key: cord-0859533-wlezt16a
authors: Veronica, Fitzpatrick; Anne, Rivelli; Christopher, Blair; Kenneth, Copeland; Jon, Richards
title: Incidence of COVID-19 recurrence among large cohort of healthcare employees
date: 2021-04-22
journal: Ann Epidemiol
DOI: 10.1016/j.annepidem.2021.04.005
sha: f5527d0d9343e6f28ba07394b80629f8c6b4c15d
doc_id: 859533
cord_uid: wlezt16a

PURPOSE: To quantify COVID-19 recurrence among clinical and non-clinical healthcare employees with SARS-CoV-2 IgG antibodies or prior COVID-19 infection. METHODS: This prospective, cohort study collected and resulted SARS-CoV-2 IgG serum samples as positive or negative from June 8 to July 10, 2020 from a convenience sample of 16,233 adult participants employed by a large Midwestern healthcare system. Documented positive PCR test results representing COVID-19 infections were recorded up to four months prior to and post-IgG testing. RESULTS: 913 (6.12%) participants, including 45 (4.93%) IgG positive participants, experienced COVID-19 infections after study initiation, representing a 51% increased risk of COVID-19 infection among IgG positive participants (IRR=1.51). Regressions adjusted for documented disparities showed no difference in COVID-19 infection by IgG status (OR=1.19; p=0.3117) but significantly greater odds in COVID-19 recurrence among participants with a prior documented COVID-19 infection (OR=1.93; p<0.0001). CONCLUSIONS: SARS-CoV-2 IgG antibodies and prior COVID-19 infection do not appear to offer meaningful protection against COVID-19 recurrence in healthcare workers. Recurrence would impact decisions regarding ongoing healthcare resource utilization. This study can inform considerations for vaccine administration to vulnerable groups.

Introduction COVID-19 was initially thought to be an immunizing, non-relapsing disease, but current research is mounting to suggest this is not the case. 1,2 Currently, we have limited understanding of the innate and adaptive immunity to the novel SARS-CoV-2 virus. 3, 4, 5 Initial false assumptions of immunity, paired with a slew of treatment and prevention challenges, served to delay efforts to recognize and understand recurrence of COVID- 19 , the disease manifestation of the SARS-CoV-2 virus. 2, 5, 6, 7 SARS-CoV-2 entered a highly-susceptible human population, resulting in its rapid and uncontrollable transmission. 1 The swift release of substandard polymerase chain reaction (PCR) tests to detect active SARS-CoV-2 ribonucleic acid (RNA) and concurrent widespread shortages of aforementioned tests contributed to misunderstanding recurrence potential. 8 Worldwide policies focused on disease containment also contributed to setbacks in documenting recurrence. 1 Evidence of recurrence of COVID-19, mostly as single or small-study case reports, are just beginning to emerge.

