Impact of COVID-19 on Fetal Outcomes in Pregnant Women: A Systematic Review and Meta-Analysis

Abstract Background: Coronavirus disease (COVID-19) is a pandemic causing respiratory symptoms, taste alterations, olfactory disturbances, and cutaneous, cardiovascular, and neurological manifestations. Recently, research interest has shifted to reproductive health to understand the factors predisposing to COVID-19 infection in pregnancy, the consequences of the infection on the fetus and on the mother, and possible vertical transmission through the placenta. Pregnancy does not increase the risk of SARS-CoV-2 infection, according to studies. However, contrary to non-pregnant women, pregnancy worsens the clinical outcome of COVID-19. Studies investigating the effects of COVID-19 on pregnancy women are heterogeneous, and the results are often conflicting. Objectives: The goal of the current work was to offer a thorough and up-to-date systematic review of, and meta-analysis on, the impact of COVID-19 on ovarian function, pregnancy, and fetal outcomes. Search strategy: This meta-analysis (PROSPERO n. CRD42023456904) was conducted using the Preferred Reporting Items for Systematic Review and Meta-Analysis (PRISMA) protocols. The search for relevant material was conducted using PubMed, Scopus, Cochrane, and Embase databases, through to 15 December 2022. Selection criteria: Original articles on fertile pregnant women or women attempting to become pregnant, with an active case of, or history of, SARS-CoV-2 infection were included, and reproductive function was compared to that of uninfected women. Data collection and analysis: The effects of COVID-19 on female reproductive function, particularly ovarian function, the profile of female sex hormones, pregnancy outcomes and fetal outcomes were the focus of our search. Quantitative analysis was performed with Comprehensive Meta-Analysis Software. The standard difference of the mean was calculated for the statistical comparison between cases and controls. Cochran’s Q test and heterogeneity (I2) indexes were used to assess statistical heterogeneity. Sensitivity analysis and publication bias tests were also performed. Main Results: Twenty-eight articles met our inclusion criteria, for a total of 27,383 patients pregnant or looking to have offspring, with active or anamnestic COVID-19, and 1,583,772 uninfected control women. Our study revealed that there was no significant difference between COVID-19 patients and the control group in terms of maternal characteristics such as age, body mass index (BMI) and comorbidities that could affect pregnancy and fetal outcomes. The risk of a miscarriage or Cesarean delivery was significantly lower, while the risk of fetal death or premature delivery was significantly higher in COVID-19 patients than in the controls. None of the included studies evaluated hormonal profiles or investigated the presence of infertility. Conclusions: Maternal comorbidities, age, and BMI do not raise the risk of COVID-19. However, pregnant women with COVID-19 had a lower risk of miscarriage and Cesarean delivery, possibly because of better prenatal care and high levels of observation during labor. COVID-19 during pregnancy increases the risk of fetal death and premature delivery.


