Hostname: page-component-7c8c6479df-fqc5m Total loading time: 0 Render date: 2024-03-29T10:50:00.670Z Has data issue: false hasContentIssue false

Estimate of undetected severe acute respiratory coronavirus virus 2 (SARS-CoV-2) infection in acute-care hospital settings using an individual-based microsimulation model

Published online by Cambridge University Press:  01 September 2022

Kasey Jones
Affiliation:
RTI International, Research Triangle, North Carolina
Emily Hadley
Affiliation:
RTI International, Research Triangle, North Carolina
Sandy Preiss
Affiliation:
RTI International, Research Triangle, North Carolina
Eric T. Lofgren
Affiliation:
Paul G. Allen School for Global Health, Washington State University, Pullman, Washington
Donald P. Rice
Affiliation:
Division of Infectious Disease, Department of Medicine, Alpert Medical School of Brown University, Providence, Rhode Island
Marie C. D. Stoner
Affiliation:
RTI International, Research Triangle, North Carolina
Sarah Rhea
Affiliation:
Department of Population Health and Pathobiology, North Carolina State University, Raleigh, North Carolina
Joëlla W. Adams*
Affiliation:
RTI International, Research Triangle, North Carolina
*
Author for correspondence: Joëlla W. Adams, E-mail: jadams@rti.org
Rights & Permissions [Opens in a new window]

Abstract

Objective:

Current guidance states that asymptomatic screening for severe acute respiratory coronavirus virus 2 (SARS-CoV-2) prior to admission to an acute-care setting is at the facility’s discretion. This study’s objective was to estimate the number of undetected cases of SARS-CoV-2 admitted as inpatients under 4 testing approaches and varying assumptions.

Design and setting:

Individual-based microsimulation of 104 North Carolina acute-care hospitals

Patients:

All simulated inpatient admissions to acute-care hospitals from December 15, 2021, to January 13, 2022 [ie, during the SARS-COV-2 ο (omicron) variant surge].

Interventions:

We simulated (1) only testing symptomatic patients, (2) 1-stage antigen testing with no confirmatory polymerase chain reaction (PCR) test, (3) 1-stage antigen testing with a confirmatory PCR for negative results, and (4) serial antigen screening (ie, repeat antigen test 2 days after a negative result).

Results:

Over 1 month, there were 77,980 admissions: 13.7% for COVID-19, 4.3% with but not for COVID-19, and 82.0% for non–COVID-19 indications without current infection. Without asymptomatic screening, 1,089 (credible interval [CI], 946–1,253) total SARS-CoV-2 infections (7.72%) went undetected. With 1-stage antigen screening, 734 (CI, 638–845) asymptomatic infections (67.4%) were detected, with 1,277 false positives. With combined antigen and PCR screening, 1,007 (CI, 875–1,159) asymptomatic infections (92.5%) were detected, with 5,578 false positives. A serial antigen testing policy detected 973 (CI, 845–1,120) asymptomatic infections (89.4%), with 2,529 false positives.

Conclusions:

Serial antigen testing identified >85% of asymptomatic infections and resulted in fewer false positives with less cost per identified infection compared to combined antigen plus PCR testing.

Type
Original Article
Creative Commons
Creative Common License - CCCreative Common License - BYCreative Common License - NCCreative Common License - ND
This is an Open Access article, distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives licence (http://creativecommons.org/licenses/by-nc-nd/4.0/), which permits non-commercial re-use, distribution, and reproduction in any medium, provided that no alterations are made and the original article is properly cited. The written permission of Cambridge University Press must be obtained prior to any commercial use and/or adaptation of the article.
Copyright
© Research Triangle Institute, d/b/a RTI International and the Author(s), 2022. Published by Cambridge University Press on behalf of The Society for Healthcare Epidemiology of America

During the recent surge of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infections attributable to the ο (omicron) variant B.1.1.529, increased attention was given to “incidental COVID-19” or hospital admissions with COVID-19 rather than COVID-19 admissions. Due to the substantial number of asymptomatic infections, particularly with newer variants, there is a high risk of admitting patients already infected with SARS-CoV-2 into healthcare settings. Reference Oran and Topol1,Reference Murray2 Therefore, policies and enhanced testing strategies to identify cases prior to inpatient admission may be needed to prevent nosocomial spread. Nosocomial spread of SARS-CoV-2 within US-based acute-care settings is likely significant; however, limited data are available to quantify it. A report from Kaiser Health News using inpatient hospital data from April to September 2020 found that >10,000 patients were diagnosed with SARS-CoV-2 after being admitted for another indication and >20% of those patients died. Reference Jewett3 In the United Kingdom, estimates for the percentage of admitted patients who acquired SARS-CoV-2 as inpatients range from 11% to 25%. Reference Read, Green and Harrison4Reference Conn and Barr6 With >60% of the US population vaccinated with at least 2 doses since the beginning of 2022 and most hospitals implementing universal masking and testing procedures for symptomatic individuals, the risk of nosocomial transmission may be decreased compared to earlier phases of the pandemic. 7,8 However, the high transmissibility of the SARS-CoV-2 ο (omicron) variant among both vaccinated individuals and those with natural immunity Reference Birhane, Bressler and Chang9 highlights the need to further examine the risk of nosocomial transmission of SARS-CoV-2 in the United States.

