Prevalence and Clinical Presentation of Health Care Workers With Symptoms of Coronavirus Disease 2019 in 2 Dutch Hospitals During an Early Phase of the Pandemic

Importance On February 27, 2020, the first patient with coronavirus disease 2019 (COVID-19) was reported in the Netherlands. During the following weeks, at 2 Dutch teaching hospitals, 9 health care workers (HCWs) received a diagnosis of COVID-19, 8 of whom had no history of travel to China or northern Italy, raising the question of whether undetected community circulation was occurring. Objective To determine the prevalence and clinical presentation of COVID-19 among HCWs with self-reported fever or respiratory symptoms. Design, Setting, and Participants This cross-sectional study was performed in 2 teaching hospitals in the southern part of the Netherlands in March 2020, during the early phase of the COVID-19 pandemic. Health care workers employed in the participating hospitals who experienced fever or respiratory symptoms were asked to voluntarily participate in a screening for infection with the severe acute respiratory syndrome coronavirus 2. Data analysis was performed in March 2020. Main Outcomes and Measures The prevalence of severe acute respiratory syndrome coronavirus 2 infection was determined by semiquantitative real-time reverse transcriptase–polymerase chain reaction on oropharyngeal samples. Structured interviews were conducted to document symptoms for all HCWs with confirmed COVID-19. Results Of 9705 HCWs employed (1722 male [18%]), 1353 (14%) reported fever or respiratory symptoms and were tested. Of those, 86 HCWs (6%) were infected with severe acute respiratory syndrome coronavirus 2 (median age, 49 years [range, 22-66 years]; 15 [17%] male), representing 1% of all HCWs employed. Most HCWs experienced mild disease, and only 46 (53%) reported fever. Eighty HCWs (93%) met a case definition of fever and/or coughing and/or shortness of breath. Only 3 (3%) of the HCWs identified through the screening had a history of travel to China or northern Italy, and 3 (3%) reported having been exposed to an inpatient with a known diagnosis of COVID-19 before the onset of symptoms. Conclusions and Relevance Within 2 weeks after the first Dutch case was detected, a substantial proportion of HCWs with self-reported fever or respiratory symptoms were infected with severe acute respiratory syndrome coronavirus 2, likely as a result of acquisition of the virus in the community during the early phase of local spread. The high prevalence of mild clinical presentations, frequently not including fever, suggests that the currently recommended case definition for suspected COVID-19 should be used less stringently.


+ Editorial + Supplemental content Introduction
Since December 2019, the world has been in the grip of the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) and the disease it causes, coronavirus disease 2019 (COVID-19). 1 On February 27, 2020, the first patient with COVID-19 was detected in the Netherlands, after a trip to northern Italy between February 18, 2020, and February 21, 2020. 2 From then until March 6, 2020, another 127 COVID-19 cases were identified in the Netherlands, including 9 health care workers (HCWs) in 2 Dutch teaching hospitals in the southern part of the Netherlands who received a diagnosis of COVID-19 between March 2, 2020, and March 6, 2020. Eight of these 9 HCWs had no history of travel to China or northern Italy, raising the question of whether undetected community circulation was occurring. Because these findings coincided with the seasonal influenza peak 3 and because SARS-CoV-2 infection in HCWs could lead to both sick leave and introduction of the virus into the hospital, this finding prompted a demand for testing HCWs. After initial observations of SARS-CoV-2 detection in persons with mild symptoms who did not meet the definition for case finding, 1 screening for SARS-CoV-2 was implemented to determine the prevalence and the clinical presentation of COVID-19 among HCWs with self-reported fever or respiratory symptoms in these 2 hospitals.

