Asymptomatic Pre-Operative COVID-19 Screening for Essential and Elective Surgeries: Early Results of Universal Screening at a Midwestern Academic Medical Center

Abstract

Background

During the novel Coronavirus 2019 (COVID-19) worldwide pandemic, viral testing has largely focused on patients presenting with fever and respiratory symptoms. Although Centers for Disease Control has reported 1,551,095 cases in the United States as of May 21, 2020, asymptomatic infection rates remain unknown within the U.S., especially in geographically disparate regions.

Methods

On April 7, 2020 our hospital established universal SARS-CoV-2 screening using RT-PCR RNA detection from nasopharyngeal swabs from asymptomatic patients prior to essential and elective surgeries. This study included 1,997 asymptomatic patients undergoing surgical procedures and 1,797 admitted for medical management at a Midwestern academic hospital between April 7, 2020 and May 21, 2020.

Results

As of May 21, asymptomatic testing for SARS-CoV-2 infection had been completed for 1,997 surgical patients and 1,797 non-surgical patients. Initial testing was positive in 26 patients, with an additional four positive tests occurring during repeat testing when greater than 48 hours had elapsed since initial testing. Overall asymptomatic infection rate was 0.79%. Asymptomatic infection rate was significantly lower in surgical patients (0.35% vs. 1.28%, p=0.001). Surgical patients tended to be older than non-surgical patients, although this was not statistically significant (51, IQR 27-65 vsx 46, IQR 28-64, p=0.057). Orthopedic surgery patients were significantly younger than those from other surgical services (42 vs. 53 yrs, p<0.001), however orthopedic and non-orthopedic surgical patients had similar asymptomatic infection rates (0.70% vs. 0.25%, p=0.173).

Conclusion

Among asymptomatic patients tested at a Midwestern academic medical center, 0.79% were infected with SARS-CoV-2 virus. These findings will help guide screening protocols at medical centers while providing essential and elective procedures during the COVID-19 pandemic. While the asymptomatic infection rate was low, this data substantiates the threat of asymptomatic infections and potential for community viral spread. These results may not be generalizable to large urban population centers or areas with high concentrations of COVID-19, each region must use available data to evaluate the risk-benefit ratio of universal testing vs universal contact precautions.

Level of Evidence: IV

Introduction

During the emergence of the worldwide novel coronavirus disease 2019–2020 pandemic (COVID-19), state, national, and international health agencies have recommended aggressive social distancing and isolation measures to minimize spread of the disease. Following a brief period of near complete closure of elective hospital surgical volume, the nation grappled with the magnitude of the domestic disease burden, essential surgical procedures have gradually resumed as local and regional conditions have allowed. The resumption of surgical and clinical services has been coordinated in a manner to minimize unnecessary consumption of personal protective equipment (PPE) and unnecessary exposure to patients and healthcare workers while allowing for timely treatment of patients with time-sensitive indications for surgical interventions.

Consensus recommendations have strongly encouraged screening of asymptomatic patients for SARS-CoV-2 prior to surgery or hospital admission to reduce asymptomatic transmission. In addition to identifying patients with asymptomatic SARS-CoV-2 infection, screening allows for conservation of PPE in critically short supply, such as filtering respirators, when caring for non-infected patients. In many areas without population testing, these screening efforts add critical information to rates reported by state and national public health officials. Addition of these screening test results may improve the accuracy of available estimates regarding community prevalence of COVID-19 infection.

Methods

This study underwent formal review by local Institutional Review Board and was determined to be exempt.

Participants

A total of 3,794 asymptomatic patients admitted for medical management or surgical procedures from April 7, 2020 to May 21, 2020 underwent screening for recent symptoms and testing for SARS-CoV-2 with RT-PCR within 48 hours of surgery of hospital admission. Upon presentation, all patients were identified as symptomatic or asymptomatic via screeing by healthcare staff for the presence of symptoms includeing fever, cough, chest pain, or shortness of breath within the previous 24 hours; patients answering yes to any of these symptoms were identified as potential COVID-19 patients, whereas those without self-reported symptoms were identified as asymptomatic. SARS-CoV-2 testing was ordered within the electronic health record according to the presence or absence of symptoms during initial screening: testing for patients endorsing symptoms was ordered as “NOVEL CORONAVIRUS (COVID-19)” whereas testing for patients denying symptoms was ordered as “COVID-19 ASYMPTOMATIC SCREEN BY PCR.” Both test orders required the ordering licensed independent practitioner to select radio buttons within the order to confirm the patient met criteria for the selected viral test.