As of December 2021, Centers for Disease Control and Prevention (CDC) guidance recommends testing anyone with mild symptoms, regardless of vaccination status, but states that “performance of a preprocedure or preadmission viral testing is at the discretion of the facility.” 8 The rationale of this policy is that the number of cases identified through preadmission testing for asymptomatic infection will be low when performed on vaccinated individuals or in areas with low-to-moderate transmission, 8 a conclusion supported by recent research. Reference Penney and Doron10,Reference Tande, Pollock, Shah, Binnicker and Berbari11 Based on this recommendation, US hospitals vary in their testing policies.

No standard guidance has been established on the appropriate testing strategy for preadmission testing. We used an individual-based model of North Carolina healthcare facilities to simulate the admission of individuals from the community into acute-care hospitals and to estimate the number of undetected cases of SARS-CoV-2 admitted as inpatients under 4 testing approaches during a period of exponential growth in cases within the community.

Methods

Analytic overview

We used the Modeling Infection Diseases in Healthcare Network (MInD-Healthcare), an individual-based microsimulation, to compare hospital testing policies over 1 month. We selected a 1-month time frame to provide insights on how healthcare facilities can prepare for future periods of exponential pandemic growth. We simulated and compared 4 testing policies for patients admitted to acute-care facilities: (1) only testing symptomatic patients (ie, no preadmission screening of asymptomatic admissions); (2) 1-stage antigen testing of asymptomatic admissions; (3) 1-stage antigen testing followed by confirmatory PCR for negative results; and (4) serial antigen testing (ie, repeated antigen test for negative results performed 2 days after the initial test) (Fig. 1). As the main analysis, we ran model scenarios under conditions reflective of conditions in North Carolina from December 15, 2021, to January 13, 2022 (ie, the SARS-CoV-2 ο (omicron) [B.1.1.529] variant was dominant). We assumed that the testing policies were implemented uniformly across all hospitals.

Fig. 1. Simulated policies on preadmission SARS-CoV-2 screening for asymptomatic patients admitted to an acute-care hospital.

Model overview

Originally built to model the natural history of C. difficile infection within a regional health system, Reference Rhea, Hilscher and Rineer13 MInD-Healthcare was adapted in 2020 to forecast hospitalizations for North Carolina during the COVID-19 pandemic. Only North Carolina residents were modeled (ie, patients admitted from other states were not simulated). We used historical reported case counts by county and day from COVID-19 ActNow, which nearly matched case counts reported by the state and were more readily available. 14 The case counts were smoothed using a 10-day rolling average because cases are generally not reported on weekends or holidays. Cases were then multiplied by a case multiplier (8× within the main analysis) to adjust the reported number of cases to reflect actual infections within the community. Measurement bias related to underreporting of positive at-home antigen tests, asymptomatic infection, lack of testing supplies, and other factors have made the use of a case multiplier standard practice in modeling studies. Values range from 4× to 8×. 15,Reference Paltiel, Zheng and Walensky16 We explored the impact of lower case multipliers within a sensitivity analysis.

Bayesian equations were used to assign cases to individuals within the ABM based on age and vaccination status and to determine severity of infection (susceptible or immune, asymptomatic, symptomatic, and recovered). Symptomatic infections were further classified into mild to moderate; severe (requiring an inpatient admission for SARS-CoV-2); and critical (requiring mechanical ventilation). Severity of infection was informed by surveillance data and large population-based cohort studies Reference Fowlkes, Gaglani, Groover, Thiese, Tyner and Ellingson17Reference Ferguson21 as well as hospitalization data. Reference Cecil22 We focused on asymptomatic screening because we assumed that those with mild and moderate infections would be identified through verbal screening and that those with severe and critical infections would be admitted for COVID-19 treatment. 8

Testing

Although testing occurs at the discretion of the facility, the current testing algorithm shared by the CDC for screening for asymptomatic infection in congregate settings begins with antigen testing. 23 Although the CDC recommends antigen testing with follow-up polymerase chain reaction (PCR) for anyone with symptoms who tests negative, 23 a follow-up PCR is not recommended for asymptomatic individuals with a negative antigen test. However, the individual is advised to quarantine if they were a close contact of someone with SARS-CoV-2. 23 Serial testing (ie, series of at least 2 antigen tests) for asymptomatic individuals is recommended if the individual has a high likelihood of SARS-CoV-2 infection, is not up-to-date on their vaccines, and has not had a SARS-CoV-2 infection in the prior 90 days. 23 However, retesting strategies will increase the potential for false positives or a positive test result for someone who is not actively infected. In practice, several testing approaches are implemented by hospital systems due to testing availability, cost, and other factors.