Detection of active SARS-CoV-2 RNA after clinical disease recovery could indicate a multitude of scenarios, including: persistent illness, prolonged viral RNA shedding, variation in collection technique, specimen handling or storage conditions affecting test performance, or, most alarming, genuine recurrence of disease. 5 Current research has raised two pathophysiological hypotheses underlying demonstrated COVID-19 recurrence: viral reactivation or viral reinfection. 7 Despite case studies offering preliminary evidence of COVID-19 recurrence from China, Korea, Italy, and the US, [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] circumstances have made it difficult to document recurrence. First, the difference between viral reactivation and viral reinfection is not well-defined, convoluting this area of knowledge. 7, 9 Additionally, PCR test shortages limited SARS-CoV-2 testing, discounting true prevalence of all infections and limiting capacity to understanding recurrence. Insufficient resources also reduced abilities to preserve or test samples for reinfection using genetic testing. Finally, experts have maintained that the immune system will function as it should, developing immunity to this virus as it has with prior viruses. 19 These circumstances have perpetuated beliefs that COVID-19 recurrence is very rare and natural herd immunity is possible; 19 however, contradictory evidence supports both the hypothesis of recurrence rarity and, in some studies, recurrence commonality. 20 Subsequently, there remains a prominent gap in COVID-19 recurrence knowledge and the role that SARS-CoV-2 antibodies may play in protection against COVID-19 recurrence. 4, 21 The asymptomatic transmission of the virus has raised barriers to documenting recurrence using PCR testing for active SARS-CoV-2 RNA alone. 5 Research has asserted that serologic tests are alternative measures of exposure where difficult to isolate or is no longer present. 18 As COVID-19 continues to sweep the US, the use of PCR testing for COVID-19 following SARS-CoV-2 serologic test is an alternative approach to capturing incidence of recurrence after prior exposure to or infection with the SARS-CoV-2 virus. 5, 7 Documenting SARS-CoV-2 recurrence Serology, or antibody, tests are one method to assess proteins made by the body's immune system to fight antigens in response to a prior infection. 22,23 A positive IgG serology test result indicates prior exposure to the SARS-CoV-2 virus and subsequent response in the form of antibody development. 24 While antibody presence to SARS-CoV-2 has not been evidenced to suggest any length of protection from recurrencea gap this study will attempt to addressit does indicate response and recovery of prior exposure. 20

The immune response to COVID-19 cannot fully be understood without more data on immunity and recurrence. 5 The primary objective of this study was to establish the incidence of COVID-19 infection, as measured by a documented positive PCR test result, among healthcare employees tested for SARS-CoV-2 IgG antibodies. This study will provide overall incidence of COVID-19 infection and incidence rate ratio (IRR) per 100 person-days among participants with positive IgG status relative to negative IgG status. This study will also determine adjusted odds ratios (AdjOR) of COVID-19 recurrence among healthcare employees using two definitions of prior exposure to SARS-CoV-2, including: 1) positive SARS-CoV-2 IgG status and 2) prior documented COVID-19 infection status. For the purposes of this study, recurrence refers to the larger umbrella term encompassing reactivation and reinfection. This study will not attempt to delineate between reactivation and reinfection, but instead will address SARS-CoV-2 recurrence, defined as documented COVID-19 infection after positive IgG status (primary analysis) or after prior documented COVID-19 infection (secondary analysis).

This study builds off a prior study by the same authors that documented disparities in seroprevalence and determined the seroprevalence of SARS CoV-2 IgG antibodies was 3.83% among 16,233 healthcare employees. 26

This prospective cohort study recruited healthcare employees across a large Midwestern healthcare system, which consists of about 70,000 employees across 26-hospitals and over 500 sites of care in Illinois and Wisconsin. SARS-CoV-2 IgG was measured in serum specimens obtained from all participants at study initiation using the SARS-CoV-2 IgG Abbott Architect assay. Performance characteristics of the SARS-CoV-2 IgG assay were validated at ACL Laboratories, determining a sensitivity of 98.7% and specificity of 99.2%. [27] [28] [29] SARS-CoV-2 RNA, as detected by a positive PCR test representing COVID-19 infection, was measured from isolated and purified nasopharyngeal, oropharyngeal and nasal swab specimens and obtained from individuals who met COVID-19 clinical and/or epidemiological criteria and opted to undergo PCR testing within the healthcare system using the Aptima Panther SARS-CoV-2 assay. 27 Both assays were approved for use under Emergency Use Authorization in US laboratories certified under the Clinical Laboratory Improvement Amendments of 1988. 29 Prior to recruitment, this study obtained approval by the healthcare system's Institutional Review Board (#20-168E). This study was authorized to enroll up to 20,000 participants or complete SARS-CoV-2 IgG assays until July 10, 2020. Participants' positive SARS-CoV-2 RNA results were recorded by the healthcare system's Employee Health Department and collected by the research team until October 10, 2020.