Introduction
Coronaviruses are a group of viruses that primarily affect humans through zoonotic transmission.They include the Middle East respiratory syndrome coronavirus (MERS-CoV) and the severe acute respiratory syndrome coronavirus (SARS-CoV), which emerged in 2012 and 2003, respectively [1].In 2019, a novel severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) etiological agent of a severe acute respiratory syndrome called coronavirus disease (COVID-19) [2] emerged in Wuhan, China.COVID-19 has rapidly become a global pandemic because SARS-CoV-2 has proven to be highly infectious with a high reproductive number (R 0 ) and airborne transmission [3].
SARS-CoV-2 has a high predilection for upper and lower respiratory tract cells where it can replicate, causing a variety of symptoms ranging from mild (such as the common cold) to more severe (such as pneumonia and acute respiratory distress syndrome) [4].In addition to respiratory symptoms, other manifestations are caused by SARS-CoV-2 infection, including taste alterations and smell disorders, skin manifestations, and cardiovascular and neurological manifestations.Furthermore, COVID-19 causes thromboembolic events, as well as a multisystem inflammatory syndrome in children [5].
The disease has also gained attention in obstetrics and gynecology, as pregnant women have weakened immune systems and undergo changes to their cardiopulmonary systems [6].The prevalence of SARS-CoV-2 infection in pregnant women has varied significantly according to geographical location and country income levels.High rates of 19% [95% confidence interval (CI) 12% to 27%] were reported in Latin America and the Caribbean region, while the lowest rates were reported in East Asia and in the Pacific region (0.4%, 95% CI 0% to 2%) [7].
Concerns about COVID-19 are related to the infection of the mother, placental cells, and the fetus through vertical transmission.Studies have shown that pregnancy does not increase the risk of SARS-CoV-2 infection [8].However, it exacerbates the clinical outcome of COVID-19 compared to non-pregnant women [9].Furthermore, pregnant women were more likely to be admitted to an intensive care unit, require invasive ventilation, extracorporeal membrane oxygenation, and die compared to non-pregnant women [6].Numerous studies have reported that SARS-CoV-2 is capable of infecting placental cells, causing inflammation, infractions, and alterations to maternal and fetal vascular perfusion [10].Vertical transmission to the fetus, causing intrauterine infection and adverse effects on the conceptus, has also been reported.However, this type of transmission is rare, probably due to the low levels of viremia [11].
On these premises, this study aimed to provide a comprehensive and updated systematic review of, and meta-analysis on, the effects of SARS-CoV-2 infection on ovarian function, pregnancy, and fetal outcomes.

Search Strategy
This systematic review and meta-analysis (PROSPERO registration n.CRD42023456904) was carried out according to the Preferred Reporting Items for Systematic Review and Meta-Analysis (PRISMA) protocols.The literature search was performed up to 15 December 2022.An extensive search of PubMed, Scopus, Cochrane, and Embase databases was performed, focusing on the effects of SARS-CoV-2 on female reproductive function, including ovarian function, female sex hormone profile, and pregnancy outcomes.

Selection Criteria
Articles were assessed for eligibility using the Population, Exposure, Comparison/ Comparator, Outcome, and Study Type (PECOS) model system [12].Specifically, we included articles on fertile pregnant women or women attempting to become pregnant, with an active case or history of SARS-CoV-2 infection and compared their reproductive function to that of uninfected women.Unhealthy women with previously diagnosed reproductive tract diseases (early menopause, hysterectomies, urogenital infections, etc.) that could affect the results were excluded.Original human studies were included, while animal studies, case reports, and non-original studies, such as reviews or comments, were excluded from the analysis.The selection criteria are detailed in Table 1.The selection of the articles was performed independently by two authors (M.M. and R.K.S).The titles and abstracts of the studies were first independently screened for inclusion.Each decision was reviewed by two non-blinded reviewers (R.C. and S.L.V.).Reviewers resolved disagreements between the authors, while the senior author (A.E.C.) resolved disagreements between the reviewers.Finally, the eligible articles underwent data extraction.

Data Extraction
The following parameters were collected: information on the first author, year of publication, study design, duration of infection (if available), age, body mass index (BMI), maternal comorbidities (if any), LH, follicle-stimulating hormone (FSH), 17β-estradiol (E 2 ), progesterone, anti-Müllerian hormone (AMH), anovulation, menstrual irregularities, presence of infertility, pregnancy complications, miscarriages, fetal death, live births, preterm delivery, Cesarean section, gestational age at delivery, small for gestational age (SGA), chorioamnionitis, and birth weight.In particular, data on BMI and maternal comorbidities were extracted to evaluate differences between infected and non-infected women.
When a value was available in a different unit of measure, it was converted according to the conversion tables.For each parameter, the number of women (COVID-19 positive/COVID-19 negative), the mean value, the standard deviation (SD), the median value, and the interquartile range (IQR) range were extracted.For studies expressing data as median and IQR, the formula by Wan and colleagues [13] was used to estimate the mean and SD.Two authors independently extracted the data.Differences between reviewers were discussed until a consensus was reached.

Quality Assessment
The quality of evidence (QoE) in the studies was assessed by R.S.K.In detail, all studies were assessed using the Cambridge Quality Checklists [14].As no randomized control trials were included, no further evaluation was needed.

