key: cord-0829936-5fwstsm3 authors: Munsch, N.; Gruarin, S.; Martin, A.; Nateqi, J.; Lutz, T.; Aberle, J.; Knapp, B. title: COVID-19 symptom frequency comparison: non-hospitalised positively and negatively tested persons with flu-like symptoms in Austria date: 2021-02-26 journal: nan DOI: 10.1101/2021.02.24.21252426 sha: 7060687217dea32640bb2ba3bffd4d52a066e939 doc_id: 829936 cord_uid: 5fwstsm3 Background A large number of studies report COVID-19 symptom frequencies but most studies focus on hospitalized patients. Therefore reported symptom frequencies vary and their applicability to the general population is limited. Here we report COVID-19 symptom frequencies for the general population of a central European country. Methods In a collaboration between the Vienna Social Fund (FSW) and the AI-biotech company Symptoma we report symptom frequencies based on the COVID-19 chatbot of the city government of Vienna and corresponding PCR-test results. Chatbot users answered 13 yes/no questions about symptoms and provided information about age and sex. Subsequently a medically trained professional came to their address to take a sample and PCR results were obtained. Findings Between November 2 and January 5, a total of 3011 persons experiencing flu-like symptoms had a PCR-test by a medical professional at home and completed the chatbot session prior to the test, 816 (27.1%) of them were COVID-19 positive. We compared the symptom frequencies between COVID-19 positive and negative users, and between male and female users. The symptoms (sorted by frequency) of users with positive PCR-test are malaise (81.1%), fatigue (72.9%), headache (64.1%), cough (57.7%), fever (50.7%), sore throat (40.7%), close contact with COVID-19 cases (34.9%), rhinorrhea (31.0%), sneezing (28.4%), dysgeusia (27.1%), hyposmia (26.5%), dyspnea (11.4%) and diarrhea (10.9%). Among these, cough, fever, hyposmia, dysgeusia, malaise, headache, close contact with COVID-19 case and fatigue are significantly (P < 0.01) increased in COVID-19 positive persons while dyspnea, diarrhea and sore throat are significantly (P < 0.01) decreased in COVID-19 positive persons. There was no significant difference for rhinorrhea and sneezing. Introduction COVID-19 symptom frequencies are a valuable information for authorities during this pandemic (e.g. screening the population and rapidly identifying cases). Symptom frequencies are used in applications including triage recommendation [1] and COVID-19 diagnostic [2] [3] . However, the reported symptom frequencies vary considerably across studies [1] . An example is "fever" for which symptom frequencies were reported between 7% and 91% [1] . Causes for these discrepancies include the over-representation of patients in hospital settings [2] and the lack of clarity on how symptoms were collected [1] . This study describes the COVID-19 symptom frequencies observed in the non-hospitalized Austrian population experiencing flu-like symptoms and compares them with COVID-19 negatively tested persons also experiencing flu-like symptoms. The data was collected systematically via a standardized web interface and the COVID-19 status determined by PCR-tests. From November 2020, the City of Vienna's digital COVID-19 symptom checker provided inhabitants of Austria's capital with an initial COVID-19 risk assessment, as well as possible options for further action, e.g. a PCR test. The aim was to offer, in addition to the non-emergency medical health line "1450", a fast to use, always available, and accurate digital channel, available via https://coronavirus.wien.gv.at/site/symptomchecker/ . Fonds Soziales Wien (Vienna Social Fund) and the company Symptoma developed the chatbot based on database, artificial intelligence, and algorithms of Symptoma [3] [4] [5] together. During . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted February 26, 2021. ; the chatbot conversation, the user is answering a series of 13 yes/no questions about symptoms. These include fever, cough, dyspnea, sneezing, rhinorrhea, sore throat, malaise, fatigue, diarrhea, headache, hyposmia, dysgeusia and close contact with COVID-19 case. Each user is asked the same questions and answers are recorded. The statistics reported in this paper are based on the combined information of the chatbot conversation and the results of a PCR result carried out by a medical professional. A total of 3011 users were screened this way between the 2nd of November 2020 and 5th of January 2021 . All data were anonymised prior to this analysis. Only sex, age, the answers to the questions and the result of the PCR-test were used for this study. Users who did not provide a sex information were exlcuded from the symptom frequencies comparison between female and male. All Symptom frequencies among C19+ and C19-users Summary statistics of participants and numerical details can be found in Table 1 . Our study cohort consists of 3011 non-hospitalised persons experiencing flu-like symptoms of which 816 tested positive for COVID-19 (C19+) and 2142 tested negative for COVID-19 (C19-). In Figure 1 , we compared the symptom frequencies between C19+ and C19-. The symptoms most . