This is an open-access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in JMIR Public Health and Surveillance, is properly cited. The complete bibliographic information, a link to the original publication on https://publichealth.jmir.org, as well as this copyright and license information must be included.
The COVID-19 symptom-monitoring apps provide direct feedback to users about the suspected risk of infection with SARS-CoV-2 and advice on how to proceed to prevent the spread of the virus. We have developed the CoronaCheck mobile health (mHealth) platform, the first free app that provides easy access to valid information about the risk of infection with SARS-CoV-2 in English and German. Previous studies have suggested that the clinical characteristics of individuals infected with SARS-CoV-2 vary by age, gender, and viral variant; however, potential differences between countries have not been adequately studied.
The aim of this study is to describe the characteristics of the users of the CoronaCheck mHealth platform and to determine country-specific and sociodemographic associations of COVID-19–related symptoms and previous contacts with individuals infected with COVID-19.
Between April 8, 2020, and February 3, 2022, data on sociodemographic characteristics, symptoms, and reports of previous close contacts with individuals infected with COVID-19 were collected from CoronaCheck users in different countries. Multivariable logistic regression analyses were performed to examine whether self-reports of COVID-19–related symptoms and recent contact with a person infected with COVID-19 differed between countries (Germany, India, South Africa), gender identities, age groups, education, and calendar year.
Most app users (N=23,179) were from Germany (n=8116, 35.0%), India (n=6622, 28.6%), and South Africa (n=3705, 16.0%). Most data were collected in 2020 (n=19,723, 85.1%). In addition, 64% (n=14,842) of the users were male, 52.1% (n=12,077) were ≥30 years old, and 38.6% (n=8953) had an education level of more than 11 years of schooling. Headache, muscle pain, fever, loss of smell, loss of taste, and previous contacts with individuals infected with COVID-19 were reported more frequently by users in India (adjusted odds ratios [aORs] 1.3-8.3, 95% CI 1.2-9.2) and South Africa (aORs 1.1-2.6, 95% CI 1.0-3.0) than those in Germany. Cough, general weakness, sore throat, and shortness of breath were more frequently reported in India (aORs 1.3-2.6, 95% CI 1.2-2.9) compared to Germany. Gender-diverse users reported symptoms and contacts with confirmed COVID-19 cases more often compared to male users.
Patterns of self-reported COVID-19–related symptoms and awareness of a previous contact with individuals infected with COVID-19 seemed to differ between India, South Africa, and Germany, as well as by gender identity in these countries. Viral symptom–collecting apps, such as the CoronaCheck mHealth platform, may be promising tools for pandemics to support appropriate assessments. Future mHealth research on country-specific differences during a pandemic should aim to recruit representative samples.
The emergence of COVID-19, caused by SARS-CoV-2 infection, has led to a major ongoing public health crisis worldwide [
In the beginning, the CoronaCheck app was intended to enable symptom monitoring to offer quick advice to those affected. This advice needed to follow the official recommendations of the Robert Koch Institute as the National German Public Health Institute so that those affected could assess their symptoms to the best of the current knowledge and to complement the official support hotlines. Another goal was to provide certain segments of the population with easier access to information than through a telephone hotline. These included adolescents, people with hearing problems who may not fully understand the telephone voice, and migrants or tourists without sufficient language skills for a hotline in the local language.
CoronaCheck is an open-science mHealth platform with 109,603 installations and 88,537 completed questionnaires as of February 3, 2022. The app was developed in the first months of the COVID-19 pandemic in Germany in collaboration with university partners from Bavaria and software companies. The app is based on the TrackYourHealth platform [
The study was approved by the Ethics Committee of the University of Würzburg (ethical approval no. 71/20-me) and the university’s data protection officer and was carried out in accordance with the General Data Protection Regulations of the European Union [
CoronaCheck was developed in a scientific collaboration between 2 German university partners from Bavaria and software companies and is based on the openly available information about SARS-CoV-2 and corresponding recommendations of the Robert Koch Institute, the German national health authority [
On April 8, 2020 (10 weeks after the first COVID-19 cases were reported in Germany), data collection from this app began in Germany, followed by the first data from South Africa and India on April 24 and 28, 2020, respectively.
We collected data on age (years, categorized in 10-year intervals up to 79 years, ie, ≤9, 10-19, 20-29, 30-39, 40-49, 50-59, 60-69, 70-79, ≥80 years), gender identity (male, female, diverse, not specified), educational level (defined by years of schooling in 4 categories: ≤9 years, 10-11 years, ≥12 years, and “not reported” if users did not answer this question), and the information for whom the questionnaire was filled out (oneself, another person, not specified).
