AB1347 LABORATORY PARAMETERS ASSOCIATED WITH CODIV-19 CYTOKINE STORM DEVELOPMENT

Background Since the end of 2019, physicians became more and familiar with SARS-CoV-2 infection variety forms in which it may present evolve. There have been a lot studies trying to understand predict why some patients develop dysregulation immune response, an exaggerated release pro-inflammatory cytokines, called cytokine storm (1–4). is scarce evidence Romania regarding this aspect. Objectives This study aims verify correlation between laboratory parameters development cohort over 200 admitted tertiary hospital from Romania, hoping that early identification these risk factors progression severe form disease can bring considerable benefit patient care. Methods analytical, observational, case-control includes 219 (all COVID-19 hospitalized on Internal Medicine 3 department Colentina Clinical Hospital, Bucharest, 01 March 2020 1 April 2021). A series data were collected, being most important, including: albumin, lymphocyte (percentage), neutrophil (absolute value), aspartate aminotransferase, alanine D-dimers, lactate dehydrogenase (LDH), anionic gap, chloremia, potassium BUN:creatinine ratio (BUN - blood urea nitrogen). The used for statistical analysis represent average values first 7 days hospitalization those who did not storm, respectively until day its development, others. Patients classified into groups, developed complication taking account clinical paraclinical criteria (impairment respiratory function, elevations certain markers 2-3 times above upper limit normal, died as result infection). Then Binary Univariate Logistic Regression was applied order individual impact every parameter development. Furthermore, all subsequently included multivariate analysis, using backward selection technique achieve model predictive possible. Results We mention demographic previously performed, showing no statistically significant relationship gender, age or comorbidities (history neoplasm, lung diseases, cardiac pathology, obesity, type II diabetes hypertension) their evolution storm. After performing binary univariate logistic regression we concluded 8 13 analyzes had change groups (ferritin, PCR, Lymphocyte, Neutrophils, TGO, LDH, ratio). Only 150 then analysis. variables lost significance, final including C-reactive protein, neutrophilia, ferritin ratio. correctly predicts 88% cases. Conclusion High are should be monitored evolution. References [1]Pedersen SF et all. SARS-CoV-2: raging [2]Mehta P al. COVID-19: consider syndromes immunosuppression [3]Hu B COVID-19. [4]Caricchio R Preliminary Acknowledgements: NIL. Disclosure Interests None Declared.