site in Fig. 1. Alpha-blocker use varied among the 15 ED sites considerably, ranging from 25.0% to 90.9%, (p b 0.001). This rep- resents more than a 3-fold difference between the use of alpha-blockers between the site with the lowest rate of alpha-blocker prescription and that of the highest. Especially notable is alpha-blocker use in for those with large stones, who are the most likely to benefit. Two sites did not provide alpha-blockers, whereas 7 sites provided alpha-blockers in 100% of participants.

Limitations

White

African American Asian

Native American Pacific Islander More than one Hispanic Refused Educational level Elementary High school College Graduate School

253 (48.2)

81 (15.6)

39 (7.4)

10 (1.9)

2 (0.4)

13 (2.6)

125 (23.8)

1 (0.2)

85 (16.2)

144 (27.4)

128 (24.4)

167 (31.8)

This is a secondary analysis of data obtained from a randomized trial, the Study of Ultrasonography vs. Computed Tomography for Suspected Nephrolithiasis, and thus some limitations exist for assessing data on therapy. One limitation that should be noted is the time of data collection, which ended in 2013. We feel that more recent data would be desirable, mainly to reflect current practice. Also, a second limitation in this study is likely decreased generaliz- ability to community emergency medicine practice, as the parent trial was conducted at academic emergency departments, chosen for their experience with point-of-care ultrasound to be included in

Insurance coverage 361 (68.8)

pain level (IQR) 9 (8-10)

History of prior stone 230 (43.9)

Stone location

the parent randomized trial. While we cannot comment on commu- nity emergency medicine practice, this is a 15-center study, with representation from a number of settings - urban, rural, university

Proximal Mid-ureter Distal Stone size 0-5 mm

6-10 mm

N 10 mm

Urologic consultation in ED

No Yes

Unknown

Urologic intervention

Within 30 days

Within 180 days

59 (11.3)

216 (41.2)

249 (47.5)

358 (68.3)

154 (29.3)

12 (2.3)

364 (69.5)

113 (21.6)

47 (9.0)

43 (8.2)

65 (12.4)

based and safety net hospitals, and is likely to reflect academic emer- gency medicine practice. Additionally, we did not collect data re- garding allergies to medications, or other contra-indications to MET. Despite these limitations, we feel that this data source is supe- rior to other administrative data sources in many respects. The data from this trial was prospectively collected from 15 ED across the United States, and contained information regarding CT findings, in- cluding stone size and location. Other data sources, such as NHAMCS, are more generalizable to community emergency medicine settings, but the most recent NHAMCS survey available is from 2013 as well. Also, our methodology is likely more valid than prior methods that have used NHAMCS or insurance claims data, as we are able to

Table 3

Adjusted odds ratios of predictors of alpha-blocker underutilization among participants with distal ureteral stones (N = 249)

	Overall	Alpha-blockers	No alpha-blockers	AOR
	N = 249	N = 179	N = 70	(95%CI)
Female	174 (70.0)	127 (71.0)	47 (67.4)	1.3 (0.4-1.6)
Age	40.9 (12.8)	41.2 (12.9)	40.0 (12.6)	1.0 (0.97-1.02)
Race
White	123 (49.4)	91(50.3)	32 (45.7)	Ref
African American	37 (14.9)	27 (13.4)	10 (14.3)	0.4 (0.1-1.1)
Other	33 (13.3)	26 (13.6)	7 (10.0)	0.5 (0.2-1.6)
Hispanic	56 (22.5)	35 (23.5)	21 (30.0)	0.9 (0.3-2.3)
Educational level
Graduate School	90 (31.8)	75 (41.9)	15 (21.4)	Ref
College	57 (24.6)	37 (22.7)	20 (28.6)	2.4 (0.9-6.3)
High school	65 (27.4)	47 (26.1)	18 (25.7)	1.7 (0.6-4.3)
Elementary	37 (16.2)	20 (11.2)	17 (24.3)	3.8 (1.2-12.1)
Insurance coverage	170 (69.1)	128 (71.9)	42 (61.8)	1.3 (0.5-3.2)
Stone size
0-5 mm	202 (68.4)	143 (66.7)	59 (72.7)	Ref
5-10 mm	44 (29.3)	34 (31.7)	10 (23.3)	0.6 (0.2-1.5)
N 10 mm ED site^a	3 (2.3)	2 (1.6)	1 (4.0)	3.0 (0.1-42.3)
5	43	40 (93.0)	3 (7.0)	Ref
1	7	6 (85.7)	1 (14.3)	2.7 (0.2-32.3)
2	12	7 (58.3)	5 (41.7)	7.7 (1.3-46.9)
3	20	11 (55.0)	9 (45.0)	27.9 (4.6-169.7)
4	33	25 (75.8)	8 (24.2)	4.6 (1.0-21.3)
6	12	8 (66.7)	4 (33.3)	6.5 (1.1-36.4)
7	16	7 (43.8)	9 (56.3)	20.3 (3.8-107.7)
8	16	13 (81.2)	5 (18.8)	3.9 (0.6-25.2)
9	13	11 (84.6)	2 (15.4)	2.8 (0.4-20.0)
10	14	11 (78.6)	3 (21.4)	4.6 (0.7-28.5)
11	6	5 (83.3)	1 (16.7)	3.0 (0.2-38.3)
12	19	16 (84.2)	3 (15.8)	3.0 (0.4-20.7)
13	15	12 (80.0)	3 (20.0)	2.9 (0.5-17.5)
14	6	4 (66.7)	2 (33.3)	11.8 (1.4-102.5)
15	17	3 (17.7)	14 (82.4)	57.2 (8.6-379.4)

^a Row percentages provided to display prevalence of alpha-blocker use at each site.

identify those participants with ureteral stones on CT, as well as specify stone location and size. Thus, we can comment on the appropriate use of MET in specific subgroups of patients, whereas prior studies are limited in this respect.

