Explore chapters and articles related to this topic
Meta-Analysis for Continuous Data
Published in Ding-Geng (Din) Chen, Karl E. Peace, Applied Meta-Analysis with R and Stata, 2021
Ding-Geng (Din) Chen, Karl E. Peace
Continuous data are commonly reported as endpoints in clinical trials and other studies. In this chapter, we discuss the meta-analysis methods for this type of data. For continuous data, the typical reported summary statistics are the means and standard deviations (or standard errors) along with the sample sizes for each study.
Hierarchical Models and Longitudinal Data
Published in Gary L. Rosner, Purushottam W. Laud, Wesley O. Johnson, Bayesian Thinking in Biostatistics, 2021
Gary L. Rosner, Purushottam W. Laud, Wesley O. Johnson
Why take a hierarchical approach? The advantages for a meta-analysis include relaxing the assumption of exchangeability of patients across studies carried out at different times and with possibly different populations, the ability to make prediction not just for a future patient but also for a future study [281], and the ability to estimate study-specific effects with shrinkage towards the overall mean, to name but a few. The data frequently take the form of a summary statistic obtained from each study, such as a mean, odds ratio, or relative risk, and the associated estimate of uncertainty, such as the standard deviation or standard error.
STATISTICAL EVALUATION OF THE RISK OF CANCER MORTALITY AMONG INDUSTRIAL POPULATIONS
Published in Richard G. Cornell, Statistical Methods for Cancer Studies, 2020
Michael J. Symons, John D. Taulbee
indicated above, directly adjusted rates also have limitations as a summary statistic because they reflect only the adjusted crude mortality and not the age-specific patterns. Second, and especially relevant to mortality studies focusing on a site-specific cancer, several age groups may have very few or no deaths observed. The age-specific rates then are not stable, since they can fluctuate widely if one death, more or less, is observed in those age groups. For rare causes of death the SMR is available, and therefore used, essentially by default. However, it should be remembered that the SMR is based on the total number of deaths for the cause of interest only and does not use the distribution of these events over the categories of standardization.
Superior rotational stability and lower re-ruptures rate after combined anterolateral and anterior cruciate ligament reconstruction compared to isolated anterior cruciate ligament reconstruction: a 2-year prospective randomized clinical trial
Published in The Physician and Sportsmedicine, 2023
Ștefan Mogos, Dinu Antonescu, Ioan-Cristian Stoica, Riccardo D’Ambrosi
Summary statistics are presented as mean and standard deviation (SD) or absolute frequency and percentage. To assess the differences between ACL and ACL+ALL, a t-test was carried out, whereas for categorical variables, a chi-square test or a Fisher's exact test was performed after testing the distribution of continuous variables. A mixed model was used to test the possible score changes over time because it allows the testing of the covariance structure among repeated measures. The compound symmetry, autoregressive, Toeplitz, Huynh-Feldt, and unstructured covariance structures were tested; the unstructured covariance structure was evaluated as the best covariance structure using the likelihood ratio test and Akaike information criterion. Wilcoxon signed rank test was used in the case of ordinal scores. Bonferroni adjustment was applied for multiple comparisons. A Fisher's exact test was used to assess the difference in the rate of re-ruptures between groups; the correlation among variables was analyzed based on variable distribution using Pearson or Spearman correlations, whichever is appropriate. All the tests were two-sided and a p-value of less than 0.05 was considered statistically significant. Statistical analyses were conducted in R version 4.1.1 and SAS/STAT 9.3.
Efficacy and safety of caplacizumab in the treatment of thrombotic thrombocytopenic purpura: a systematic review and meta-analysis
Published in Expert Review of Hematology, 2023
Jingyi He, Jiaqian Qi, Haohao Han, Xiaoyan Xu, Xueqian Li, Xiaofei Song, Yue Han
We assessed heterogeneity across studies by Q and I2 statistics (25%, 50%, and 75% are consistent with cutoffs for low, moderate, and high heterogeneity, respectively). If I2 > 50%, heterogeneity was considered statistically significant, and we used a random-effects model. Otherwise, a fixed-effects model was used. If studies reported medians and ranges, we calculated means and standard deviations from the available summary statistics. It’s inappropriate to perform funnel plots to determine publication bias as the total number of included studies was only eight (<10). Subgroup analysis was performed to evaluate the potential impacts of study design, setting, sample size, publication year, caplacizumab as first-line (i.e. using caplacizumab as part of the initial treatment of an acute iTTP episode use at the same time or within the very first days after the first TPE) or second-line (i.e. using caplacizumab after an exacerbation or refractoriness). The published meta-regressions had a p-value of<0.05, indicating the presence of heterogeneity. Sensitivity analysis was conducted further to assess the stability and reliability of the results we performed by deletion of every single investigation. All statistical analyses were performed using Review Manager version 5.4 and R-4.0.3 software.
Burden of asthma-like symptoms and a lack of recognition of asthma in Vietnamese children
Published in Journal of Asthma, 2023
Hong H. T. C. Le, Pham Le An, Nguyen Nhu Vinh, Robert S. Ware, Dung Phung, Phong K. Thai, Sarath Ranganathan, Tran Ngoc Dang, Phan Hoang Thuy Dung, Do Thi Hoai Thuong, Hai Phung, To Thi Hien, Peter D. Sly
Summary statistics are described as mean and standard deviation for continuous data and as frequency and percentage for categorical data. The association between sex and location (inner area/suburban area) with symptoms was investigated using logistic regression with robust standard errors to account for the possible non-independence of responses from children who attended the same school. First univariable analyses were undertaken, and then multivariable analyses, where models were adjusted for the potentially confounding variables sex, age, location, child’s smoking status, tobacco smoke exposure, food allergy and whether there are pets in the child’s residence. The results of these analyses are expressed as adjusted odds ratios (aOR) and 95% confidence intervals (95%CI). All statistical analyses were performed using Stata version 16.0 (StataCorp, College Station, TX, USA).