key: cord-1037560-3wuob3xl
authors: Liu, Xiyao; Chen, Miaomiao; Wang, Yu; Sun, Lin; Zhang, Jun; Shi, Yuan; Wang, Jianhui; Zhang, Hua; Sun, Guoqiang; Baker, Philip N.; Luo, Xin; Qi, Hongbo
title: Prenatal anxiety and obstetric decisions among pregnant women in Wuhan and Chongqing during the COVID‐19 outbreak: a cross‐sectional study
date: 2020-06-24
journal: BJOG
DOI: 10.1111/1471-0528.16381
sha: 5a0574e715560440e6b9068c796daed1a1f07771
doc_id: 1037560
cord_uid: 3wuob3xl

OBJECTIVES: To investigate the mental status of pregnant women and to determine their obstetric decisions during the COVID‐19 outbreak. DESIGN: Cross‐sectional study. SETTING: Two cities in China, Wuhan (epicentre) and Chongqing (a less affected city). POPULATION: 1947 pregnant women. METHODS: We collected demographic, pregnancy, and epidemic information from our pregnant subjects, along with their attitudes towards COVID‐19 (using a self‐constructed five‐point scale). The Self‐Rating Anxiety Scale (SAS) was used to assess anxiety status. Obstetric decision‐making was also evaluated. The differences between cities in all of the above factors were compared, and the factors that influenced anxiety levels were identified by multivariable analysis. MAIN OUTCOME MEASURES: Anxiety status and its influencing factors. Obstetric decision‐making. RESULTS: Differences were observed between cities in some background characteristics and women's attitudes towards COVID‐19 in Wuhan were more extreme. More women in Wuhan felt anxious (24.5% vs 10.4%). Factors that influenced anxiety also included household income, subjective symptoms, and attitudes. Overall, obstetric decisions also revealed city‐based differences; these decisions mainly concerned hospital preference, time of prenatal care or delivery, mode of delivery and infant feeding. CONCLUSIONS: The outbreak aggravated prenatal anxiety, and the associated factors could be targets for psychological care. In parallel, key obstetric decision‐making changed, emphasising the need for pertinent professional advice. Special support is essential for pregnant mothers during epidemics.

In late December 2019, a cluster of pneumonia caused by a novel coronavirus (2019-nCoV) was reported. 1 This pathogen was eventually named SARS-CoV-2, 2 with its associated disease By the first quarter of 2020, the COVID-19 epidemic had become established in China. As of 3 May, 82,877 cases and 4,633 deaths were confirmed by the Chinese authorities. 4 Wuhan, as the epicentre in China, accounted for 60.7% of all cases and 83.5% of deaths. 5 Currently, SARS-CoV-2 has undoubtedly raised "a very high level of global risk" and become the "public enemy number one". 6, 7 The latest data confirmed over 3 million cases and more than 0.2 million deaths outside of China. 8 The fight against COVID-19 has become a global priority.

The rapid transmission and life-threatening characteristics of COVID-19 have been reported. The public, influenced by both accurate and erroneous information, are stressed. 9 All provinces in mainland China have adopted first-level public health emergency (PHE) responses, including travel bans and executive orders on daily life. 10 Consequently, the Chinese New Year holiday was disrupted and public anxiety was further aggravated.

Pregnant women, as a vulnerable population, 11 may be of a particular concern, as anxiety has been described as a common psychological problem during pregnancy. 12 Recent discussions of pregnancies during the COVID-19 outbreak have mainly focused on the therapeutic aspects; [13] [14] [15] little is known regarding mental status and psychological needs.

Prenatal care is vital to a healthy pregnancy. 16 The emergency traffic bans have made some medical resources inaccessible and may deter women from attending routine prenatal care. 17 Obstetricians have observed a dramatic decline in prenatal care attendance and births, as well as an increase in the caesarean section rate, all of which could threaten pregnancy outcomes. Relevant risks include ectopic pregnancy, delayed detection of foetal congenital anomalies, uncontrolled hypertension and pre-eclampsia, post-term delivery and dystocia. [18] [19] [20] These adverse events may have significant consequences, possibly greater than the COVID-19 infection itself. 9, 21, 22 Greater science preparedness for pregnant women during public health emergencies has been advocated, 23 but we know little about women's decisions during the COVID-19 and other PHEs.

