During the COVID-19 pandemic surge in New York, several hospitals in New York City and Long Island began testing all women presenting to the labor and delivery units for SARS-CoV-2. They found that 14% of asymptomatic women tested positive.Reference Sutton, Fuchs, D’Alton and Goffman1,Reference Vintzileos, Muscat and Hoffmann2 Unidentified, these asymptomatic women were at risk of infecting their newborns following birth, hospital staff, as well as other patients. It is unclear, however, whether the high rate of asymptomatic infections in New York is a reflection of a particularly high prevalence of SARS-CoV-2 during that time period in New York or a more generalizable phenomenon applicable to other high-prevalence areas. Boston followed New York as another high-prevalence metropolitan area (1,628 cases per 100,000 residents vs 2,046 in New York City as of May 1, 2020). We therefore report on the prevalence of asymptomatic SARS-CoV-2 in women presenting to the labor and delivery units in Boston, another high-prevalence community in the United States.
Methods
On April 18, 2020, 2 academic and 2 community hospitals affiliated with Mass General Brigham Health began universally testing all women admitted to their labor and delivery units for SARS-CoV-2 using RT-PCR 53 (nasopharyngeal swab). Prior to this intervention, multiple infection control strategies in addition to those routine in our facilities had been implemented in response to the COVID-19 pandemic: (1) symptom and exposure screening of all patients with implementation of immediate isolation if symptom screen is positive and testing for SARS-Cov-2, (2) universal masking of employees, patients, and visitors on facility premises,Reference Klompas, Morris, Sinclair, Pearson and Shenoy3 (3) daily employee symptom attestation with exclusion from work and referral for testing if symptom screen positive; and (4) deferral of all nonessential in-person visits and elective procedures.
Demographic and SARS-CoV-2 test results were abstracted from the electronic medical record for all women admitted to the labor and delivery units between April 18, 2020, and May 5, 2020. All records for women with positive tests on admission were independently reviewed by 2 physicians (I.T.G. and D.K.) to confirm symptom status based on established symptom screening including fever (subjective or documented), new cough, shortness of breath, sore throat, muscle aches, new rhinorrhea, or new anosmia). The descriptive data are presented as frequencies.
Results
The 4 major hospitals affiliated with Mass General Brigham Health provide maternity care to ~14,750 women per year. Over 18 days of universal testing on the labor and delivery units, 763 women were admitted and 757 (99.2%) were tested. Of those, 139 had symptoms possibly consistent with COVID-19. Of symptomatic women, 11 of 139 (7.9%) tested positive. Among asymptomatic women, 9 of 618 (1.5%) tested positive (Fig. 1). Thus, 9 of 20 patients positive for SARS-CoV-2 at admission (45%) had no symptoms of COVID-19 at presentation. The percentage of asymptomatic women who tested positive varied by hospital: 2.7% and 1.5% in the 2 academic hospitals, 1.8% and 0.6% in the 2 community hospitals. Across the 4 hospitals, none of the positive asymptomatic women developed COVID-19 symptoms during the delivery hospitalization and all 9 newborns tested negative for SARS-CoV-2.
All women tested for SARS-CoV-2 on the labor and delivery units.
Discussion
In a large healthcare system in metropolitan Boston, we identified a low prevalence of COVID-19 infection among asymptomatic pregnant women presenting for admission to the labor and delivery units. The incidence of asymptomatic infection amongst women admitted to the labor and delivery units in greater Boston was substantially lower than that of New York City despite similar case counts per capita. Notably, the 1%–2% incidence of asymptomatic infection in our population more closely mirrors asymptomatic infection rates in other areas.Reference Ng, Marimuthu and Chia4,Reference Hoehl, Rabenau and Berger5 Several theories may explain the lower prevalence of asymptomatic infection in Boston compared to New York City: (1) we began testing >30 days after physical distancing orders were placed by the state and hence were sampling at a time with declining community transmission, (2) the overall population density of greater Boston is lower than New York City, perhaps leading to less community-based transmission, and (3) some New York hospitals transiently stopped or considered stopping birth partners from attending deliveries, which could have led to some women underreporting symptoms.
Universal testing of women presenting for labor and delivery, as one element of a multipronged approach to reducing the risk of SARS-CoV-2 transmission in healthcare facilities, is likely to remain a core strategy for the foreseeable future to inform both clinical care and infection control operations. Universal testing in this specific patient population is an especially important public health priority given the implications of SARS-CoV-2 on maternal and newborn care at the time of birth and during the postpartum and neonatal period. In addition, testing the asymptomatic obstetric population provides a window into the community prevalence of infection which in turn can inform the timing and effect of when, where, and how to enhance versus relax social distancing measures. Assessing the community-based COVID-19 prevalence rates must take into account the possibility of local clustering of disease where a community lies within the pandemic curve and the status of contemporaneous mitigation strategies. These data may, therefore, guide decision making about moving between mitigation versus containment measures and thoughtfully resuming both healthcare and nonhealthcare operations.
Acknowledgments
The authors would like to acknowledge Karen E. Lynch, BSN of Massachusetts General Hospital, Laboratory of Computer Science for assistance with data abstraction and analysis.
Financial support
No financial support was provided relevant to this article.
Conflicts of interest
All authors report no conflicts of interest relevant to this article.
Open access
