Editor—Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) transmission is thought to be through fomites, droplets, and droplet nuclei (aerosols).1van Doremalen N. Bushmaker T. Morris D.H. et al.Aerosol and surface stability of SARS-CoV-2 as compared with SARS-CoV-1.N Engl J Med. 2020; 382: 1564-1567Crossref PubMed Scopus (6034) Google Scholar Aerosol-generating medical procedures are commonly performed and are associated with increased risk of infection of healthcare workers.2Tran K. Cimon K. Severn M. Pessoa-Silva C.L. Conly J. Aerosol generating procedures and risk of transmission of acute respiratory infections to healthcare workers: a systematic review.PLoS One. 2012; 7e35797Crossref PubMed Scopus (1223) Google Scholar Some clinicians are using barriers such as transparent plastics and Plexiglas boxes to reduce aerosol spread.3Matava C.T. Yu J. Denning S. Clear plastic drapes may be effective at limiting aerosolization and droplet spray during extubation: implications for COVID-19.Can J Anaesth. 2020; 67: 902-904Crossref PubMed Scopus (122) Google Scholar, 4Cubillos J. Querney J. Rankin A. Moore J. Armstrong K. A multipurpose portable negative air flow isolation chamber for aerosol-generating procedures during the COVID-19 pandemic.Br J Anaesth. 2020; 125: e179-e181Abstract Full Text Full Text PDF PubMed Scopus (46) Google Scholar, 5Lang A.L. Shaw K.M. Lozano R. Wang J. Effectiveness of a negative-pressure patient isolation hood shown using particle count.Br J Anaesth. 2020; (Advance Access published on May 15)https://doi.org/10.1016/j.bja.2020.05.002Abstract Full Text Full Text PDF PubMed Scopus (13) Google Scholar, 6Begley J.L. Lavery K.E. Nickson C.P. Brewster D.J. The aerosol box for intubation in COVID-19 patients: an in-situ simulation crossover study.Anaesthesia. 2020; 75: 1014-1021Crossref PubMed Scopus (166) Google Scholar, 7Yang S.S. Zhang M. Chong J.J.R. Comparison of three tracheal intubation methods for reducing droplet spread for use in COVID-19 patients.Br J Anaesth. 2020; 125: e190-e191Abstract Full Text Full Text PDF PubMed Scopus (13) Google Scholar However, these barriers may limit access to the patient and mobility of the clinician.8Kovatsis P.G. Matava C.T. Peyton J.M. More on barrier enclosure during endotracheal intubation.N Engl J Med. 2020; 382: e69Crossref PubMed Scopus (23) Google Scholar An alternative to barriers that may reduce aerosol spread is directed high flow air extraction. A high flow air extractor combines high flow suction and a high-efficiency particulate (HEPA) filter. We conducted a study to determine if high flow air extraction reduces aerosol exposure of clinicians. We designed an experimental model that determined the efficacy of removal of particles similar in size to human aerosols. We used two particles to simulate aerosols, essential oil particles ranging in size from 1 nm to 1 μm, and ISO 12103-1 A1 Ultrafine test dust (Powder Technologies Inc., Arden Hills, MN, USA) ranging in size from 1 to 20 μm. We simulated human breathing using an essential oil diffuser as a continuous aerosol source. Human cough aerosols range in size from 0.58 to 5.42 μm with 80% in the 0.74–2.12 μm range.9Yang S. Lee G.W. Chen C.M. Wu C.C. Yu K.P. The size and concentration of droplets generated by coughing in human subjects.J Aerosol Med. 2007; 20: 484-494Crossref PubMed Scopus (321) Google Scholar For coughing experiments, a manikin (Electripod ET/J10 Tracheal Intubation model; TUQI, Shanghai, China) was used (Supplementary 1a, b). We applied 500 mg of A1 Ultrafine test dust to the oropharynx and distal trachea of the manikin