Significant heterogeneity in the performance of antigen testing may exist for detecting the SARS-CoV-2 ο (omicron) variant, Reference Bekliz, Perez-Rodriguez and Puhach24 and this variance may be time varying, with decreased sensitivity in the first 1–2 days after exposure. Reference Adamson, Sikka, Wyllie and Premsrirut25 A Canadian government report and several preprint studies report antigen test sensitivity for detecting the SARS-CoV-2 ο (omicron) variant as ranging from 37.1 to 97.6%. Reference Drain26Reference Jüni, Baert and Corbeil30 Within the main analysis, we assumed a test sensitivity of 67.4% (the middle of this range) for detecting the SARS-CoV-2 ο (omicron) variant in asymptomatic individuals using antigen testing and performed a sensitivity analysis decreasing this number. PCR testing was assumed to have a test sensitivity of 77% for asymptomatic infection. Reference Butler-Laporte, Lawandi and Schiller31,Reference Hellewell, Russell and Matthews32 We assumed a test specificity of 98%, the negative percentage agreement required for approval by the FDA for both antigen and PCR testing. Reference Drain26 We additionally assumed that 30% of individuals who had a SARS-CoV-2 infection within the previous 90 days but had since recovered would test positive using PCR testing (ie, persistent positivity) due to persistent viral shedding rather than true reinfection. Reference Landi, Carfì and Benvenuto33,Reference Aldhaeefi, Tahir, Cote, Izzy and El Khoury34 Although we did not conduct a formal cost-effectiveness analysis, we accounted for testing costs based on Centers for Medicare and Medicaid Services (CMS) data ($44 per antigen test and $112 per PCR) to estimate the cost per identified infection and number needed to screen (NNS) to detect 1 infection.

Vaccination

Vaccination was included as a time-invariant agent variable that decreases the likelihood of infection, increases the likelihood of asymptomatic infection, Reference Birhane, Bressler and Chang9,Reference Fowlkes, Gaglani, Groover, Thiese, Tyner and Ellingson17 and decreases the likelihood of hospitalization.

Recent studies have reported that although receipt of a booster dose confers ∼50% vaccine effectiveness against infection with the SARS-CoV-2 ο (omicron) variant, marked waning of immunity against infection was seen for those without a booster (ie, not up to date) with, on average, only 7% protection against infection. 35Reference Spensley, Gleeson and Martin37 As of January 5, 2022, 37% of the vaccinated population of North Carolina had received a booster dose. 38 Therefore, we implemented a weighted vaccine effectiveness against infection parameter (24%), which accounted for 37% of the vaccinated population having a booster. This calculation resulted in 50% effectiveness against infection and 63% having 7% effectiveness against infection.

Vaccination also increased the likelihood of being asymptomatic if infected. Reference Klompas39 For reported cases, we assumed that 5% of the unvaccinated and 25% of the vaccinated population were asymptomatic. Reported asymptomatic cases would represent cases detected through work-based surveillance or through school- or office-based testing for close contacts. Given the lack of evidence for nonreported cases, we assumed that 25% of the unvaccinated and 50% of the vaccinated population were asymptomatic. Due to uncertainty around this parameter, we increased the proportion of asymptomatic cases by 50% in a sensitivity analysis.

Vaccination decreased the likelihood of developing severe illness and hospitalization. Using data on hospital admissions for COVID-19, which included vaccination status, we replicated historical admissions. Overall, 69% of individuals hospitalized with SARS-CoV-2 were unvaccinated, in line with national trends.

Movement from community into hospital settings

For this analysis, we simulated the movement from the community by agents representing patients into 104 short-term acute-care hospitals over 1 month. At each daily time step, agents had a probability of entering as a patient for COVID-19 or a different primary indication. The probability of hospitalization for a non–COVID-19 indication was informed by the agent’s age group, sex, and comorbidities based on statewide hospital discharge data. Reference Cecil22 Parameterizations of length of stay, hospital capacity, rate of transfer, and patient demographic information were informed by statewide hospital discharge data. Reference Cecil22 Key parameters and data sources are summarized in Table 1, and full details are available in the Supplementary Materials.

