Key Words: kidney; nephrolithiasis; tomography, x-ray computed; ultrasonography; emergency service, hospital

Transcription

1 Patient and Institutional Characteristics Associated with Initial Computerized Tomography in Children Presenting to the Emergency Department with Kidney Stones Justin B. Ziemba, Douglas A. Canning, Jane Lavelle, Angela Kalmus and Gregory E. Tasian* From the Division of Urology, Department of Surgery, The Perelman School of Medicine at University of Pennsylvania (JBZ, DAC, GET) and Division of Urology, Department of Surgery (DAC, AK, GET), Department of Emergency Medicine (JL) and Center for Pediatric Clinical Effectiveness (GET), Children s Hospital of Philadelphia, Philadelphia, Pennsylvania Abbreviations and Acronyms AUA ¼ American Urological Association CT ¼ computerized tomography ED ¼ emergency department PHIS ¼ Pediatric Hospital Information System Accepted for publication September 2, Study received institutional review board approval. Supported by National Institutes of Health Grant T32HD (GET). * Correspondence: Department of Surgery, Division of Urology, Wood Center, 3rd Floor, Children s Hospital of Philadelphia, 34th St. and Civic Center Blvd., Philadelphia, Pennsylvania (telephone: ; FAX: ; tasiang@chop.edu). Purpose: Professional associations recommend ultrasound as the initial imaging study in children with suspected nephrolithiasis but computerized tomography remains frequently used. We identified patient and institutional characteristics associated with computerized tomography as the first imaging study in children with nephrolithiasis diagnosed in the emergency department. Materials and Methods: We performed a cross-sectional study of children 2 to 18 years old with nephrolithiasis who were referred to a freestanding pediatric hospital from 2003 to We identified the imaging modality first used to evaluate the child. Medical directors at the emergency department where children were first evaluated were sent a questionnaire to ascertain emergency department characteristics. Multivariate hierarchical logistic regression models were used to determine patient and institutional characteristics associated with initial computerized tomography. Results: Of 536 eligible children 323 (60.2%) were evaluated at emergency departments from which surveys were returned. Of the 323 children 238 (71%) underwent computerized tomography as initial imaging. Ultrasound was available at all emergency departments. Older patient age was associated with higher initial computerized tomography use (OR 1.09, 95% CI 1.04e1.16). A more recent year of diagnosis (OR 0.80, 95% CI 0.72e0.88) and a clinical care pathway that used ultrasound as initial imaging (OR 0.29, 95% CI 0.01e0.38) were associated with lower initial computerized tomography use. Conclusions: A clinical care pathway in the emergency department was the only institutional characteristic associated with lower computerized tomography use. Future studies are needed to determine whether care pathways using ultrasound for initial imaging in children with suspected nephrolithiasis would decrease inappropriate computerized tomography and improve adherence to national guidelines. Key Words: kidney; nephrolithiasis; tomography, x-ray computed; ultrasonography; emergency service, hospital THE incidence of nephrolithiasis in adolescents increased approximately 10% per year during the last 20 years. 1e3 During this period CT use in children diagnosed with nephrolithiasis at pediatric hospitals 1848 j /15/ /0 THE JOURNAL OF UROLOGY 2015 by AMERICAN UROLOGICAL ASSOCIATION EDUCATION AND RESEARCH,INC. Vol. 193, , May 2015 Printed in U.S.A.

