The role of eosinophil as a diagnostic, differential, and prognostic marker for Covid-19 during pregnancy: Case-control study from tertiary center

Aysegul Atalay,Selcan Sinaci,Berhan Besimoglu,Onur Ozkavak,Ecem Kaya,Ozgur Kara,Dilek Sahin
DOI: https://doi.org/10.1111/jog.15713
Abstract:Purpose: To evaluate the level of eosinophil count in Covid-19 pregnant patients as a cost-effective tool for aiding in diagnosis, differential diagnosis, and prognosis and, to assess whether eosinopenia could be similar or superior to lymphopenia in pregnancy with Covid-19. Methods: Retrospective case-control study was conducted including pregnant women tested simultaneously for SARS-CoV-2 by RT-PCR and complete blood count (CBC). Eosinophil counts (EOS), lymphocytes (LYM), neutrophils (NEU) with calculation of EOS/LYM ratio, EOS/NEU ratio, NEU/LYM ratio, LYM/NEU ratio (LNR), eosinopenia, and lymphopenia were compared between the groups. ROC analysis was performed for determination of optimal cut-off values and comparative analysis between AUC were performed using the paired sample design. Logistic regression was used to determine the factors affecting the categorical variables. Results: Four thousand two hundred sixteen pregnant women were included in the final analysis of which (n = 845) as healthy control group, (n = 1482) as non Covid-19 patient group and (n = 1889) as Covid-19 group. Covid-19 group was divided into three subgroups based on severity. To diagnose and distinguish Covid-19 from other infectious conditions, EOS had better performance with higher area under curve (AUC) (0.769 and 0.757 AUC, respectively, p < 0.001). Categorically, eosinopenia had better diagnostic (Covid-19 vs. healthy), prognostic (severe-critical vs. mild-moderate), and differential diagnostic (Covid-19 vs. non Covid-19) performance than lymphopenia (OR = 5.5 vs. 3.4, OR = 3.4 vs. 1.8, OR = 5.4 vs. 2.7 p = 0.000). Conclusion: Eosinopenia is a low-cost, reliable, and convenient marker for Covid-19 not only in diagnosis, but also in prognosis by acting as an early marker for predicting severe-critical cases.
What problem does this paper attempt to address?