key: cord-0692945-ufe8kmgm
authors: Johnson-Agbakwu, Crista E; Eakin, Cortney M; Bailey, Celeste V; Sood, Shelly; Ali, Nyima; Doehrman, Pooja; Bhattarai, Bikash; Chambliss, Linda; Coonrod, Dean V
title: SARS-CoV-2: A Canary in the Coal Mine for Public Safety Net Hospitals
date: 2021-03-21
journal: AJOG global reports
DOI: 10.1016/j.xagr.2021.100009
sha: e5146a9a39fedd326f2505358da1d563cf000270
doc_id: 692945
cord_uid: ufe8kmgm

Background The COVID-19 pandemic has exposed disproportionate health inequities among underserved populations, including refugees. Public safety net health care systems play a critical role in facilitating access to care for refugees, and informing coordinated public health prevention and mitigation efforts during a pandemic crisis. Objective To evaluate the prevalence of SARS-CoV-2 among refugee women admitted for delivery relative to non-refugee parturient patients. We suspect the burden of infection is disproportionately distributed across refugee communities which may act as sentinels for community outbreaks. Study Design A cross-sectional study was performed examining parturient women admitted to the maternity unit between May 6 and July 22, 2020, when universal testing for SARS-CoV-2 was first employed. Risk factors for SARS-CoV-2 positivity were ascertained, disaggregated by refugee status, and other clinical and socio-demographic variables examined. Prevalence ratios (PR) were calculated and comparisons made to county level community prevalence over the same time period. Results The percent positive at the County level during this study period was 21.6%. Of 350 women admitted for delivery, 33 (9.4%) screened positive for SARS-CoV-2. When disaggregated by refugee status, 45 (12.8%) were refugees, of whom 8 (17.8%) tested positive, compared to 25 (8.19%) non-refugee patients testing positive, PR 2.16 (95%CI 1.04-4.51). Seven of the SARS-CoV-2 positive tests were among refugees from Central Africa. Conclusion The SARS-CoV-2 outbreak has disproportionately affected refugee populations. This study highlights the utility of universal screening in mounting a rapid response to an evolving pandemic and how we can better serve the refugee community. Focused response may help achieve more equitable care related to SARS-CoV-2 among vulnerable communities. Identification of such populations may help mitigate spread and facilitate a timely, culturally and linguistically enhanced public health response.

the country have adopted universal testing of pregnant women, reporting rates among asymptomatic women of 0.0-13.5%. [5] [6] [7] [8] [9] [10] [11] While there is great geographic variation in prevalence of COVID-19, universal testing has been advocated for high prevalence regions to inform public health surveillance, care coordination, personal protective equipment (PPE) conservation, and to protect health care workers. 12 However, data is limited on the utility of universal testing in public safety net health care systems serving a higher proportion of racial/ethnic minorities and other marginalized communities, including refugees.

Arizona ranks among the top U.S. states welcoming newly arrived refugee populations, resettling nearly 83,000 refugees since 1975. 13 Valleywise Health Medical Center (VHMC), the public safety net health care system for the greater Phoenix Area (Maricopa County), has served over 9,000 refugees since 2008 from over 60 countries across Sub-Saharan Africa, South-East Asia and the

We suspected the burden of infection was significantly higher in the refugee community. As the pandemic evolved, initial data after the first four weeks of universal testing on L&D at VHMC demonstrated an alarming testpositive percentage of 27.3% in the refugee patients compared with 3.7% in nonrefugee patients. We considered this phenomenon the proverbial 'canary in the coal mine'. As the pandemic unfolded over the ensuing summer months, it became clear that the refugee population had been a sentinel for the subsequent disease burden. The test-positive percentage rates for Maricopa County subsequently grew to become among the highest in the world, portended by the early outbreaks within the refugee community. Our objective was to evaluate the prevalence of SARS-CoV-2 among refugee relative to non-refugee parturient patients alongside this growing pandemic.

A cross-sectional study was performed of all patients admitted to the L&D Unit for delivery at VHMC from May 6, 2020 to July 22, 2020, which was the first trial of universal rapid testing for all patients admitted to the Unit. VHMC is a tertiary academic teaching institution in Maricopa County, Arizona that performs approximately 2,000 deliveries annually.