2 INITIAL COMPUTERIZED TOMOGRAPHY IN CHILDREN PRESENTING WITH KIDNEY STONES 1849 increased 26% to 45%. 4 Increasing CT use is concerning because of the risk of malignancy associated with ionizing radiation, which may be higher in children than in adults due to longer life expectancy and greater sensitivity of developing tissue to the effects of radiation. 5 In response to this potential harm from radiation the AUA and EAU (European Association of Urology) developed imaging guidelines for children with suspected nephrolithiasis in 2012 and 2013, respectively. 6,7 The 2 groups recommend ultrasound as first line imaging with noncontrast CT reserved when ultrasound is nondiagnostic. 6,7 Currently CT is the most frequently used modality to evaluate abdominal pain and in children with suspected nephrolithiasis in the ED. 8,9 This is concerning because children with stones often undergo multiple CTs during a single stone episode and many of them experience recurrences. 4,10 There is substantial variability in CT use across hospitals participating in the PHIS after controlling for demographic and socioeconomic factors, which suggests that other factors affect CT use. 4 However, our knowledge of institutional characteristics associated with CT use is limited because prior studies were done using administrative databases. 2,4,8,9,11 We determine patient and institutional characteristics associated with CT as the first imaging study in children with nephrolithiasis diagnosed in the ED at pediatric and nonpediatric hospitals. We hypothesized that institutional factors would have the strongest association with initial CT use. Our secondary aim was to characterize the patterns of diagnostic imaging at these hospitals. MATERIALS AND METHODS Study Population We performed a cross-sectional study of all children 2 to 18 years old diagnosed with symptomatic kidney stones who were referred to the Division of Urology, Children s Hospital of Philadelphia, a 480-bed freestanding pediatric hospital. The Division of Urology has a prospectively maintained database of patients with nephrolithiasis who were initially evaluated at or referred to the hospital for continued treatment. We queried this database for patients with renal or ureteral calculi who underwent initial evaluation in the ED from 2003 to Excluded from analysis were patients with neuropathic bladder and anatomical urinary tract abnormalities. A kidney stone episode was defined as the time from initial symptomatic presentation in the ED to stone passage or surgical management. We retrospectively reviewed all CTs and ultrasounds performed during a kidney stone episode. Since at least one of these studies was done in all patients, additional imaging such as plain x-ray or excretory urogram was not reviewed. The ED visit note for each child was not reviewed because these records were not available. This study was approved by our institutional review board. Survey Instrument We developed a survey to identify current characteristics of the EDs where children were diagnosed with nephrolithiasis. Questionnaire completion required approximately 5 minutes. The questionnaire consisted of a maximum of 29 questions (supplementary material, jurology.com/). It assessed hospital characteristics with answer choices for ED size and volume determined by reviewing national ED data. 12,13 The questionnaire also assessed the availability of imaging resources and consultants. To determine quality initiatives designed to deliver evidence-based therapy to patients we asked about a clinical care pathway in the ED to manage suspected nephrolithiasis in children. If a pathway was present, the respondent was asked to indicate the first line imaging modality. The survey was pretested by administering it to a focus group of ED physicians to ensure that the questions comprehensively and accurately ascertained all ED characteristics potentially associated with imaging use. We revised the questionnaire based on the feedback. We compiled a list of 110 local hospitals within the geographical referral base of our hospital. Each hospital was contacted by telephone to obtain the contact information of the ED medical director, which was used to distribute the survey by . A reminder to complete the survey was sent 1 and 2 weeks after the initial invitation. Survey data were developed, collected and managed with REDCap (Research Electronic Data Capture) 14 hosted at University of Pennsylvania. REDCap is a secure, web based application designed to support data capture for research studies. Outcome Definition and Predictor Variables The primary outcome was CT as the initial imaging study during the ED visit when the stone was diagnosed. The secondary outcome was the order of imaging during the ED visit. We determined the frequency of ultrasound and CT use during the stone episode and identified the first imaging study done in each patient using the radiology report date and time stamp. The primary predictor was a clinical care pathway using ultrasound for initial imaging in children with suspected nephrolithiasis. Covariates assessed for inclusion in the regression model included patient demographics and institutional characteristics identified in the survey. Statistical Analysis Differences in age of the patients who underwent CT vs ultrasound as initial imaging were determined using the rank sum test after confirming nonparametric distribution. The chi-square test was used to evaluate differences in all categorical patient and institutional characteristics between patients who first underwent CT vs ultrasound. We used hierarchical logistic regression models to estimate the association of patient and institutional characteristics with initial CT. Regression models were built using manual backward selection of covariates. Included in the final model were all covariates associated with CT at p <0.2 on univariate analysis, those with a priori defined face validity and those showing significant