4. Discussion

We conducted a multicenter study using prospectively collected data to describe MET use in ED patients with urolithiasis. We found

Fig. 1. Proportion of eligible subjects who received an alpha-blocker, stratified by ED site and stone size (b5 mm, 5-10 mm, and N 10 mm).

that in this diverse cohort of patients presenting to academic emergency departments, MET was prescribed in approximately 70% of those with a ureteral stone on CT scan, a significantly higher rate than previously re- ported [3,11,12]. When a MET agent was prescribed, it was almost ex- clusively an alpha-blocking medication, tamsulosin. We found that the pattern of emergency physicians alpha-blocker use reflects American Urologic Association guidelines (which recommends those with a ure- teral stone b 10 mm receive MET), as N 70% of those with distal, mid-ure- teral, and proximal ureteral stones b 10 mm in size received an alpha- blocker. Some participants with ureteral stones N 10 mm received tamsulosin; in the majority of these cases, a urologist was consulted, suggesting that the decision to provide MET was made jointly.

This study provides an updated understanding of alpha-blocker use by emergency physicians. Prior studies using ICD9 codes to identify eli- gible patients reported infrequent use of alpha-blockers for kidney stone. According to a national survey of kidney stone management in 2000, alpha-blockers were absent from a list of the 20 most frequently prescribed medications for urolithiasis [21]. In a national survey of claims data from 2000 to 2006, the overall prevalence for MET use was 2.5% [11]. A subsequent study using data from the National Hospital Ambulatory Medical Care Survey found that alpha-blockers were used in 14% of participants in the years 2007-2009 [3]. Our study, which de- fined eligibility for MET based on stone location and size on CT scan, suggests that alpha-blocker use is now widespread in academic emer- gency departments. It is unclear as to which factors explain this differ- ence in the reported prevalence from prior studies, and this study. The results of this study, while unlikely to represent community practice, is likely to be internally valid, as we are able to accurately identify those who are actually eligible for MET based on criteria which mirror those of clinical trials.

We identified independent predictors of alpha-blocker use in eligi- ble patients. The strongest predictor of underutilization is ED site, with substantial variation between the lowest and highest-using sites. This likely reflects the local practice of each ED faculty’s practice patterns and beliefs regarding alpha-blocker efficacy. Especially striking is the pattern of alpha-blocker use in 5-10 mm distal ureteral stones (the sub- group of participants who would benefit the most from tamsulosin) in which we found that 2 sites did not prescribe any alpha-blocker, and 7 sites prescribed alpha-blockers in 100% of subjects. Based on recent clin- ical trial and meta-analysis data, we believe that patients with large ure- teral stones should receive tamsulosin, unless the patient has a contraindication, such as pregnancy, an allergic reaction to alpha- blockers, previous alpha-blocker or beta-blocker use, postural hypoten- sion, or need for immediate intervention. We believe that while most academic emergency physician practices frequently use MET, continu- ing education and awareness of the role of MET for a common ED prob- lem would result in improved standardization and quality of care.

In summary, we report current patterns of alpha-blocker use by emergency physicians at 15 academic emergency department sites across the United States. The main agent that emergency physicians prescribed was tamsulosin, as opposed to other alpha-blockers, calcium channel blockers, or steroids. We found that alpha-blockers are fre- quently prescribed for patients diagnosed with ureteral stones on CT scan, which are indicated in large ureteral stones.

Funding/Support

This study was supported by funding from the Agency of Healthcare Research and Quality AHRQ Grant # K08 HS02181 (RCW) and a Multi- disciplinary K12 Urologic Research Career Development Program, Grant # K12-DK-07-006 (TC).

Conflict of interest

The authors declare no conflicts of interest.

Author contributions

Contributors: RCW conceived the work, collected data, performed data cleaning and statistical analysis, and drafted and critically revised the manuscript. NA participated in data analysis and manuscript revi- sion. TC helped with study conception, design, participated in data col- lection and manuscript revision. MM participated in data collection, cleaning, analysis and manuscript revision. CM helped with study con- ception, design, participated in data collection and manuscript revision. SS helped to perform statistical analysis and manuscript revision. RSB helped with study conception, design, participated in data analysis and manuscript revision. All authors had full access to the data, take re- sponsibility for the integrity of the data and have approved the manu- script. The data were collected, results analyzed, and the manuscript prepared without influence from funding agencies. RCW takes responsi- bility for the manuscript as a whole.

Appendix 1. Patient flow diagram

References

AJEM