Giving birth under hospital visitor restrictions: Heightened acute stress in childbirth in COVID-19 positive women

As the novel coronavirus (COVID-19) has spread globally, a significant portion of women have undergone childbirth while possibly infected with the virus and also under social isolation due to hospital visitor restrictions. Emerging studies examined birth outcomes in COVID-19 positive women, but knowledge of the psychological experience of childbirth remains lacking. This study survey concerning childbirth and mental health launched during the first wave of the pandemic in the US. Women reporting confirmed/suspected COVID-19 during childbirth were matched on various background factors with women reporting COVID-19 negative. We found higher prevalence of clinically significant acute stress in birth in COVID-19 positive women. This group was 11 times as likely to have no visitors than matched controls and reported higher levels of pain in delivery, lower newborn weights, and more infant admission to neonatal intensive care units. Visitor restrictions were associated with these birth outcomes. COVID-19 positive women with no visitors were 6 times as likely to report clinical acute stress in birth than COVID-19 positive women with visitors. The findings underscore increased risk for childbirth-induced psychological morbidity in COVID-19-affected populations. As hospitals continue to revise policies concerning visitor restrictions, attention to the wellbeing of new mothers is warranted.


57
The coronavirus (COVID-19) pandemic's immense scope and duration has made clear the 58 urgent need to better understand the virus' physical and psychological impacts on vulnerable 59 and the American College of Obstetricians and Gynecologists (ACOG), a better understanding of 108 the psychological childbirth experiences in COVID-19 vulnerable mothers, such as those being 109 suspected or confirmed of infection, is warranted. In the writing of this work, visitor prohibitions 110 have been largely lifted in maternity wards. However, those delivering who test positive for 111 COVID-19 may still face social isolation and not be allowed any visitors during their entire 112 hospital stay, underlining the importance of relevant research. 113 To this end, we studied a large sample of women who recently gave birth when  19 was prevalent in the United States, among them 68 women reported suspected or confirmed 115 COVID-19 positive. We matched this group on a wide range of background factors to 68 women 116 who gave birth in the outbreak of the pandemic but were negative for COVID-19. There have been 117 no studies to date that use a comprehensive matched-group analysis that could allow for better 118 understanding of the contribution of COVID-19 positivity to childbirth outcomes while controlling 119 for background factors that increase perinatal adversity. We examined whether being COVID-19 120 positive is associated with stressful psychological experiences of birth as well as obstetrical and 121 neonatal outcomes and whether having no visitors during delivery hospitalization stay was 122 associated with these outcomes. 123

124
Participants 125 This study is part of a research project that was launched on April 2 nd , 2020, in the midst 126 of the COVID-19 pandemic in the United States, with the overarching goal of understanding the 127 impact of COVID-19 on childbirth and maternal mental health. Women who had given birth in the 128 last six months were recruited through announcements on our hospital's research study platform 129 as well as via social media and postpartum professional communities; they were asked to complete 130 an anonymous survey. Partners Healthcare (Mass General Brigham) Human Research Committee 131 granted exemption for this study. The sample in this study was derived from 2,417 women who 132 gave birth since COVID-19 was prevalent in their communities and provided the childbirth date; 133 they were on average two months postpartum. We identified 68 women who reported being 134 COVID-19 positive, suspected or confirmed, during pregnancy and/or childbirth. We then 135 identified a matched control group of 68 women who reported being COVID-19 negative. The 136 groups were matched on demographic factors, primiparity, prior trauma and childbirth history, and 137 prior mental health. 138 In this sample of a total of 136 postpartum women, the vast majority delivered a healthy 139 baby at term (86.8%), had a vaginal delivery (71.3%), and around half (50.7%) were primiparas. 140 The average age of participants was 32 years old. The majority were married (89%), had at least 141 middle-class income (i.e., $100,000 per year, 66.2%), were employed (72.1%), and had at least a 142 college degree (83.1%). Participants resided in the United States (80.0%), in Canada (4.4%), 143 Europe (2.9%), Central/South America (2.9%), Asia (2.9%), and 2.2% in the Caribbean and 144 Middle East. Four participants (2.9%) did not report their geographic location. 145

146
Acute stress responses to childbirth were assessed with the commonly used Peritraumatic 147 Distress Inventory (PDI) [17]. The PDI is a 13-item self-report with good psychometric properties. 148 It assesses negative emotional responses (e.g., "I felt helpless"; "I thought I might die") 149 experienced during and/or immediately after a specified traumatic event on a 0 (not at all) to 4 150 (extremely true) scale. In this study, participants rated their responses in regard to their recent 151 childbirth experience. The PDI has been used to assess acute childbirth-related stress in postpartum 152 samples [18]. To define clinically significant acute stress response symptoms, we used the 153 suggested cutoff of 17 [19]. Reliability in the current study was high ( = 0.91). 154 All rights reserved. No reuse allowed without permission.
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted December 2, 2020. To create matched groups who share similar background characteristics between COVID-176 19 positive and negative women, we conducted a propensity score matching procedure using SPSS 177 logistic regression, the matching algorithm was nearest neighbor matching with caliper of 0.

