Search

Rationale:Early mobilization and physical rehabilitation improve functional outcomes and are essential to high quality critical care. Despite its importance, it is common for rehabilitation to be deferred in the critically ill due to a variety of barriers, including infection with SARS-CoV-2. We present a single academic center's experience providing physical and occupational therapy to critically ill patients infected with SARS-CoV-2. Methods:All patients with Coronavirus Disease 2019 (COVID-19) associated illness admitted to the intensive care unit (ICU) from March 1st to July 31st, 2020 were identified in this retrospective chart review. Patients who received at least one therapy treatment session were included in the study. Results:Three-hundred and seventy-nine physical and occupational therapy sessions were conducted with 116 patients. The majority (85%) of patients were admitted to the ICU for hypoxemic respiratory failure. The median number of treatment sessions during ICU admission per patient was 2, (IQR: 1-4). The median time from ICU admission to first PT session was 4 days (IQR, 3-5). The median percentage of ICU days with physical and occupational therapy treatment was 33% (IQR, 21-50). The median session length was 25 minutes (IQR, 25-30min). Sitting was achieved in 353 sessions, (93%) standing was achieved in 261 sessions (69%), walking was achieved in 185 sessions (48%), and sitting in the bedside chair 118 times (31%).Patients with respiratory failure completed therapy sessions while receiving mechanical ventilation (21% of sessions), high flow nasal cannula (45% of sessions), non-invasive positive pressure ventilation by helmet and facemask (7% of sessions), and ECMO (12% of sessions). Patients requiring vasoactive medications (4%) and continuous renal replacement therapy (6%) were also treated by physical and occupational therapy. Delirium, determined by confusion assessment method (CAM-ICU), was frequently encountered by the physical and occupational therapy teams and was not an absolute barrier (32%) (Table 1). Discharge destinations included: home (n=57, 61%), acute rehabilitation units (n=16, 17%), long term acute care hospitals (n=9, 10%), sub-acute care centers (n=8, 8%), and skilled nursing facilities (n=4, 4%). No members of the therapy team were diagnosed with SARS-CoV-2 during the study period. Conclusions:This report demonstrates the feasibility of conducting physical and occupational therapy in COVID-19 specific ICUs. Providing therapy services appeared to be safe for patients and members of the therapy team, as adverse events were rare and no therapist was diagnosed with COVID-19.

Rationale:Patients with COVID-19 frequently develop severe respiratory disease and may require invasive mechanical ventilation. A study of primarily white patients intubated for COVID-19 associated respiratory failure found predictors of 28-day mortality to be respiratory system compliance, age, tidal volume, arterial pH and heart rate. Little is known about the outcomes of minority populations with severe COVID-19 pneumonia. Therefore, we present an analysis of the predictors of mortality in a group of primarily African American patients with COVID-19 associated respiratory failure. Methods:All adult patients admitted to the University of Chicago COVID-19 intensive care unit receiving invasive mechanical ventilation between March 1st and June 31st, 2020 were identified. Patients were included in the study if they had at least one recorded measure of plateau airway pressure while receiving volume-controlled ventilation allowing determination of driving pressure and lung compliance. Univariable analysis was conducted comparing survivors with those who died in-hospital followed by construction of a multivariable logistic regression model predicting in-hospital mortality based on significant factors from univariable analysis, excluding colinear variables. Results:Eighty-five patients were included in this retrospective study. Patients were primarily African American (n=73, 86%). Among all study patients, median tidal volume was 6.0 cc/kg ideal body weight (IQR 5.8-6.2), PEEP was 8 cm H2O (IQR 5.0-10), and driving pressure was 14 cm H2O (IQR 11-16). Median respiratory system compliance was 27 ml/cm H2O (IQR 21-34). Salvage therapies for refractory hypoxemia in the cohort included prone positioning (27%), paralysis (27%), inhaled pulmonary vasodilators (19%), and extracorporeal membrane oxygenation (1%). In the multivariable logistic regression model, age (OR 1.077, 95% CI 1.031 to 1.125, p=0.001) and driving pressure (OR 1.174, 95% CI 1.009 to 1.366, p=0.038) were found to be independent predictors of mortality. Conclusions:In a predominantly African American patient population with COVID-19 pneumonia requiring invasive mechanical ventilation, higher driving pressure was predictive of overall mortality. These finding are consistent with the work of Botta et al (2020), who demonstrated reduced lung compliance was predictive of mortality among a largely white group of patients with severe COVID-19 pneumonia. While minority populations infected with COVID-19 have been found worse outcomes, early lung mechanics appear to be comparable to white patients. These findings support that higher driving pressures and low lung compliance are indicative of serious lung injury which may lead to death.

This paper presents neural and hybrid (symbolic and subsymbolic) applications downloaded on the distributed computer architecture ArMenX. This machine is articulated around a ring of FPGAs acting as routing resources as well as fine grain computing resources and thus giving great flexibility. More coarse grain computing resources-Transputer and DSP-tightly coupled via FPGAs give a large application spectrum to the machine, making it possible to implement heterogeneous algorithms efficiently involving both low level (computing intensive) and high level (control intensive) tasks. We first introduce the ArMenX project and the main architecture features. Then, after giving details on the computing of propagation and back-propagation of the multi-layer perceptron on ArMenX, we will focus on a handwritten digit (issued from a zip code data base) recognition application. An original and efficient method, involving three neural networks, is developed. The first two neural networks deal with the 'reading process', and the last neural network, which learned to write, helps to make decisions on the first two network outputs, when they are not confident. Before concluding, the paper presents the work of integration of ArMenX into a high level programming environment, designed to make it easier to take advantage of the architecture flexibility.

The photoelectron energy corresponding to the Fermi level of an investigated specimen and the onset energy for detection of secondary electrons are two characteristic energy values that can be used for work-function evaluation. Secondary-electron onset measurements are usually carried out by applying an acceleration voltage between the specimen and the spectrometer entrance, giving a structured secondary-electron spectrum showing three types of secondary electrons: those from the specimen, those from the entrance-slit system, and those from the X-ray tube window. For the latter electrons the electrical field built up in the sample chamber acts as an energy analyzer, imaging a discrete energy interval onto the spectrometer entrance slit. The variation of energy and intensity of these “ghost-lines” is discussed as a function of the applied acceleration voltage and the specimen tilting angle, leading to optimized experimental parameters for recording of secondary-electron spectra.