China
Africa
Explore chapters and articles related to this topic
Developing and Validating Prognostic Models of Clinical Outcomes
Published in Susan Halabi, Stefan Michiels, Textbook of Clinical Trials in Oncology, 2019
Susan Halabi, Lira Pi, Chen-Yen Lin
Discrimination describes the ability of a prognostic model to distinguish between patients with and without the outcome of interest [4]. Several useful measures can be used to report on the performance of a model [4]. A widely used measure is the c-index, which refers to the proportion of agreement between outcomes and prediction. A value of 0.5 indicates random prediction whereas a value of 1 indicates perfect discrimination. Somers’ D rank correlation is another measure and is related to the c-index and is computed as 2*(c−0.5). Both c and Somers’ D are computed using standard statistical software.
How registered nurses determine their scope of practice: a cross-sectional study
Published in Contemporary Nurse, 2018
Melanie Birks, John Smithson, Daniel Lindsay, Jenny Davis
Data were analysed using IBM SPSS statistics version 23 (SPSS Inc. Chicago, IL). Frequency distributions, means of central tendency and variability were calculated for all relevant variables. Chi-square analyses were used to examine differences in the variables of interest. As two ordinal variables were being used in all analyses, Somers’ D was calculated to estimate the effect size of the association between variables of interest. The magnitude of the Somers’ D effect was based on the values presented in Cohen (1988). To account for the running of multiple comparisons, a Bonferroni adjusted α-value of .0029 (.05/17) was used to test for significance for all Chi-square tests. Any p-values less than .0029 were therefore considered significant. As there were only a few cases with missing values in the data set, listwise deletion of missing values managed missing data (Howell, 2008). Analyses based on age were not performed, as years of practice as a registered nurse was believed to be a more relevant variable. The influence of gender on responses was not explored, given the uneven distribution of males and females in this sample. Area of practice was also not used in the analysis as the research team could not identify a mechanism for grouping related areas of practice into categories that would allow meaningful analysis.
Prediction models for risk of diabetic kidney disease in Chinese patients with type 2 diabetes mellitus
Published in Renal Failure, 2022
Ling Sun, Yu Wu, Rui-Xue Hua, Lu-Xi Zou
Model 2 was established by 19 variables using the backward elimination method, the AUC of model 2 was 0.8946, and its optimal cutoff value was 0.22, with a specificity and sensitivity of 0.797 and 0.818 (Figure 1(B)). Similarly, in this backward elimination model, the hazard ratio (HR) of age, WBC, neutrophils, α-HBDH, albumin, globulin, pre-albumin, TCH, APOB/APOA1 ratio, eGFR, BUN, UA, C-peptide, UACR, SBP, as well as RASI usage exceeded 1.0 with statistical significance, indicated that the DKD risk could positive correlated with these variables. Moreover, the value of Somers’ D arranges from −1 to 1, the Somers’ D values of both models was 0.789 in this study. The larger values of AUC and Somers’ D indicated the stronger predictive power of the model.
Does locally delivered small group continuing medical education (CME) meet the learning needs of rural general practitioners?
Published in Education for Primary Care, 2019
S. Dowling, J. Last, H. Finnegan, K. O’Connor, W. Cullen
Statistical analyses of the data were performed using SPSS version 24. An alpha value of 0.05 was chosen for significance level. The GP’s location (urban, mixed or rural practice) was the main predictor variable of interest. Comparisons between nominal and ordinal variables were made using the Chi-squared test for linear association. The ordinal measures had pre-specified directionality and therefore, the non-symmetric Somers’ delta (Somers’ D) was used as an estimate of the strength and direction of the relationship between GP location and all ordinal outcomes. The strength of association between exclusively nominal variables was assessed using the Cramer’s V statistic.