Valeur ajoutee des algorithmes de traitement automatique du langage: exemple de la prise en charge du cancer du pancreas a l'Assistance publique-Hopitaux de Paris.

Introduction: L'impact de la pandemie de COVID-19 sur le retard de prise en charge des cancers est debattue. Nous etudions l'interet des algorithmes de traitement automatique du langage (TAL) en plus des donnees medico-administratives (PMSI) dans le cancer du pancreas (CP). Methodes: Nous avons identifie dans l'entrepot de donnees de l'AP-HP les patients nouvellement adresses pour un CP entre 01/2019 et 12/2020. Nous avons analyse le PMSI, les donnees de mortalite extrahospitalieres et les comptes rendus (CR) analyses par TAL. Differents jeux de donnees ont ete annotes par un oncologue pour developper puis valider les algorithmes de TAL (selection des CR et evaluation du contenu). Resultats: Au total, 1442 patients (736 en 2019 et 706 en 2020) sont suivis jusqu'en 02/2022. Le nombre de nouveaux patients a baisse de 24 % entre mars-mai 2019 et 2020. Les taux de patients traites par chirurgie, chimiotherapie ou soins palliatifs exclusifs etaient comparables en 2020 et 2019 (32 % versus 30 %, p=0,46;49% versus 52 %, p=0,24;19 % versus 18 %, p=0,57). Parmi les 1175 patients avec traitement actif, 70 % et 21 % ont un CR de concertation pluridisciplinaire (RCP) avant puis apres le premier traitement, respectivement. Un CR scanner a baseline est trouve pour 40 %. Parmi les 344 patients operes sans traitement neoadjuvant, 95 % ont un CR d'anatomopathologie et 82 % un score pTNM. Les taux de pTNM pejoratif, pTxN1-2, (50 % versus 51 %, p=0,95), de stade metastatique (38 % versus 34 %, p=0,32) et le delai RCP-premier traitement (21 jours EIQ [12-40]) versus 21 jours EIQ [12-42], p=0,77) sont stables. Les taux de survie a 1 an des patients en 2019 ou 2020 sont comparables chez les patients operes (92 % versus 92 %, p=0,99) ou traites par chimiotherapie exclusive (60 % versus 52 % p=0,04). Discussion/Conclusion: Grace a des methodes de TAL, nous avons pu completer le PMSI et montrer que le taux d'adressage des nouveaux patients pour CP a diminue sans impact sur leur stade tumoral ni sur le delai de prise en charge initial. Mots-cles: Cancer;COVID-19;EDS;TAL;Analyse de survie Declaration de liens d'interets: Les auteurs declarent ne pas avoir de liens d'interets.Copyright © 2023

Factors associated with psychological symptoms in hospital workers of a French hospital during the COVID-19 pandemic: Lessons from the first wave.

PURPOSE: The COVID-19 pandemic has put hospital workers around the world in an unprecedented and difficult situation, possibly leading to emotional difficulties and mental health problems. We aimed to analyze psychological symptoms of the hospital employees of the Paris Saint-Joseph Hospital Group a few months after the 1st wave of the pandemic. PARTICIPANTS AND METHODS: From July 15 to October 1, 2020, a cross-sectional survey was conducted among hospital workers in the two locations of our hospital group using the Hospital Anxiety and Depression Scale (HADS) and Post-Traumatic Stress Disorder (PTSD) Checklist (PCL) to measure anxiety, depression, and PTSD symptoms. Factors independently associated with these symptoms were identified. RESULTS: In total, 780 participants (47% caregivers, 18% health administrative workers, 16% physicians, and 19% other professionals) completed the survey. Significant symptoms of anxiety, depression, and PTSD were reported by 41%, 21%, and 14% of the participants, respectively. Hierarchical regression analysis showed a higher risk of having psychological symptoms among those (1) who were infected by SARS-CoV-2 or had colleagues or relatives infected by the virus, (2) who retrospectively reported to have had an anxious experience during the first wave, and (3) with a previous history of burnout or depression. In contrast, job satisfaction appeared to be a protective factor. Overall, hospital workers showed the statistically same levels of anxiety, depression, and PTSD symptoms, regardless of their profession and whether they had worked in units with COVID-19 patients or not. CONCLUSIONS: Our cross-sectional survey of 780 hospital workers shows that after the first wave, hospital employees had a significant burden of mental health symptoms. Specific preventive measures to promote mental well-being among hospital workers exposed to COVID-19 need to be implemented, first among particularly vulnerable staff, and then, for all hospital staff for whom anxiety is detected early, and not only those who were directly exposed to infected patients.

Adjusting D-dimer to lung disease extent to exclude pulmonary embolism in COVID-19 patients (Co-LEAD)

Background : D-dimer is a safe tool to exclude pulmonary embolism (PE) but its specificity is decreased in COVID-19. Aims : Our aim was to derive a new algorithm with D-dimer threshold adjusted to CT extent of lung damage. Methods : We conducted a multicenter, retrospective cohort study among 774 COVID-19 patients with suspected PE. D-dimer threshold adjusted to CT extent of lung damage was derived in a patient set ( n = 337), and its safety assessed in an independent validation set ( n = 337). Results : According to ROC curves, D-dimer safely excluded PE, with one false negative when using a 900 ng/mL threshold when lung damage extent was <50% and 1700 ng/mL when lung damage extent was ≥50%. In the derivation set, the algorithm's sensitivity was 98.2% (95% CI: 94.7-100.0), and its specificity 28.4% (95% CI: 24.1-32.3). The negative likelihood ratio (NLR) was 0.06 (95% CI: 0.01-0.44) and the AUC was 0.63 (95% CI: 0.60-0.67). In the validation set, sensitivity and specificity were 96.7% (95% CI: 88.7-99.6) and 39.2% (95% CI: 32.2-46.1), respectively. The NLR was 0.08 (95% CI: 0.02-0.33), and the AUC did not differ from that of the derivation set (0.68 (95% CI: 0.64-0.72), P = 0.097). Using the Co-LEAD algorithm, 76/250 (30.4%) COVID-19 patients with suspected PE could have been managed without CT pulmonary angiography (CTPA). Conclusions : The Co-LEAD algorithm safely excludes PE, and allows reducing the use of CTPA in COVID-19 patients. Prospective management studies are necessary to validate this strategy.