Oral Communication ▸ PACS: clinical outcome
OC 123Long COVID phenotypes and association with SARS CoV-2 variants in the EUCARE-POSTCOVID study
Authors: F. Bai1, S. De Benedittis1, A. Tavelli1, M.M. Santoro2, F. Ceccherini-Silberstein2, M. Iannetta3, Y. Shimoni4, S. Ravid4, T. Kozlovski4, F. Köning5, N. Pfeifer5, E. Shamsara5, M.Parczewski6, A. Cozzi-Lepri7, A. d’Arminio Monforte1, G. Marchetti1
Affiliation: 1Clinic of Infectious Diseases, San Paolo Hospital, ASST Santi Paolo e Carlo, Department of Health Sciences, University of Milan, Italy, 2Department of Experimental Medicine, University of Rome Tor Vergata, Roma, Lazio, Italy, 3Infectious Disease Unit, Department of System Medicine, Tor Vergata University and Hospital, Rome, Italy, 4Healthcare Informatics, IBM Research-Haifa, Mount Carmel Haifa, Israel, 5Institute for Bioinformatics and Medical Informatics, University of Tübingen, Tübingen, Germany, 6Department of Infectious Tropical Diseases and Immune Deficiency, Pomeranian Medical University in Szczecin, Szczecin, Poland, 7Centre for Clinical Research, Epidemiology, Modelling and Evaluation (CREME), Institute for Global Health, UCL, London, UK
ABSTRACT
Background: The EUCARE-POSTCOVID study is a multicenter study enrolling recovered COVID-19 patients to investigate the prevalence and predictors of long COVID. The prevalence, risk factors and clinical presentation of long COVID aren’t well understood yet; first data showed a reduced probability of long COVID during the Omicron era compared to Delta. We aimed to investigate the association between viral variants and long COVID.
Materials and methods: Long COVID was defined according to WHO (symptoms at 3 months after the COVID-19, that last at least 2 months without an alternative diagnosis). We included patients hospitalized for acute COVID-19 in San Paolo Hospital, Milan and Policlinico Tor Vergata, Rome (24/02/2020-07/01/2022). An unselected group of hospitalized patients underwent post COVID-19 evaluations at 1-3, 6-9 and 12-15 months after the acute phase; patients with ≥1 visit have been included. At the post COVID evaluation patients filled in questionnaires about ongoing symptoms. Viral variants were approximated by country variant surveillance data that allowed to assign a predominant variant over time. Logistic regression analyses were fitted to calculate propensity scores (PS) using predictors of undergoing a post COVID visit; PS were then used to evaluate the role of viral variants in developing long COVID by means of Inverse Probability Weighting.
Results: Among 2802 hospitalized patients, 772 (27.5%) underwent at least one post COVID visit; 84.6% patients had a post COVID visit at 1-3 months visit, 13.6% at 6-9 months and 1.8% at 12-15 months. 530/772 (68.7%) had long COVID (65.4% and 29.7% had physical and psychological symptoms, respectively). Most frequent symptoms were Myalgic Encephalomyelitis, defined by fatigue, post exertional malaise and unrefreshing sleep and ≥1 symptom between impaired memory and orthostatic intolerance (57.8%), brain fog (31.5%), respiratory symptoms (30%) and musculoskeletal symptoms (25.4%). Long COVID was independently associated with female gender; no association with comorbidities or disease’s severity was found (Table 1). Omicron variant was available in 35/2802 (1.2%) patients. We did not observe any association between viral variants and long COVID, even using the weighted model; a borderline association between Delta variant (vs wild strain) and fatigue or brain fog was shown, but the association was largely attenuated in the weighted analysis (Figure 1).
Conclusions: Long COVID seems a common complication after acute COVID-19 at least during Delta circulation and is characterized more frequently by myalgic encephalomyelitis. Compared to the primary viral strain, Delta variant was slightly associated with a higher probability of fatigue and brain fog, but no association between long COVID and variants was found after adjustment for baseline unbalanced characteristics. Although there was no evidence for an association with Omicron subvariants, this needs to be reassessed as more events cumulate.
Attach