Putpose: The dissertation includes two clinical studies investigating the prognostic signilicance of specific biomarkers at the time of hospital admission in two common and potentially fatal diseases. ln the first study, at the onset of the pandemic, we sought a comprehensive analysis to lind biomarkers that could help stratify the risk for intensive care unit (lCU) admission and mortality. The second study aimed to accurately predict pancreatic necrosis and provide a detailed analysis ol a large multicenter cohort study of acute necrotizing pancreatitis (ANP). Methods: ln the first study, we performed a systematic review and meta-analysis ol the available literature in live databases. Pooled mean differences were calculated for continuous outcomes and pooled odds ratios for dichotomous outcomes. ln the second study, we used the XGBoost machine learning algorithm to process data from 2387 patients with AP Shapley Additive exPlanations (SHAP) scores were calculated to quantify the contribution of each provided variable. The analyzed dataset was collected by the Hungarian Pancreatic Study Group between 2012 and 2019. Results: The analysis of the 93 eligible studies revealed an association between lymphopenia, low CD4+ and CD8+ lymphocyte subsets and worse prognosis in COVID-19 (p<0.05 in all cases). ln the AP study, the XGBoost classifier had an AUC of 0.757 Glucose, C-reactive protein, alkaline phosphatase, sex and total white blood cell count have the greatest impact on prediction based on SHAP values. Conclusion: 0ur meta-analysis showed that admission laboratory parameters could serve as important and early prognostic factors in C0VID-19. The first Al algorithm for ANP risk estimation was developed in our second study. The predictive potential of this model is comparable to existing clinical scoring systems and the model is expected to improve with use. TriaI registration numbers: C R D 4 2020 17 6836, 2225 4 - 1/2012 I/EKU, 177 87 - 8/2020/EilG  Kulcsszavak: pancreatitis, necrosis, artificial intelligence, COVID-19, SARS-CoV-2