This article is protected by copyright. All rights reserved

We conducted a survey among pregnant women in Wuhan (the hardest-hit area) and Chongqing (a neighbouring city) during the COVID-19 outbreak, to investigate anxiety status and its influencing factors, to determine and explain key prenatal decisions, and finally to guide social and medical practice.

This is a cross-sectional study through a self-administered questionnaire. The anonymous survey questionnaire was designed with four modules to collect data regarding: (1) background demographic, pregnancy and COVID-19 status; (2) attitudes towards COVID-19; (3) anxiety status, by using Self-Rating Anxiety Scale (SAS), 24,25 which is widely used and demonstrated to have excellent reliability and validity in pregnancy; 26,27 (4) obstetric decisions, defined as those pertaining to important obstetric procedures. The content of the questionnaire was reviewed and pretested by professors in Psychiatry (XYZ and his colleagues) and Obstetrics (HBQ and XL).

Eventually, the validity was established following appropriate revisions. The main content of the questionnaire is shown in Figure 1 and the English version of the full questionnaire is detailed in Appendix S1.

The questionnaires were distributed from February 3 to 9, 2020, mainly through a widely-used big data platform for pregnant mothers (YunYiTong 28 , covering more than 250,000 WeChat users nationwide) to those registered for prenatal care in hospitals in Wuhan and Chongqing. Two distribution strategies, namely, one-by-one WeChat message and advertising on the official accounts, were used. Additional questionnaires were distributed by obstetricians in both cities to those referred for outpatient services. Of note, each participant could only fill in this questionnaire once and only completed questionnaires could be submitted.

We followed relevant guidelines to ensure that the study was voluntary and confidential. The study was approved by the ethics committee of the First Affiliated Hospital of Chongqing Medical University, and an electronic informed consent was obtained before completing the questionnaire.

This article is protected by copyright. All rights reserved

The participants were not involved in the development of our study and none of the established core outcome sets has been used.

Data collection and input were automatically conducted. All data from the questionnaires were reviewed and the following subjects were excluded: (1) maternal age <14 or >60 years; (2) non-pregnant, with the answer of "already delivered" or "<0 or >45 weeks of gestational age"; (3) answers with wrong format; (4) illogical answers (choosing two options that contradicted each other in multiple-choice questions). Figure 2 shows the flow chart of our study.

The residency was based on both the city that the subjects registered for check-up (Wuhan or Chongqing) and the region subjects actually resided in at the time of survey. As a result of the Chinese Festival travel rush, these two addresses were not always exactly the same. The registration area was used as the residency for the following analyses. As a reference, a map on the actual area of the participants is displayed in Figure 2 .

We classified those aged ≧35 years as elderly gravida. Participants were assigned into three gestational age (GA) groups: (1) the first trimester: GA <14 complete weeks, (2) the second trimester: GA from 14 to 27 +6 weeks, (3) the third trimester: GA ≧28 weeks. Parity was divided into nullipara and multipara. Other grouping standards are in accordance with the categorical options in the questionnaire (Appendix S1).

Items measuring attitudes towards COVID-19 were designed on a five-point scale from "totally disagree" to "strongly agree". Although this assessment was comprised of three sections (knowing about the COVID-19, the objective impact of COVID-19, and the assistance), these 11 questions were analysed separately (Appendix S1).

This article is protected by copyright. All rights reserved

The anxiety status was assessed using the Chinese version of the SAS, 25 and the responses to the scale were summed as a standard score and a degree of anxiety by an established method 26,27 : the scores from 20 items were calculated to obtain a raw score ranging from 20 to 80, and the standard score was calculated using the raw score multiplied by 1.25; the standard score ≧50 indicates anxiety status; specifically, the standard score 50-59, 60-69 and ≧70 were considered mild, moderate and severe anxiety, respectively.