3 1850 INITIAL COMPUTERIZED TOMOGRAPHY IN CHILDREN PRESENTING WITH KIDNEY STONES confounding of the association between a nephrolithiasis care pathway using ultrasound for initial imaging and CT in the final multivariate model. Confounding was defined as a change of 15% or greater between the unadjusted and adjusted ORs of CT. Covariates included a priori were patient age, gender, diagnosis year, CT availability in the ED, 24-hour availability of ultrasound, number of pediatric beds in the ED and number of annual ED pediatric visits. Age was modeled as a continuous variable. All other variables were categorical. This random intercept model included fixed effects for the mentioned predictor variables and random effects for the hospital to which the patient presented. Thus, this model controlled for clustering of practice patterns at the hospital level, which previous PHIS studies demonstrated were common. 4 Analysis was completed on subjects with complete data. To assess for response bias, sensitivity analysis was performed to determine differences in patient and imaging characteristics between hospitals of which the medical director did vs did not complete the survey. To explore the multicollinearity of our data we constructed a correlation matrix, eliminated covariates with a correlation coefficient >j0.4j and then refit the model. Tests were 2-sided with statistical significance considered at p <0.05. Analysis was done with StataÒ 13. RESULTS Cohort Demographics Eligibility criteria were met by 539 children, of whom 3 were excluded from study due to missing imaging information, leaving 536 in the eligible population. Median age was 13.9 years in the overall cohort, 54.2% of participants were female and 77.9% were white. When comparing children who underwent CT vs ultrasound, there was no difference in gender, race/ethnicity or a prior nephrolithiasis episode (table 1). However, there was Table 1. Characteristics of children with nephrolithiasis evaluated in ED from 2003 to 2012 Ultrasound CT Overall p Value No. pts * e Median yrs age (IQR) 13 (7e16) 14 (11e17) 14 (10e16) <0.001 No. male (%) 59 (46) 188 (46.2) 247 (45.8) 0.93 No. female (%) 70 (54) 219 (53.8) 292 (54.2) No. race/ethnicity (%): White 96 (74) 321 (79.0) 420 (77.9) Black (%) 10 (8) 29 (7.1) 39 (7.2) Hispanic (%) 4 (3) 16 (3.9) 20 (3.7) 0.43 Asian (%) 0 (0) 5 (1.2) 5 (0.9) Other (%) 13 (10) 25 (6.1) 38 (7.1) Unknown (%) 6 (5) 11 (2.7) 17 (3.2) No. prior nephrolithiasis episode: Yes 105 (81) 349 (85.7) 454 (84.2) No 23 (18) 57 (14.0) 80 (14.8) 0.38 Unknown 1 (1) 1 (0.3) 5 (1.0) Median mm stone size (IQR) 5 (2e7) 3 (2e5) 3 (2e5) <0.001 * Including 3 patients excluded from CT or ultrasound group due to missing radiographic information. a difference in age and stone size. Children who underwent CT were more likely to be older and have smaller stones (table 1). Survey Results A total of 23 ED medical directors (21%) completed the questionnaire. At hospitals from which surveys were returned 323 children (60.2%) were evaluated, including 238 (71%) with CT as initial imaging. Most EDs had pediatricians (83%) and urologists (83%) available for consultation. All EDs had ultrasound available and 83% always had ultrasound available. At 1 ED (4%) there was a clinical care pathway to evaluate children with suspected nephrolithiasis, which was used to care for 66 patients (20%). Table 2 lists full responses to the questionnaire. There was no difference in patient characteristics (data not shown). However, children evaluated at EDs where the questionnaire was not completed had a smaller stone size (3 vs 4 mm) and a higher CT rate (83% vs 71%) than those evaluated at EDs where the questionnaire was completed. Initial CT Predictors The final regression model included data on the 323 children (60.2%) for whom complete patient and institutional data were available (table 3). On multivariate regression adjusting for practice pattern clustering at hospitals, 2 patient characteristics were associated with initial CT use. Older age was associated with higher odds of CT (OR 1.09, 95% CI 1.04e1.16) and a more recent year of diagnosis than 2003 was associated with lower odds of CT (OR 0.80, 95% CI 0.72e0.88). The only institutional characteristic associated with initial CT was a clinical care pathway that used ultrasound as initial imaging in children with suspected nephrolithiasis. Children evaluated at an ED with a pathway were 71% less likely to undergo CT as the first diagnostic imaging modality than children evaluated at an ED without a pathway (OR 0.29, 95% CI 0.01e0.38, table 3). There was no significant change of the results in the sensitivity analysis. Imaging Patterns Of the eligible population 407 children (75.9%) underwent CT and 129 (24.1%) underwent ultrasound. One ultrasound or 1 CT was performed in 298 children (55.6%) and 238 (44.4%) underwent more than 1 imaging study, including ultrasound plus CT in 174 (32.4%) and 2 or more CTs in 54 (10.0%). DISCUSSION In this cross-sectional study of children with nephrolithiasis patient and institutional characteristics were associated with initial CT in the ED. Despite