Statistical Analysis
Comprehensive Meta-Analysis software (Version 3) (Biostat Inc., Englewood, NJ, USA) was used to analyze the data.Standardized mean difference (SMD) or odd ratio (OR) were used for statistical comparisons between cases and controls.Statistical significance was accepted for p-values ≤ 0.05.Statistical heterogeneity was assessed using the heterogeneity index (I 2 ) and Cochran's Q test.In detail, when I 2 was ≤50%, the variation of the studies was considered homogenous, and the pooled effect size was calculated using the fixed effect model.However, for I 2 > 50%, indicating significant heterogeneity across the studies, the random effects model was adopted.Finally, the qualitative analysis of publication bias was assessed through the asymmetry of the funnel plot, which suggested some missing studies on one side of the graph.Quantitative analysis of publication bias was instead evaluated using Egger's intercept test, which indicates the statistical significance of publication bias.In this case, the unbiased estimates were calculated using the "trim and fill" method.

Study Characteristics
Using the above search strategy, 3552 articles were retrieved.After excluding 84 editorials, book chapters, and reviews, 2894 articles on different topics, 259 articles with unproved SARS-CoV-2 infection, and 199 articles were screened.Of these, 154 were judged to be not pertinent for the purpose of this study after reading their abstracts or full texts.Additionally, 28 studies were excluded as these were animal studies or non-original articles.The remaining 28 studies were carefully read and included in the analysis (Figure 1).
These studies included 27,383 patients with an active or anamnestic COVID-19 infection and 1,583,772 age-and BMI-matched healthy controls with no prior diagnosis of gynecological or pregnancy diseases.The main characteristics of the studies selected for the meta-analysis are described in Table 2.These studies included 27,383 patients with an active or anamnestic COVID-19 infection and 1,583,772 age-and BMI-matched healthy controls with no prior diagnosis of gynecological or pregnancy diseases.The main characteristics of the studies selected for the meta-analysis are described in Table 2.

Quality of Evidence of Included Studies
All the included studies assessed with the Cambridge Quality Checklists scored >6 out of a total of 15.While this scale does not establish a precise threshold for differentiating between high-or low-quality studies, the results suggest that the included studies are of moderate to high methodological quality (Table 3).
Table 3. Quality of evidence assessment of the included studies (results of the Cambridge Quality Checklists [14] and Cochrane's risk-of-bias tool for randomized controlled trials [42]).

Study Name Type of Study
Cambridge Quality Checklists

Specific Results
The results of the analysis are summarized in Table 4.They include patient and control characteristics, ovarian function, pregnancy outcomes, and fetal outcomes.

Ovarian Function
None of the included studies reported data on FSH, LH, E 2 , progesterone, AMH, anovulation, menstrual irregularities, or the presence of infertility.Therefore, this analysis could not be performed.