The obstetric decisions pertained to: (1) online consultation; (2) hospital preference; (3) schedule on prenatal visit or delivery; (4) the mode of delivery, infant feeding and postnatal resting; (5) the five-degree subjective impact on pregnancy of the items including changing schedule, reduced activities, and possible screening examination of COVID-19 (e.g., chest CT scan). These unstructured questions were analysed one-by-one (Figure 1 and Appendix S1).

We calculated the exact numbers and proportions for all variables in Wuhan, Chongqing, and the total for the two cities. Cronbach's alpha was used to assess the reliability of the anxiety scale. To compare the distribution of background, attitude, anxiety, and obstetric decisions between the two cities, the Chi-squared test, Kruskal-Wallis test and Student's t test were used in accordance with the type of data.

All factors related to pregnant women's background and their attitude towards COVID-19 were selected as independent variables. We used stepwise logistic regression models to estimate the effect of these factors on the anxiety status.

The Statistics Analysis Software version 9.4 (SAS Institute Inc, Cary, NC, USA) was used and a significance level of P-value < 0.05 was applied. Figures presented were plotted with PRISM version 8.0 for windows (GraphPad Software Inc, San Diego, CA).

This work was supported by grants from National Natural Science Foundation of China (No.

This article is protected by copyright. All rights reserved China (No. 2016YFC1000407). The funders had no involvement in the study design, data collection and analysis, interpretation of data, and preparation of the manuscript. The corresponding authors had full access to all of the data in the study and had the final responsibility to submit for publication.

A total of 1947 valid questionnaires were received, of whom 932 and 1015 were from Wuhan and Chongqing, respectively. 866 (92.9%) in Wuhan and 934 (92.0%) in Chongqing stayed where they were registered during the study period ( Figure 2 ).

Participants reported diverse demographic, pregnancy and epidemic characteristics (Table S1) .

Firstly, the general characteristics of their demographic background was of families with middle-level income and a working pregnant mother, although distribution differences existed between cities. Secondly, most pregnant women surveyed were in their second (32.8%) or third trimester (62.9%). There were more third trimester women in Wuhan (79.8% vs 47.4%). The majority of participants in both cities were nullipara and had experienced spontaneous singleton conception without comorbidity or complication, though Chongqing had a higher proportion of multiple pregnancies (4.4% vs 1.6%). Detailed distribution of comorbidity and complication is as detailed in Table S2 . Thirdly, information on COVID-19 from official media were widely accepted during this period in both cities. The proportions of self-reported symptoms were statistically the same in both cities, but the exposure history to diagnosed or suspected cases was more prevalent in Wuhan (4.7% vs 0.1%).

Attitudes towards COVID-19 were more extreme in Wuhan, with statistically more responses of "totally disagree" or "strongly agree" (Figure 3 ). Overall, four fifths of mothers felt nervous about the objective impacts of COVID-19, such as epidemic control, outdoor activity and person-to-person contact. Over 90% of the participants considered themselves vulnerable to this

This article is protected by copyright. All rights reserved epidemic. Women in both cities held comparatively positive attitudes towards online medical consultation and psychological counselling, but differences still existed between cities (P < 0.001).

The Cronbach's alpha for the SAS was indicative of moderate-to-good internal reliability: 0.78.

Specific details are shown in Figure S1 .

As shown in Table 1 , the participants in Wuhan had significantly higher anxiety scores. The mean standard score for anxiety was 43.97 (SD 8.71) for pregnant mothers in Wuhan, with a quarter of them scoring 50 or more. In Chongqing, an average score of 40.37 (SD 7.15) was reported, among whom about 90% scored lower than 50. The overall prevalence of anxiety during this period was 17.2%. Pregnant women in the epidemic hardest-hit area (Wuhan) were much more anxious, 18.8% and 5.7% of whom underwent mild, and moderate/severe anxiety; the corresponding proportions were 9.4% and 1.1% in Chongqing, respectively. The effect of the different cities on the SAS standard score was small-to-medium (effect size [ES]: Cohen's d 0.44).