4 INITIAL COMPUTERIZED TOMOGRAPHY IN CHILDREN PRESENTING WITH KIDNEY STONES 1851 Table 2. Survey responses from 23 respondents at referring EDs Question No. Respondents (%) Time to pediatric hospital (mins): Less than (57) Greater than (43) Practice location: Urban 8 (36) Suburban 14 (64) Practice environment: Academic 10 (44) Nonacademic 13 (56) Pediatric hospital affiliate: Yes 8 (35) No 15 (65) Trauma center: Yes 6 (26) No 17 (74) No. ED beds: Less than (48) Greater than (52) Dedicated pediatric ED: Yes 11 (48) No 12 (52) No. pediatric ED beds:* Less than 5 2 (18) Greater than 5 9 (82) General consulting pediatrician available: Yes 19 (83) No 4 (17) General consulting urologist available: Yes 19 (83) No 4 (18) Consulting pediatric urologist available: Yes 4 (21) No 15 (79) CT available in ED: Yes 16 (70) No 7 (30) Ultrasound available in ED: Yes 23 (100) No 0 Ultrasound available in ED 24 hrs/day: Yes 19 (83) No 4 (17) Radiologist read ED films: Yes 22 (96) No 1 (4) Radiologist available 24 hrs/day: Yes 15 (71) No 6 (29) Total No. pts/yr evaluated in ED: Less than 50, (55) Greater than 50, (45) No. pediatric pts/yr evaluated in ED: Less than 10, (45) Greater than 10, (55) Clinical care pathway to evaluate suspected nephrolithiasis: Yes 1 (4) No 22 (96) * Total of 11 respondents with dedicated pediatric ED. Total of 19 respondents. Total of 21 respondents. Total of 22 respondents. variability in ED volume and academic affiliation, the only ED characteristic associated with decreased initial CT use was a clinical care pathway that used ultrasound for initial imaging in children with suspected nephrolithiasis. Table 3. Final multivariate logistic regression model for initial CT use in the ED Predictor OR (95% CI) p Value Age at diagnosis 1.09 (1.04e1.16) Gender 0.98 (0.56e1.69) 0.94 Diagnosis yr (referent 2003) 0.80 (0.72e0.88) <0.001 Pediatric ED No. beds 0.48 (0.52e4.45) 0.52 ED CT scanner 0.38 (0.02e5.87) 0.49 Ultrasound available 24 hrs/day 0.54 (0.35e8.46) 0.66 No. pediatric ED visits/yr 1.99 (0.84e4.72) 0.12 Nephrolithiasis clinical care pathway 0.29 (0.01e0.38) The AUA and EAU published imaging guidelines on children with suspected nephrolithiasis in 2012 and 2013, respectively. 6,7 The 2 groups recommended ultrasound as the first line imaging modality with CT reserved for when ultrasound was nondiagnostic and clinical suspicion remained high. 6,7 Additionally, in 2007 Image GentlyÒ began a campaign to decrease radiation exposure in children who need diagnostic imaging. 15 These recommendations were based on the harms of CT and the acceptable diagnostic performance of ultrasound. Although ultrasound is less sensitive and specific than CT, 16,17 it identifies most stones in children. In a recent prospective cohort study using CT as the gold standard the sensitivity and specificity of ultrasound for kidney stones in children was 76% and 100%, respectively. 17 Importantly the discrepancy between ultrasound and CT was clinically important in a minority of cases. 17,18 Given the variability in diagnostic imaging use for nephrolithiasis observed among PHIS hospitals, 4 we hypothesized that significant variability would exist among EDs where children were first diagnosed with kidney stones before referral to a urology clinic at a PHIS hospital. Ultrasound was universally available and at almost all EDs a pediatrician and a urologist were available for consultation. This suggests that lack of availability of ultrasound or specialist consultants is not a significant factor in determining CT use. Older patient age and earlier year of diagnosis were associated with increased odds of initial CT. This might reflect increased willingness to perform CT in older patients and temporal trends of decreasing CT use. 4 We observed that the ultrasound first pathway in the ED was associated with a 71% lower odds of undergoing initial CT for nephrolithiasis. Although only 1 hospital had a clinical care pathway, this was the only institutional characteristic associated with decreased CT even when accounting for other measures of care such as academic affiliation, and ED volume and size. However, given the study design and the fact that only 1 clinical care pathway was present, we cannot attribute causality to this association. A pathway