Upon admission to the L&D Unit, all women were screened for symptoms of SARS-CoV-2 (fever or chills, cough, shortness of breath, fatigue, muscle or body aches, loss of taste or smell, sore throat, congestion, nausea or vomiting or diarrhea). SARS-CoV-2 testing was performed via a nasopharyngeal swab with the rapid Cepheid Xpert ® Xpress SARS-CoV-2 polymerase chain reaction (PCR) assay. This assay has a reported sensitivity of 98% and results were available within one hour. 14, 15 If the rapid test was reported as negative but there was a high degree of clinical suspicion, the patient also underwent the Abbott RealTime SARS-CoV-2 assay which has a sensitivity of 93% and specificity of 100%. 16 Hospital practice is to allow asymptomatic positive and mildly symptomatic patients the option of rooming in with their newborn following delivery. Viral testing was performed on all neonates born to SARS-CoV-2 positive mothers within 48 hours after birth. Patients who refused testing or left against medical advice before testing could be completed were excluded from the final analysis. Only women who were admitted and delivered were included in the study. Any non-delivered patients were excluded since these might comprise a heterogeneous group of subjects.

All hospital personnel wore full PPE (defined as an N95 respirator covered by a surgical mask, face shield and/or goggles, disposable head/shoe coverings, isolation gown and gloves) in all clinical encounters while awaiting test results

and continued for patients who tested SARS-CoV-2 positive. Hospital infection control protocols were employed for all patients who tested positive for SARS-CoV-2. All patients along with one allowable support person were given face masks to wear throughout their hospitalization. Support personnel were screened for symptoms of SARS-CoV-2 but did not undergo testing.

SARS-CoV-2 positive patients were categorized based on presence or absence of symptoms. In addition, data was collected regarding mode of delivery, need for admission to Infectious Disease Unit (IDU), neonatal Apgar scores and birth weight. (The IDU is the VHMC critical care unit dedicated to SARS-CoV-2 infected patients.) To determine whether the refugee population displayed a higher apparent prevalence, patient demographics were reviewed for linkage with a local refugee resettlement agency, language spoken, and mother"s nativity were ascertained. Refugee status was designated by any of the following: patient received prenatal care in VHMC"s Refugee Women"s Health Clinic (RWHC), patient is from a refugee source country and/or country of first asylum. 17 Other risk factors examined included age, race/ethnicity, gravity/parity, gestational age at delivery, body mass index (BMI), and insurance status.

General demographic and clinical features of the patients were described with proportions for count variables. Measures of central tendency and dispersion were calculated for continuous variables at 4-weeks and 11-weeks (at completion of universal screening period). Chi-square tests for factors and Wilcoxon ranksum test for non-normal continuous variables were used to compare between test-positive and negative groups across demographic strata. We examined severe maternal morbidity (SMM) indicators as defined by the CDC and corresponding ICD-10 codes during delivery hospitalizations during the study period to determine whether refugee status was associated with co-morbidity. 18 To calculate the percent positive per day in Maricopa County, the number of confirmed positive cases per day was divided by the number of SARS-CoV-2 PCR tests performed per day for the study period. A summative proportion was also calculated for the study period. These calculations were performed using publicly available data through the Arizona Department of Health Services Data Dashboard.

This study was approved by the Valleywise Health Institutional Review Board (IRB protocol # 2020-044).

During the study period, 416 SARS-CoV-2 tests were performed on Labor & Delivery (L&D) at VHMC. Only patients screened for SARS-CoV-2 and admitted for delivery and who ultimately delivered were included (N = 350). No patient refused testing. Patient characteristics, delivery and neonatal outcomes are summarized in Table 1 . The statistical analysis demonstrated refugee patients were older, had a lower BMI, more likely to deliver at term, less likely to have Federal Emergency Services coverage and more likely to have a low Apgar score at delivery (Table 1) .

A total of 310 (88.57%) women delivered full-term and 40 (11.43%) were preterm (<37 completed weeks of gestation). There were 3 intrauterine fetal demises and 1 second-trimester loss. The primary cesarean section rate was 9.3% while the repeat cesarean section rate was 19.4%.

At the completion of the study period (11-weeks) , the total number of positive SARS-CoV-2 tests within the general population during the study period was 110,301 in Maricopa County, with an overall SARS-CoV-2 test-positive percentage of 21.6% ( Figure 1 ). 19 Of the 350 patients tested at time of admission to L&D, 33 women were SARS-CoV-2 positive, resulting in a testpositive percent of 9.43%. Seven of the 33 women who tested positive were symptomatic for SARS-CoV-2 infection. A total of 32 neonates were tested for SARS-CoV-2 after delivery, one stillborn delivery to a SARS-CoV-2 positive mother was excluded. Two (6.3%) of the 32 neonates tested were positive. One Table 2) . A disproportionate burden of co-morbidities was not identified between refugee and non-refugee patients as co-morbidities were too sparse for valid statistical comparisons.

To our knowledge this is the first reported study examining SARS-CoV-2 prevalence among refugee women receiving maternity care in the United States.