Table 1. Estimates for Key Model Parameters NC MInD-Healthcare ABM

a Actual reported December 15, 2021, through January 13, 2022.

b Assumes vaccine effectiveness of 50% with booster and 7% for 2 doses or unvaccinated.

c Vaccination is modeled as a time-invariant agent state and does not differ by vaccine manufacturer, number of received doses, or time since last dose.

d Varies by age group, informed by hospital inpatient data.

e Varied by county.

Sensitivity analyses

We varied the following elements in sensitivity analyses: (1) We changed the case multiplier from 8× to 4× and 6×. (2) We decreased antigen testing sensitivity for asymptomatic infection from 67.4% to 37%. (3) We increased the probability of asymptomatic infection by 50%. (4) We increased vaccine effectiveness against infection from 37% to 75%. And (5) we lowered community COVID-19 incidence from 6,550 to 820 and 200 per 100,000 population weekly.

We ran each model scenario 100 times with 5 million agents. Model output was scaled to represent 10 million agents to represent the population of North Carolina. The MInD-Healthcare model was approved by the RTI International Institutional Review Board and was programmed using Python software. The full model code is available online (https://doi.org/10.5281/zenodo.6141066) and is accompanied by scripts and instructions for recreation of all model results.

Results

Model scenarios initialized with agents representing the population of North Carolina. We modeled 104 acute-care settings with 65% of non-ICU beds and 50% of ICU beds filled at model initialization. Over 1 month within the main analysis (ie, reflecting conditions from December 15, 2021–January 13, 2022), there were 77,980 (credible interval [CI], 75,126–80,875) hospital admissions; 10,724 (13.7%) for COVID-19 (ie, patients with severe or critical SARS-CoV-2 disease status); 3,386 (4.3%) with but not for COVID-19 (ie, had mild or asymptomatic infection); and 63,869 (82.0%) for non–COVID-19 indications and without SARS-COV-2 infection.

Without screening for asymptomatic infection prior to admission, 1,089 (CI, 946–1,253), or 7.72%, of total SARS-CoV-2 infections went undetected (Table 2). In a scenario in which hospitals implemented 1-stage antigen screening, 64,958 tests were performed and 734 (CI, 638–845) total asymptomatic infections (67.4%) were detected, with 1,277 false positives. In a combined antigen and PCR testing scenario, 127,631 tests were performed and 1,007 (CI, 875–1,159) total asymptomatic infections (92.5%) were detected. However, in addition to 2,529 false positives, 3,049 patients tested positive with a PCR due to a prior SARS-CoV-2 infection in the previous 90 days (ie, persistent PCR positivity), despite a negative rapid result. In all, 5,587 (CI, 5,374–5,810) total admissions (7.15%) were incorrectly assumed to be currently infectious. A serial antigen testing scenario resulted in 127,631 tests performed and 973 (CI, 845–1,120) total asymptomatic infections (89.4%) were detected, with 2,529 false positives.

Table 2. Simulated Outcomes for SARS-CoV-2 Testing Policies for Acute-Care Settings

a PCR is assumed to be falsely positive for 30% of individuals who had a SARS-CoV-2 in the past 90 days but who are no longer within the 7- to 10-day infectious period.

b Key parameters for the main analysis: case multiplier (adjustment to reported cases to account for underreporting) = 8x; PCR test sensitivity for asymptomatic cases = 77%; antigen test sensitivity for asymptomatic cases = 67.4%; vaccine effectiveness against infection = 24%; percent asymptomatic for reported cases = 5% if unvaccinated and 25% if vaccinated; percent asymptomatic for unreported cases = 25% if unvaccinated and 50% if vaccinated.

In a sensitivity analysis lowering the case multiplier, we found that a lower percentage of admitted patients with SARS-CoV-2 were asymptomatic; therefore, fewer infections were detected with asymptomatic screening. When antigen test sensitivity was decreased from 67.4% to 37%, the number of detected asymptomatic infections decreased to 403 (CI, 350–464) for 1-stage antigen, to 931 (CI, 809–1,071) for combined antigen and PCR, and to 656 (CI, 571–756) for serial antigen testing. A sensitivity analysis increasing the probability of asymptomatic infection by 50% resulted in 1,638 (CI, 1,441–1,811) admitted patients (11.6%) with asymptomatic SARS-CoV-2. The number of asymptomatic infections detected ranged from 1,104 for the one-stage antigen testing to 1,515 for the combined antigen and PCR testing. With increased vaccine effectiveness, 679 (CI, 547–821) patients (6.1%) with SARS-CoV-2 were asymptomatic. A scenario increasing vaccine effectiveness decreased total hospital admissions and averted 2,912 (CI, 2,558–3,093) total admissions (∼20%) with SARS-CoV-2 because we assumed that vaccinated individuals had a greater probability of “blocking” SARS-CoV-2 infection. Testing policies did not perform qualitatively different; between 458 (one-stage antigen) and 628 (combined antigen plus PCR) asymptomatic cases were detected. Sensitivity analyses lowering community incidence resulted in fewer admissions with or for COVID-19 and increased the number of tests with a higher ratio of false positives. Overall, the percentage of asymptomatic cases detected ranged from 37% for 1-stage antigen testing with decreased sensitivity to 92.5% for combined antigen plus PCR testing.

