In reply:

Abstract

Overall, we wish to congratulate the authors on their study that addresses an important issue and appreciate the opportunity to respond to their concerns about our external validation study. We agree that the STONE score stratifies patients with suspected ureteral stone into low-, moderate-, and high-risk groups and that it was more specific than physician gestalt for identifying ureteral stone. In the randomized trial published in the New England Journal of Medicine that we used to perform our external validation, the test characteristics of computed tomography (CT) were determined by using the clinical diagnosis of stone confirmed by stone passage or stone retrieval as the reference standard.1Smith-Bindman R. Aubin C. Bailitz J. et al.Ultrasonography versus computed tomography for suspected nephrolithiasis.N Engl J Med. 2014; 371: 1100-1110Crossref PubMed Scopus (386) Google Scholar The CT result was considered positive (stone was identified, including intrarenal stones) in approximately 58% of subjects. In contrast, to perform the external validation of the STONE score, we used the finding of ureteral stone on CT to mirror the outcome used in the original derivation. We found that the prevalence of ureteral stone on CT scan in our cohort was 39%, and the prevalence of intrarenal stones without ureteral stones was 17%, higher than the 10% prevalence mentioned in the letter to the editor. We have rechecked the reliability of the initial data collection by performing a second blinded assessment of CT scan interpretations at 2 of the 9 sites and found near perfect agreement (κ=0.94). As the authors of the letter to the editor suggest, the differences in prevalence may have been caused by differing enrollment criteria and patient mix. We acknowledge that these prevalence differences may have affected the STONE score screening performance. We also agree that race is a strong epidemiologic factor in the occurrence of kidney stone. In a univariate analysis of our cohort, white race was significantly associated with ureteral stone compared with black race. However, this association became less strong when controlling for the other STONE score predictors, in particular duration of pain (Appendix Table E2).2Wang R.C. Rodriguez R. Moghadassi M. et al.External validation of the STONE score, a clinical prediction rule for ureteral stone: an observational multi-institutional study.Ann Emerg Med. 2015; (http://dx.doi.org/10.1016/j.annemergmed.2015.08.019)Google Scholar The timing of patient presentations after symptom onset may vary according to a number of factors, aside from the acuity or severity of their pain. In our multi-institutional cohort, we found that black subjects presented after a longer duration of pain compared with white subjects, which may reflect racial disparities in terms of access to ED care. We suggest that this question be analyzed in a diverse cohort prospectively. Finally, we agree that the low prevalence of important alternative diagnoses is an important and promising finding. However, the upper limit of the 95% confidence interval was 3.6%, which may be a higher prevalence of important alternative diagnoses (eg, appendicitis, diverticulitis, ovarian torsion) than many emergency physicians would tolerate. Furthermore, the predictive values of a test may be misleading because they vary with disease prevalence, as opposed to sensitivity and specificity.3Green S.M. Schriger D.L. Yealy D.M. Methodologic standards for interpreting clinical decision rules in emergency medicine: 2014 update.Ann Emerg Med. 2014; 64: 286-291Abstract Full Text Full Text PDF PubMed Scopus (55) Google Scholar Thus, we question whether patients with a high STONE score should have imaging deferred. We look forward to studies incorporating noninvasive testing. Alternatively, a prospective validation with a larger sample of the disease of interest would allow more precise point estimates with narrower 95% confidence intervals3Green S.M. Schriger D.L. Yealy D.M. Methodologic standards for interpreting clinical decision rules in emergency medicine: 2014 update.Ann Emerg Med. 2014; 64: 286-291Abstract Full Text Full Text PDF PubMed Scopus (55) Google Scholar and may enhance acceptability to clinicians. External Validation of the STONE ScoreAnnals of Emergency MedicineVol. 67Issue 2PreviewAs authors of the original derivation and validation of the STONE score,1 we were pleased to see an external validation of it by Wang et al.2 Moreover, we found their results to be reassuring in that the STONE score did allow stratification of patients into low-, moderate-, and high-risk groups, and that it improved prediction of ureteral stone beyond physician gestalt. Full-Text PDF