From Other Journals December 2021.

This study recruited 800 patients from 31 French centres between 2015 and 2020 and randomly assigned them to receive either IV hydrocortisone (200 mg/day via continuous infusion for 4–7 days, with tapering over 8 or 14 days) or an identical IV saline placebo infusion. Apart from study interventions, patients received treatment based on intensive care unit (ICU) clinician discretion. Eligibility required at least one of four criteria: the initiation of mechanical ventilation (invasive or non-invasive) with PEEP ≥5 cm H2O; initiation of high flow nasal oxygen with FiO2 ≥50% and associated PaO2:FiO2 ratio <300; for patients wearing a non-rebreathing mask, a PaO2:FiO2 ratio <300; or Group V Pneumonia Severity Index score of >130. Mortality at 28 days was 6.2% (24/400 patients) for hydrocortisone and 11.9% (49/395 patients) for placebo (absolute risk reduction 5.6%, P = 0.006, number needed to treat 18). Hydrocortisone also improved secondary outcomes, reducing need for intubation and commencing vasopressors among patients not requiring either of these at enrolment. There was no increase in ICU-acquired infections but patients requiring insulin in the first week received higher daily doses in the hydrocortisone group. These impressive gains may have been overestimated as the trial was stopped early just prior to the COVID pandemic, and on second planned interim analysis, when only 70% of target recruitment (795/1145) had been reached. The study cohort was highly selective: 5148 of 5948 (86%) patients with severe community-acquired pneumonia (CAP) requiring ICU care were deemed ineligible, of which 1200 had septic shock at baseline that was already routinely treated with corticosteroids. Another notable exclusion was concurrent influenza infection. The authors do not detail whether severe CAP patients with new or chronic hypercapnia, including those requiring BiPAP, were enrolled. With only a third of study patients having chronic obstructive pulmonary disease (COPD), hydrocortisone's benefit in severe CAP could well be mediated by systemic anti-inflammatory and immunomodulatory benefits in addition to reducing severe bronchoconstriction in COPD. Time-critical early administration of steroids in this study (median 15 h), or the use of hydrocortisone rather than methylprednisolone could account for the benefit in severe CAP that other studies have failed to demonstrate. This study convincingly demonstrates that hydrocortisone should be started as soon as possible for adult patients with hypoxic respiratory failure and severe CAP, without influenza co-infection. The optimal type and dosing regimen for corticosteroids in severe CAP remain unclear. It remains to be seen whether the convenience of easy-to-administer hydrocortisone boluses could provide similar benefits as an infusion in the treatment of severe CAP. (Dequin P-F et al. N. Engl. J. Med. 2023; 388: 1931–41) This randomised non-inferiority trial found that replacing chest X-ray (CXR) with ultra-low dose computed tomography (CT) in the diagnostic work-up of ED patients with suspected non-traumatic pulmonary disease did not change short-term patient outcomes but identified more incidental findings. Two thousand four hundred and eighteen consecutive patients presenting to one of two Dutch EDs with suspected non-traumatic pulmonary disease requiring chest imaging as part of their work up were cluster randomised to receive CXR or CT. There was no significant difference in the primary outcome of functional health at 28 days after ED presentation, as measured by the physical component summary scale score of the SF-12 questionnaire. More CT patients required follow-up for incidental findings such as pulmonary nodules (8.3% compared to 1.2% in the CXR group); however, this did not appear to worsen mental health at 28 days as measured by the mental component summary scale score of the SF-12. There were no significant differences between the groups for other outcomes including hospital admission, length of stay or mortality within 28 days. In terms of radiation exposure, the median CT radiation dose was 0.2 mSv, compared to 0.02 mSv for portable CXR and 0.05 mSv for standard 2 view CXR. However, additional imaging within 28 days was performed in 39.3% of CT patients and 53.9% of CXR patients, making it difficult to draw conclusions about the impact of this protocol on medical imaging radiation exposure. More patients in the CT group were diagnosed with CAP and influenza than in the CXR group, consistent with known improved sensitivity on CT. This trial recruited patients in 2017–2018, so may not be applicable in the COVID era. While CT may have improved diagnostic accuracy, it came at the cost of increased incidental findings and made no difference to patient-reported health scores at 28 days. There is no information here about impacts on healthcare costs, overall radiation exposure and longer-term health impacts related to increased diagnosis of incidental findings. There may be cases where the increased sensitivity of CT is desirable but for routine presentations, stick with CXR. (van den Berk IAH et al. Thorax 2023; 78: 515–22) This double-blind randomised controlled trial comparing 4-factor prothrombin complex concentrate (4F-PCC) to placebo showed no benefit in reducing 24-h blood product consumption and an increased rate of thromboembolic events. Traumatic coagulopathy remains an important cause of mortality in trauma. Current trauma resuscitation is aimed at transfusing coagulation factors as soon as possible. Observational