When estimating the cost of testing, the lowest cost per identified infection was 1-stage antigen testing at $3,890 (range, $3,488–4,360) followed by serial antigen testing at $5,770 (range, $5,180–6,480) (Table 3). One-stage antigen plus confirmatory PCR testing ($9,810; range, $8,800–11,020) was the most expensive testing policy. The NNS ranged from 89 for 1-stage antigen testing to 131 for serial antigen testing within the main analysis. Community incidence impacted the estimated cost per identified infection and NNS. The NNS for the 3 testing policies ranged from 726 to 1,082 when community incidence was lowered to 820 per 100,000 and from 2,033 to 3,034 at 200 per 100,000. Cost per identified infection ranged from $89,440 to $227,630.

Table 3. Estimated Costs and Number Needed to Screen for SARS-CoV-2 Testing Policies for Acute-Care Settings

Note. N/A, not available.

a No. needed to screen (NNS) calculated as no. of asymptomatic tests performed/no. of asymptomatic infections identified.

b Estimated total cost of testing policy calculated as (no. of asymptomatic tests performed × cost of test). Cost of rapid antigen test, $45 and PCR, $112. Cost estimates based on Centers for Medicare & Medicaid Services (https://www.cms.gov/files/document/mac-covid-19-test-pricing.pdf).

c Estimated cost per identified infection calculated as (estimated total cost of testing policy/# of asymptomatic infections identified).

Discussion

In an individual-based microsimulation of acute-care settings in North Carolina, we estimated that nearly 10% of SARS-CoV-2 infections (∼1,000 cases) went undetected when no preadmission screening for asymptomatic infection was performed over a month of increased community incidence attributable to the SARS-CoV-2 ο (omicron) variant. When asymptomatic screening was implemented with either a combined antigen plus PCR test or serial antigen testing, >85% of asymptomatic infections were detected. However, combined antigen plus PCR testing resulted in a significant number of false positives due to prior SARS-CoV-2 infection among admitted patients and had the highest associated costs. Our modeling results suggest that implementing serial antigen testing during periods of high community transmission may be an effective approach to detecting asymptomatic infection among hospital admissions. Detection of asymptomatic infections and subsequent implementation of appropriate infection prevention measures could reduce SARS-CoV-2 nosocomial transmission risk, along with other infection control measures including screening for symptoms and universal masking.

In a retrospective point-prevalence study of preadmission testing from January to August 2021 during a period of low community incidence, Tande et al Reference Tande, Pollock, Shah, Binnicker and Berbari11 reported a low rate of positivity in both asymptomatic vaccinated and unvaccinated patients (0.30% and 1.23%, respectively). Within our analysis, positivity rates were 1.51% and 2.55% for vaccinated and unvaccinated individuals, respectively. The positivity rate during this period was likely higher than our simulated values, but in the absence of comprehensive data in North Carolina, we chose a more conservative approach. Indeed, if the positivity rate had been higher, the modeled testing policies would have been even more efficient in identifying asymptomatic cases. When community incidence is low, preadmission testing may be of limited value; Reference Penney and Doron10 this assertion is supported by sensitivity analyses showing that the NNS and cost per identified infection notably increased at the threshold level for CDC community level moving from low to medium risk (200 per 100,000 weekly incidence). Likewise, the ability of tests to ascertain active infection is crucial to the utility of preadmission testing. Persistent positives limit the value of a combined antigen and PCR testing policy.

Within our analysis, serial antigen testing detected the most asymptomatic infections with fewer false positives compared to the combined antigen plus PCR strategy. Ramifications of false positives include unnecessary treatment cancellation or postponement for elective procedures, financial losses related to self-isolation, psychological distress to the patient, and cohorting with actively infected patients. Reference Surkova, Nikolayevskyy and Drobniewski43 Evidence suggesting that rapid antigen testing will not detect infection with SARS-CoV-2 ο (omicron) variant in the first 2 days despite individuals being infectious Reference Adamson, Sikka, Wyllie and Premsrirut25 further supports the need for serial antigen testing.