Pregnancy Outcomes
Data on pregnancy and live birth rates could not be analyzed since they were not reported in the included studies.
The analysis showed the absence of inter-study heterogeneity, as indicated by the I 2 test (I 2 = 0%, p = 0.56, Q-value = 2060), and of publication bias, supported by the symmetry of the funnel plots and Egger's test (intercept 0.54, 95% CI −2.64, 3.71; p = 0.27).However, one study was found to be sensitive enough to alter the results [27].Indeed, its removal resulted in a non-significant risk of miscarriage in patients compared with controls (OR 0.74; CI 95% 0.25; 2.24; p = 0.59) (Supplementary Figure S9).The analysis showed the absence of inter-study heterogeneity, as indicated by the I 2 test (I 2 = 0%, p = 0.56, Q-value = 2060), and of publication bias, supported by the symmetry of the funnel plots and Egger's test (intercept 0.54, 95% CI −2.64, 3.71; p = 0.27).However, one study was found to be sensitive enough to alter the results [27].Indeed, its removal resulted in a non-significant risk of miscarriage in patients compared with controls (OR 0.74; CI 95% 0.25; 2.24; p = 0.59) (Supplementary Figure S9).
Gestational age at delivery: Seven articles [17,19,31,33,36,40,44] reported data on gestational age at delivery, and were therefore included in the analysis; no difference between the patients and controls was found (SMD 0.049; CI 95% −0.147; 0.245; p = 0.62) (Figure 7).However, inter-study heterogeneity was identified by the I 2 test (I 2 = 70.7%,p = 0.002) and the Q-test (Q-value = 20.509), in the absence of publication bias, as shown by the funnel plot symmetry and Egger's test (intercept 2.08, 95% CI −1.09, 5.26; p = 0.08).In fact, no study was found to be sensitive enough to alter the results (Supplementary Figure S14).However, inter-study heterogeneity was identified by the I 2 test (I 2 = 70.7%,p = 0.002) and the Q-test (Q-value = 20.509), in the absence of publication bias, as shown by the funnel plot symmetry and Egger's test (intercept 2.08, 95% CI −1.09, 5.26; p = 0.08).In fact, no study was found to be sensitive enough to alter the results (Supplementary Figure S14).
Small for gestational age: An analysis of the eight studies [15,19,21,33,35,39,44] that evaluated the prevalence of children born with SGA in infected and non-infected women found no difference in the risk of SGA between patients and controls (OR 1,0; CI 95% 0.880; 1.149; p = 0.9) (Figure 8).The analysis did not show inter-study heterogeneity, as demonstrated by the I 2 test (I 2 = 26%, p = 0.219, Q-value = 9497), but did identify the presence of publication bias, as indicated by the funnel plot symmetry and Egger's test (intercept −0.62, 95% CI −1.93, 0.70; p = 0.15).No study was found to be sensitive enough to alter the results (Supplementary Figure S15).

Discussion
The evolution of a new virus, such as SARS-CoV-2, carries a considerable public health concern, particularly for high-risk groups such as pregnant women.Viral infection during pregnancy puts the mother at risk and has the potential to interfere with

Discussion
The evolution of a new virus, such as SARS-CoV-2, carries a considerable public health concern, particularly for high-risk groups such as pregnant women.Viral infection during pregnancy puts the mother at risk and has the potential to interfere with pregnancy outcomes and fetus wellbeing, through vertical transmission.During the COVID-19 pandemic, several studies addressed this issue.Therefore, this systematic review and metaanalysis aimed to evaluate the association between COVID-19 infection during pregnancy and pregnancy outcomes.We conducted a thorough literature search and extracted and analyzed the data.Overall, 28 studies were selected and included.We searched for maternal characteristics, including age, BMI, diabetes (both gestational and pre-gestational), hypertension (chronic hypertension and pre-eclampsia), and pregnancy and fetal outcomes.
The results of our study demonstrated that maternal characteristics influencing pregnancy and fetal outcomes were not significantly different between COVID-19 patients and the controls.Therefore, the results of the primary endpoints are not biased by differences in age, BMI, or comorbidities between infected patients and non-infected women.This conclusion is in agreement with what is known so far, namely, that pregnancy or pregnancy-related comorbidities are not risk factors for SARS-CoV-2 infection [43].
Preterm delivery and fetal death were found to be significantly higher in patients with COVID-19.However, the miscarriage rate and the number of Cesarean deliveries were significantly lower.The high prevalence of preterm delivery and fetal death in SARS-CoV-2-infected pregnancies is likely a consequence of the vertical transmission of the virus [48].
Miscarriage and Cesarean section rates were lower in patients with COVID-19, and this could be attributed to some bias in clinical care towards COVID-19 patients, offering them a higher level of care including careful observation, meticulous assistance, and targeted treatment.Regarding the lack of a difference found for birth weight, it should be noted that multiple factors influence this outcome, including parity [49]; comorbidities such as maternal diabetes [50] and hypertension [51]; genetics [52]; and BMI [53].However, none of the studies conducted multiple regression analyses to evaluate the roles of these factors.
The sensitivity analyses showed that one study [27] for miscarriage and three studies [23,26,32] for birth weight were sensitive enough to skew the results.In the study by Bortoletto and colleagues [27], the cohort included women who became pregnant after the use of assisted reproduction techniques who were compared with historical control patients enrolled one year before the pandemic.The other three studies looked at pregnant women who were asymptomatic or had negative COVID-19 test results.These differences may in part explain why these studies could bias the results.In particular, for birth weight, a lack of symptoms or even negative tests for COVID-19 do not exclude totally the chance of viral infection.
Forty-one meta-analyses on maternal COVID-19 infection and pregnancy outcomes have been published to date.These studies include a high diversity of analyzed data and often reach contradictory conclusions.The number of included studies varies considerably, as well as the types of studies, since some meta-analyses included not only cohort and casecontrol studies but also case reports [54], case series [55], and brief reports [56].Some other meta-analyses did not include a control group and the data were presented as prevalence and not as an odd ratio or standard deviation of the mean [55,57,58].
The strengths of our study are several.First, we conducted a systematic literature search nearly two years after the onset of the COVID-19 pandemic.This allowed us to collect a large and comprehensive number of studies.Furthermore, we only included controlled observational studies and assessed their quality, a process which was lacking in most of the previous meta-analyses [54,59].Our study also includes a sensitivity analysis, contrary to most meta-analyses published so far.Moreover, the analysis of comorbidities included all cases of diabetes in pregnancy (pre-gestational and gestational) and the analysis of hypertension included both chronic hypertension in pregnancy and preeclampsia.We combined these etiologies, as most studies did not include a clear definition of these outcomes.The same limitation applied to low birth weight and Cesarean deliveries.Indeed, most studies did not correctly define low birth weight (below 10th percentile/3rd percentile/less than 2500), and for Cesarean deliveries, there was no difference between urgent Cesarean delivery and elective ones.