Our experience with SARS-CoV-2 universal testing demonstrates the utility of incorporating routine screening on L&D to identify high-risk populations. In the first four weeks of universal testing, the test-positive percentage on L&D was similar to Maricopa County (7.8% and 9.0% respectively). During this time, however, the prevalence ratio of positive status among refugee patients was 7 times higher than their non-refugee counterparts (27.3% versus 3.7% respectively). After completion of 11 weeks of universal testing, the total testpositive percentage for Maricopa County grew to parallel that of the refugee population tested on L&D (21.6% and 17.8% respectively). Implementation of universal testing for SARS-CoV-2 enabled early identification of a high proportion of SARS-CoV-2 among refugee communities during the early days of a rapidly evolving pandemic crisis, thus encapsulating the proverbial 'canary in the coal mine', which portended the gravity of an emerging calamity. Consequently, a more robust and coordinated public health response was mounted to support innovative, culturally and linguistically appropriate strategies to enhance testing, contact tracing, isolation and educational outreach specifically in the refugee community.

Universal testing allows for rapid identification of SARS-CoV-2 positive mothers, identification of asymptomatic carriers and the subsequent allocation of limited inpatient resources that would otherwise not be possible with symptomatic screening alone. While universal testing has the potential to identify marginalized populations, asymptomatic pregnant women may not be representative of the community prevalence of asymptomatic viremia, thus limiting generalizability. 5, 20 Nevertheless, this study"s demonstrated overall test-positive percentage on L&D was only slightly higher than county-level statistics. We also recognize our small sample size over 11 weeks of data collection at a single institution. However, our findings of profound disparity in the refugee population justifies timely release of preliminary findings to guide further research and to mobilize enhanced public health responses.

We found that refugee women were at much higher risk of infection.

Systemic racism and structural inequities disproportionately affect communities of color and have profound implications for achieving health equity. 21, 22 Among marginalized communities, additional factors such as economic instability, poor health literacy and environmental disparities may further exacerbate inequalities.

SMM has been shown to be higher in migrant populations. 23 Furthermore, there is an increased risk of disease morbidity associated with SARS-CoV-2 infection in populations with underlying medical co-morbidities. 24, 25 However, in our study, the presence of co-morbidities was too sparse to compare refugee and nonrefugee groups. The disproportionate burden of SARS-CoV-2 infection in the refugee population may be more indicative of factors related to acquisition of disease due to social determinants of health rather than underlying comorbidities. [26] [27] [28] [29] [30] The COVID-19 pandemic has further illuminated such disparities. 2 Table 3 and must be considered when coordinating a public health response to the COVID-19 pandemic.

Our finding of a higher proportion of Central African refugees among our SARS-CoV-2 parturients argues for the need for greater specificity in race/ethnicity data collection to include language and/or nativity in order to 13 capture hidden communities of color who would otherwise not be captured. The umbrella terms "Black" or "African American" encompasses not only U.S.-born Blacks who are direct ancestral descendants, possessing the cultural and historical legacy of enslaved Africans, but also immigrants from the African Diaspora who have arrived to the U.S. in recent decades from Africa, the Caribbean and Europe. Consideration must be made in balancing the ethics of universal testing while avoiding unintentional harm in singling out particular ethnic/cultural groups already stigmatized due to their refugee status, as identification of these vulnerable and marginalized communities may have profound implications for mounting a culturally and linguistically inclusive public health response. 7, 36, 37 Our findings also have wider generalizability to bolstering public health responses in vulnerable, otherwise hidden, communities across the United States. Relatively higher risk among patients receiving care in a public safety net health care system and further higher risk among refugees also elucidates the importance of focused testing among vulnerable communities which may act as sentinels for community outbreaks. Public health policies should consider prioritizing universal testing in public safety net hospitals in order to guide a more targeted response in marginalized populations. Possible strategies within refugee communities may include the deployment of mobile testing units, utilization of multilingual community health workers to facilitate contract tracing and education, and prioritization of family-centered isolation in transitional housing for large, multigenerational families. Furthermore, public health infrastructure should foster coordination of medical care in partnership with refugee resettlement and social service agencies. Following these early findings, Maricopa County is now implementing several of these recommendations in attempts to reduce the spread of COVID-19 within the refugee population.

The major strength of this paper is that the current literature on the COVID-19 pandemic has yet to describe the burden of disease felt by the refugee population in the United States. The generalizability of this study is limited due to our primary focus on quantifying SARS-CoV-2 infection specifically among parturient patients. Hence our findings are not estimates of true prevalence in the community but instead, a demonstration of disparity in the refugee population which will inform public health interventions. However, since presenting for delivery is not likely to be influenced by the presence or absence of infection, it might represent a less biased estimate of the difference in the populations. The statistical analysis is likely weakened by the small total number of patients included. We hope that ongoing data collection will improve the strength of this analysis.

Public safety net health care systems are playing a crucial role in facilitating access to care for underserved populations throughout this COVID-19 pandemic.

