Local and Overall Deviance R-Squared Measures for Mixtures of Generalized Linear Models.

In generalized linear models (GLMs), measures of lack of fit are typically defined as the deviance between two nested models, and a deviance-based is commonly used to evaluate the fit. In this paper, we extend deviance measures to mixtures of GLMs, whose parameters are estimated by maximum likelihood (ML) via the EM algorithm. Such measures are defined both locally, i.

SIMEU position paper on non-invasive respiratory support in COVID-19 pneumonia.

The rapid worldwide spread of the Coronavirus disease (COVID-19) crisis has put health systems under pressure to a level never experienced before, putting intensive care units in a position to fail to meet an exponentially growing demand. The main clinical feature of the disease is a progressive arterial hypoxemia which rapidly leads to ARDS which makes the use of intensive care and mechanical ventilation almost inevitable. The difficulty of health systems to guarantee a corresponding supply of resources in intensive care, together with the uncertain results reported in the literature with respect to patients who undergo early conventional ventilation, make the search for alternative methods of oxygenation and ventilation and potentially preventive of the need for tracheal intubation, such as non-invasive respiratory support techniques particularly valuable.

Correlation between continuous Positive end-expiratory pressure (PEEP) values and occurrence of Pneumothorax and Pneumomediastinum in SARS-CoV2 patients during non-invasive ventilation with Helmet.

Background: Acute Hypoxemic Respiratory Failure is a common complication of SARS-CoV2 related pneumonia, for which non-invasive ventilation (NIV) with Helmet Continuous Positive Airway Pressure (CPAP) is widely used. The frequency of pneumothorax in SARS-CoV2 was reported in 0.95% of hospitalized patients in 6% of mechanically ventilated patients, and in 1% of a post-mortem case series.

Multivariate generalized hidden Markov regression models with random covariates: Physical exercise in an elderly population.

A time-varying latent variable model is proposed to jointly analyze multivariate mixed-support longitudinal data. The proposal can be viewed as an extension of hidden Markov regression models with fixed covariates (HMRMFCs), which is the state of the art for modelling longitudinal data, with a special focus on the underlying clustering structure. HMRMFCs are inadequate for applications in which a clustering structure can be identified in the distribution of the covariates, as the clustering is independent from the covariates distribution.