Intravenous thrombolysis in wake-up stroke: real-world data from the Austrian Stroke Unit Registry.

BACKGROUND AND PURPOSE: Data on real-world experience with intravenous thrombolysis (IV tPA) in wake-up stroke (WUS) are limited. The aim of this study was to examine the efficacy and safety of IV tPA in patients with WUS included in the Austrian Stroke Unit Registry. METHODS: Data from a large nationwide stroke unit registry including initial stroke severity, vascular risk factors, comorbidities, treatment with IV tPA, symptomatic intracerebral haemorrhage (sICH) and functional outcome were extracted and analysed. Patients with WUS were compared with patients with known-onset stroke (KOS) regarding the frequency of IV tPA treatment, neurological improvement (National Institutes of Health Stroke Scale score ≥4), sICH and 3-month functional outcome by modified Rankin Scale score using standard statistical tests. RESULTS: A total of 107 895 stroke patients entered the analysis, including 12 534 with WUS and 91 899 with KOS. Altogether, 904 (7.2%) patients with WUS received IV tPA as compared with 16 694 (18.2%) patients with KOS. Patients with WUS who received IV tPA treatment had twofold higher initial National Institutes of Health Stroke Scale score (median 8 vs. median 4) as compared with patients with KOS. There was no statistical difference in functional outcome by modified Rankin Scale score 0-1 at 3 months between patients with WUS and patients with KOS treated with IV tPA (adjusted odds ratio, 1.08; 95% confidence interval, 0.9-1.31). Also, the rate of sICH did not differ (4.1% vs. 4%, P = 0.852). CONCLUSIONS: In this large non-randomized comparison, the safety and efficacy of IV tPA in patients with WUS in the real-world setting seems to be comparable to patients with KOS.

Biological assessment of robust noise models in microarray data analysis.

MOTIVATION: Although several recently proposed analysis packages for microarray data can cope with heavy-tailed noise, many applications rely on Gaussian assumptions. Gaussian noise models foster computational efficiency. This comes, however, at the expense of increased sensitivity to outlying observations. Assessing potential insufficiencies of Gaussian noise in microarray data analysis is thus important and of general interest. RESULTS: We propose to this end assessing different noise models on a large number of microarray experiments. The goodness of fit of noise models is quantified by a hierarchical Bayesian analysis of variance model, which predicts normalized expression values as a mixture of a Gaussian density and t-distributions with adjustable degrees of freedom. Inference of differentially expressed genes is taken into consideration at a second mixing level. For attaining far reaching validity, our investigations cover a wide range of analysis platforms and experimental settings. As the most striking result, we find irrespective of the chosen preprocessing and normalization method in all experiments that a heavy-tailed noise model is a better fit than a simple Gaussian. Further investigations revealed that an appropriate choice of noise model has a considerable influence on biological interpretations drawn at the level of inferred genes and gene ontology terms. We conclude from our investigation that neglecting the over dispersed noise in microarray data can mislead scientific discovery and suggest that the convenience of Gaussian-based modelling should be replaced by non-parametric approaches or other methods that account for heavy-tailed noise.