Risk factors for diagnosis of Escherichia coli infection after flexible ureteroscope holmium laser lithotripsy by imaging information technology under Nomogram mathematical model

• Main Authors: Yuelong Zhang, Qi Zhang, Jia Lv, Dahong Zhang, Meriem Kaddouri, Saeed Hameed Kurdi Al Dulaimi
• Format: Article
• Language: English
• Published: Elsevier 2021-06-01
• Series: Results in Physics
• Subjects: Nomogram mathematical model; Imaging information technology; Flexible ureteroscope holmium laser lithotripsy; Escherichia coli
• Online Access: http://www.sciencedirect.com/science/article/pii/S2211379721004599

The evaluation value of Nomogram mathematical model was investigated for risk factors of Escherichia coli infection after flexible ureteroscope holmium laser lithotripsy (f-URL) under the diagnosis of imaging information technology. A total of 124 cases with upper urinary tract calculi (UUTC) were selected as the research objects, and they were rolled into an infection group (41 cases) and a control group (83 cases) according to whether Escherichia coli infection occurred after the surgery. The difference of surgical indicators between the two groups was compared, and the Nomogram mathematical prediction model was established based on Logistic regression risk factors. Besides, the predictive ability of the Nomogram mathematical model was analyzed with indicators such as the integrated discrimination improvement (IDI), the area under the curve (AUC), and GiViTI calibration curve band. The results found that age, diabetes, stone size, surgical time, and antibiotic use time were all risk factors for Escherichia coli infection after surgery. The AUC of the different risk factors of the Nomogram mathematical model was between 0.752 and 0.814, the sensitivity was 57.45%, the specificity was 97.96%, the C-index was 0.734 (95% confidence interval (CI): 0.631–0.837). Both the 80% and 95% CI regions of the Giviti calibration curve band did not cross the 45° diagonal bisector, and P = 0.518. Moreover, the Nomogram model showed an increase of 8.9% (95% CI: 0.061–1.1874 and P = 0.043) compared with the Logistic regression analysis model. Therefore, these results indicated that the Nomogram mathematical model can markedly improve the prediction ability of Escherichia coli infection risk after f-URL, which had great value in the assessment and early diagnosis and treatment of Escherichia coli infection after f-URL.