Performance and feasibility of universal PCR admission screening for SARS‐CoV‐2 in a German tertiary care hospital

Anamnestic screening of symptoms and contact history is applied to identify coronavirus disease 2019 (COVID‐19) patients on admission. However, asymptomatic and presymptomatic patients remain undetected although the viral load may be high. In this retrospective cohort study, all hospitalized patients who received polymerase chain reaction (PCR) admission testing from March 26th until May 24th, 2020 were included. Data on COVID‐19‐specific symptoms and contact history to COVID‐19 cases were retrospectively extracted from patient files and from contact tracing notes. The compliance to the universal testing protocol was high with 90%. Out of 6940 tested patients, 27 new severe acute respiratory syndrome coronavirus‐2 infections (0.4%) were detected. Seven of those COVID‐19 cases (26% of all new cases) were asymptomatic and had no positive contact history, but were identified through a positive PCR test. The number needed to identify an asymptomatic patient was 425 in the first wave of the epidemic, 1218 in the low incidence phase. The specificity of the method was above 99.9%. Universal PCR testing was highly accepted by staff as demonstrated by high compliance. The costs to detect one asymptomatic case in future studies need to be traded off against the costs and damage caused by potential outbreaks of COVID‐19.

Universal testing for SARS-CoV-2 using nucleic-acidamplification tests has been described for distinct patient groups such as pregnant women, stroke patients, or before interventions [7][8][9][10][11][12] as well as for a smaller sample size 13 but to our knowledge not yet for the entire patient population of a large hospital with several thousand patients.

| Aim of the study
The aim of this study is to evaluate the potential and feasibility of universal testing to detect SARS-CoV-2 infections among asymptomatic patients.

| SARS-CoV-2 testing
Detection of SARS-CoV-2-RNA by real-time reverse transcriptase PCR was conducted routinely at the Institute for Virology and Immunobiology using primers and probes located in the SARS-CoV-2 E-gene as described by Corman et al. 14

| Study design and data collection
In this retrospective, observational study hospital admissions between March 26th and May 24th were included. Data on all SARS-CoV-2 tests performed in the virological laboratory were merged with patient admission data using unique case numbers representing one admission of a patient. Oropharyngeal swabs taken within two days before or after a patient's admission were considered to be admission testings as swabs were taken one or two days before elective admissions in some departments, on the day of admission and missed testing swabs were caught up on during the Calculation of costs for PCR assays was conducted using the German doctors' fee schedule EBM. 18

| Statistical analysis
During the study period, the German government started implementing extensive social distancing measures for the population on March 21st. Considering a maximum incubation period of 14 days for SARS-CoV-2 infection, the measures were estimated to take full effect on the infection rate by April 4th. [19][20][21][22] Thus, the interval from the beginning of the policy until April 4th was assigned to the first wave of the COVID-19 pandemic in Germany while the interval after April 4th represented a low incidence setting. The positivity rate in tested patients during the first wave and in the low incidence setting was compared by Fisher's exact test.

| Ethics approval and consent to participate
The Ethics committee of the University of Würzburg after viewing the design of the study waived the need to formally apply for ethical clearance because routine data of the clinical information system were to be retrospectively evaluated (University of Würzburg, Ethics committee, #20200507 06).

| Study population
As expected, the age distribution in the hospital admission cohort differed substantially from the general Bavarian population. As observed elsewhere, the COVID-19 incidence was highest in the Bavarian age strata 80-89 and above 90 years (see Figure 1).    [15][16][17] Accordingly, the weekly positivity rate of universal testing results declined from 2.1% in the first week 13 of the program to 0% in week 21 (see Figure 3).

| Accordance with the epidemiology in the general population
Statistical comparison of the frequency of positive results before and after this intervention shows a significantly higher SARS-CoV-2 positivity rate from March 26th till April 3rd compared to the testing period after April 4th (1.41% vs. 0.25% positive test results). The number needed to identify an asymptomatic patient was 425 in the first wave of the pandemic, and 1218 in the low incidence phase (see Figure 4).
In the hospital's catchment area Lower Franconia, the incidence rate declined from more than 60 weekly cases per 100,000 population to less than 5 weekly cases per 100,000 population.

| Costs of the universal PCR testing on admission
In this setting, the average testing costs were approximately 58,469  in hospitals support the concept of screening for clinical criteria and contact history as well. 28,29 Still, one in three detected cases was asymptomatic and identified only through universal PCR testing. In particular, in the second half of the study period, asymptomatic COVID-19 patients represented the majority of newly detected cases. The reason for this is unclear and may be related to different age groups affected, seasonal effects or the more widespread use of face coverings. Nevertheless, there is substantial evidence that viral shedding and disease transmission occur during the asymptomatic or presymptomatic period of  Transmissions by undetected SARS-CoV-2-positive staff and visitors may also cause nosocomial infections. Additional tests of patients who become symptomatic despite testing SARS-CoV-2 negative on admission may detect a proportion of these. In a high incidence setting, even regularly (e.g. weekly) universal testing of all patients may be useful, but the benefit must be traded off for the additional costs.
Limitations of our data arise from the identification of testing data: As admission swabs were not marked as such, all oropharyngeal swabs taken within two days before and two days after admission were identified as belonging to the universal testing. Data on symptoms, and previous COVID-19 contacts were only evaluated for SARS-CoV-2 positive patients as the systematic electronic recording of this information had just started in the middle of the study period and this data had to be extracted manually from medical notes. As this study was conducted in a low incidence country with a high

CONFLICT OF INTERESTS
The authors declare that there are no conflict of interests.

AUTHOR CONTRIBUTIONS
Ulrich Vogel and Manuel Krone designed and coordinated the study.

DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.