Although this analysis simulated hospitals in North Carolina, results are likely generalizable to US hospitals. North Carolina has ∼10 million residents with diversity across age, race and ethnicity, and socioeconomic status. In addition, urban and rural counties were represented, with wide variance in vaccination rates. 38

This study had several limitations. Our model simulated entry into the hospital environment based on discharge data and therefore only included inpatient stays. Individuals entering the hospital environment for outpatient or emergency department care or visitation were not modeled. Testing was less likely for these individuals and relies on other mitigation strategies to decrease the risk of nosocomial transmission. We assumed that between 5% (unvaccinated) and 25% (vaccinated) of infected individuals would be asymptomatic. Initial studies of the SARS-CoV-2 ο (omicron) variant suggest that the proportion of asymptomatic infection may be much higher, with estimates up to 80%. Reference Murray2 However, this evidence is still preliminary, and we chose to incorporate a more conservative assumption and to test a higher rate of asymptomatic infection within a sensitivity analysis. We did not model differing vaccine effectiveness dependent on timing of doses or immunosuppression. In addition, the pediatric population was grouped within the “under 50” age group, even though certain ages were ineligible to receive a vaccine. Lastly, we did not model the transmission of SARS-CoV-2; therefore, we could not simulate cases of SARS-CoV-2 resulting from a nosocomial outbreak.

Patients admitted to an acute-care hospital are likelier to be older, immunocompromised, and frailer than other community members and therefore at higher risk of SARS-CoV-2–related morbidity and mortality. Undetected asymptomatic infection poses a preventable risk within acute-care settings. These model results suggest that serial antigen testing for asymptomatic infection may be the optimal approach for detecting asymptomatic infection in acute-care settings when community levels of transmission are higher.

Supplementary material

To view supplementary material for this article, please visit https://doi.org/10.1017/ice.2022.174

Acknowledgments

We are grateful for the support and input from Stacy Endres-Dighe and Georgiy Bobashev of RTI International and from our UNC Health collaborators and our public health collaborators. This activity was based on a model originally developed through support from the CDC Modeling Infectious Disease in Healthcare (MInD-Healthcare) Network.

Financial support

This work was supported by the National Science Foundation (grant nos. 2110109 and 2027802) and the Centers for Disease Control and Prevention (grant no. 75D30121P10548).

Conflicts of interest

All authors report no conflicts of interest relevant to this article.