5 1852 INITIAL COMPUTERIZED TOMOGRAPHY IN CHILDREN PRESENTING WITH KIDNEY STONES that uses ultrasound as initial imaging in children with suspected kidney stones may possibly decrease CT use by directly influencing physician decisions. Alternatively, it is possible that a pathway is a surrogate for other processes and quality initiatives that affect imaging use. Future randomized trials should be done to assess whether clinical pathways decrease inappropriate CT use as they do for other surgical diseases of childhood such as appendicitis. 19 Previous PHIS series demonstrated that more than 50% of children with nephrolithiasis undergo at least 2 imaging studies during a kidney stone episode. 4 We observed similar imaging patterns in a broader sample of EDs that included freestanding pediatric hospitals and local hospitals. We found that 44% of children underwent more than 1 imaging study, in 32% imaging was done with multiple modalities and 10% underwent 2 or more CTs. Repeat CT during a single kidney stone episode is disturbing because a single pediatric stone protocol CT is estimated to result in 2 or 3 radiation induced cancers per 1,000 naturally occurring cancers. 5 Although this rate is low, the recurrent nature of nephrolithiasis and the likelihood of future scans compound the adverse effects of radiation. We acknowledge the limitations of this study. It was performed at a single referral hospital and, thus, may not be generalizable to other areas of the country. However, we included children who were first evaluated at local hospitals, which provides insight into practices and determinants of imaging that is not possible by analyzing administrative databases. We could not determine imaging in children in whom kidney stones were suspected but who were ultimately found to have an alternative diagnosis. It is possible that the choice of imaging was influenced by presenting symptoms, vital signs and laboratory values, which affected the pretest probability of nephrolithiasis. 20 We also did not assess technologist or radiologist comfort with performing and interpreting imaging in children or whether there were barriers to a study, such as the need to call in a technologist from home. Furthermore, although most EDs had a pediatrician and urologist available for consultation, we do not know whether these resources were used for a given child. Future prospective studies must consider these variables. Our study has the limitation of any self-reported survey, including response bias. Although all EDs were initially contacted by telephone, questionnaires were completed at only 21%. The 79% of hospitals with no response to the survey may possibly have significantly different ED characteristics. We observed greater CT use in children at hospitals where the survey was not completed (83% v 71%). Therefore, the addition of institutional characteristics of these nonresponders, especially in regard to a clinical care pathway, could have modified the association with CT use. Finally, the survey only assessed the current state of ED characteristics, which may have changed during the study course. CONCLUSIONS Although ultrasound is widely available in the ED, it is performed less frequently than CT for initial imaging in children with nephrolithiasis. A clinical care pathway in the ED was the only institutional characteristic associated with lower CT use. Future studies are needed to determine whether care pathways that use ultrasound for initial imaging in children with suspected nephrolithiasis would decrease inappropriate CT. REFERENCES 1. Dwyer ME, Krambeck AE, Bergstralh EJ et al: Temporal trends in incidence of kidney stones among children: a 25-year population based study. J Urol 2012; 188: Routh JC, Graham DA and Nelson CP: Epidemiological trends in pediatric urolithiasis at United States freestanding pediatric hospitals. J Urol 2010; 184: Sas DJ, Hulsey TC, Shatat IF et al: Increasing incidence of kidney stones in children evaluated in the emergency department. J Pediatr 2010; 157: Routh JC, Graham DA and Nelson CP: Trends in imaging and surgical management of pediatric urolithiasis at American pediatric hospitals. J Urol 2010; 184: Kuhns LR, Oliver WJ, Christodoulou E et al: The predicted increased cancer risk associated with a single computed tomography examination for calculus detection in pediatric patients compared with the natural cancer incidence. Pediatr Emerg Care 2011; 27: Fulgham PF, Assimos DG, Pearle MS et al; American Urological Association Guideline: Clinical Effectiveness Protocols for Imaging in the Management of Ureteral Calculous Disease: AUA Technology Assessment. Linthicum: American Urological Association T urk C, Knoll, Petrik A et al: Guidelines on Urolithiasis. Arnhem, The Netherlands: European Association of Urology Fahimi J, Herring A, Harries A et al: Computed tomography use among children presenting to emergency departments with abdominal pain. Pediatrics 2012; 130: e Johnson EK, Graham DA, Chow JS et al: Nationwide emergency department imaging practices for pediatric urolithiasis: room for improvement. J Urol 2014; 192: Hahn B, Dima J, Hirschorn D et al: Incidence and pathology of repeat computed tomography of the abdomen and pelvis in a pediatric emergency department population. Pediatr Emerg Care 2013; 29: Tasian G, Pulido J, Keren R et al: Use of and regional variation in initial CT for kidney stones. Pediatrics 2014; 134: 909.

