2.1. Patient File Selection

The institutional review board approved this study. We retrospectively reviewed the institutional imaging software (PACS, version 11, Minnetonka, MN, USA) for all NCCT scans performed in the ED between March 2016 and May 2017. Inclusion criteria included detection of a single ureteral stone by NCCT, body temperature ≤ 37.8 °C, and sufficient background data as described below. Our intention was to differentiate between patients who will require surgical intervention and those who can be managed with a conservative approach. The combination of ureterolithiasis and fever generally requires urgent drainage and we therefore excluded these patients. We intentionally selected patients with known, single ureteral stones identified by NCCT, focusing on identification of risk factors specific to those patients that ultimately required surgical intervention.

From these patient files, we collected demographic, clinical and laboratory data that were available to the treating ED physician on all such patients. This included age, gender, duration of symptoms, pain level at presentation according to Visual Analog Scale score, nausea and/or vomiting, history of urolithiasis and associated intervention, and laboratory tests including blood count and creatinine level. Urinalysis is routinely assessed for all patients in the ED, but precise data on many patients with microhematuria were lacking; in any case, this marker is considered weak and less sensitive for ureterolithiasis [20]. Moreover, while microhematuria may be useful for diagnosis of ureterolithiasis, its role in prediction of surgical intervention for patients with ureteral stones appears generally less relevant.

The cohort was divided into patients who required surgical intervention (Group 1), and patients who were successfully managed conservatively (Group 2). We considered conservative management to be successful for patients who fulfilled all of the following: (1) did not require any surgical intervention, (2) no symptoms for at least two months (recognizing the definition for successful conservative management for ureterolithiasis in previous studies, which stipulated at least one month free of symptoms and with no need for surgical intervention [7]), (3) lack of evidence of hydronephrosis or stone on follow-up imaging, and (4) normal creatinine level. This practical endpoint has been used successfully elsewhere, and negates the use of repeat NCCT in many patients [7]. For Group 1 patients, surgical interventions included either drainage of the affected renal unit or primary ureteroscopy. The decision whether or not to intervene was left to the discretion of the attending urologist according to commonly accepted indications, such as intractable pain, elevated creatinine levels, large and proximal ureteral stones, presence of peri-renal urinoma and recurrent ED visits.

Subsequently, for the Group 1 patients, we obtained data on intervention for the ureteral stone using a unique, integrated hospital-community electronic medical record database (OFEK). This system provides data on the patient’s family doctor visits, hospital outpatient visits, laboratory and imaging results, both at the medical center, and in the community. Visits to other medical centers are also documented. Depending on the health care provider, the completeness of data varied. Therefore, when data on surgical intervention were unclear, we contacted the patient by phone. Cases were excluded if the outcome could not be determined clearly from the medical file, and the patient could not be contacted.