This study enrolled and tested a convenience sample of 16,293 participants for SARS-CoV-2 IgG assay results between June 8, 2020 and July 10, 2020 and followed them until October 10, 2020 for positive PCR test results representing COVID-19 infection documented in the system's Electronic Medical Records (EMR). This study also recorded positive PCR results up to four months prior to and post-SARS-CoV-2 IgG testing, which established study initiation. For study inclusion, English-and Spanish-speaking adults ages ≥ 18 employed by the healthcare as of the study initiation date were eligible. Team members who met study inclusion criteria and completed a lab blood draw to test for SARS-CoV-2 IgG were participants in this study. Primary analysis 

On June 6, 2020, a detailed recruitment email was sent to all team members' work email addresses. The email provided instructions for participation in the study, including an alteration of consent and a study-specific passcode required for study registration. Interested team members were instructed to register for a study-related IgG assay in their active online health portal. Team members became participants in this study once they voluntarily had their blood drawn for the IgG assay, not at registration.

Data gathered for this study included demographics, SARS-CoV-2 IgG assay result and any documented positive PCR test results for COVID-19 infection, documented in Epic, the healthcare system's EMR. Age was grouped into quantiles (ages 18-31; 32-41; 42-51; 52-82) for analysis to evaluate risk by increasing age and to avoid underrepresentation of the oldest-age participants. Race included White-, Black-, Asian-, or American Indian-only or Mixed-race (those who identified as two or more races), with 332 (2.23%) total missing values. 

Data management and analysis were performed by the study research team and conducted using SAS statistical software (Version 9.4; SAS Institute, Cary, NC).

Univariate analyses are reported as counts (%) or means (standard deviation) and median (interquartile range), as appropriate. Bivariate analyses highlight variables across IgG status (exposure). Corresponding measures of association represent differences in participants who were IgG positive status relative to IgG negative status and include mean differences for days to infection and age, and odds ratios (OR) for age quantiles, sex, race, ethnicity and clinical role category. OR represents the ratio of odds of COVID-19 infection among those who were IgG positive relative to IgG negative at each variable level relative to the reference level of the same variable. Variable reference levels were chosen based on lowest presumed risk. Corresponding pvalues were generated from Student's T-tests for continuous variables and Pearson Chi-square (or Fisher's Exact Tests when any cell size(s) was < 5) to represent infection differences.

Incidence of COVID-19 infection after IgG test (outcome) was calculated as number of participants who experienced the outcome by person-days contributed to follow-up before outcome was experienced. Logistic regressions were performed to estimate adjusted odds of COVID-19 infection after exposure. Adjusted logistic regressions were adjusted for IgG status in primary analysis and documented COVID-19 infection prior to study initiation in secondary analysis, as well as age (as quantiles), race, ethnicity and clinical role category in both analyses. Due to a missing race variable, 332 and 391 participants were excluded from adjusted regressions in primary and secondary analyses, respectively.

This study was funded internally. The healthcare system had no influence over the study design, conduct, results, or dissemination of findings. The authors received no direct financial support for the research, authorship, and/or publication of this article. The authors have no competing interests to declare. Table 3 for complete results of secondary analyses.

The results of this study contribute noteworthy data regarding recurrence of COVID-19 among 

This prospective study enrolled and followed a large cohort of individuals to determine incidence of recurrence of COVID-19, the largest cohort of individuals monitored for subsequent COVID-19 infection to date. This study offered free IgG testing, which eliminated a well-documented barrier to testing in vulnerable populations. This study utilized two objective exposure measures -IgG test status and prior COVID-19 infection statusto more comprehensively document potential for recurrence. The healthcare system-affiliated lab performed all IgG tests using the same assay and methods; furthermore, given the large size and breadth of the health system conducting this study, it is likely that most, if not all, assays for COVID-19 infection were performed within system-affiliated labs, resulting in performance and reporting consistency. All data was stored in the EMR and extracted by the system's Analytics Team, resulting in data collection consistency.