The multivariable analysis (Table 2) showed the impact of each of the selected background or attitude factors on anxiety when controlling for all other factors. Firstly, pregnant women from middle-level income families were about half as likely to report anxiety than those earning an extremely high or low wage. Secondly, women in Wuhan were about twice as likely to develop anxiety (OR 1.83, 95% CI 1.38 to 2.41). Thirdly, those who had fever, cough, diarrhoea or symptoms of suspected infection were five times as likely to have anxiety than otherwise healthy women (OR 4.92, 95% CI 1.84 to 13.17). Furthermore, the attitudes towards COVID-19 were associated with anxiety status. Those with relatively more knowledge about COVID-19 and with rational risk perception (not too nervous about epidemic control or going out), were less likely to be anxious. Additionally, positive attitudes towards online medical consultation demonstrated a protective feature from anxiety (OR 0.65, 95% CI 0.57 to 0.73), while those who opted for psychological consultation showed the opposite effect (OR 1.31, 95% CI 1.14 to 1.52).

The participants' obstetric decisions are summarized in Table 3 .

This article is protected by copyright. All rights reserved Online consultation was requested by more than 70% of the participants, a higher proportion of whom were in Wuhan (75.4% vs 69.5%). Absolute differences could be found between the two cities in hospital preference during this period. Of pregnant women in Wuhan, 41.9% reported refusal of going to any hospital recently, compared with 27.7% in Chongqing. Questionnaire responses revealed a general trust in previous (53.0%) and specialized (29.0%) hospital among mothers, although differences existed in the proportion of that trust between cities (P < 0.0001).

Inconvenience caused by traffic bans raised significant concern, and as a result, 80.4% and 42.4% of the 1947 participants would defer their appointments for prenatal care and hospitalized delivery.

These responses were more common in Wuhan than in Chongqing (92.4% vs 72.9% and 47.3% vs 31.0%, respectively). Fear of infection was another reason for delaying plans. With respect to prenatal care, a minority (16.3% in general) reported "as planned" visit. Very few mothers in Wuhan chose to complete their scheduled check online (N=4) or face-to-face (2.6%). In Chongqing, however, 2.8% and 21.6% were willing to undertake their prenatal checks on time via the Internet and face-to-face. When it came to hospitalized delivery, 27.9% of all participants chose "ahead of time"; 15.2% wanted to be hospitalized earlier to wait for the onset of labour while 12.8% wanted to have a caesarean in advance. Only 25.2% of all women reported an "as planned" hospitalized delivery, and this proportion was higher in Chongqing (39.1% vs 19.2%).

12.7% of pregnant women in Wuhan would change delivery mode from vaginal delivery to caesarean section consequent upon the epidemic, while this proportion in Chongqing was halved (6.0%). However, the reverse change, from caesarean to vaginal delivery, was smaller (Wuhan 5.6%, Chongqing 3.1%). These city-based differences and uneven changes in mode could also be seen in choices of infant feeding and postnatal resting. Overall, there were more women who preferred caesarean section, bottle feeding and postnatal rest at home during this period than before, especially in Wuhan.

Over 90% of pregnant women believed that there had been negative impact on their pregnancy through changing schedules and reduced activities, and these subjective impacts were deemed to be more significant in Wuhan (P <0.0001). Slightly more than half of the women (50.1%) thought

This article is protected by copyright. All rights reserved the chest CT scan would significantly influence their pregnancy, and it was more severe in Chongqing (51.6% vs 48.4%, P = 0.0187).

We report the first large cross-sectional study of pregnant women's anxiety status and obstetric decision-making during the outbreak of COVID-19. We focused on the epicentre (Wuhan) and a neighbouring city (Chongqing), and involved 932 and 1015 participants, respectively.

