In this study, the symptoms and laboratory results of pregnant women admitted to the hospital with suspected COVID-19 were investigated, and the distinctive features of the PCR-confirmed COVID-19 patients studied. Also, a model was suggested to facilitate the identification of these patients.

This descriptive cross-sectional study was carried out among pregnant women admitted to our hospital between 1 April 2020 and 31 May 2020 with suspected COVID-19. The data about patients was obtained from patient records and the hospital automation system. Ethical approval was received from the Research Council and the Ethical Committee of the Kharkiv Medical Academy of Postgraduate Education, No16.0220p. The patients gave informed consent to participate. The study report was compiled as per the STROBE guidelines [9].

During the study period, 5,217 pregnant women applied to the hospital. The symptoms and a history of contact with people having the disease suggested COVID-19 in 64 of them. Therefore, they were hospitalized as suspected COVID-19 cases. The patients were divided into two groups based on PCR results: PCR (+) (n = 33) and PCR (-) (n = 31) (Figure 1).

Figure 1.

— Participant flow chart.

The primary outcome variable of the study was the D-dimer levels. Secondary outcome variables were the presence/absence of cough, shortness of breath, headache, fever, weakness, proteinuria, diarrhoea, haematuria, loss of taste, hypertension, and gestational diabetes mellitus (GDM). Besides, the data on the following numerical secondary outcome variables was collected: age (years), pregnancy history (gravidity, parity, normal spontaneous deliveries (NSD), and C-sections (C/S)), body-mass index (kg/m²) (BMI), gestational age (weeks), haemoglobin levels (g/dL), white blood cell (WBC) count, c-reactive protein (CRP) (mg/L)), procalcitonin (ug/L), neutrophil count, lymphocyte count, glomerular filtration range (GFR) (mL/min/1.73 m²), urea (mg/dL), creatinine (mg/dL), systolic arterial blood pressure (mm Hg), diastolic arterial blood pressure (mm Hg), pulse rate/min, and, peripheral venous oxygen saturation (SPO2) (%).

The Statistical Package for Social Sciences (SPSS) program (SPSS for Windows, Version 25.0, Chicago, IC, USA) was used for statistical analysis. The results were presented as means and standard deviations for numerical variables, and frequencies and percentages for categorical data. The conformity of numerical variables to normal distribution was evaluated with skewness values and histograms.

The numerical variables fitting normal distribution were compared through an independent samples t-test; the analysis of those not conforming to normal distribution was carried out by means of a Mann-Whitney U test; and the Chi-Squared (or Fisher’s exact test) test was used for comparing categorical variables. Depending on their distribution, the relations between numerical variables were determined through either Spearman or Pearson correlation analysis. For multivariate examinations, logistic regression analysis with the entered model was used. A p-value of < 0.05 was taken as sufficient for statistical significance.

The prevalence of headache was higher in PCR (+) pregnant women than the PCR (-) ones. On the other hand, those with PCR (-) results had higher D-dimer levels and fever.

Lack of differential diagnosis of the PCR (-) women and the absence of long follow-up results can be considered limitations of this study.

In December 2019, a new type of coronavirus (SARS-CoV-2) causing severe acute respiratory syndrome was first isolated in the patients diagnosed with pneumonia of unknown origin in China's Hubei province [10-12]. It was also demonstrated with mathematical models that the infectious property of this new virus is very high [13, 14]. Due to their changing physiology, susceptibility to infections, and mechanical and immunological differences, pregnant women are believed to be a group that should be handled more carefully in such a pandemic [15, 16].

In a review of 108 pregnant women with COVID-19, it was reported that women who presented with fever (68%) and cough (34%) in the third trimester [17]. Reportedly, 70% of those women had high C-reactive protein, 59% had lymphocytopenia, and 91% delivered by C-section. In another study, symptoms including fever, cough, shortness of breath, and anosmia were found to be significantly associated with COVID-19 in pregnancy [18].

Laboratory test results examined in a study comparing pregnant and non-pregnant women with positive COVID-19 test results demonstrated significantly higher levels of inflammatory markers such as white blood cell count, neutrophil count, C-reactive protein, procalcitonin, and D-dimer. The mean lymphocyte percentage was significantly lower in pregnant women than in the non-pregnant ones [19].

The incidence rates of symptoms in our study were similar to the previous reports. However, it showed for the first time that fever and D-dimer levels were lower in PCR (+) pregnant women compared to PCR (-) pregnant patients. The reason for this finding may be the inclusion of suspected COVID-19 patients. Indeed, fever is a symptom that leads one to suspect a COVID-19 infection. However, it is unclear why D-dimer was lower in PCR (+) pregnant women. Perhaps, it can be attributed to the presence of an undetectable infection with the symptoms similar to COVID-19 in the PCR (-) pregnant women. The situation can be clarified with further studies based on a similar methodology.

Acute renal damage was observed in up to 25% of severe COVID-19 patients (20), similarly to the 33-year-old pregnant woman receiving COVID-19 therapy in Iran [21]. However, no data were obtained during our study to demonstrate any relation to PCR positivity.

The real-time polymerase chain reaction (RT-PCR) method of sampling from the respiratory tract (nasopharyngeal and/or throat swab) is a gold standard for the conclusive diagnosis of COVID-19 infection [7, 22]. However, in the light of such factors as the shortage of personnel, time constraints, and economic reasons, many models have been proposed to predict the diagnosis and severity of COVID-19 infections [23-27]. Unfortunately, these failed to show the desired performance [23]. Furthermore, our research shows there is no such prediction model for pregnant women. In this study, it was claimed that PCR positivity in COVID-19 could be inferred in a specific group of pregnant women based on the symptoms, e.g., headache, fever, and D-dimer, with a relatively high sensitivity but low specificity. This way, potentially harmful imaging methods such as computed tomography can be avoided by relying on simple and accessible data.