22 October 2018 in JMIR mHealth and uHealth
Mobile and tablet apps, ubiquitous and pervasive computing, wearable computing and domotics for health.
2 June 2017 in Pharmaceutical Statistics
OCOn behalf of the COMBACTE‐Net consortium
On behalf of the COMBACTE‐Net consortium
1 July 2014 in The Journal of Hand Surgery
16 October 2018 in Biometrical Journal
31 July 2018 in European Journal of Cardio-Thoracic Surgery
This systemic review of the literature and meta-analysis examined the current state of the evidence in long-term outcomes for and/or against aortic valve reimplantation (RAV) versus composite valve graft (CVG) intervention in patients with an acute type A dissection. Descriptive statistics were used to summarize the baseline characteristics of patients across studies. A random-effects metaregression was performed across study arms with logit-transformed proportions weighted by the study size for each of these outcomes. The results are presented as odds ratios with the RAV procedure as compared to the CVG procedure, including 95% confidence intervals (CIs) and P-values. Further outcomes are summarized with medians, interquartile ranges and the range and number of patients at risk. A total of 27 retrospective studies that included a combined 3058 patients were analysed. In-hospital mortality was in favour of the RAV procedure, which was 2% vs 8% for the CVG procedure. Survival rate at midterm was 98.8% (95% CI 91.7-100%) for RAV and 81.3% (CI 78.5-83.9%) for CVG. Freedom from valve-related reintervention was 100% (CI 93.7-100%) for RAV and 94.6% (CI 86.7-99.1%) for CVG. For an acute type A aortic dissection in the mid-term period, RAV provides a superior outcome over CVG, both in terms of aortic-valve-related reintervention and survival rate.
1 May 2015 in Statistical Science
Bayesian model selection poses two main challenges: the specification of parameter priors for all models, and the computation of the resulting Bayes factors between models. There is now a large literature on automatic and objective parameter priors in the linear model. One important class are $g$-priors, which were recently extended from linear to generalized linear models (GLMs). We show that the resulting Bayes factors can be approximated by test-based Bayes factors (Johnson [ Scand. J. Stat. 35 (2008) 354–368]) using the deviance statistics of the models. To estimate the hyperparameter $g$, we propose empirical and fully Bayes approaches and link the former to minimum Bayes factors and shrinkage estimates from the literature. Furthermore, we describe how to approximate the corresponding posterior distribution of the regression coefficients based on the standard GLM output. We illustrate the approach with the development of a clinical prediction model for 30-day survival in the GUSTO-I trial using logistic regression.