Universal testing for SARS-CoV-2 on maternity units in these systems provide Table 2 Footer: Data are n (%) unless otherwise specified. Table 3 

Detail Economic Stability and Employment -Financial constraints resulting in obligation to work despite symptoms or known positive status -Need for financial assistance with rent, utilities and food expenses if employment is lost -Disconnected cell phone service causing barriers to contact tracing -Inability to socially distance at work -Insufficient access to PPE -Frontline, essential worker, unable to work from home -Lack of paid sick time from employer -Fear of disclosing positive status or contacts with positive status due to fear of loss of employment Environment and Neighborhood -Shared communal spaces -Common retail stores without adequate PPE or distancing measures -Population density, high number of residential units per building -Lack of independent transportation -Refusal of transportation agencies to provide services for symptomatic patients -Lack of access to food delivery services if symptomatic/quarantined Health Care and Health Literacy -Lack of established preventative care with primary care providers -Limited understanding of impact of positive status -Limited understanding of possible infectivity despite lack of symptoms -Limited understanding of social distancing needs -Limited understanding of CDC self-quarantine guidelines if exposed -Seeking care at more advanced stages of disease and/or symptomatology -Obtaining medical information from unreliable social media sources -Decreased access to quality care due to uninsured/underinsured status Language -Delay in reporting results to patient due to language barriers -Lack of multilingual contact tracers -Lack of health provider linguistic or cultural competency -Failure to trust health care providers due to impaired understanding from incomplete/inadequate translation services Household and Cultural -Large household membership within small residential housing units -Multigenerational households -High value in familial relations resulting in reluctance to selfquarantine -Myths regarding immunity to illness -Perception that the current crisis does not pertain to them -Hyper-religiosity that God alone will protect them without taking necessary preventive precautions

MMWR -Hospitalization Rates and Characteristics of Patients Hospitalized with Laboratory-Confirmed Coronavirus Disease 2019 -COVID-NET, 14 States

The Color of Coronavirus: COVID-19 Deaths by Race and Ethnicitiy in the

The fullest look yet at the racial inequity of coronavirus. The New York Times

The COVID Racial Data Tracker

Clinical implications of universal severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) testing in pregnancy

Concerns of stigmatization resulting in challenges to contact tracing and expanded testing of identified high-risk areas

Is universal testing for severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) needed on all Labor and Delivery units?

Universal screening for SARS-CoV-2 in women admitted for delivery

Screening all pregnant women admitted to Labor and Delivery for the virus responsible for COVID-19

Prevalence of SARS-CoV-2 among patients admitted for childbirth in Southern Connecticut

Universal testing for COVID-19 in pregnant women admitted for delivery: prevalence of peripartum infection and rate of asymptomatic carriers at four New York hospitals within an integrated healthcare system

American College of Obstetricians & Gynecologists

Study Raises Questions About False Negatives From Quick COVID-19 Test

Multi-center evaluation of cepheid xpert® xpress SARS-CoV-2 point-of-care test during the SARS-CoV-2 pandemic

Validation and verification of the Abbott RealTime SARS-CoV-2 assay analytical and clinical performance

Health and health care of African refugees: An underrecognized minority

How Does CDC Define Severe Maternal Morbidity? Center for Disease Control and Prevention

Arizona Department of Health Services. Arizona"s emergency response to the SARS-COV-2 outbreak: Data Dashboard

The Relationship between Status at Presentation and Outcomes among Pregnant Women with COVID-19

Racism and health: evidence and needed Research

Structural racism and health inequities in the USA: evidence and interventions

Severe maternal morbidity associated with maternal birthplace in three high-immigration settings

Clinical Update on COVID-19 in pregnancy: A review article

Coronavirus disease 2019 in pregnancy was associated with maternal morbidity and preterm birth

Outbreak of SARS-CoV-2 infection at a large refugee shelter in Toronto

Potential Impact of COVID-19 on Recently Resettled Refugee Populations in the United States and Canada: Perspectives of Refugee Healthcare Providers

Impact of COVID-19 on Resettled Refugees

COVID-19 and Immigrant Essential Workers: Bhutanese and Burmese Refugees in the United States

Critical Considerations for COVID-19 Vaccination of Refugees, Immigrants, and Migrants

COVID-19 exacerbating inequalities in the US

Associations between built environment, neighborhood socioeconomic status, and SARS-CoV-2 infection among pregnant women in New York City

Healthcare barriers of refugees post-resettlement

COVID-19 and U . S . -Based Refugee Populations : Commentary

Maine has nation"s worst COVID-19 racial disparity

Consolidated principles for screening based on a systematic review and consensus process

Testagem universal de COVID-19 na população obstétrica: impactos para a saúde pública