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted February 26, 2021. and hyposmia (26.5%). Dyspnea (11.4%) and diarrhea (10.9%) were rarely reported ( Figure 1A ). C19+ users significantly ( P < 0.001) more often experienced cough, fever, hyposmia, dysgeusia, malaise, headache and close contact with COVID-19 case. On the contrary, C19users significantly ( P < 0.001) more often experienced sore throat, diarrhea and dyspnea. ( P < 0.001). No significant difference between C19+ and C19-was found for rhinorrhea (P = 0.19) and sneezing (P = 0.11) ( Figure 1A ). The largest relative increase in C19+ persons can be found in hyposmia (+63% or +16 percent points), dysgeusia (+46% or +12 percent points), cough (+39% or +22 percent points) and fever (+39% or +20 percent points) while the largest relative decrease in C19+ persons can be found in dyspnea (-56% or -6 percent points) and diarrhea (-60% or -7 percent points) ( Figure 1B ). . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted February 26, 2021. ; https://doi.org/10.1101/2021.02.24.21252426 doi: medRxiv preprint In the C19+ group, fever is more frequently present for men than for women ( P = 0.010) while sneezing is more frequently present for women than for men ( P = 0.0012). In the C19-group, the fever frequency is also higher for men than for women (P < 0.001) but for sneezing the difference is insignificant ( P = 0.27). In the C19-group, women more frequently experience headache and sore throat (P < 0.001) (Figure 2) . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted February 26, 2021. ; The results also show that all the symptom frequencies reported here (with the exception of rhinorrhea and sneezing) are different between C19+ from C19-users. This agrees with other studies [7] [8] that also reported hyposmia, dysgeusia and fever as significantly increased in C19+ persons. Considering the high relative difference of hyposmia and dysgeusia frequencies for C19+ users, our study suggests that hyposmia and dysgeusia are specific but not sensitive, i.e. their presence likely leads to C19+ but no conclusion can be drawn from their absence. Finally, results show symptoms were reported in different frequencies by men and women, which could be caused by sex-specific differences in the clinical course [9] , [10] . One strength of our study is the systematic symptom collection i.e. each user was asked the same questions by the chatbot which allows for higher data consistency than handwritten reports. Additionally, the result of the PCR-test is automatically associated with the reported . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted February 26, 2021. ; symptoms, instead of being communicated by the users themselves which would risk the inclusion of falsely reported PCR-test results. Another strength is that most symptom frequencies reported in the literature are based on hospitalized patients [1] while this study reports symptom frequencies for the general population in central Europe. However, our study also has limitations: there is a selection bias for participants as old people are unlikely to use a chatbot (see Table 1 and [11] ) and there is an overrepresentation of female participants in this study (45.0% male and 55.0% female). Also, the dyspnea frequency and difference reported here might be due to a sample bias as (a) dyspnea is often a late symptom of an infection while chat-bot users might rather be at an earlier stage of an infection and (b) dyspnea can be a distressing symptom and affected individuals might rather call an emergency hotline instead of using a chatbot [12] . All relevant data is reported within the study. . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted February 26, 2021. ; https://doi.org/10.1101/2021.02.24.21252426 doi: medRxiv preprint All individuals using the chatbot agreed that their data will be used in an anonymised way Signs and symptoms to determine if a patient presenting in primary care or hospital outpatient settings has COVID-19 disease Real-time tracking of self-reported symptoms to predict potential COVID-19 An artificial intelligence-based first-line defence against COVID-19: digitally screening citizens for risks via a chatbot Diagnostic Accuracy of Web-Based COVID-19 Symptom Checkers: Comparison Study From symptom to diagnosis-symptom checkers re-evaluated : Are symptom checkers finally sufficient and accurate to use? An update from the ENT perspective COVID-19 signs, symptoms and severity of disease: A clinician guide Wearable sensor data and self-reported symptoms for COVID-19 detection Self-reported symptoms from exposure to Covid-19 provide support to clinical diagnosis, triage and prognosis: An exploratory analysis Male sex identified by global COVID-19 meta-analysis as a risk factor for death and ITU admission Sex differences in severity and mortality from COVID-19: are males more vulnerable? Digital Health Strategies to Fight COVID-19 Worldwide: Challenges, Recommendations, and a Call for Papers An official American Thoracic Society statement: update on the mechanisms, assessment, and management of dyspnea