|
The details of statistical analysis are necessary, and include: (1) sample size estimation and follow- up/data collection; (2) softwares and methods for data entry and management, (3) statistical analysis softwares and schemes used; (4) expression and description of data: (x ± s) describes quantitative data obeying normal distribution; Median (P25，P75) describes quantitative data with skewed distribution; vertical and horizontal headings should be reasonably arranged and the meaning of data should be clearly present in tables; the type of statistical figures used should match the material properties, and the scale value of axes should meet mathematics principles; the relative number should be reported with the absolute number, for example, 27% (68/250); when the denominator is less than 20, fraction is recommended to use instead of percentage, for example 5/7, and the difference between relative number of intensity (percentage) and constituent ratio (the proportion) should cause more attention; (5) selection of statistical analysis methods: for quantitative data, appropriate methods should be selected according to the design of studies , available data, and purpose of analysis, blind use of t test or single factor analysis of variance should be avoided; for qualitative data, appropriate methods should be selected based on the design of studies, nature of qualitative variables, frequency distribution, and purpose of analysis, blind use of chi-square test should be avoided. For regression analysis, appropriate analysis methods should be selected according to professional knowledge and scatter diagrams, and avoid indiscriminate use of linear regression analysis. The data from multiple measurements at various time points on one indicator from a same subject, which means repeated measurement data, should not be simply processed, and repeated- measures analysis of variance is recommended if necessary; multivariate statistical analysis on the basis of univariate analysis should be applied for data with multiple factors or indicators to realize comprehensive and reasonable evaluation of associations and interactions among the multiple factors. (6) statistical results: both statistical values (for example, t = 3.45, χ2 = 4.68, F = 6.79) and P values should be provided; significant results and 95% confidence intervals are necessary when talking about a population parameter (for example, overall mean or overall rate). Any research should follow the principles of medical ethics. When reporting experiments on human subjects or animals, authors should indicate whether the procedures followed were in accordance with the ethical standards of the responsible committee on human experimentation (institutional, regional or national) and with the Helsinki Declaration, and provide the hard or scanned copy of the approval document from the Institutional Review Board (IRB) (the ID of the approval document should be present in the content) and copy of informed consent signed by the subjects or their relatives. When researches were conducted on animals, authors should indicate whether the institutional and national guidelines for the care and use of laboratory animals were followed. Animal welfare and experimental procedures should be carried out strictly in accordance with the Guide for Care and Use of Laboratory Animals (National Research Council of USA, 1996). The Journal would preferentially consider publishing clinical trials that have been registered with a clinical trial registry certified by WHO. A registration number, which is a global unique registration number, should be provided at the end of the Abstract with registration authority and number under the title of 'Trial registration'. Abstracts of original articles on randomized controlled clinical trials should contain the essential items included in the latest version of CONSORT (Consolidated Standards of Reporting Trials, http://www.)
|
|
|