6 INITIAL COMPUTERIZED TOMOGRAPHY IN CHILDREN PRESENTING WITH KIDNEY STONES Burt CW and McCaig LF: Staffing, capacity, and ambulance diversion in emergency departments: United States, Adv Data 2006; 376: Schappert SM and Bhuiya F: Availability of pediatric services and equipment in emergency departments: United States, Natl Health Stat Reports 2012; 47: Harris PA, Taylor R, Thielke R et al: Research electronic data capture (REDCap)da metadatadriven methodology and workflow process for providing translational research informatics support. J Biomed Inform 2009; 42: Goske MJ, Applegate KE, Boylan J et al: The Image Gently campaign: working together to change practice. AJR Am J Roentgenol 2008; 190: Johnson EK, Faerber GJ, Roberts WW et al: Are stone protocol computed tomography scans mandatory for children with suspected urinary calculi? Urology 2011; 78: Passerotti C, Chow JS, Silva A et al: Ultrasound versus computerized tomography for evaluating urolithiasis. J Urol 2009; 182: Eshed I and Witzling M: The role of unenhanced helical CT in the evaluation of suspected renal colic and atypical abdominal pain in children. Pediatr Radiol 2002; 32: Russell WS, Schuh AM, Hill JG et al: Clinical practice guidelines for pediatric appendicitis evaluation can decrease computed tomography utilization while maintaining diagnostic accuracy. Pediatr Emerg Care 2013; 29: Persaud AC, Stevenson MD, McMahon DR et al: Pediatric urolithiasis: clinical predictors in the emergency department. Pediatrics 2009; 124: 888. EDITORIAL COMMENT Previous investigations comparing CT to ultrasound in children with urolithiasis have been limited to single institutions or administrative data. These authors designed a study that aimed to bridge the knowledge gap by exploring patient and hospital characteristics at a level of detail not possible using administrative data alone. This approach is important to consider as we aim to verify and contextualize the findings of recent investigations in urology that use administrative data. A notable limitation of this study is that only 1 hospital had a clinical care pathway. Thus, it is difficult to draw definitive conclusions regarding whether the clinical care pathway or some other feature of the hospital accounted for the higher rates of ultrasound. Nonetheless, clinical care pathways are a key strategy to improve the quality of emergency care. 1e3 This finding certainly warrants further investigation. Emilie K. Johnson Department of Urology Ann and Robert H. Lurie Children s Hospital of Chicago Northwestern University Feinberg School of Medicine Chicago, Illinois REFERENCES 1. Ball CG, Dixon E, MacLean AR et al: The impact of an acute care surgery clinical care pathway for suspected appendicitis on the use of CT in the emergency department. Can J Surg 2014; 57: Lougheed MD, Olajos-Clow J, Szpiro K et al: Multicentre evaluation of an emergency department asthma care pathway for adults. CJEM 2009; 11: Iyer PM, McNamara PH, Fitzgerald M et al: A seizure care pathway in the emergency department: preliminary quality and safety improvements. Epilepsy Res Treat 2012; 2012: REPLY BY AUTHORS Prior studies of administrative data sets such as PHIS and NEDS (Nationwide Emergency Department Sample) preclude analysis of processes of care that may cause CT use to deviate from the imaging recommendations of the AUA and other professional organizations. Our current study addressed this knowledge gap and provides preliminary data on institutional characteristics associated with initial CT in children with kidney stones. We agree that a limitation of our study was that only 1 of the 23 responding regional hospitals had a clinical care pathway. The presence of only a single clinical care pathway highlights that these tools are likely underused in the community 1 despite their clinical effectiveness for other surgical diseases of childhood (reference 19 in article). However, given the study design, we cannot attribute a causal relationship to the association between a care pathway and a decreased probability of inappropriate CT. Further prospective studies

7 1854 INITIAL COMPUTERIZED TOMOGRAPHY IN CHILDREN PRESENTING WITH KIDNEY STONES using a cluster randomized trial design would define the effect of clinical pathways 2 on delivering the best quality of care to children with suspected nephrolithiasis. It is time that we as a pediatric urology community expand our research endeavors beyond the analysis of existing administrative data sets and retrospective institutional cohort studies. REFERENCES 1. Phillips E, Kieley S, Johnson EB et al: Emergency room management of ureteral calculi: current practices. J Endourol 2009; 23: Costantini M, Romoli V, Leo SD et al: Liverpool Care Pathway for patients with cancer in hospital: a cluster randomised trial. Lancet 2014; 383: 226.