There are several limitations to this study that could not be circumvented. This study's focus was on recurrence and, thus, did not address the relationship between documented prior COVID -19 infection and IgG status. It is important to note that individuals with no prior documented infection, and thus categorized as no exposure in secondary analysis, may not have a negative exposure history, but instead did not have RNA testing performed or documented, at all or at a system-affiliated lab.

Given the changing landscape of the COVID-19 outbreak, this paper provides much needed data to the emerging body of literature. This paper's COVID-19 recurrence findings may suggest a couple scenarios, one behavioral and one biological. First, it is possible that, after receiving antibody test results, individuals with positive IgG status behaved more recklessly by exposing themselves to potential infection sooner and more frequently, due to a misconception of immunity. Alternatively, individuals with positive IgG status may be more susceptible to COVID-19 infection due to their IgG status or a previous COVID-19

infection.

Considering these two potential scenarios, public health efforts should continue to widely disseminate the importance of infection-prevention measures, including but not limited to social distancing, mask-wearing, and hand hygiene. Furthermore, messaging should convey that previous exposure to SARS-CoV-2 or prior infection with COVID-19 does not ensure immunity to subsequent COVID-19 infections.

Without immunity, individuals are capable of spreading the disease and resource utilization still requires vigilance. Lastly, and perhaps most importantly, natural herd immunity appears unachievable, so policy efforts should challenge this narrative and advocate for universal vaccine uptake, when available.

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

The authors declare the following financial interests/personal relationships which may be considered as potential competing interests. 

Coronavirus protective immunity is short-lasting. medRxiv

The possibility and cause of relapse after previously recovering from COVID-19: a systematic review

Long-term existence of SARS-CoV-2 with antibody response in COVID-19 patients

Convergent antibody responses to SARS-CoV-2 infection in convalescent individuals

COVID-19 and post-infection immunity: limited evidence, many remaining questions

Scientific consensus on the COVID-19 pandemic: we need to act now. The Lancet

Clinical recurrences of COVID-19 symptoms after recovery: Viral relapse, reinfection or inflammatory rebound?

Questions about COVID-19 test accuracy raised across the testing spectrum. NCBnews

Setting the criteria for SARS-CoV-2 reinfection -six possible cases

Characteristics of COVID-19 recurrence: a systematic review and meta-analysis

Clinical characteristics of severe acute respiratory syndrome coronavirus 2 reactivation

Post-discharge surveillance and positive virus detection in two medical staff recovered from coronavirus disease 2019 (COVID-19), China

Recurrence of COVID-19 after recover: a case report from Italy. Infection

South Korea reports more recovered coronavirus patients testing positive again. Reuters

COVID-19 re-infection by a phylogenetically distinct SARS-coronavirus-2 strain confirmed by whole genome sequencing

Asymptomatic reinfection in two healthcare workers from India with genetically distinct SARS-CoV-2

Genomic evidence for reinfection with SARS-CoV-2: a case study

an overview of their replication and pathogenetic

Coronavirus reinfections are real but very, very rare. The New York Times

Antibody Status and Incidence pf SARS-CoV-2 Infection in Health Care Workers

Antibody (serology) testing for COVID-19: information for patients and consumers. U.S. Food & Drug Administration (FDA)

Rady Children's Hospital San Diego (RCHSD)

SARS-CoV-2 immunoassays: advancing diagnostics of COVID-19

Immunoglobulin Antibodies in a Large Midwestern Health Care System

Comparison of Commercially Available and Laboratory Developed Assays for in vitro Detection of SARS CoV-2 in Clinical Laboratories

SARS-CoV-2 IgG for use with Architect. U.S Food & Drug Administration

19-emergency-use-authorizations-medical-devices

0001 for Pearson Chi square tests (or Fisher's Exact test) if categorical or Student's T-test if continuous *Statistically significant at p<0.05 or Pearson Chi square tests (or Fisher's Exact test) if categorical or Student's T-test if continuous REF