Introduction: Numerous studies suggest that alterations in blood parameters, such as changes in platelet, lymphocyte, hemoglobin, eosinophil, and basophil counts; increased neutrophil counts; and elevated neutrophil/lymphocyte and platelet/lymphocyte ratios, signal COVID-19 infection and predict worse outcomes. Leveraging these insights, our study seeks to create a predictive mortality model by assessing age and crucial laboratory markers. Materials and Methods: Patients were categorized into two groups based on their hospital outcomes: 130 survivors who recovered from their Intensive Care Unit (ICU) stay (Group 1) and 74 who died (Group 2). We then developed a predictive mortality model using patients’ age, neutrophil-to-lymphocyte ratio (NLR), platelet-to-lymphocyte ratio (PLR), procalcitonin levels, and ferritin lactate (FL) index results. Results: A total of 204 patients were included. Patients in Group 2 had a notably higher mean age compared to those in Group 1 (76 ± 11 vs. 66 ± 15 years) (p < 0.001). Using specific cut-off values, our analysis revealed varying effectiveness in predicting COVID-19 mortality: Those aged over 73 years showed 74% sensitivity and 60% specificity, with an area under the curve (AUC) of 0.701. Procalcitonin levels above 0.35 ng/mL balanced true-positive and -negative identifications well, achieving an AUC of 0.752. The FL index, with a threshold of 1228 mg/dL, had 68% sensitivity and 65% specificity with an AUC of 0.707. A PLR higher than 212 resulted in 48% sensitivity and 69% specificity, with an AUC of 0.582. An NLR higher than 5.8 resulted in 55% sensitivity and 63% specificity, with an AUC of 0.640, showcasing diverse predictive accuracies across parameters. The statistical analysis evaluated the effects of age (>73), procalcitonin levels (>0.35), FL > 1228, PLR > 212, and NLR > 5.8 on mortality variables using logistic regression. Ages over 73 significantly increased event odds by 2.1 times (p = 0.05), procalcitonin levels above 0.35 nearly quintupled the odds (OR = 5.6, p < 0.001), high FL index levels more than tripled the odds (OR = 3.5, p = 0.003), a PLR > 212 significantly increased event odds by 3.5 (p = 0.030), and an NLR > 5.8 significantly increased event odds by 1.6 (p = 0.043). Conclusions: Our study highlights significant predictors of mortality in COVID-19 ICU patients, including advanced age, elevated procalcitonin, FL index levels, the PLR, and the NLR.

Introduction: Carbapenem-resistant Enterobacterales (CRE) cause serious and life-threatening infections with limited treatment options. The most common causes of carbapenem resistance are carbapenemases. We aimed to determine the most prevalent carbapenemase genes; bla _NDM , bla _KPC , bla _IMP-1, bla _VIM , bla _OXA-48 like genes among Klebsiella pneumoniae and Escherichia coli strains in the southeast region of Turkey. Methods: Eighty-nine isolates (74 K. pneumoniae, 15 E. coli) were included in the study. The isolates were found as carbapenem-resistant by BD Phoenix automated system and Kirby Bauer disk diffusion test. Antibiotic susceptibility testing was performed by BD Phoenix automated system. Combination disc method (CDM) was also carried out as phenotypic method of carbapenemase detection. The presence of bla _NDM , bla _KPC , bla _IMP-1, bla _VIM and bla _OXA-48-like genes were investigated by Xpert CARBA-R (Cepheid, USA) multiplex PCR commercial system. Results: Antibiotic resistance rates by Phoenix were 48.3 97.8, 80.9, 96.6, 93.3, 96.6, 97.8, 69.7 and 97.8% for amikacin, aztreonam, cefepime, ceftazidime, ciprofloxacin, levofloxacin, moxifloxacin, trimethoprim sulfamethoxazole and piperacillin-tazobactam, respectively. The bla _OXA-48-like gene was detected in 65 (73%); bla _NDM gene in 4 (4.5%); co-production of bla _OXA-48-like and bla _NDM in 6 (6.7%); co-production of bla _OXA-48-like, bla _VIM and bla _IMP-1 genes in 1(1.2%) isolate. None of the bla _OXA-48-like, bla _NDM , bla _VIM , bla _IMP-1, bla _KPC genes were detected in 13 (14.6%) of the isolates. The sensitivity and specificity of CDM were calculated as 80 and 85% respectively. Conclusion: We detected bla _OXA-48-like gene most frequently in our region. To our knowledge, this is the first report of K. pneumoniae-co-producing bla _OXA-48-like, bla _VIM and bla _IMP-1 genes. The coexistence of these genes is alarming and causes both infection control and treatment problems. Effective infection control measures are essential to prevent the spread of antibiotic resistance.