The global estimated prevalence of prenatal anxiety fluctuates between 14% and 24%. [29] [30] [31] Unfortunately, during the study period, the overall anxiety rate was 17.2%, with a higher rate in Wuhan (24.5%) than in Chongqing (10.4%). Such a difference was probably attributed to the higher exposure and the stricter restrictions in Wuhan.

We found that women with medium-level incomes were protected from anxiety compared with those with high-or low-level wages. Other demographic characteristics or factors related to pregnancy were not found to be related to anxiety, although previous studies suggested that age, education, occupation, parity, and gestational age would exert effects. [31] [32] [33] Living in the epidemic centre and suffering subjective symptoms had major impacts on anxiety levels. We noted the high proportion of women who obtained their information through official media channels during the outbreak (84.3%), though the multivariable analysis did not show any association of information source with anxiety. Exposure to COVID-19 did appear with a higher frequency in the epicentre (Wuhan), although there were no differences in symptoms across cities (Table S1 ). However, symptoms but not exposure were an independent influencing factor of anxiety.

Important attitudes towards COVID-19, which were easy to be psychologically manipulated (discussed blow), were found to be associated with anxiety.

This article is protected by copyright. All rights reserved Our univariable analysis identified other potential factors that correlated with anxiety (Table S3) .

We believe that the mechanism of epidemic to anxiety is unique, so regional differences greatly mask the effect of these conventional factors, which further reminds us to pay attention to prenatal anxiety during a PHE.

Similar to anxiety, obstetric decisions revealed city-based differences. Key choices changed, including the time of prenatal care or delivery, mode of delivery and infant feeding, etc. These choices translated to eventual outcomes, according to our clinical observations (Table S5 ).

Although the factors that influence obstetric decisions are difficult to fully determine, we conjecture that the severity of the epidemic is likely to be the dominant factor, although requests for online consultation and the subjective impact of chest CT scans were sensitive to gestational age (Table 3 and Table S4 ).

Our hypothesis (Figure 1) is supported by evidence that anxiety can change prenatal decisions. 17, 34 A mediation analysis should be conducted to better understand the pathway of prenatal anxiety to obstetric decisions during a PHE.

Wuhan as the epidemic centre was included in our study. Chongqing, a nearby city, is a good comparison, not only for its epidemic condition (less affected but still in the outbreak, Figure S2 ), but also for its relatively high data accessibility and demographic comparability. Nevertheless, some limitations should be considered. Firstly, the study design suggests the possibility of self-report bias. Secondly, bias might exist as data from uncompleted questionnaires were inaccessible. Thirdly, the baseline was not completely balanced across cities. Efforts have been made to minimize biases: multiple centres were involved, the two cities provided a large and similar sample size, and sensitivity analysis was performed.

Our participants in Wuhan and Chongqing, generated relatively reliable and representative samples. More importantly, we hope that the real data from these two cities can serve as a reference and be promoted to more regions, especially to emerging epidemic areas worldwide.

This article is protected by copyright. All rights reserved Appropriate adjustments are needed before cross-region application. In the comparative results, Chongqing was hit harder than other regions far away from Wuhan, which may weaken the impact gap of COVID-19. Fortunately, descriptive parameters were recorded in each city and are likely to be accessible in cities that are the subjects of future studies.

Anxiety levels may have been underestimated. Pregnant women who are not registered on the platform or the hospital did not participate, and they may have a lower socio-economic status and higher anxiety levels. 33 The inclusion of only completed questionnaires may also have led to an underestimation. In addition, the majority of pregnant women were in the middle and late trimester, while the highest level of anxiety reported previously was in early pregnancy. 33 This was a cross-sectional study, and long-term cohort studies on post-traumatic stress disorder or postpartum mental state are merited.

During an unpredictable period, such as a PHE, comprehensive recommendations on prenatal care are needed. Our findings can serve as a reference in the following aspects:

Prenatal anxiety, which may affect the pregnancy outcomes, 21, 22, 35, 36 should be considered carefully. Psychological intervention and corresponding public health measures during PHEs are necessary, and need multidisciplinary cooperation. 9 These interventions could be: (1) to encourage the information from authoritative sources; (2) to correctly understand the susceptibility to disease;

(3) to perceive the risk rationally (avoid extremely perceiving symptoms); (4) to use online consultation and counselling.