References

Oran, DP, Topol, EJ. The proportion of SARS-CoV-2 infections that are asymptomatic. Ann Intern Med 2021;174:655662.CrossRefGoogle ScholarPubMed
Murray, CJL. COVID-19 will continue but the end of the pandemic is near. Lancet 2022;399:417419.CrossRefGoogle Scholar
Jewett, C. Patients went into the hospital for care after testing positive there for COVID, some never came out. Kaiser Health News website. https://khn.org/news/article/hospital-acquired-covid-nosocomial-cases-data-analysis/. Published November 4, 2021. Accessed August 19, 2022.Google Scholar
Read, JM, Green, CA, Harrison, EM, et al. Hospital-acquired SARS-CoV-2 infection in the UK’s first COVID-19 pandemic wave. Lancet 2021;398:10371038.CrossRefGoogle ScholarPubMed
Oliver, D. David Oliver: deaths from hospital-acquired COVID are everyone’s problem. BMJ 2021;373:n1492.CrossRefGoogle ScholarPubMed
Conn, D, Barr, C. NHS faces questions over COVID infections contracted in hospital. The Guardian website. https://www.theguardian.com/world/2021/feb/26/nhs-faces-questions-over-covid-infections-contracted-in-hospital. Published February 26, 2021. Accessed August 19, 2022.Google Scholar
US coronavirus vaccine tracker. USA Facts website. https://usafacts.org/visualizations/covid-vaccine-tracker-states/. Accessed March 2, 2022.Google Scholar
Interim infection prevention and control recommendations for healthcare personnel during the coronavirus disease 2019 (COVID-19) pandemic. Centers for Disease Control and Prevention website. https://www.cdc.gov/coronavirus/2019-ncov/hcp/infection-control-recommendations.html. Published 2020. Accessed March 2, 2022.Google Scholar
Birhane, M, Bressler, S, Chang, G, et al. COVID-19 vaccine breakthrough infections reported to CDC—United States, January 1–April 30, 2021. Morb Mortal Wkly Rep 2021;70:792793.Google Scholar
Penney, JA, Doron, SI. Finding the off-ramp: rethinking severe acute respiratory coronavirus virus 2 (SARS-CoV-2) preoperative screening. Infect Control Hosp Epidemiol 2022;43:918919.CrossRefGoogle ScholarPubMed
Tande, AJ, Pollock, BD, Shah, ND, Binnicker, M, Berbari, EF. mRNA vaccine effectiveness against asymptomatic severe acute respiratory coronavirus virus 2 (SARS-CoV-2) infection over seven months. Infect Control Hosp Epidemiol 2022;43:393395.CrossRefGoogle ScholarPubMed
Calderwood, MS, Deloney, VM, Anderson, DJ, et al. Policies and practices of SHEA Research Network hospitals during the COVID-19 pandemic. Infect Control Hosp Epidemiol 2020;41:11271135.CrossRefGoogle Scholar
Rhea, S, Hilscher, R, Rineer, JI, et al. Creation of a geospatially explicit, agent-based model of a regional healthcare network with application to Clostridioides difficile infection. Health Security 2019;17:276290.CrossRefGoogle ScholarPubMed
US COVID risk and vaccine tracker. Covid Act Now website. https://covidactnow.org/?s=29951854. Accessed March 2, 2022.Google Scholar
Estimated COVID-19 burden. Centers for Disease Control and Prevention website. https://www.cdc.gov/coronavirus/2019-ncov/cases-updates/burden.html. Published 2021. Accessed March 2, 2022.Google Scholar
Paltiel, AD, Zheng, A, Walensky, RP. Assessment of SARS-CoV-2 screening strategies to permit the safe reopening of college campuses in the United States. JAMA Network Open 2020;3:e2016818.CrossRefGoogle ScholarPubMed
Fowlkes, A, Gaglani, M, Groover, K, Thiese, MS, Tyner, H, Ellingson, K. Effectiveness of COVID-19 vaccines in preventing SARS-CoV-2 infection among frontline workers before and during B.1.617.2 (delta) variant predominance—eight US locations, December 2020–August 2021. Morb Mortal Wkly Rep 2021;70:11671169.CrossRefGoogle ScholarPubMed
Tartof, SY, Slezak, JM, Fischer, H, et al. Effectiveness of mRNA BNT162b2 COVID-19 vaccine up to 6 months in a large integrated health system in the USA: a retrospective cohort study. Lancet 2021;398:14071416.CrossRefGoogle Scholar
COVID-19 case surveillance public use data with geography. Centers for Disease Control and Prevention website. https://data.cdc.gov/Case-Surveillance/COVID-19-Case-Surveillance-Public-Use-Data-with-Ge/n8mc-b4w4. Accessed March 2, 2022.Google Scholar
Espenhain, L, Funk, T, Overvad, M, et al. Epidemiological characterisation of the first 785 SARS-CoV-2 omicron variant cases in Denmark, December 2021. Eurosurveillance 2021;26:2101146.CrossRefGoogle ScholarPubMed
Ferguson, N. Report 49: Growth and immune escape of the omicron SARS-CoV-2 variant of concern in England. Imperial College London website. https://www.imperial.ac.uk/media/imperial-college/medicine/mrc-gida/2021-12-16-COVID19-Report-49.pdf2021/12/15/. Published December 15, 2021. Accessed August 19, 2022.Google Scholar
Cecil, G. Short term acute-care hospital discharge data—patient characteristics. Summary data for all hospitals. Sheps Center for Health Services Research University of North Carolina at Chapel Hill website. https://www.shepscenter.unc.edu/wp-content/uploads/2020/05/ptchar_all_and_by_hosp_2018_and.pdf. Accessed March 2, 2022.Google Scholar
Interim guidance for antigen testing for SARS-CoV-2. Centers for Disease Control and Prevention website. https://www.cdc.gov/coronavirus/2019-ncov/lab/resources/antigen-tests-guidelines.html. Accessed March 2, 2022.Google Scholar