Participants were asked about the presence of 11 different COVID-19–related symptoms in the past 24 hours (yes/no): fever (defined as a temperature of 37.5 °C or more), sore throat, stuffed or runny nose, cough, loss of smell, loss of taste, shortness of breath, headache, muscle or joint pain, diarrhea, and general weakness. In the case of diarrhea, the symptom was assessed only in the first 2 months (April-June 2020), before it was deleted in accordance with the recommendations of the Robert Koch Institute.
We asked participants whether they or the person for whom they completed the questionnaire had had close contact with a person with confirmed COVID-19 in the past 14 days (yes, no, or not specified). We did not ask whether this confirmed case was based on a polymerase chain reaction (PCR) test, a medical diagnosis (including clinical radiology tests), or other information.
The app was developed based on the TrackYourHealth platform [
App illustration: screenshots of (1) Start Check, (2) Questionnaire, (3) Feedback Module, (4) Tips, and (5) Detailed Tip.
Only participants who met the following criteria were included in the final analysis: provided consent to use data for research purposes; answered questions about age, user status, and gender identity; provided GPS data; and answered questions about previous contacts (in the past 14 days) with a person infected with COVID-19 and the presence or absence of the 11 symptoms listed before. In case the app was used multiple times by a single person, only the first recording was used. It should be noted that every time a questionnaire is filled out, the GPS location is saved, provided that the user gives their consent. The location permission is regulated by the operating system (Android or iOS). We do not store the raw location, but we process the data we receive from the operating system to make it more coarse-grained, and only store a value with an accuracy of 11.1 km. The level of detail with which we store the GPS sensor data was determined in discussions with all stakeholders, the institutional review board, and relevant laws and regulations. Further, only data from countries with more than 3000 users (Germany, India, and South Africa) were included in the statistical analyses. Data from other countries were comparatively small (each <600 users). The educational level was categorized into 3 groups (<12 years, ≥12 years, missing). Since most reports were collected in 2020 and the first variants of concern emerged around the turn of 2020/2021 [
Descriptive statistics were conducted to describe the CoronaCheck user characteristics. To compare sociodemographic characteristics (gender identity, age, education), calendar year, and user status among app users from Germany, India, South Africa, and other countries, chi-square tests were applied. Chi-square tests and Fisher exact tests were performed to compare the proportion of users who reported COVID-19–related symptoms and contact with a person infected with COVID-19 between 3 countries (Germany, India, and South Africa), gender identity (3 categories), educational level (3 categories), calendar years (2020 vs 2021/2022), and user status (symptoms reported for oneself vs another person). Univariate logistic regressions were calculated to assess the associations between reporting COVID-19–related symptoms and reporting contact with a person who tested positive for COVID-19 with country, gender, age (in 10-year categories), calendar year, educational level, and user status. Multivariable logistic regression was conducted to adjust the data for country, gender identity, age, calendar year, education level, and user status. Odds ratios (ORs) and their 95% CIs were estimated to assess the statistical uncertainty. All statistical analyses were performed using SPSS version 26 (IBM Corp), and all tests were 2-tailed.
From April 8, 2020, to February 3, 2022, we received a total of 88,537 recordings. Recordings with missing GPS data (n=29,449, 33.3%), missing consent that data can be used for research purposes (n=30,836, 34.8%), or missing information about age (n=2001, 2.3%), gender identity (n=1927, 2.2%), or user status (n=10,039, 11.3%) were excluded. Only the first recording per person was included in the case of multiple use. After applying these exclusion criteria, 23,179 recordings remained. All recordings included answers on the questions about previous contact (in the past 14 days) with a person infected with COVID-19 and the presence or absence of the 11 symptoms listed before. Most recordings were from Germany (n=8116, 35.0%), India (n=6622, 28.6%), and South Africa (n=3705, 16.0%). Recordings from 131 other countries ranged from 1 (0.004%) to 595 (2.6%) per country and were summarized into 1 group, “other countries” (n=4736, 20.4%).
The characteristics of the sample stratified by country are summarized in
The 4736 (20.4%) app users from “other countries” were excluded from further analysis, yielding a final analytic cohort of 18,443 (79.6%) individuals.
Characteristics of CoronaCheck app users stratified by country.