To prevent irreversible adverse events, deferral of prenatal visits or delivery should not be standard care. 37 Our findings also suggested that caesarean delivery and bottle feeding may have increased as women sought to avoid vertical transmission. For researchers, as new evidence comes to light, 14,15,38 such transmission is likely to be rejected. Still, lessons could be learnt from similar PHEs (SARS and MERS). [39] [40] [41] For obstetricians, the contingency plans should include the capacity

This article is protected by copyright. All rights reserved to increase caesarean deliveries and manage potential perinatal infections; these plans need to be developed and implemented quickly. 42 Importantly, authoritative information on delivery and infant feeding should be available to prenatal mothers. The education about the radiography during pregnancy should focus on its comparative safety and diagnostic necessity, to decrease subjects' concerns.

In our study, more than 70% of participants requested for online support. Digital health merits increased investment.

The restrictive measures imposed, as a consequence of the crisis, have dual effects during PHEs, and a balance between epidemic control and mental pressure is required, whilst considering the availability of medical resources. 12, 31, 43 Meanwhile, the rationale of opening "green channel" for pregnant women has merit, as does support for this special group in the quarantine zones.

The outbreak aggravated prenatal anxiety, and the associated factors could be targets for psychological care. In parallel, key obstetric decision-making changed, emphasising the need for pertinent professional advice. Special support is essential for pregnant mothers during epidemics.

Here, we would like to thank Xinyu Zhou and his colleagues, professors in Psychiatry, for their advice on the questionnaire design and analysis. We also want to thank Professor Xiaoni Zhong, from the School of Public Health and Management in Chongqing Medical University, for her indispensable contributions to the statistical design and analysis. We would like to thank Dr. Kim C. Liu from University of Cambridge for his proof-reading and comments on the manuscript.

This article is protected by copyright. All rights reserved

The authors report no conflict of interest. Completed disclosure of interest forms are available to view online as supporting information.

HBQ, XL and XYL contributed to the protocol design. MMC and GQS collected data. LS and YW analysed data. JZ, YS, HZ and PB contributed to the interpretation of results. XYL drafted the manuscript and JHW proofread the manuscript. PB also assisted with detailed editing of the manuscript to get it to appropriate language. HBQ and XL revised the final version and are guarantors of this manuscript. All authors made substantial contributions to the paper and read and approved the final manuscript.

The study was approved by the ethics committee of the First Affiliated Hospital of Chongqing Medical University on 14 February, 2020 (ID:20200501), and an electronic informed consent was obtained before completing the questionnaire.

This work was supported by grants from National Natural Science Foundation of China (No. 81771614 and No. 81771613) , and the National Key Research and Development Program of China (No. 2016YFC1000407).

The data used and analysed during the current study are available from the corresponding author on reasonable request. Comparations were conducted across the two cities.

A, Data are mean±SD and median (IQR). Student's t test was used.

B, Data are n (%). Kruskal-Wallis test was used. § The Cohen's d for the SAS standard score was indicative of small-to-medium effect: 0.44. OR, odds ratio; CI, confidence interval. CNY, Chinese yuan. §, The attitudes ranged from "totally disagree" to "strongly agree".

This article is protected by copyright. All rights reserved 

Physical activity and blood pressure during pregnancy: Mediation by anxiety symptoms

National Health Commission of the People's Republic of China. Notice on strengthening maternal disease treatment and safe midwifery during the prevention and control of new coronavirus pneumonia

This article is protected by copyright. All rights reserved [https://www.who.int/dg/speeches/detail/who-director-general-s-opening-remarks-at-the-media-briefing-on-covid-19---28-february-2020]. Accessed 3 Mar 2020.

This article is protected by copyright. All rights reserved