Study Design, Setting, and Population
The study was reviewed by the Ethics Committee Brabant, the Netherlands. The study was judged to be beyond the scope of the Medical Research Involving Human Subjects Act, and a waiver of written informed consent was granted. Oral informed consent was obtained from all HCWs for SARS-CoV-2 testing and from SARS-CoV-2-infected HCWs for data collection. Data were deidentified before analysis. This study follows the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline.
A cross-sectional study was conducted in 2 teaching hospitals (700-bed Amphia Hospital, Breda, the Netherlands; 800-bed Elisabeth-TweeSteden Hospital, Tilburg, the Netherlands) employing 9705 HCWs (Figure 1). Between March 7, 2020, and March 12, 2020, HCWs with selfreported fever or (mild) respiratory symptoms in the last 10 days were tested voluntarily for SARS-CoV-2 infection, following the local infection control policy during outbreaks.

Procedures
A semiquantitative real-time reverse transcriptase-polymerase chain reaction (45 cycles) targeting the SARS-CoV-2 E-gene with high analytical sensitivity and specificity was performed on self-collected oropharyngeal samples, as described previously 4 and in eAppendix 1 in the Supplement. Structured interviews were conducted between March 12, 2020, and March 16, 2020, to document symptoms for all HCWs with confirmed COVID-19, including those who received a diagnosis before March 7, 2020 (eAppendix 2 in the Supplement). Data were collected with EDC data collection software version 2020.1 (Castor). Recovery was defined as being without symptoms for more than 24 hours.

Statistical Analysis
Given the descriptive nature of this study, sample size calculations and analysis for statistical significance were not performed. Continuous variables were expressed as medians and ranges.
Categorical variables were summarized as counts and percentages. There were no missing data. All analyses were performed with SPSS statistical software version 25.0 (IBM Corp). Data analysis was performed in March 2020.

Results
Of The median real-time reverse transcriptase-polymerase chain reaction cycle threshold value (ie, the number of cycles at which the fluorescence exceeds the threshold) was 27.0 (range, 14.5-38.5).
Within the limited resolution in time since the onset of symptoms, cycle threshold values tended to be higher in HCWs who were tested later in the course of the disease (Figure 3). Cycle threshold values were

Discussion
Two weeks after the first Dutch patient with COVID-19 was reported, the prevalence of COVID-19 in HCWs with self-reported fever or respiratory symptoms in 2 Dutch hospitals in the southern part of the Netherlands was 6%, representing 1% of all HCWs employed. This unexpected high prevalence  explain the vast majority of cases coming from more than 50 different departments in 2 hospitals.
The low percentage of men among HCWs with COVID-19 (17%) reflects that among the source population of HCWs in the 2 participating hospitals (18%).
Most HCWs experienced mild disease compared with the clinical presentation and outcomes reported for hospitalized patients so far. 5,6 Notably, fever or a feverish feeling was frequently not

Limitations
This study has several limitations. First, the screening of HCWs was based on the presence of fever or respiratory symptoms in the last 10 days, and no data were collected for HCWs without these symptoms. The observed 1% prevalence in all HCWs is, thus, a minimal estimate. The lack of data for asymptomatic HCWs also precluded estimates of the sensitivity and specificity of the reported symptoms. Second, oropharyngeal swabs were used for testing, which may have slightly lower sensitivity than a nasopharyngeal swab. 7  In Dutch hospitals, sick leave has no personal financial consequences. Underreporting is, therefore, not expected to be substantial in this group of professionals with a high sense of responsibility. In addition, recall bias is unlikely to have affected the reporting of fever or respiratory symptoms.
Health care workers were not aware of their SARS-CoV-2 infection status at the time of testing, and the recall period was short (up to 10 days). At the time of the interview, however, participants had knowledge of their SARS-CoV-2-positive test, and recall bias could thus have influenced the spectrum of symptoms reported. Prospective studies using diaries and, if possible, documenting symptoms while masking participants for test results may overcome such bias.

Conclusions
During the containment phase and within 2 weeks after the first Dutch case was detected, a substantial proportion of HCWs with self-reported fever or respiratory symptoms were infected with SARS-CoV-2, likely as the result of acquisition of the virus in the community during the early phase of local spread. This observation confirms the insidious nature of SARS-CoV-2 spread, given the high prevalence of mild clinical presentations that may go undetected. 8 The spectrum of mild symptoms present in HCWs with COVID-19, frequently not including fever, suggests that the currently recommended case definition for suspected COVID-19 should be used less stringently.