Characteristics | Country | Total (N=23,179), n (%) | ||||||||||
|
Germany (n=8116, 35.0%), n (%) | India (n=6622, 28.6%), n (%) | South Africa (n=3705, 16.0%), n (%) | Other countries (n=4736, 20.4%), n (%) |
|
|||||||
|
||||||||||||
|
Man | 4769 (58.8) | 4686 (70.8) | 2082 (56.2) | 3305 (69.8) | 14,842 (64.0) | ||||||
|
Woman | 3308 (40.8) | 1893 (28.6) | 1611 (43.5) | 1406 (29.7) | 8218 (35.5) | ||||||
|
Diverse | 39 (0.5) | 43 (0.6) | 12 (0.3) | 25 (0.5) | 119 (0.5) | ||||||
|
||||||||||||
|
< 30 | 1964 (24.2) | 4649 (70.2) | 1982 (53.5) | 2507 (52.9) | 11,102 (47.9) | ||||||
|
30-49 | 2722 (33.5) | 1556 (23.5) | 1469 (39.6) | 1480 (31.4) | 7227 (31.2) | ||||||
|
≥50 | 3430 (42.3) | 417 (6.3) | 254 (6.9) | 749 (15.8) | 4850 (20.9) | ||||||
|
||||||||||||
|
<12 | 3827 (47.2) | 1849 (27.9) | 891 (24.0) | 1372 (29.0) | 7939 (34.3) | ||||||
|
≥12 | 3030 (37.3) | 2047 (30.9) | 1906 (51.4) | 1970 (41.6) | 8953 (38.6) | ||||||
|
Missing | 1259 (15.5) | 2726 (41.2) | 908 (24.5) | 1394 (29.4) | 6287 (27.1) | ||||||
|
||||||||||||
|
2020 | 7084 (87.3) | 5805 (87.7) | 3094 (83.5) | 3740 (79.0) | 19,723 (85.1) | ||||||
|
2021/2022 | 1032 (12.7) | 817 (12.3) | 611 (16.5) | 996 (21.0) | 3456 (14.9) | ||||||
|
||||||||||||
|
Reported for oneself | 7499 (92.4) | 5222 (78.9) | 3314 (89.4) | 4032 (85.1) | 20,067 (86.6) | ||||||
|
Reported for another person | 617 (7.6) | 1400 (21.1) | 391 (10.6) | 704 (14.9) | 3112 (13.4) |
Of the 18,443 recordings collected in Germany, India, and South Africa, the 3 most frequent self-reported symptoms were headache (n=5351, 29.0%), cough (n=5219, 28.3%), and general weakness (n=4592, 24.9%), followed by muscle pain (n=4131, 22.4%), runny nose (n=3984, 21.6%), fever (n=3815, 20.7%), and sore throat (n=3646, 19.8%). Shortness of breath (n=2852, 15.5%), loss of smell (n=2224, 12.1%), loss of taste (n=2161, 11.7%), and diarrhea (n=136, 0.7%) were reported less frequently.
In 5105 (27.7%) of the 18,443 recordings collected in the 3 countries, the app users reported close contact with someone infected with SARS-CoV-2 in the past 14 days.
We found differences between the 3 countries in all symptoms (
aORs and their 95% CIs stratified by the country of the CoronaCheck app users. aOR: adjusted odds ratio.
Our results showed gender identity differences for all symptoms except diarrhea (
aORs and their 95% CIs stratified by the gender identity of the CoronaCheck app users. aOR: adjusted odds ratio.
Older persons were less likely to report headache, cough, general weakness, runny nose, fever, sore throat, and loss of smell (aORs 0.86-0.96, 95% CI 0.84-0.99, per 10-year category), whereas they were more likely to report muscle pain (aOR 1.03, 95% CI 1.01-1.05, per 10-year category;
Unadjusted ORsa and aORsb by 10-year age category (N=18,445).c
Variable | Unadjusted OR (95% CI) | aOR (95% CI) |
Headache | 0.84 (0.82-0.86) | 0.87 (0.85-0.89) |
Cough | 0.85 (0.84-0.87) | 0.90 (0.89-0.92) |
General weakness | 0.87 (0.86-0.89) | 0.96 (0.94-0.99) |
Muscle pain | 0.96 (0.94-0.98) | 1.03 (1.01-1.05) |
Runny nose | 0.87 (0.85-0.89) | 0.86 (0.84-0.88) |
Fever | 0.76 (0.74-0.77) | 0.92 (0.90-0.94) |
Sore throat | 0.84 (0.82-0.86) | 0.87 (0.85-0.89) |
Shortness of breath | 0.91 (0.89-0.93) | 0.99 (0.97-1.02) |
Loss of smell | 0.81 (0.79-0.83) | 0.95 (0.92-0.98) |
Loss of taste | 0.83 (0.81-0.86) | 0.98 (0.95-1.02) |
Diarrhea | 1.17 (1.08-1.28) | 0.92 (0.84-1.01) |
Contact with infected person | 0.71 (0.70-0.73) | 0.87 (0.85-0.89) |
aOR: odds ratio.
baOR: adjusted odds ratio.
cThe multivariable regression model was adjusted for country, gender identity, calendar year, education, and user status.