All patients were tested with Nasopharyngeal swabs collected by clinical staff according to standard practices. Samples were processed to extract nucleic acids according to Centers for Disease Control and Prevention (CDC) 2019 Novel Coronavirus (2019-nCoV) Real-Time RT-PCR Diagnostic Panel.¹ Extracted specimens underwent nucleic acid extraction amplification using reverse transcriptase polymerase chain reaction. Diagnostic panels included a no template control, SARS-CoV-2 positive control, and human specimen control.

Study Design

This retrospective study evaluated asymptomatic patients admitted for medical management or surgical procedures at a Level 1 Midwestern Academic Medical Center from April 7 to May 21, 2020. We aimed to determine the proportion of positive tests among asymptomatic patients undergoing mandatory preadmission or preoperative testing for SARS-CoV-2 infection. All patients scheduled for surgery underwent testing unless delay for testing was believed to place the patient at unacceptable risk for serious adverse outcome including loss of life or limb. Patients indicated for emergency surgery that could not be delayed for testing were excluded from testing requirement prior to surgery and all staff involved in care of these patients observed airborne, contact precautions and eye protection including face shield and N95 mask use during and after the procedure until testing was completed.

Statistical Analysis

All statistical analyses were performed using SAS statistical software version 9.4 (SAS Institute, Inc., Cary, NC). Demographics were compared to assess differences between patients with positive and negative SARS-CoV-2 tests. Between-group means were compared using t-tests and frequencies were compared using Fischer’s exact test where appropriate. Statistical significance was defined as p < 0.05.

Results

Nasopharyngeal swabs were obtained from 3,794 asymptomatic patients prior to surgery or hospital admission and tested for SARS-CoV-2 nucleic acid. Table 1 provides a comparative breakdown of demographic characteristics among patients presenting for non-surgical admissions and those undergoing surgical procedcures. Patients presenting for surgical procedures are further separated into non-orthopedic surgery patients and orthopedic surgery patients. Among all asymptomatic patients, 1,997 (52.6 %) underwent surgical procedures and 1,797 (47.4%) were admitted for non-surgical management. Among patients undergoing surgical procedures, 1,569 patients had non-orthopedic procedures (78.6%), while 428 had orthopedic procedures (21.4%).

Table 1.

Demographic Characteristics of Patients Presenting for Non-Surgical Admission and Surgical Procedures

Variable	Non-Surgical Variable Admission (n=1797)	Surgical Procedures (n=1997)	p-value	Non-Orthopedic Procedures (n=1569)	Orthopedic Procedures (n=428)	p-value
Age: median (IQR)	46 (28-64)	51 (27-65)	0.057	53 (30-67)	42 (22-60)	<0.001
Min-max	0-98	0-97		0-91	0-97
BMI: median (IQR)	27.3 (22.7-32.8)	27.9 (23-33.6)	0.073	27.8 (23.0-33.6)	28.3 (23.4-33.5) 12.1-106.2	0.244
Min-max	10.7-71.0	8.5-119.1		8.5-119.1	(n=7 missing)
	(n=67 missing)	(n=11 missing)		(n=4 missing)
Gender (n, % female)	940 (52.3%)	1025 (51.3%)	0.545	825 (52.6%)	200 (46.7%)	0.032
Coronary Artery Disease	89 (5.0%)	116 (5.8%)	0.244	102 (6.5%)	14 (3.3%)	0.011
Hypertension	418 (23.3%)	563 (28.2%)	<0.001	470 (30.0%)	93 (21.7%)	<0.001
Congestive Heart Failure	28 (1.6%)	58 (2.9%)	0.005	53 (3.4%)	5 (1.2%)	0.016
Asthma	111 (6.2%)	101 (5.1%)	0.134	87 (5.5%)	14 (3.3%)	0.057
COPD	72 (4.0%)	96 (4.8%)	0.231	82 (5.2%)	14 (3.3%)	0.098
Type II Diabetes	188 (10.5%)	232 (11.6%)	0.257	193 (12.3%)	39 (9.1%)	0.068

