Skip to main content
. 2019 Feb 5;123(6):959–967. doi: 10.1111/bju.14673

Table 5.

Multivariable regression analyses for the prediction of high‐grade cancer in the follow‐up of 321 patients with previous non‐muscle‐invasive bladder cancer

Multivariable logistic regression analysis for positive Bladder EpiCheck Multivariable logistic regression analysis for continuous Bladder EpiScore
OR 95% CI P OR 95% CI P
Last pathological stage (Ta reference category)
Tis 3.96 0.42–40.7 0.23 5.44 0.49–69.0 0.99
T1 3.43 0.63–23.1 0.17 5.62 0.88–46.9 0.17
High grade 1.00 0.18–4.97 0.99 0.64 0.09–3.54 0.08
Age (continuous) 0.98 0.93–1.05 0.63 0.97 0.91–1.04 0.62
Female gender 0.77 0.14–3.26 0.73 0.93 0.16–4.36 0.39
Time from last recurrence to urine collection 0.82 0.66–0.96 0.03 0.82 0.65–0.99 0.93
Ongoing intravesical therapy (never reference category)
No 0.95 0.22–4.09 0.94 0.82 0.06–2.33 0.81
Yes 0.40 0.06–2.07 0.29 0.40 0.06–2.33 0.32
AUC 72.7% (95% CI 62.7–76.6)
Positive EpiCheck 78.3 19.2–547 <0.001 EpiScore (continuous) 1.08 1.05–1.12 <0.001
AUC 94.9% (95% CI 88.3–96.8) AUC 96.1% (95% CI 90.9–97.9)

AUC, area under the curve; OR, odds ratio.