App users with at least 12 years of school education reported headache, general weakness, muscle pain, runny nose, and sore throat more often compared to users with 11 or fewer years of education (aORs 1.09-1.39, 95% CI 1.00-1.51;
aORs and their 95% CIs stratified by educational level (years of schooling). aOR: adjusted odds ratio.
Most COVID-19–associated symptoms were more likely to be reported by app users in 2021/2022 compared to 2020 (aORs 1.11-1.33, 95% CI 1.01-1.46;
The reporting of close contact with someone infected with SARS-CoV-2 in the past 14 days was more likely in 2021/2022 compared to 2020 (aOR 1.37, 95% CI 1.23-1.52).
aORs and their 95% CIs stratified by the year of using the CoronaCheck app. aOR: adjusted odds ratio.
We found differences between persons for whom symptoms were reported for all symptoms except diarrhea (
Symptoms as well as close contact with a person positive for SARS-CoV-2 in the past 14 days were more often reported when the data were entered for another person compared to for oneself (aORs 1.28-2.09, 95% CI 1.17-2.31).
aORs and their 95% CIs stratified by user status (ie, user-reported symptoms for another person or for oneself). aOR: adjusted odds ratio.
Our study showed that self-reported COVID-19–related symptoms and reports of previous contact with individuals infected with COVID-19 vary by country, gender identity, age, calendar year, education, and user status.
Overall, the most frequently reported symptoms were headache, cough, and general weakness. Symptoms commonly included in COVID-19 case definitions, such as shortness of breath and fever [
Our findings regarding the occurrence of common symptoms, such as headache, general weakness, and runny nose, are consistent with previous studies. Specifically, previous studies in individuals infected with PCR-confirmed SARS-CoV-2 have shown that fatigue, headache, malaise, myalgia, and upper respiratory symptoms (sore throat, cough, sneezing, rhinitis) occur early after symptom onset, whereas symptoms considered more characteristic of COVID-19, such as lower respiratory and chemosensory symptoms, occur later [
Our study showed differences in self-reported COVID-19 symptoms primarily between Germany, India, and South Africa. App use from India and South Africa was associated with a disproportionately high likelihood of experiencing COVID-19–related symptoms and knowledge of close contact with an infected person. Specifically, the likelihood of reporting contact with a person infected with COVID-19 was 8.3-fold higher among users from India compared to Germany. Most symptoms were also more likely to be reported by users from India than by German users. World Health Organization (WHO) official data on confirmed COVID-19 cases show conflicting results, with higher numbers of confirmed cases and deaths per capita in Germany than in India and South Africa [
Gender identity differences were found in this study. Several symptoms were more likely to be reported by women and gender-diverse users than by men. Exceptions to the general assumption that women are more likely to report symptoms than men were observed for fever. Previous studies of individuals testing positive for COVID-19 support the assumption that women are more likely than men to be affected by headache and sore throat, as well as rhinitis [
The impact of COVID-19 on COVID-19 symptoms in gender-diverse individuals has not been well captured; however, the more than 5-fold increase in the likelihood of gender-diverse users reporting prior contact with an individual who tested positive for COVID-19, as well as the higher odds of reporting symptoms observed in this study, is supported by the notion that gender-diverse individuals are more vulnerable to COVID-19, which extends to the areas of increased barriers to COVID-19 transmission prevention strategies, decreased physical and mental health, and barriers to health services [
Older age was found to be associated with a lower likelihood of reporting contact with a person infected with COVID-19. In support of this finding, previous studies have found an association between older age and self-reported adherence to COVID-19 public health measures [
The increased odds of reporting contact with someone who tested positive for SARS-CoV-2 in 2021/2022 compared to 2020 is supported by the increased transmission of the variants of concern, which predominated from 2021 onward. The Alpha variant (B.1.1.7), first identified in the United Kingdom at the end of 2020; the Beta variant (B.1.351), first identified in South Africa at the end of 2020; the Delta variant (B.1.617.2), first identified in India at the end of 2020; and the Omicron variant (B.1.1.529), first identified in South Africa at the end of 2021, have roughly 50% increased transmission [
In this study, the person for whom the symptom check was used affected the outcome: the likelihood of reporting contact with a person positive for COVID-19 and various COVID-19–related symptoms was higher when the app was used to check symptoms for another person. Thus, it appears that the app was used more frequently for other individuals when there was a specific suspicion of infection. Differences in reporting systems have been observed previously, with a higher percentage of symptoms reported in self-reports than in physician reports [
Our study had several strengths. First, data were collected over a long time interval that began early in the pandemic. Data were collected across different countries, with considerable differences in health care systems, economic status, and governmental regulations to combat the spread of the virus, therefore increasing the generalizability of the findings. The number of gender-diverse individuals among app users also allowed us to identify the specifics of symptoms reported by COVID-19 among gender minorities. In addition, it was possible to develop the app in a relatively short period and adapt it to the MDR, which meant that, among other things, special consideration was given to the risks involved in using the app. The app made use of GPS measurements when users allowed it, which shows that such measurements are also welcomed by many users. Finally, the content of the app was based on the recommendations of the Robert Koch Institute, the national public health institute in Germany. We therefore regularly adapted the app to the changing official recommendations.