Patients undergoing surgical procedures had slightly higher median age than those admitted for non-surgical management (51, IQR 27-65 years vs. 46, IQR 28-64 years), however this did not reach statistical significance (p=0.057). There was no significant difference in the distribution of gender between surgical and nonsurgical cohorts, with females accounting for 52.3% of non-surgical patients and 51.3% of surgical patients. Patients undergoing surgical procedures had significantly greater rates of hypertension and congestive heart failure (28.2% and 2.9%, respectively) compared to non-surgical patients (23.3% and 1.6%, respectively) (p<0.001 and p=0.005 respectively). There were no significant differences between surgical and non-surgical cohorts for rates of coronary artery disease, asthma, COPD, or type II diabetes mellitus (Table 1).

Patients undergoing surgical procedures were classified by whether they underwent orthopedic surgery procedures or non-orthopedic surgical procedures (Table 1). Within the surgical cohort, patients undergoing orthopedic procedures were significantly younger than those undergoing non-orthopedic procedures (42, IQR 22-60 vs. 53, IQR 30-67, p<0.001). females accounted for a significantly smaller portion of orthopedic patients compared to non-orthopedic surgical patients (46.7% vs. 52.6% respectively, p=0.032). Compared to patients undergoing non-orthopedic procedures, patients undergoing orthopedic surgery had significantly lower rates of coronary artery disease, hypertension, congestive heart failure, asthma, and type II diabetes mellitus (Table 1). There was no difference in rate of COPD between orthopedic and non-orthopedic patients (Table 1).

There were 30 positive tests for SARS-CoV-2, suggesting an overall asymptomatic infection rate of 0.79% in this case series (Table 2). Initial SARS-CoV-2 testing resulted in 26 positive tests (0.69%). Testing was repeated for 422 patients due to passing of greater than 48 hours since initial negative tests, 4 of which were subsequently positive (0.95%) for a total of 30 patients with positive SARS-CoV-2 tests. A third RT-PCR test was performed for 96 patients, with all 96 testing negative for SARS-CoV-2 RNA. The asymptomatic infection rate among the 1,997 surgical patients was 0.35% (7 of 1,997), which was significantly lower than asymptomatic infection rate of 1.28% within the non-surgical cohort (p=0.001). Asymptomatic infection rates among orthopedic surgery patients and non-orthopedic surgical patients were statistically similar (0.25% vs. 0.70%, p=0.173).

Table 2.

Asymptomatic Test Results of Patients Presenting for Non-Surgical Admission or Surgical Procedures

Measure	Non-Surgical Admission	Surgical Procedures	p-value	Service		p-value
				Orthopedic Procedures	Non-Orthopedic Procedures
Asymptomatic Tests	1797	1997	0.001	428 (21.4%)	1569 (78.6%)	0.173
Asymptomatic Positive Tests	23	7		3 (42.9%)	4 (57.1%)
Positive Test Rate	1.28%	0.35%		0.70%	0.25%

When all patients were grouped by SARS-CoV-2 test results, patients with positive tests were significanly younger than those with negative SARS-CoV-2 tests (31, IQR=18-56 yrs vs. 49, IQR=28-65 yrs, p=0.005). The youngest patient testing positive for SARS-CoV-2 was <1 year old and the oldest was 56 years old. Among patients younger than 18 years of age, the positive test rate was 1.4%, compared to 0.7% among patients 18 or older (p=0.118). There were no differences in distribution of gender or BMI between patients with positive or negative SARS-CoV-2 testing (Table 3). Prevalence of coronary artery disease, hypertension, congestive heart failure, asthma, COPD, and type II diabetes mellitus were statistically similar between patients with positive or negative SARS-CoV-2 testing (Table 4).