and simulated a cough using a medical air gun connected to the distal trachea and fired for 0.4 s. The researchers placed their hand 2–3 cm from the mouth of the manikin to simulate a covered cough. The high-flow air extractor Epurair HA-500 (Industrie Orkan Inc., Montreal, Quebec, Canada) was placed 25–30 cm above the manikin's head. We quantified aerosols with the following sensors (Supplementary 1a, b). Two dust aerosol calibrated DustTrak DRX (TSI, Shoreview, MN, USA) units using four chambers placed near the source and the clinician's head. Two wide-range aerosol spectrometers, miniWRAS 1371 (Grimm Aerosol Technik, Ainring, Germany) each with 41 bins and calibrated to an oil aerosol were similarly placed. To determine the vertical and horizontal variation in concentrations, 10 DC1700 optical particle monitors (Dylos, Riverside, CA, USA) were placed at predetermined positions (Supplementary 1a, b). To eliminate inter-monitor variation, monitors were co-located for 10 min after the experiments and reported concentrations corrected by the deviation from the mean concentration of each monitor. The high-flow air extractor is a portable high efficiency filtration unit allowing up to 235 L s−1 (500 ft3 min−1) that can be used to transform a regular room into a negative pressure room. It contains a HEPA filter that removes 99.97% of all airborne pathogens of 0.3 μm or greater. The filtered air can be adapted to an existing exhaust system or vented outside. We operated the device with a calibrated booster fan to maintain a continuously measured flow of 142 L min−1 for the experiments (Supplementary 2). Each experiment was completed in triplicate, and mean concentration values were used for analysis. Our primary outcome was to determine the reduction of aerosols at the source. A 99% reduction in the aerosol concentration near the source would be consistent with the Centre for Disease Prevention and Control's (CDC) requirements for air exchanges between patient encounters.10Jensen P.A. Lambert L.A. Iademarco M.F. Ridzon R. Guidelines for preventing the transmission of Mycobacterium tuberculosis in health-care settings.MMWR Recomm Rep. 2005; 54 (2005): 1-141PubMed Google Scholar Secondary outcomes included reduction of aerosol concentrations at the level of the clinician's head with the high-flow air extractor 'on' during a cough and an obstructed cough. The effectiveness, H, was calculated by subtracting the ratio of 'high-flow air extractor on' to 'high-flow air extractor off' mean particle concentration measured by each aerosol quantification device from unity. The high-flow extractor device was 99% effective at removing aerosols near the source, resulting in no levels detected at the clinician's head (Fig. 1a and Supplementary 3 online video). During an uncovered cough, the high-flow extractor had a 97% effectiveness in reducing the aerosols detected near the clinician's head (Fig. 1b). In these first two scenarios, aerosols were effectively removed at source and did not contaminate the room or reach the clinician's head. However, when the cough was covered by the provider's hand there was only a 52% reduction in aerosols detected at the clinician's head; the absolute concentration was very low because of less aerosols reaching the clinician's head as a result of covering the cough (Fig. 1c). The covered cough resulted in a higher concentration of aerosols at sensors placed lateral to the patient (Supplementary 4). This was likely because aerosols were diverted away from the device's intake but subsequently reached the clinician's head. The effectiveness of the