Bekliz, M, Perez-Rodriguez, F, Puhach, O, et al. Sensitivity of SARS-CoV-2 antigen-detecting rapid tests for omicron variant. medRxiv 2022. doi: 10.1101/2021.12.18.21268018.CrossRefGoogle Scholar
Adamson, B, Sikka, R, Wyllie, AL, Premsrirut, P. Discordant SARS-CoV-2 PCR and rapid antigen test results when infectious: a December 2021 occupational case series. medRxiv 2022. doi: 10.1101/2022.01.04.22268770.CrossRefGoogle Scholar
Drain, PK. Rapid diagnostic testing for SARS-CoV-2. N Engl J Med 2022;386:264272.CrossRefGoogle ScholarPubMed
Goodall, BL, LeBlanc, JJ, Hatchette, TF, Barrett, L, Patriquin, G. Investigating sensitivity of nasal or throat (ISNOT): a combination of both swabs increases sensitivity of SARS-CoV-2 rapid antigen tests. medRxiv 2022. doi: 10.1101/2022.01.18.22269426.CrossRefGoogle Scholar
Schrom, J, Marquez, C, Pilarowski, G, et al. Comparison of SARS-CoV-2 reverse transcriptase polymerase chain reaction and BinaxNOW rapid antigen tests at a community site during an omicron surge: a cross-sectional study. Ann Intern Med 2022;175:682690.CrossRefGoogle Scholar
de Michelena, P, Torres, I, Ramos-García, A, et al. Real-life performance of a COVID-19 rapid antigen detection test targeting the SARS-CoV-2 nucleoprotein for diagnosis of COVID-19 due to the omicron variant. J Infect 2022;84:e64e66.CrossRefGoogle ScholarPubMed
Jüni, P, Baert, S, Corbeil, A, et al. Use of rapid antigen tests during the omicron wave. Ontario COVID-19 Science Advisory Table website. https://covid19-sciencetable.ca/sciencebrief/use-of-rapid-antigen-tests-during-the-omicron-wave/. Accessed March 2, 2022.CrossRefGoogle Scholar
Butler-Laporte, G, Lawandi, A, Schiller, I, et al. Comparison of saliva and nasopharyngeal swab nucleic acid amplification testing for detection of SARS-CoV-2: a systematic review and meta-analysis. JAMA Intern Med 2021;181:353360.CrossRefGoogle ScholarPubMed
Hellewell, J, Russell, TW, Matthews, R, et al. Estimating the effectiveness of routine asymptomatic PCR testing at different frequencies for the detection of SARS-CoV-2 infections. BMC Med 2021;19:106.CrossRefGoogle ScholarPubMed
Landi, F, Carfì, A, Benvenuto, F, et al. Predictive factors for a new positive nasopharyngeal swab among patients recovered from COVID-19. Am J Prevent Med 2021;60:1319.CrossRefGoogle ScholarPubMed
Aldhaeefi, M, Tahir, Z, Cote, DJ, Izzy, S, El Khoury, J. Comorbidities and age are associated with persistent COVID-19 PCR positivity. Front Cell Infect Microbiol 2021;11:650753.CrossRefGoogle ScholarPubMed
COVID-19 vaccine surveillance report, week 4. UKHSA, UK Health Security Agency website. https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/1050721/Vaccine-surveillance-report-week-4.pdf. 2022. Accessed August 19, 2022.Google Scholar
Buchan, SA, Chung, H, Brown, KA, et al. Effectiveness of COVID-19 vaccines against omicron or delta infection. medRxiv 2022. doi: 10.1101/2021.12.30.21268565.CrossRefGoogle Scholar
Spensley, K, Gleeson, S, Martin, P, et al. Comparison of vaccine effectiveness against the omicron (B.1.1.529) variant in patients receiving haemodialysis. Kidney Int Rep 2022;7:14061409.CrossRefGoogle Scholar
Response NC. Vaccinations Dashboard. North Carolina Department of Health and Human Services COVID-19 Response. https://covid19.ncdhhs.gov/dashboard/vaccinations. 2022. Accessed March 2, 2022.Google Scholar
Klompas, M. Understanding breakthrough infections following mRNA SARS-CoV-2 vaccination. JAMA 2021;326:20182020.CrossRefGoogle ScholarPubMed
Ferguson, N, Laydon, D, Nedjati Gilani, G, et al. Report 9: Impact of nonpharmaceutical interventions (NPIs) to reduce COVID-19 mortality and healthcare demand. Imperial College London website. https://www.imperial.ac.uk/media/imperial-college/medicine/sph/ide/gida-fellowships/Imperial-College-COVID19-NPI-modelling-16-03-2020.pdf. Published March 16, 2020. Accessed August 19, 2022.Google Scholar
Jansen, L. Investigation of a SARS-CoV-2 B.1.1.529 (omicron) variant cluster—Nebraska, November–December 2021. Morb Mortal Wkly Rep 2021;70:17821784.CrossRefGoogle ScholarPubMed
Hay, J, Kissler, S, Fauver, JR, et al. Viral dynamics and duration of PCR positivity of the SARS-CoV-2 omicron variant. Digital Access to Scholarship at Harvard (DASH) website. https://dash.harvard.edu/handle/1/37370587. Published 2022. Accessed August 19, 2022.Google Scholar
Surkova, E, Nikolayevskyy, V, Drobniewski, F. False-positive COVID-19 results: hidden problems and costs. Lancet Respir Med 2020;8:11671168.CrossRefGoogle ScholarPubMed
Figure 0

Fig. 1. Simulated policies on preadmission SARS-CoV-2 screening for asymptomatic patients admitted to an acute-care hospital.

Figure 1

Table 1. Estimates for Key Model Parameters NC MInD-Healthcare ABM

Figure 2

Table 2. Simulated Outcomes for SARS-CoV-2 Testing Policies for Acute-Care Settings

Figure 3

Table 3. Estimated Costs and Number Needed to Screen for SARS-CoV-2 Testing Policies for Acute-Care Settings

Supplementary material: PDF

Jones et al. supplementary material

Jones et al. supplementary material

Download Jones et al. supplementary material(PDF)
PDF 470 KB