Table 3.

Demographic Characteristics Among Patients Testing Positive or Negative for SARS-CoV-2 Virus

	Total	Negative SARS-	Positive SARS-	P-value
	(n=3,794)	CoV-2 (n=3,764)	CoV-2 (n=30)	(+ vs. -)
Age	49 (0-98)	49 (0-98)	31 (0-78)	0.005
	IQR: 28-65	IQR: 28-65	IQR: 18-56
Gender (Female)	1,954 (51.8%)	1,950 (51.8%)	15 (50.0%)	0.843
BMI	27.6 (8.5-119.1)	27.6 (8.5-119.1)	25.8 (12.8-46.3)	0.237
	IQR: 22.9-33.3	IQR: 22.9-33.3	IQR: 18.4-33.7
	(n=3,716)

Table 4.

Medical Comorbidity Prevalence Among Patients Testing Positive or Negative for SARS-CoV-2 Virus

	Negative SARS-CoV-2	Positive SARS-CoV-2	P-value
	(n=3,764)	(n=30)
Coronary Artery Disease	205 (5.5%)	0 (0.0%)	0.406
Hypertension	976 (25.9%)	5 (16.7%)	0.248
Congestive Heart Failure	86 (2.3%)	0	1.000
Asthma	212 (5.6%)	0	0.411
Chronic Obstructive Pulmonary Disease	167 (4.4%)	1 (3.3%)	1.000
Type II Diabetes Mellitus	419 (11.1%)	1 (3.3%)	0.246

Discussion

Asymptomatic community transmission of SARS-CoV-2 has been identified as a potential major obstacle to containment of the ongoing COVID-19 pandemic.² To our knowledge, this is one of the first domestic reports of SARS-CoV-2 infections rate among a large cohort of asymptomatic patients without known prior exposure to an infected patient. In addition, we were able to compare asymptomatic infection rates among patients undergoing orthopedic procedures with patients under going non-orthopedic procedures and patients presenting for non-surgical admissions. The three key findings from this study include the asymptomatic infection rates, conversion of multiple patients who initially tested negative, and the younger age of patients with asymptomatic infections.

Mandatory screening prior to essential surgeries was adopted following consensus recommendations from multiple national and international healthcare agencies. Results of our hospital-wide SARS-CoV-2 asymptomatic testing policy suggest an overall 0.69% rate of asymptomatic or pre-symptomatic patients tested at a Midwestern academic tertiary referral center. However, the asymptomatic infection rate was significantly lower among patients undergoing surgical procedures compared to those asymptomatic patients admitted for non-surgical management. We also identified a non-statistically significant trend toward lower asymptomatic infection rates among patients undergoing orthopedic procedures compared to those undergoing non-orthopedic procedures.

Previous domestic reports of SARS-CoV-2 infection rates have been limited to targeted high risk populations, such as nursing home residents, persons with known exposure to patients with confirmed COVID-19, or those with symptoms consistent with the disease.^3,4 The positive test rate from endemic areas have ranged from 13.7% among asymptomatic women admitted for delivery to 46.8% of 12,594 patients tested at New York University.^4,5 These rates are markedly different than those observed in the current study, demonstrating marked variability among geographically disparate regions.

In addition to differences among geographically disparate regions, the asymptomatic SARS-CoV-2 infection rates we have reported are markedly lower than the 16.6% positive test rate reported by the Iowa Department of Public Health (IDPH) that representing a similar geographic area: As of April 29, IDPH reported 6,843 positive cases, along with an additional 34,494 negative tests.⁶ Although it is unknown how many of the test patients reported to IDPH were asymptomatic, it is likely that the majority of these tests represent patients presenting with symptoms concerning for active COVID-19 infection or close contact with a confirmed or suspected COVID-19 positive patient given the ongoing shortage of community-based screening in this region. Since March 1, 2020, a similar cumulative positive SARS-CoV-2 test rate of 18.4% has been reported to CDC by state and local public health, clinical, and commercial laboratories across the nation.⁷