studies show that early administration of 4F-PCC (coagulation factors II, VII, IX and X, proteins S and C) reduces blood product consumption, but at the risk of increasing thromboembolic events. This study recruited adult trauma patients in 12 Level 1 trauma centres in France at risk of massive transfusion, defined as transfusion of at least 1 U of packed red blood cell concentrate (PRBC) during prehospital care or within 1 h of admission, and either an assessment of blood consumption score of at least 2 or assessment by the attending physician of risk of massive transfusion. Exclusion criteria included traumatic cardiac arrest or patients with anticipated unsurvivable injuries. One hundred and fifty-nine patients received the active intervention of 4F-PCC at a dose of 25 IU factor IX per kg (1 mL/kg). One hundred and forty-nine patients received the placebo of 1 mL/kg of 0.9% saline solution. Patients were resuscitated with restricted fluid expansion and early transfusion of blood products with a PRBC: fresh frozen plasma (FFP) ratio between 1:1 and 2:1, fibrinogen concentrate if fibrinogen concentration less than 1.5 g/L or viscoelastic criteria showing a functional deficiency and platelets to maintain count higher than 50 × 109/L at all times. The primary outcome was the total number of all blood product units (RBC, FFP and platelet concentrate) consumed within the first 24 h. Thromboembolic events were recorded on day 28. Ninety-five per cent of patients received 4F-PCC or placebo within the first hour of admission. There was no difference in median 24-h total blood product consumption in the 4F-PCC group versus the placebo group (12 U vs 11 U) or the consumption of individual components. The number of patients with at least one thromboembolic event was greater among those who received 4F-PCC versus placebo (56 vs 37). Trauma resuscitation has made significant progress through the implementation of restrictive volume resuscitation and massive transfusion protocols, along with the recognition of traumatic coagulopathy. In this study, the addition of 4F-PCC to massive transfusion did not demonstrate meaningful benefits in patient outcomes or blood product reduction. It raises the question of whether we have reached the point of diminishing returns, where further gains in patient outcomes and reduced blood product usage may be challenging to achieve. (Bouzat P et al. JAMA 2023; 329: 1367–75) In this study of trauma patients requiring intubation, delayed sequence intubation (DSI) using ketamine was associated with reduced peri-intubation hypoxia and higher first-pass success compared to rapid sequence intubation (RSI) but several factors may limit the external generalisability of their findings. Intubations were performed by second-year anaesthetics trainees in a single trauma centre in India. Notably, adults with anticipated difficult airways, facial trauma, aspiration and cardiac arrest were excluded. Patients in the DSI arm were given up to 1.5 mg/kg of ketamine in aliquots of 0.5 mg/kg until dissociation was achieved, followed by preoxygenation for 3 min prior to paralysis with succinylcholine and intubation. RSI patients were preoxygenated for 3 min then given ketamine 1.5 mg/kg and succinylcholine together. Both arms were allowed to preoxygenate with spontaneous respirations on 10 L/min oxygen. SaO2 was measured at 1-min intervals from start of preoxygenation to 1 min post-intubation. The study randomised 236 patients, the majority of whom had head injury with a median Glasgow Coma Scale of 6. Following randomisation, 24 patients in the RSI group and 12 patients in the DSI group had unanticipated difficult airways (defined as difficulty with bag mask ventilation, laryngoscopy or tracheal intubation) and were excluded from the primary analysis. This uneven distribution makes us wonder if timing of ketamine administration may have had a causal effect on intubation difficulty that should be explored further. The incidence of hypoxia (defined as SaO2 <93% at any point in the preoxygenation and intubation period) was 8% in the DSI group and 35% in the RSI group. Desaturation events were more common early in the preoxygenation period in both groups. First-pass success was 83% for DSI compared to 69% for RSI. Excluding patients with anticipated difficult airways from eligibility and unanticipated difficulty airways from post-recruitment analysis leaves readers in the dark about the patients we might be most interested in. Baseline agitation was not measured which may have helped explain which patients in the DSI group had the most benefit from this technique. We do not know what, if any oxygen was administered prehospital in this cohort which may impact external generalisability of these findings to settings with different ED prearrival supplemental oxygen use. Despite the limitations of this study, emergency clinicians may already be using DSI in trauma patients based on evidence in medical patients and the pragmatic appeal of this technique. Although this study does not strengthen the evidence base for the practice, it should not discourage clinicians from utilising DSI specifically in agitated trauma patients who require preoxygenation. (Bandyopadhyay A et al. Anesth. Analg. 