Group matching 196
An overall balance test [23] indicated that the balance of the matching was high, χ 2 (27) = 197 18.21, p = .90, such that each group comprised 68 women. 198

COVID-19 positive birth-related outcomes 199
Percentages of birth-related outcomes are presented in Table 1; mean differences are 200 presented in Table 2. 201 All rights reserved. No reuse allowed without permission.
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted December 2, 2020. To examine whether no visitors during delivery hospitalization account for the differences 219 between the COVID-19 positive and negative groups (i.e. in pain in delivery, NICU admission, 220 infant weight, rooming in, and acute stress response to birth), we compared COVID-19 negative 221 women who had visitors (n = 66), with COVID-19 positive women who had visitors (n = 51) and 222 those with no visitors (n = 17). Percentages are presented in Table 3; mean differences are 223 presented in Table 4. 224 225 All rights reserved. No reuse allowed without permission.
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted December 2, 2020. The analyses indicated that COVID-19 positive women who had no visitors reported 235 significantly greater pain in delivery (see Figure 1) and delivered infants with lower weights (see 236 Figure 2). In addition, their infants were more likely to get admitted to the NICU and less likely to 237 be in the same room with their mothers during the hospital stay. Finally, COVID-19 positive 238 All rights reserved. No reuse allowed without permission.
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted December 2, 2020. ; https://doi.org/10.1101/2020.11.30.20241026 doi: medRxiv preprint women with no visitors had much higher prevalence of acute stress responses at a clinical level 239 (see Figure 3). 240 Figure 1. Pain in delivery by study group. Red dots represent the mean. 241 All rights reserved. No reuse allowed without permission.
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted December 2, 2020. ; https://doi.org/10.1101/2020.11.30.20241026 doi: medRxiv preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

263
The COVID-19 pandemic offers a rare opportunity to examine the experience of childbirth 264 under stressful conditions such as social isolation. As infectious disease outbreaks continue, it is 265 critical that we generate new knowledge to inform preparations and guidelines of perinatal care 266 during these outbreaks. 267 Our study sought to examine the childbirth experiences of women who had delivered The main study findings show that nearly 50% of suspected or confirmed COVID-19 279 positive women reported clinically significant acute stress symptoms in birth. They were as much 280 as two times more likely to experience acute stress than non-affected women and to perceive higher 281 degrees of pain in childbirth even though no differences were found in factors such as obstetrical 282 complications, medication for pain, or delivery mode between COVID-19 positive and negative 283 cases. These findings underscore how childbirth can become a traumatic experience and evoke an 284 acute stress response for women with the novel coronavirus. 285 All rights reserved. No reuse allowed without permission.
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted December 2, 2020. ; https://doi.org/10.1101/2020.11.30.20241026 doi: medRxiv preprint We further document increased exposure to salient social stressors surrounding childbirth 286 in affected women. As might be expected, the results reveal that hospital policies enforcing visitor 287 restrictions were frequently implemented with delivering women suspected or confirmed of 288 COVID-19 infection. As much as 25% of COVID-19 positive women had no visitors during their 289 delivery hospitalization stay. This group was 11 times as likely not to be permitted a support person 290 to accompany them than women negative for COVID-19. COVID-19 positive women were also 291 much more likely to experience physical separation from their newborn. In accord with previous 292 studies [7], the newborns were nearly four times as likely to be admitted to the NICU. In the writing 293 of this manuscript, the CDC has updated its guidance and currently recommends rooming-in for a 294 COVID-positive mother and her newborn and acknowledges that the decision should be 295 determined by the family. 296 Our findings reveal that social isolation surrounding childbirth may increase risk for 297 maternal morbidity. We found heightened clinically significant acute stress in COVID-19 positive 298 women who had no visitors. They were 6 times as likely to report acute stress symptoms than 299 COVID-19 affected women who were permitted visitors during their delivery hospitalization. We 300 also found that COVID-19 positive women who did not have a support person experienced greater 301 pain in delivery, delivered newborns with lower weight, and had elevated NICU admission rates. 302 These findings accord with the evidence of emotional comfort and support in birth being associated 303 with improved birth outcomes [14], and suggest how a diminished sense of support may increase 304 maternal stress and subsequent adversity. Psychological traumatic morbidity in birth has been 305 shown to result in maternal mental illness during the postpartum period based on pre-COVID 306 samples [24,25] and has also been documented in women who gave birth since the pandemic [26]. 307 All rights reserved. No reuse allowed without permission.
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. as those with stable symptoms in accord with routine care are likely to be quickly discharged and 318 face social isolation during the postpartum period, which is considered a time of heightened 319 psychological vulnerability [27,28]. 320 Shortcomings of this study include reliance on anonymous self-report measures that 321 allowed for conducting a study swiftly during the initial heights of the pandemic but not for 322 inclusion of patients' medical records. We rely on respondents accurately reporting their COVID-323 19 infection status, and their receiving accurate information from COVID-19 testing protocols at 324 the hospitals where they delivered. Additionally, we do not have information on the severity of 325 respondents' COVID-19 symptoms, only their infection status. We cannot rule out that acutely ill 326 women were those not permitted visitors. Also, while we used a well-validated measure to assess 327 acute stress which has shown good correspondence with clinician assessments, we did not include 328 diagnostic measures. Retrospective assessments could be prone to recall bias and hence the 329 All rights reserved. No reuse allowed without permission.
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted December 2, 2020. ; https://doi.org/10.1101/2020.11.30.20241026 doi: medRxiv preprint importance of the use of matched controls. This convenient internet sample introduces a bias 330 towards women from a certain socioeconomic class. 331

332
In conclusion, we find that confirmed or suspected COVID-19 positive women experience 333 increased psychological morbidity surrounding childbirth compared to delivering women without 334 COVID-19. We find that COVID-19 positive women experience increased levels of pain during 335 delivery and give birth to newborns of lower weight which are more likely to be separated from 336 their mothers and sent to the NICU. This increased adversity appears especially heightened in 337 cases where a support person is not allowed in the maternity unit. As hospitals around the world 338 continue to update their delivery protocols for COVID-19 positive women and determine risk and 339 benefits of visitor restriction policies, more research is needed to optimize maternal care during 340 these unprecedented times. 341 All rights reserved. No reuse allowed without permission.
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.