Objective:To explore the predictive value of the neutrophil-to-lymphocyte ratio (NLR) combined with blood glucose at admission for a positive blood culture for sepsis.Methods:A single-center retrospective cohort study was conducted. According to the 2016 American Society of Critical Care/European Society of Critical Care Medicine (SCCM/ESICM) and diagnostic criteria for sepsis and septic shock-3.0 (sepsis-3.0), patients with sepsis were admitted to the Emergency Department of Fujian Medical University Union Hospital for more than 24 h from January 2019 to December 2021 were enrolled. Age, gender, sequential organ failure assessment, source of infection, NLR, and blood culture results were recorded. Based on the blood culture results, patients were divided into a blood culture positive group (Gram-positive group, Gram-negative group) and blood culture negative group, and the differences between the groups were compared. The risk factors for a positive blood culture were analyzed using multivariate logistic regression. A receiver operating characteristic analysis was performed for the NLR combined with the blood glucose measurement.Results:A total of 265 patients with sepsis were included, of which 62 were positive in blood culture (15 Gram-positive patients, 37 Gram-negative patients and 10 fungal patients). The positive rate of blood culture was 23.4%. The number of patients with history of diabetes, neutrophil count, procalcitonin, blood glucose, and NLR in the positive blood culture group were significantly higher than those in the negative blood culture group (all P<0.001). Multivariate logistic regression analysis revealed that random admission blood glucose ( OR=1.116, 95% CI: 1.051~1.186, P<0.001) and NLR ( OR=1.039, 95% CI: 1.015~1.064, P=0.001) were independent risk factors for blood culture positivity in sepsis patients. For patients with blood culture positive, and with Gram-negative bacterial bloodstream infections, the AUC of the NLR combined with the admission blood glucose level was 0.819 (95% CI: 0.761-0.877, P<0.001) and 0.871 (95% CI: 0.813-0.928, P<0.001), respectively. Conclusions:The combination of NLR and random admission blood glucose could provide a good predictive value for blood culture positive and gram-negative bacterial bloodstream infections in sepsis patients.

Objective To analyze the clinical value of digital breast tomosynthesis (DBT) compared with digital mammography (DM) and ultrasound for diagnosing non-calcified masses in dense breasts.Methods Images taken with DBT,DM and ultrasound of 1144 patients with non-calcified masses in dense breasts were retrospectively analyzed using breast imaging reporting and data system (BI-RADS).Taking histopathologic results as golden standards,the detection rate and diagnostic accuracy,sensitivity,specificity,false negative and BI-RADS category were evaluated and compared statistically.Results The detection rate of DBT,DM and ultrasound for non-calcified massed in dense breasts was 86.62％ (991/1 144),77.80％ (890/1 144) and 99.65％ (1 140/1 144),respectively (P＜0.05),while the diagnostic accuracy was 83.92％ (960/1 144),75.00％ (858/1 144) and 94.67％ (1 083/1 144),respectively (P＜0.01).The sensitivity of DBT,DM and ultrasound was 89.39％ (312/349),79.93％ (231/289) and 92.70％ (432/466),the specificity was 81.51％ (648/795),73.33％ (627/855) and 96.02％ (651/678),while the false negative rate was 10.60％(37/349),20.07％ (58/289) and 7.30％ (34/466),respectively.No significant difference was found for benign lesions among three examination methods (P=0.75),while there was significant difference for malignant lesions among three examination methods (P＜0.01),and the differences of ultrasonography with DM and DBT,DBT and DM in the for BI-RADS category of malignant lesions were statistically significant (all P＜0.016 7).Conclusion For suspected masses in dense breasts,DBT shows significant advantage than DM,while DBT has the similar advantage compared with ultrasound for the detection and diagnosis of non-calcified masses in dense breasts.