high-flow air extractor was high for larger particles (>1 μm) emitted from the simulated cough, and generally low for small particles (<1 μm) (Supplementary 5a, b). The following is the supplementary data related to this article:eyJraWQiOiI4ZjUxYWNhY2IzYjhiNjNlNzFlYmIzYWFmYTU5NmZmYyIsImFsZyI6IlJTMjU2In0.eyJzdWIiOiI5NjM3NTIyMDVkYTllYzk0ODNlMTJiNjM4YmNhMGNiYiIsImtpZCI6IjhmNTFhY2FjYjNiOGI2M2U3MWViYjNhYWZhNTk2ZmZjIiwiZXhwIjoxNjg4MzY0Njk5fQ.khN1bVz7GnwLDGDL6k63nf7VI5RON9zuyc6-Wp_FIrCF9HgjkfIPJiZFX7UaHyFrDBLR2p190QHEyRrhHZPb74UqfWuGYh9neTl1bykIoYi_FBTMzGCG_b9A3xXZFovNXQGcBWOUOjvqxJrwToizfL5Ea7pbk0F9Or0BwunarqPsF1K5InfKxE87fQTuJCvMc8vRRwuSF0PvY8OV-lPr27Sg8HUFO0f8-PS6SRSiNlROyW5sAUr_g1JNxs6LaFARl8DQz_NUMvckIyIJmSvZAugKnYQrgIomGDoT2VYSJFQp4EPAJM3kpZWg2qjBSlnFsC5JsOYjqVKZNchRObSrTg(mp4, (17.14 MB) Download video eyJraWQiOiI4ZjUxYWNhY2IzYjhiNjNlNzFlYmIzYWFmYTU5NmZmYyIsImFsZyI6IlJTMjU2In0.eyJzdWIiOiI5NjM3NTIyMDVkYTllYzk0ODNlMTJiNjM4YmNhMGNiYiIsImtpZCI6IjhmNTFhY2FjYjNiOGI2M2U3MWViYjNhYWZhNTk2ZmZjIiwiZXhwIjoxNjg4MzY0Njk5fQ.khN1bVz7GnwLDGDL6k63nf7VI5RON9zuyc6-Wp_FIrCF9HgjkfIPJiZFX7UaHyFrDBLR2p190QHEyRrhHZPb74UqfWuGYh9neTl1bykIoYi_FBTMzGCG_b9A3xXZFovNXQGcBWOUOjvqxJrwToizfL5Ea7pbk0F9Or0BwunarqPsF1K5InfKxE87fQTuJCvMc8vRRwuSF0PvY8OV-lPr27Sg8HUFO0f8-PS6SRSiNlROyW5sAUr_g1JNxs6LaFARl8DQz_NUMvckIyIJmSvZAugKnYQrgIomGDoT2VYSJFQp4EPAJM3kpZWg2qjBSlnFsC5JsOYjqVKZNchRObSrTg(mp4, (17.14 MB) Download video Our study shows that a high-air flow extractor is effective in removing aerosols during simulated continuous breathing and a simulated cough. However, simply covering a cough with a gloved hand resulted in the escape of aerosols and subsequent detection at the clinician's head. Removal of aerosols may enhance the safety of healthcare workers and improve operational efficiencies. Currently, a minimal air exchange rate of 15–20 h−1 is recommended for operating room air decontamination. At this rate 18–28 min is required to reduce airborne contaminants by 99%.10Jensen P.A. Lambert L.A. Iademarco M.F. Ridzon R. Guidelines for preventing the transmission of Mycobacterium tuberculosis in health-care settings.MMWR Recomm Rep. 2005; 54 (2005): 1-141PubMed Google Scholar This delay causes workflow inefficiencies and the extractor can be used to accelerate air decontamination. A limitation of this study is the difference between airflows in the test environment and actual operating rooms. Compared with the test environment, operating rooms have higher air exchange rates (15–20 vs 0.75 h−1), which may cause turbulence, interfere with the extractor exhaust plume, and decrease capture efficiency. We have shown that the high-flow air extractor is highly effective at reducing aerosol concentrations at the source. This has potentially large-scale implications for clinical practice and warrants translation into high-risk clinical areas in order to minimise clinician exposure. Furthermore, this technique is consistent with current recommendations from the CDC to augment room air exchanges. Conceptualisation: CM, TE, VC, PF, JF Methodology: CM, TE, VC, PF, JS, SD Visualisation: CM, TE, VC, PF, JS, TL, BD Software: PF, JS, TL, BD Analysis: PF, JS, TL, BD Original draft preparation: CM, TE, VC, PF, JS, SD, TL, BD Review and editing of the manuscript: CM, TE, VC, PF, JS, SD, TL, BD, JF We acknowledge the contributions of Theo Tackey, Rachelle and Jesse Matava. The authors declare no that they have no conflicts of interest. National Sciences and Engineering Research Council of Canada (Grant RGPIN-2014-06698) and the Canada Foundation for Innovation (Grant 32319) to JS.

The authors report no conflict of interest.

