Supporting more than one patient with a single mechanical ventilator: useful last resort or unjustifiable risk?

Predicting intubation for intensive care units patients: A deep learning approach to improve patient management

OBJECTIVE

For patients in the Intensive Care Unit (ICU), the timing of intubation has a significant association with patients' outcomes. However, accurate prediction of the timing of intubation remains an unsolved challenge due to the noisy, sparse, heterogeneous, and unbalanced nature of ICU data. In this study, our objective is to develop a workflow for pre-processing ICU data and to develop a customized deep learning model to predict the need for intubation.

METHODS

To improve the prediction accuracy, we transform the intubation prediction task into a time series classification task. We carefully design a sequence of data pre-processing steps to handle the multimodal noisy data. Firstly, we discretize the sequential data and address missing data using interpolation. Next, we employ a sampling strategy to address data imbalance and standardize the data to facilitate faster model convergence. Furthermore, we employ the feature selection technique and propose an ensemble model to combine features learned by different deep learning models.

RESULTS

The performance is evaluated on Medical Information Mart for Intensive Care (MIMIC)-III, an ICU dataset. Our proposed Deep Feature Fusion method achieves an area under the curve (AUC) of the receiver operating curve (ROC) of 0.8953, surpassing the performance of other deep learning and traditional machine learning models.

CONCLUSION

Our proposed Deep Feature Fusion method proves to be a viable approach for predicting intubation and outperforms other deep learning and classical machine learning models. The study confirms that high-frequency time-varying indicators, particularly Mean Blood Pressure (MeanBP) and peripheral oxygen saturation (SpO2), are significant risk factors for predicting intubation.

Development and assessment of the performance of a shared ventilatory system that uses clinically available components to individualize tidal volumes

OBJECTIVES

To develop and assess a system for shared ventilation using clinically available components to individualize tidal volumes.

DESIGN

Evaluation and in vitro validation study SETTING: Ventilator shortage during the SARS-CoV-2 pandemic.

PARTICIPANTS

The team consisted of physicians, bioengineers, computer programmers, and medical technology professionals.

METHODS

Using clinically available components, a system of ventilation consisting of two ventilatory limbs was assembled and connected to a ventilator. Monitors for each limb were developed using open-source software. Firstly, the effect of altering ventilator settings on tidal volumes delivered to each limb was determined. Secondly, the impact of altering the compliance and resistance of one limb on the tidal volumes delivered to both limbs was analysed. Experiments were repeated three times to determine system variability.

RESULTS

The system permitted accurate and reproducible titration of tidal volumes to each limb over a range of ventilator settings and simulated lung conditions. Alteration of ventilator inspiratory pressures, of respiratory rates, and I:E ratio resulted in very similar tidal volumes delivered to each limb. Alteration of compliance and resistance in one limb resulted in reproducible alterations in tidal volume to that test lung, with little change to tidal volumes in the other lung. All tidal volumes delivered were reproducible.

CONCLUSIONS

We demonstrate the reliability of a shared ventilation system assembled using commonly available clinical components that allows titration of individual tidal volumes. This system may be useful as a strategy of last resort for Covid-19, or other mass casualty situations, where the need for ventilators exceeds supply.

Automatic air volume control system for ventilation of two patients using a single ventilator: a large animal model study

The COVID-19 pandemic outbreak led to a global ventilator shortage. Hence, various strategies for using a single ventilator to support multiple patients have been considered. A device called Ventil previously validated for independent lung ventilation was used in this study to evaluate its usability for shared ventilation. We performed experiments with a total number of 16 animals. Eight pairs of pigs were ventilated by a ventilator or anesthetic machine and by Ventil for up to 27 h. In one experiment, 200 ml of saline was introduced to one subject's lungs to reduce their compliance. The experiments were analyzed in terms of arterial blood gases and respiratory parameters. In addition to the animal study, we performed a series of laboratory experiments with artificial lungs (ALs). The resistance and compliance of one AL (affected) were altered, while the tidal volume (TV) and peak pressure (Ppeak) in the second (unaffected) AL were analyzed. In addition, to assess the risk of transmission of pathogens between AL respiratory tracts, laboratory tests were performed using phantoms of virus particles. The physiological level of analyzed parameters in ventilated animals was maintained, except for CO₂ tension, for which a permissive hypercapnia was indicated. Experiments did not lead to injuries in the animal's lungs except for one subject, as indicated by CT scan analysis. In laboratory experiments, changes in TV and Ppeak in the unaffected AL were less than 11%, except for 2 cases where the TV change was 20%. No cross-contamination was found in simulations of pathogen transmission. We conclude that ventilation using Ventil can be considered safe in patients undergoing deep sedation without spontaneous breathing efforts.