Our asymptomatic infection rate of 0.69% may be most similar to a subset of patients presented in the recently published results of population screening in Iceland whichh reported 13 of 2,283 randomly sampled persons from the population tested positive (0.57%).⁸ This same study also reported results of targeted high risk screening, with 1,221 of 9,199 high risk persons testing positive for SARS-CoV-2 (13.27%).⁸ Of note, 11.9% of patients in the randomly selected Icelandic population screening group reported having symptoms, with 6 of 13 patients with positive tests reporting symptoms (46.2%). While the presence of symptoms in over 10% the Icelandic population screening group is in contrast to our asymptomatic cohort, it is likely more important to consider the sampling methods and disparate sociodemographic tendencies that might drive exposure and infection rates both domestically and abroad. Results from both our study and the population screening results from Iceland suggest an asymptomatic infection rate of less than 1%, whereas targeted high risk screening and results from endemic areas have reported infection rates ranging from 13.7% to 46.8%.^4-8

During this study, we identified four of 422 patients with initial negative test results that later converted to positive tests upon repeat testing (0.95%), suggesting a possible association between repeated hospital encounters and positive test rates. All four of these patients converted on a second test, with none of the 96 patients in our case series undergoing a third test converting to a positive test. None of the four patients who converted on repeat testing exhibited typical symptoms, such as fevers, cough, or dyspnea. Based on these four cases, our experience strongly supports repeated testing of patients prior to surgical procedures even in the absence of clinical symptoms.

Our internal testing policy has evolved throughout the COVID-19 pandemic. Based on initial recommendations, we required a negative test within 7 days of a surgical procedure. However, as case reports of patients converting to positive SARS-CoV-2 tests became available this policy was adjusted to testing within 48 hours of procedures even if the patient had quarantined since a previous test. Following conversion of three patients at our medical center, our testing policy now requires SARS-CoV-2 testing within 24 hours of planned surgeries. Identification of asymptomatic or pre-symptomatic infections is particularly critical for protecting healthcare providers during procedures including endotracheal intubation or other aerosol generating procedures, as these are particularly high risk for viral transmission.

In addition to guiding precautions taken by healthcare providers, testing of patients prior to elective procedures enables surgeons to delay elective surgeries for patients testing positive for SARS-CoV-2. Patients with asymptomatic infection experience markedly increased rates of severe pulmonary disease including 44.1% ICU admission rate and 20.5% mortality rate.⁹ Recent Clinical Practice Guidelines released by the International Consensus Group (ICM) and Research Committee of the American Association of Hip and Knee Surgeons has strongly recommended risk stratification including preoperative RT-PCR testing for SARS-CoV-2 prior to elective cases and delaying elective surgery in patients with active COVID-19 until recovery from infection.¹⁰

Limitations

There are several limitations to these findings: Nearly all patients in this cohort were from a largely rural geographic area including the entire state of Iowa and surrounding. However, these results can be interpreted in the context of regional and national tests results reported from IDPH and CDC. Our hospital-wide screening policy took effect early in on during the emergence of the COVID-19 pandemic within the United States. We have observed increasing rates of positive tests in asymptomatic patients that parallel the number of confirmed cases within our region. These results are also limited by a relatively small cohort patients compared to the overall regional population. Surveillance data from a forthcoming TestIowa.com statewide testing effort will help expand upon these early results and help to identify further areas of concentrated COVID-19 cases to focus surveillance and mitigation efforts.

Conclusion

We report an asymptomatic or pre-symptomatic SARS-CoV-2 infection rate of 0.69% in patients screened prior to essential surgeries at a regional tertiary referral center, with asymptomatic infection rate increased to 0.95% among patients with repeat testing during subsequent healthcare encounters. We also observed that patients testing positive for the virus were significantly younger than those with negative testing. Although asymptomatic or pre-symptomatic infection rates remain low, these results reiterate the risk of asymptomatic community spread, and should reinforce the need for preventive measures including social distancing and meticulous hygiene practices. These findings strongly support repeat testing after 48 hours lapse from prior tests in advance of surgical procedures or hospital admissions to identify infected patients without symptoms and mitigate asymptomatic SARS-CoV-2 transmission.