Our study also had some limitations that need to be noted. First, objective data on whether app users tested positive for COVID-19 or not were not available. Thus, we were not able to validate the subjectively reported information about COVID-19 test positivity. Moreover, no objective data were available if the infected contact person of the user had a positive COVID-19 test; however, the corresponding question asked specifically about contact with a person who was a confirmed case.
Second, all information about specific symptoms was self-reported by the users, which was not validated by, for example, doctors’ diagnoses. In those cases where the questionnaire was filled out for another person, the validity of the data is further limited. This especially applied to the youngest and oldest app users, where a higher proportion of reported information was not entered by themselves (51.5% among children ≤9 years old and 21.1% among the elderly ≥80 years old vs 12.1% among 10-79-year-old persons).
Third, participants’ socioeconomic status was limited to the information about the years of education in school. This information was further limited by the high number of missing entries as well as by differences in the educational systems between countries.
Fourth, we had to adjust the app several times to meet the changing official recommendations of the German public health authority, the Robert Koch Institute, over the course of the pandemic. Therefore, during the overall time of app usage, especially in the early phase of the pandemic, some information was slightly differently presented to users. For example, the question regarding diarrhea was only administered for the first 2 months before it was removed.
Furthermore, as data collection was based on a mobile app, results might not be generalized to individuals with lower affinity for digital health apps. In concrete terms, this also means that information and selection bias must be taken into account. However, observer and recall bias can be reduced to some extent by an app in terms of patient-reported outcomes. Therefore, app-based data have strengths as well as weaknesses that need to be considered. Moreover, mHealth apps are always not just tools but also social phenomena that create a feedback loop between app and user, which means that users continuously adjust their behavior as they use the app. However, since the CoronaCheck app has collected a comparatively large amount of data in different cultures compared to other mHealth apps, certain differences may have been balanced out in the sense of generalization. Finally, it should be mentioned that apps that are made available on multiple mobile operating systems will also always generate differences in the resulting data, as the user interface is never completely identical.
Our app-based results indicated country-specific differences in COVID-19–related symptom patterns. It seems that self-reported COVID-19–related symptoms and awareness about previous contacts with individuals infected with COVID-19 are more likely among app users from India and South Africa (compared to users from Germany), as well as in gender-diverse (compared to male) users. Symptom assessment tools could play a positive role in the context of widespread infectious diseases. In general, with COVID-19, we saw that the way symptoms were reported and the way infection and COVID-19 deaths were defined were different [
Future mHealth research on country-specific differences during a viral pandemic should aim to recruit representative samples. If pseudonymous assessments are possible, studies should collect information about symptoms, confirmed testing, and adherence to public health recommendations longitudinally.
Supplementary tables.
adjusted odds ratio
Medical Device Regulations
mobile health
odds ratio
polymerase chain reaction
We are grateful to all app users for providing data for research purposes. We thank Rachel Dale for proofreading the manuscript and Marc Schickler for his help in visualizing the results.
The data presented in this study are only available on request from the corresponding author. The data are not publicly available, because participants’ informed consent did not cover public deposition of data.
EH analyzed the data and drafted the manuscript. TK was the coprincipal investigator, supervised the development of the app regarding the medical content, planned and developed the questionnaire, and participated in the planning of the statistical analyses and writing of the manuscript. CS participated in the development of the study design and questionnaire and updated the medical content and corresponding recommendations by state and nationwide health authorities. WS and MW participated in the development of the study design. PH spent resources and provided advice. MW participated in the statistical analyses. TP analyzed the data and drafted the manuscript. RP was the principal investigator, was responsible for the study design and technical development, and drafted the manuscript. All coauthors have reviewed and approved the manuscript for submission. The corresponding author attests that all listed authors meet authorship criteria and that no others meeting the criteria have been omitted.
None declared.