How Can a Deep Learning Algorithm Improve Fracture Detection on X-rays in the Emergency Room?

The growing need for emergency imaging has greatly increased the number of conventional X-rays, particularly for traumatic injury. Deep learning (DL) algorithms could improve fracture screening by radiologists and emergency room (ER) physicians. We used an algorithm developed for the detection of appendicular skeleton fractures and evaluated its performance for detecting traumatic fractures on conventional X-rays in the ER, without the need for training on local data. This algorithm was tested on all patients (N = 125) consulting at the Louis Mourier ER in May 2019 for limb trauma. Patients were selected by two emergency physicians from the clinical database used in the ER. Their X-rays were exported and analyzed by a radiologist. The prediction made by the algorithm and the annotation made by the radiologist were compared. For the 125 patients included, 25 patients with a fracture were identified by the clinicians, 24 of whom were identified by the algorithm (sensitivity of 96%). The algorithm incorrectly predicted a fracture in 14 of the 100 patients without fractures (specificity of 86%). The negative predictive value was 98.85%. This study shows that DL algorithms are potentially valuable diagnostic tools for detecting fractures in the ER and could be used in the training of junior radiologists.

Reducing pain by using venous blood gas instead of arterial blood gas (VEINART): a multicentre randomised controlled trial.

INTRODUCTION: Venous sampling for blood gas analysis has been suggested as an alternative to arterial sampling in order to reduce pain. The main objective was to compare pain induced by venous and arterial sampling and to assess whether the type of sampling would affect clinical management or not. METHODS: We performed an open-label randomised multicentre prospective study in four French EDs during a 4-week period. Non-hypoxaemic adults, whose medical management required blood gas analysis, were randomly allocated using a computer-generated randomisation list stratified by centres with an allocation ratio of 1:1 using random blocks to one of the two arms: venous or arterial sampling. The primary outcome was the maximal pain during sampling, using the visual analogue scale. Secondary outcomes pertained to ease of sampling as rated by the nurse drawing the blood, and physician satisfaction regarding usefulness of biochemical data. RESULTS: 113 patients were included: 55 in the arterial and 58 in the venous sampling group. The mean maximal pain was 40.5 mm±24.9 mm and 22.6 mm±20.2 mm in the arterial group and the venous group, respectively, accounting for a mean difference of 17.9 mm (95% CI 9.6 to 26.3) (p<0.0001). Ease of blood sampling was greater in the venous group as compared with the arterial group (p=0.02). The usefulness of the results, evaluated by the prescriber, did not significantly differ (p=0.25). CONCLUSIONS: Venous blood gas is less painful for patients than ABG in non-hypoxaemic patients. Venous blood gas should replace ABG in this setting. TRIAL REGISTRATION NUMBER: NCT03784664.

Hereditary angioedema, emergency management of attacks by a call center.

OBJECTIVE: Hereditary angiœdema (HAE) is a rare autosomal dominant disease characterized by recurrent, unpredictable, potentially life-threatening swelling. Objective is to assess the management of the acute HAE attacks in the real life setting through a call center in France. METHODS: A pre-specified ancillary study of SOS-HAE, a cluster-randomized prospective multicenter trial, was conducted. HAE patients were recruited from 8 participating reference centers. The outcome of interest was the rate of hospitalization. RESULTS: onerhundred patients were included. The median (quartile) age was 38 (29-53) years, and 66 (66%) were female. Eighty (80%) patients had HAE type I, 8 (8%) had HAE type II and 12 (12%) patients had FXII-HAE. Fifty-one (51%) patients had experienced at least one time the call center during the follow-up. Nine over 166 (5%) attacks for 9 different patients resulted in hospital admission to the hospital (in the short-stay unit, ie, <24 h) during the follow-up period. During 2 years, there were 166 calls to call center for 166 attacks. All attacks were treated at home after call center contact. CONCLUSIONS: Use of emergency departments and hospitalizations are reduced by the use of a coordinated national call center in HAE after therapeutic education program that promoted self-administration of specific treatment and use of call to call center. TRIAL REGISTRATION: clinicalTrials.gov identifier: NCT01679912.