Ethical approval was obtained from the Swedish Ethical Review Authority (No.2014/136).

The study was performed using a previously reported stone passage databank, in which stone expulsion rates for the whole cohort in relation to stone size and location, but not RSSI, were reported [6]. A retrospective review was carried out of 1,824 consecutive patients who presented at our emergency department with flank pain and underwent NCCT performed between April 2012 and September 2014. The inclusion criterion was a solitary ureteral stone > 2 mm in diameter in the axial plane. Exclusion criteria inclusive numbers are shown in the flowchart in Fig. 1.

Flowchart showing exclusion criteria with numbers

We calculated the required sample size with regard to UWT based on the findings from previous studies, in terms of both the proportion of the non-events and the estimated standard deviation of UWT in the population. To achieve 80% power with an effect size of 1-mm difference in UWT between the event and non-event groups, a sample size of at least 150 subjects was needed. Because of the high probability of SSP in distal ureteral stones, only stones in the upper and middle ureter were included, resulting in a study population of 160 subjects.

The CT examinations were intermediate-dose non-contrast enhanced scans performed on two different CT scanners: 67 patients were examined using a 40-detector row CT scanner (Brilliance, Philips Medical Systems, Best, The Netherlands) with a low-dose NCCT protocol for the urinary tract (120 kV, 70 mAs/slice, CTDI 4.9 mGy), and 93 patients were examined with a 2 × 128-channel scanner (Somatom Definition Flash, Siemens, Erlangen, Germany) (120 kVp, 70 mAs/slice CTDI 4.7). Manual measurements were performed on both axial 1-mm slices and 3-mm axial, sagittal and coronal reformats, which were generated in the main axes of the patient.

Patient-related data such as age, sex, stone laterality and C-reactive protein (CRP) at diagnosis or interventions were retrieved from the medical records. Stone-related data were obtained from CT scans using the integrated PACS measurement tool (Sectra IDS7, Linköping, Sweden).

Stone size was measured according to the methodology previously described by Jendeberg et al. (i.e. independently by three readers in the axial, coronal and sagittal reformations in a soft-tissue window) [6]. Stone length was defined as the largest of the three reformation measurements. The mean value of three readers was used. Our cohort included only upper and middle ureteral stones, which were defined as being located in the ureter segment between the ureteropelvic junction and the lower edge of the sacroiliac joint.

UWT was measured at the spot of greatest soft-tissue thickness (ureteral wall + periureteral oedema) both on axial 1-mm slices and 3-mm reformations around the stone circumference at the level of its largest axial diameter (Fig. 2).

UWT measured on 1-mm axial slice (A) and 3-mm axial slice (B)

Ureter diameter (UD) was measured one slice below (UDBS) and above (UDAS) the stone on all reformations (including 1- and 3-mm axial slices) at its widest place. At the same spot as the UD, the average ureter attenuation was measured both above (UAAS) and below (UABS) the stone, manually placing a circular region of interest (ROI) within the ureter covering up to 2/3 of the surface in all reformations (Fig. 3).

UAAS and USAS measured on all reformations: A axial 3 mm; B coronal 3 mm; C axial 1 mm; D sagittal 3 mm. UDAS and UDBS were measured analogously but below the stone

UWT, UDs, UAAS and UABS were measured in a standardised soft-tissue window (L50/W400) by four independent readers, of which two were radiologists (JJ, KS) and two were urologists (MP, PG). The readers were not aware of the spontaneous passage status at the time of measurement. A median value of all readers was used for further analysis. Stone length, UWT and UDs were reported in millimetres to one decimal place.

The presence of hydronephrosis was independently graded as 0–3 (0 = no, 1 = mild, 2 = moderate, 3 = severe) by MP. Renal pelvis diameter (RPD) was measured on 1-mm axial slices between the anterior and posterior wall at its widest place (anteroposterior diameter) by one reader (MP). The presence of a rim sign (i.e. a soft tissue rim around the stone on the axial planes) was assessed by two readers (MP and JJ) independently, and only concordant assessments were taken for further analysis as a positive rim sign.

All radiological examinations were reviewed up to 26 weeks after diagnosis with regard to SSP or intervention. SSP was defined as absence of a stone on follow-up imaging after conservative treatment including analgesics and/or medical expulsive therapy (MET), without any need for surgical intervention, such as shock wave lithotripsy (SWL), ureteroscopy (URS) or drainage (double pigtail catheter or nephropyelostomy tube). Patients who underwent surgical interventions were included in the analysis as failed SSP. However, no standardised protocol for indication to surgical intervention was utilised due to retrospective nature of this study. The decision to intervene surgically was made individually by the responsible urologist based on best clinical practice and current guidelines. Follow-up imaging in the SSP group included intravenous urography (IVU) (n = 69), NCCT (n = 19) or contrast-enhanced CT (CECT) (n = 9). According to the local routine, follow-up imaging was first performed after 4–6 weeks if the patient qualified for conservative treatment. Additional follow-up imaging was usually advised after 4–6 weeks if the stone was present at the first control and there was still no indication of a need for surgical intervention.

As described previously, passage rates in the short and long term were determined [6]. A short-term subgroup was identified, including patients with conservative follow-up imaging or surgical intervention within 28 ± 14 days. Similarly, the long-term outcome group included all the patients who were managed conservatively or with surgical intervention during the period of up to 140 days (20 weeks).

The statistical analysis was performed using IBM SPSS v27.0.1.0 (SPSS Inc., Chicago, IL, USA). Between-group comparisons were performed using Pearson’s chi-square or Fisher’s exact test for quantitative variables and Student’s t-test or the Mann–Whitney U test for continuous variables. Correlations between predictors were assessed with the Pearson or Spearman correlation coefficients. Because of high correlation (|r| > 0.5) and no significant difference in prediction accuracy between UWT, UAAS, UABS, UDAS and UDBS measured on different reformats, only measurements performed on 1-mm axial slices were selected for further analysis and are reported in this article. To detect potential multicollinearity among continuous variables, we calculated the variance inflation factor (VIF) prior to multivariable analysis. VIF values > 5 were considered to indicate a high multicollinearity. Multivariable analysis was conducted with binary logistic regression using SSP as the dependent variable. Receiver operating characteristic (ROC) curves were calculated for stone length separately and in combination with stone impaction variables (UDAS, UDBS and UWT) using probabilities from logistic regression. Furthermore, to determine the reproducibility of the measurements, we investigated both reliability and inter-observer agreement. Reliability was assessed by computing the intra-class correlation coefficient (ICC) using analysis of variance (ANOVA; two-way mixed model with absolute agreement). Values close to 1 indicate high reliability. Agreement plots were created in which the difference between the reader’s measurement and the mean measurement (y-axis) was plotted against the mean measurement [13]. A two-sided p < 0.05 was considered statistically significant.