2023; 136: 913–9) In this single centre, US study of patients with mild to moderate diabetic ketoacidosis (DKA), protocol based fast-acting subcutaneous (SQ) insulin showed similar efficacy and safety to management with IV insulin infusion while reducing ED length of stay (LOS). Adults with mild to moderate DKA (pH >7 and a bicarbonate level >10) were eligible for the protocol. Patients with confusion, serious infection, active comorbidities and need for surgery were excluded. The one-page SQuID (subcutaneous insulin in diabetic ketoacidosis) protocol is included in the manuscript. Insulin lispro, a rapid acting short duration insulin analogue also known as Humalog, was used at doses of 0.1–0.2 unit/kg subcutaneously alongside IV fluids with and without dextrose titrated based on two-hourly fingerstick blood glucose levels. Clinicians were prompted to initiate the protocol by a best practice advisory embedded within the electronic medical record linked to a prepopulated electronic order set. To understand the impact of the protocol they used a prospective quasi-experimental (pre–post) study design. The SQuID cohort was compared to a contemporaneous group with mild to moderate DKA managed on a traditional insulin infusion and two historical cohorts, one immediately preintervention and one pre-COVID. There was no difference in adherence to timely blood glucose checks and use of rescue dextrose for hypoglycaemia. Median SQuID cohort LOS was 8.9 h compared to 11.9 h in the contemporaneous insulin infusion cohort, 10.3 h in preintervention cohort and 12.5 h in the pre-COVID cohort. Hospital and patient-specific issues impact ED LOS and disposition decisions in the management of DKA and thus the impact of a protocol like this is likely to vary significantly across institutions. The authors provide a safe and effective roadmap for those interested in exploring alternatives to traditional IV insulin effusion for DKA in their ED. (Griffey RT et al. Acad. Emerg. Med. 2023; doi: 10.1111/acem.14685) In this cluster-randomised trial, the use of video consultations for acutely ill children in rural and community EDs resulted in less frequent interfacility transfers than consultations done by telephone. The study enrolled 16 rural and community EDs in Northern California, with each ED randomised to 1 of 4 unbalanced (3:1) crossover treatment assignment sequences of either telemedicine (synchronous audio–video communication with visual assessment and physical examination of the patient) or a standard telephone consultation with University of California Davis Children's Hospital Paediatric ICU specialists. Six hundred and ninety-six children aged 14 years or younger who presented to a participating ED with an acute medical condition that required referral centre consultation were included. The overall transfer rate was 84.0% in the telemedicine arm and 90.6% in the telephone arm. In intention-to-treat analysis, the adjusted risk of transfer was lower in the telemedicine arm than the telephone arm (RR 0.93), with the number-needed-to-treat with telemedicine to avoid transfer being 16.5 patients. The risk of transfer among patients in the telemedicine arm remained significantly lower in the adjusted treatment-received and per-protocol analyses despite the high proportion of crossover in patients allocated to telemedicine (only 42.5% received telemedicine consults). Consultations for patients who arrived at the presenting ED by ambulance were more likely to be conducted by telephone when allocated to telemedicine (67.9%) than for patients who did not arrive by emergency medical services (54.2%). The reason for this is unclear. It is possible that these patients were sicker, and the treating clinicians felt it was faster to use the telephone rather than the video system. As outcomes for patients who were discharged from the presenting ED or hospital were not assessed, it is possible that some of these children may have had adverse events from not being transferred. Adding video to paediatric interfacility transfer consultations could reduce the number of transfers. It remains to be seen whether this approach is safe and acceptable to families and clinicians. (Marcin JP et al. JAMA Network Open 2023; 6: e2255770) This randomised non-inferiority trial suggests that using non-sterile gloves and dressings in the closure of wounds in the ED may be as safe as sterile equipment. Traumatic wounds arise in non-sterile skin, in a non-sterile environment and are caused by non-sterile objects, and should therefore be considered contaminated. Sterile gloves and dressings are commonly used for repair despite a paucity of evidence supporting this. This study was a multicentre, randomised controlled non-inferiority trial of non-sterile gloves and dressings versus sterile gloves, dressings and drapes for suturing of traumatic wounds. Adult patients presenting to the ED with a traumatic wound requiring sutures were eligible. All patients returned to the ED for suture removal and review. Primary outcome was the rate of wound infection. Seven hundred and forty-seven and 733 patients were randomised to the sterile and non-sterile groups. The wound infection rate in the sterile group was 6.8% versus 5.7% in the non-sterile group, mean difference −1.1% (95% CI −3.7% to 1.5%). Infection was more likely in wounds on the lower extremities or in patients on immunosuppressants. The study was stopped early, and therefore underpowered. Rates of infection were also considerably higher than what is normally reported. Nonetheless, this trial suggests that there is unlikely to be a large difference between non-sterile and sterile gloves and dressings for suturing of traumatic wounds in the ED. Non-sterile treatment is of value in health settings where sterile materials are not readily available, as well as potentially creating less environmental waste. (Zwaans JJM et al. Emerg. Med. J. 2022; 39: 650–4)