To the Editor: A 6-year-old, 22-kg girl diagnosed with sickle cell disease (SCD) was presented to the emergency department of our institution with 1-month history of headache and dizziness, and a new onset of vertical diplopia and isolated right cranial nerve IV palsy. The magnetic resonance imaging (MRI) findings were suggestive of acute disseminated encephalomyelitis. Seven weeks before she experienced a left lower lobe pneumonia of unidentified etiology followed by a rhinovirus upper respiratory tract infection. She was therefore admitted and treated with IV methylprednisolone 650mg for 5 days. She was discharged 4 days later with a prednisone taper over 2 weeks. One month later she returned to the hospital with a 4-day history of fever and 24 hours of right arm weakness. During an urgent head MRI, she suddenly deteriorated (Glasgow Coma Score from 15 to 7) due to a left hyperacute subarachnoid and intraparenchymal hemorrhage (Fig. 1). She was first transferred to the intensive care unit where she received saline 3% and mannitol 25%, and then she underwent an emergent decompressive craniotomy. Upon opening the dura, the brain parenchyma massively herniated and there was active bleeding challenging to control. Available packed red blood cells and fresh frozen plasma were given, and the massive transfusion protocol was activated. The initial laboratory investigations showed severe coagulopathy (Table 1). Despite the ongoing transfusions, diffuse oozing and brain herniation persisted, and coagulopathy worsened. Under the current circumstances, tranexamic acid (TXA) was started (50 mg/kg bolus +infusion of 10mg/kg/h). Soon, the surgeon was able to complete the operation with better hemostasis. The estimated blood loss was 3.5 L. A postoperative MRI on postoperative day 3 revealed disseminated cortical and subcortical lesions. Acute hemorrhagic leukoencephalitis,1,2 variant of the previous encephalomyelitis,2 was suspected and therapy started. The patient was discharged after 7 weeks. Three months later, she was alert, oriented, and able to communicate and to walk independently despite the right hemiplegia. The pathology report showed widespread intracapillary sickling but did not confirm the leukoencephalitis. In SCD children, the risk of stroke is 200 times higher than in the general pediatric population, with an ischemic to hemorrhagic events ratio of 7:1.3,4 A hemorrhagic stroke may be secondary to cerebral aneurysm rupture, use of steroids, or remains unexplained.3,4 In pediatrics, the antifibrinolytic TXA decreases blood loss during surgeries and mortality after trauma.5 In the context of intracranial bleeds, such as adult aneurismal subarachnoid hemorrhage, the reduced risk of rebleeding with TXA may be offset by an increased incidence of cerebral ischemia.6 The use of TXA in SCD is only reported in adults.7 Theoretically, TXA may increase the thrombotic risk in these patients. Our patient also experienced a severe coagulopathy, which may have been secondary to the massive bleeding. The disruption of the cerebral parenchyma may have also been a contributing factor. Bain injury, in fact, can cause coagulopathy through different pathways (ie, massive release of tissue factor, maladaptive protein C response), although it rarely appears before 24 hours from the injury.8 The drop of platelets (> 50% within 24 h), the coagulopathy, and the hemorrhage raised the suspicion of disseminated intravascular coagulation. Therefore TXA can potentially enhance thrombogenesis unless fibrinolysis predominates over coagulation.9 As fibrinolysis products were not measured, a definite diagnosis is not warranted. We justified the use of TXA by the lack of signs of thrombosis and the massive, uncontrollable bleeding.

CASES OF CARDIAC ARREST after the administration of a neuromuscular reversal agent have been reported in the literature. However, no clear explanation of these near-fatal events has been formulated. In the authors' institution, 2 additional cases of cardiac arrest have been experienced. Therefore, this study reviewed the characteristics of all available patients and, in conjunction with the latest studies on heart transplant patients, speculated on the possible causes of these events.

Retrograde intubation is part of the difficult airway algorithm in the American Society of Anesthesiologists, but its usage is rare in small pediatric patients with the advent of new intubation techniques. We present our experience of retrograde intubation for a 4-month-old patient who presented for laryngeal cleft repair on cardiopulmonary bypass. This case highlights the unique place for retrograde intubation in small patients in the current era.