Conclusions
This systematic review and meta-analysis provides a moderate to high level of evidence that maternal age and comorbidities do not increase the risk for SARS-CoV-2 infection.COVID-19 during pregnancy increases the risk of preterm delivery and fetal death.However, pregnant women infected with SARS-CoV-2 had a lower chance of miscarriage and Cesarean delivery, probably due to more careful prenatal care and a higher level of observation during labor.

Supplementary Materials:
The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/jpm13091337/s1, Figure S1: Forest plot of the age in patients and controls.Figure S2: Funnel plot (A) and sensitivity analysis (B) of the age in patients and controls.Figure S3: Forest plot of the body mass index in patients and controls.Figure S4: Funnel plot (A) and sensitivity analysis (B) of the body mass index in patients and controls.Figure S5: Forest plot of the risk for diabetes in patients and controls.Figure S6: Funnel plot (A) and sensitivity analysis (B) of the risk for diabetes in patients and controls.Figure S7.Forest plot of the risk for hypertension in patients and controls.Figure S8.Funnel plot (A) and sensitivity analysis (B) of the risk for hypertension in patients and controls.Figure S9.Funnel plot (A) and sensitivity analysis (B) of the risk for miscarriage in patients and controls.Figure S10.Funnel plot (A) and sensitivity analysis (B) of the risk for chorioamnionitis in patients and controls.Figure S11.Funnel plot (A) and sensitivity analysis (B) of difference in birth weight in patients and controls.Figure S12.Funnel plot (A) and sensitivity analysis (B) of difference in pre-term delivery in patients and controls.Figure S13.Funnel plot (A) and sensitivity analysis (B) of the risk for cesarean delivery in patients and controls.Figure S14.Funnel plot (A) and sensitivity analysis (B) of the risk for cesarean delivery in patients and controls.Figure S15.Funnel plot (A) and sensitivity analysis (B) of the risk for small for gestational age in patients and controls.Figure S16.Funnel plot (A) and sensitivity analysis (B) of the risk for fetal death in patients and controls.

Figure 1 .
Figure 1.PRISMA flow diagram of the literature screening.

Table 2 .
Characteristics of the 28 studies included in the meta-analysis.
Figure 1.PRISMA flow diagram of the literature screening.

Table 2 .
Characteristics of the 28 studies included in the meta-analysis.

Table 4 .
Summary of the results of the present meta-analysis.