Use of a novel "Split" ventilation system in bench and porcine modeling of acute respiratory distress syndrome

Split ventilation (using a single ventilator to ventilate multiple patients) is technically feasible. However, connecting two patients with acute respiratory distress syndrome (ARDS) and differing lung mechanics to a single ventilator is concerning. This study aimed to: (1) determine functionality of a split ventilation system in benchtop tests, (2) determine whether standard ventilation would be superior to split ventilation in a porcine model of ARDS and (3) assess usability of a split ventilation system with minimal specific training. The functionality of a split ventilation system was assessed using test lungs. The usability of the system was assessed in simulated clinical scenarios. The feasibility of the system to provide modified lung protective ventilation was assessed in a porcine model of ARDS (n = 30). In bench testing a split ventilation system independently ventilated two test lungs under conditions of varying compliance and resistance. In usability tests, a high proportion of naïve operators could assemble and use the system. In the porcine model, modified lung protective ventilation was feasible with split ventilation and produced similar respiratory mechanics, gas exchange and biomarkers of lung injury when compared to standard ventilation. Split ventilation can provide some elements of lung protective ventilation and is feasible in bench testing and an in vivo model of ARDS.

Removal of a Giant Cyst of the Left Ovary from a Pregnant Woman in the First Trimester by Laparoscopic Surgery under Spinal Anesthesia during the COVID-19 Pandemic

This paper reports a case of a 21 year old primigravida at 6 weeks gestation, suffering from important abdominal pain, who was admitted into the medical center with a giant cyst of 28 × 20 cm on her left ovary. A torsion of the ovarian cyst was suspected. Her COVID-19 status was unknown. In view of the emergency of the situation and the COVID-19 pandemic, laparoscopy in spinal anesthesia was performed. The patient remained conscious during the surgical intervention and tolerated it well apart from a slight dyspnea, which was easily eliminated by changing her body position and decreasing the pneumoperitoneum pressure. The ovarian cyst was removed by enlarging the trocar incision. The patient recovered with neither incident nor pregnancy loss. COVID-19-related complications can induce adverse pregnancy outcomes. Under general anesthesia, patients with COVID-19 are at risk of severe pneumonia and of passing their infection to the medical personnel. To avoid such complications in non-specialized centers, laparoscopy should be performed in regional anesthesia. Laparoscopy in spinal anesthesia can be performed safely on pregnant patients by placing them in the proper position, using a low pneumoperitoneum, and monitoring the hemodynamics. During early pregnancy, general anesthesia induces a higher risk of teratogenic effects and of miscarriage.

Sharing Mechanical Ventilator: In Vitro Evaluation of Circuit Cross-Flows and Patient Interactions

During the COVID-19 pandemic, a shortage of mechanical ventilators was reported and ventilator sharing between patients was proposed as an ultimate solution. Two lung simulators were ventilated by one anesthesia machine connected through two respiratory circuits and T-pieces. Five different combinations of compliances (30–50 mL × cmH₂O⁻¹) and resistances (5–20 cmH₂O × L⁻¹ × s⁻¹) were tested. The ventilation setting was: pressure-controlled ventilation, positive end-expiratory pressure 15 cmH₂O, inspiratory pressure 10 cmH₂O, respiratory rate 20 bpm. Pressures and flows from all the circuit sections have been recorded and analyzed. Simulated patients with equal compliance and resistance received similar ventilation. Compliance reduction from 50 to 30 mL × cmH₂O⁻¹ decreased the tidal volume (V_T) by 32% (418 ± 49 vs. 285 ± 17 mL). The resistance increase from 5 to 20 cmH₂O × L⁻¹ × s⁻¹ decreased V_T by 22% (425 ± 69 vs. 331 ± 51 mL). The maximal alveolar pressure was lower at higher compliance and resistance values and decreased linearly with the time constant (r² = 0.80, p < 0.001). The minimum alveolar pressure ranged from 15.5 ± 0.04 to 16.57 ± 0.04 cmH₂O. Cross-flows between the simulated patients have been recorded in all the tested combinations, during both the inspiratory and expiratory phases. The simultaneous ventilation of two patients with one ventilator may be unable to match individual patient’s needs and has a high risk of cross-interference.

2020 Year in Review: Shared Ventilation for COVID-19

COVID-19 resulting from severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) has resulted in a pandemic of respiratory failure previously unencountered. Early in the pandemic, concentrated infections in high-density population cities threatened to overwhelm health systems, and ventilator shortages were predicted. An early proposed solution was the use of shared ventilation, or the use of a single ventilator to support ≥ 2 patients. Spurred by ill-conceived social media posts, the idea spread in the lay press. Prior to 2020, there were 7 publications on this topic. A year later, more than 40 publications have addressed the technical details for shared ventilation, clinical experience with shared ventilation, as well as the numerous limitations and ethics of the technique. This is a review of the literature regarding shared ventilation from peer-reviewed articles published in 2020.

Planning for the Pandemic: A Community Hospital Story

When the Covid 19 pandemic affected New York State, Federal and mostly State, mandates were given to hospitals to prepare for the expected influx of patients. This is a community hospital's planning journey that includes preparing for placing patients, educating caregivers, matching the abilities of the available caregivers with the needs of the patients, securing needed equipment and supplies, and caring for the caregivers. Planning for patient placement resulted in a phased-in guide, accommodating seriously and critically ill affected patients. Education and training were initial and ongoing, rapidly changing as new information became available. Effective care delivery models that focused on team were modified depending on the needs of patients and staff competence. Securing and maintaining equipment and supplies were challenging and caring for the caregivers was a priority. Working as a team, this community hospital developed a road map that was effective in planning for the surge and allowed the hospital to maintain a safe environment for staff and patients who received quality care in difficult time.