ABSTRACT
OBJECTIVE@#To investigate the clinical value of hemoglobin to serum creatinine ratio (Hb/SCr) combined with blood uric acid (SUA) in predicting in-hospital mortality after emergency percutaneous coronary intervention (PCI) in patients with acute myocardial infarction (AMI).@*METHODS@#The clinical data of AMI patients who underwent emergency PCI in the First Affiliated Hospital of Kangda College of Nanjing Medical University from January 2017 to December 2021 were retrospectively analyzed. The general information, underlying medical history, blood routine, liver and kidney function, blood coagulation routine, SUA and other indicators were collected from patients. The primary composite endpoint was defined as in-hospital death, including in-hospital all-cause death during PCI and 15-day post-procedure hospitalization. Multivariate Logistic regression was used to analyze factors associated with in-hospital death after emergency PCI in patients with AMI. Multivariate Logistic regression was used to analyze the independent related factors and construct a risk prediction model. The Hosmer-Lemeshow method and receiver operator characteristic curve (ROC curve) were used to test the goodness of fit and predictive effect of the model and correlates, respectively.@*RESULTS@#A total of 1 976 patients were enrolled, 92 died in hospital and 1 884 survived. SUA was higher in the death group than that in the survival group (μmol/L: 476.88±132.04 vs. 354.87±105.31, P < 0.01), and the Hb/SCr ratio was significantly lower than that in the survival group (13.84±5.48 vs.19.20±5.74, P < 0.01). Pearson analysis showed a linear negative correlation between SUA and Hb/SCr ratio (r = -0.502, P < 0.01). Logistic regression risk model analysis finally included age [odds ratio (OR) = 0.916], Hb/SCr ratio (OR = 0.182), white blood cell count (WBC, OR = 2.733), C-reactive protein (CRP, OR = 3.611), SUA (OR = 4.667), blood glucose (Glu, OR = 2.726), homocysteine (Hcy, OR = 2.688) 7 factors to construct a risk prediction model, which were independent correlation factors for in-hospital death in AMI patients after emergency PCI (all P < 0.05). Hosmer-Lemeshow test verified the fitting effect of the model, and the result showed P = 0.447. The area under the ROC curve (AUC) of the model for predicting in-hospital death in AMI patients after emergency PCI was 0.764 [95% confidence interval (95%CI) was 0.712-0.816, P = 0.001]. When the cut-off value was 0.565 8, the sensitivity was 70.7%, the specificity was 70.2%, and the Yoden index was 0.410. When Hb/SCr ratio+SUA, SUA, Hb/SCr ratio, Hb and SCr were used to predict in-hospital death in AMI patients after emergency PCI, the AUC of Hb/SCr ratio+SUA was the largest, which was 0.810. When the optimal cut-off value was -0.847, the sensitivity was 77.7%, the specificity was 74.5%, and the Youden index was 0.522.@*CONCLUSIONS@#Age, SUA, Hb/SCr ratio, WBC, CRP, Glu, and Hcy are independent risk factors for in-hospital death after emergency PCI in AMI patients. The lower the Hb/SCr ratio and the higher the SUA at admission, the higher the risk of in-hospital death after emergency PCI in AMI patients. Hb/SCr ratio combined with SUA has a higher predictive value for in-hospital death after emergency PCI in AMI patients than single index, which is helpful for early identification of high-risk patients.
Subject(s)
Humans , Hospital Mortality , Uric Acid , Creatinine , Percutaneous Coronary Intervention , Retrospective Studies , Myocardial Infarction/therapy , PrognosisABSTRACT
Objective To build the index about the training effect of oncology specialist nurse and test the reliability and validity.Methods By consulting literature and the methods of Delphi,which was consulted by the 9 experts and the two rounds of 35 experts,the original index was estabhshed.The correlation coefficient and validity of the index was evaluated.Results A total of the index was consist of 106 first-l,second-land third-llevel items.Questionnaire recovery rate was 100％ (35/35).The level of authority was 0.816.Expert opinion concentration in various purposes mean variation coefficient was less than 0.244 2.The coordination level of expertise was 0.302.The overall reliability Cronbach α was 0.963.Adopting the method of double variable Spearman calculated the correlation coefficient was above 0.428.Cronbach a of second-and third-level indexes were 0.512-0.895.The second-land third-level indexes correlation coefficient were 0.428-0.931.Conclusions The index about the training effect of oncology specialist nurse has the high authority and representative.The reliability and validity of the index are better,which can be used to evaluate the efficacy of the evaluation about the training of oncology specialist nurse.
ABSTRACT
A design of ECG monitor was presented in this paper. It is based on the latest MCU and BLE4.0 technologies and can interact with multi-platform smart devices with extra low power consumption. Besides, a clinical expansion part can realize functions including displaying the real-time ECG and heart rate curve, reading abnormal ECG signals stored in the monitor, and setting alarm threshold. These functions are suitable for follow-up use.
Subject(s)
Humans , Electrocardiography , Equipment Design , Heart Rate , Signal Processing, Computer-AssistedABSTRACT
Objective To evaluate the value of ultrasonic quantitative method in the diagnosis of liver fibrosis in chronic hepatitis B (CHB) patients. Methods Ultrasonography was performed in 186 CHB patients who underwent liver biopsies. Fifteen indices including liver capsule thickness and fourteen texture parameters of gray level co-occurrence matrix were extracted from standard sonograms and compared with fibrosis stages by histopathology. The status of liver fibrosis was divided into five stages from S0 to S4 by histopathology based on the disease severity. ANOVA and Spearman correlation analysis were applied to analyze the differences and relationships between these indices and pathological stage, respectively. Then discriminant analysis models were established based on the indices for quantitative diagnosis of liver fibrosis. Results Among the fifteen indices, including liver capsule thickness, only the variance (F=0. 55, r=0. 06; both P＞0. 05), sum average (F=0.61, both r=0.05 ; P＞0.05), sum entropy (F=1.68, r=0.09; both P≥0.05) and entropy (F=1.39,r=0.12; both P＞0.05) were not significantly associated with the stages and not manifested linear correlation. Using biopsy results as gold standard, the correct rank rate of discriminant analysis model analysis in the patients staged from S0 to S4 were 80. 0%, 64. 9%, 61.3%, 74. 1% and 80.6 %, respectively. There were 73.1% of cross-validated cases who were accurately classified by the model analysis. The sensitivity, specificity and accuracy in patients with stage ≥ 1 were 97. 6%,80.0% and 91.9%, respectively; those in patients with stage≥2 were 92.1%, 89.7% and 90.9%,respectively; those in patients with stage≥3 were 94.8%, 96.1% and 95.7%; and those in patients with stage 4 were 80. 6%, 97.4 % and 94.6%, respectively. When considered S0 as no fibrosis, S1 as mild fibrosis, S2 and S3 as moderate to severe fibrosis and S4 as early cirrhosis, the consistence rates between discriminant analysis model and biopsy result were 81.7%, 78. 4%, 56. 9% and 90.3%,respectively. There were 74.7% of cross-validated cases who were correctly classified. The sensitivity, specificity and accuracy of the models for determining the fibrosis severity in patients≥mild fibrosis were 97.6%, 81.7% and 92.5%, respectively; those in patients ≥ moderate to severe fibrosis were 83. 1%, 94.8% and 89.2%, respectively; those in patients with early cirrhosis were 90.3%, 93.5% and 93.0%, respectively. Conclusion As a novel and noninvasive method, ultrasonic texture analysis could quantitatively determine liver fibrosis in CHB patients and is worthy of further investigation.