JPH JMIR Public Health Surveill JMIR Public Health and Surveillance 2369-2960 JMIR Publications Toronto, Canada v9i1e40036 36692925 10.2196/40036 Original Paper Original Paper Syndromic Surveillance Using Structured Telehealth Data: Case Study of the First Wave of COVID-19 in Brazil Sanchez Travis Mavragani Amaryllis Yan Sun Afari-kumah Eben Tran Ngoc Ha Boaventura Viviane S MD, PhD 1
Laboratório de Doenças Infecciosas Transmitidas por Vetores Instituto Gonçalo Moniz, Fundação Oswaldo Cruz Rua Waldemar Falcao, 121 Salvador, 40296-710 Brazil 55 71 9619 4245 viviane.boaventura@fiocruz.br
2 https://orcid.org/0000-0002-7241-6844 Grave Malú DSc 3 4 5 https://orcid.org/0000-0002-7697-0658 Cerqueira-Silva Thiago MD, PhD 2 6 https://orcid.org/0000-0003-4534-2509 Carreiro Roberto MSc 7 https://orcid.org/0000-0002-0356-0076 Pinheiro Adélia MD, PhD 8 https://orcid.org/0000-0003-2992-2167 Coutinho Alvaro DSc 3 https://orcid.org/0000-0002-4764-1142 Barral Netto Manoel MD, PhD 2 6 https://orcid.org/0000-0002-5823-7903 Laboratório de Doenças Infecciosas Transmitidas por Vetores Instituto Gonçalo Moniz, Fundação Oswaldo Cruz Salvador Brazil Faculdade de Medicina Federal University of Bahia Salvador Brazil Department of Civil Engineering Federal University of Rio de Janeiro Rio de Janeiro Brazil Fundação Oswaldo Cruz Salvador Brazil Department of Civil Engineering Fluminense Federal University Niterói Brazil Instituto Gonçalo Moniz, Fundação Oswaldo Cruz Salvador Brazil Centre for Data and Knowledge Integration for Health Instituto Gonçalo Moniz, Fundação Oswaldo Cruz Salvador Brazil Departamento de Ciências da Saúde Universidade Estadual de Santa Cruz Salvador Brazil Corresponding Author: Viviane S Boaventura viviane.boaventura@fiocruz.br 2023 24 1 2023 9 e40036 2 6 2022 14 11 2022 24 11 2022 27 12 2022 ©Viviane S Boaventura, Malú Grave, Thiago Cerqueira-Silva, Roberto Carreiro, Adélia Pinheiro, Alvaro Coutinho, Manoel Barral Netto. Originally published in JMIR Public Health and Surveillance (https://publichealth.jmir.org), 24.01.2023. 2023

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.

 Background

Telehealth has been widely used for new case detection and telemonitoring during the COVID-19 pandemic. It safely provides access to health care services and expands assistance to remote, rural areas and underserved communities in situations of shortage of specialized health professionals. Qualified data are systematically collected by health care workers containing information on suspected cases and can be used as a proxy of disease spread for surveillance purposes. However, the use of this approach for syndromic surveillance has yet to be explored. Besides, the mathematical modeling of epidemics is a well-established field that has been successfully used for tracking the spread of SARS-CoV-2 infection, supporting the decision-making process on diverse aspects of public health response to the COVID-19 pandemic. The response of the current models depends on the quality of input data, particularly the transmission rate, initial conditions, and other parameters present in compartmental models. Telehealth systems may feed numerical models developed to model virus spread in a specific region.

 Objective

Herein, we evaluated whether a high-quality data set obtained from a state-based telehealth service could be used to forecast the geographical spread of new cases of COVID-19 and to feed computational models of disease spread.

 Methods

We analyzed structured data obtained from a statewide toll-free telehealth service during 4 months following the first notification of COVID-19 in the Bahia state, Brazil. Structured data were collected during teletriage by a health team of medical students supervised by physicians. Data were registered in a responsive web application for planning and surveillance purposes. The data set was designed to quickly identify users, city, residence neighborhood, date, sex, age, and COVID-19–like symptoms. We performed a temporal-spatial comparison of calls reporting COVID-19–like symptoms and notification of COVID-19 cases. The number of calls was used as a proxy of exposed individuals to feed a mathematical model called “susceptible, exposed, infected, recovered, deceased.”

 Results

For 181 (43%) out of 417 municipalities of Bahia, the first call to the telehealth service reporting COVID-19–like symptoms preceded the first notification of the disease. The calls preceded, on average, 30 days of the notification of COVID-19 in the municipalities of the state of Bahia, Brazil. Additionally, data obtained by the telehealth service were used to effectively reproduce the spread of COVID-19 in Salvador, the capital of the state, using the “susceptible, exposed, infected, recovered, deceased” model to simulate the spatiotemporal spread of the disease.

 Conclusions

Data from telehealth services confer high effectiveness in anticipating new waves of COVID-19 and may help understand the epidemic dynamics.

 telehealth telemedicine disease surveillance mathematical model COVID-19 prediction cases detection monitoring surveillance computational modeling spread transmission disease infectious diseases syndromic Introduction

Telehealth encompasses the distinct ways of interaction between patients and their health care providers. The growing popularity of virtual visits stems from the possibility of faster contact, elimination of transportation time [1], and by providing initial care in remote areas without adequate health care [2], making health care more efficient. Telehealth services can be used for diagnosis, treatment, follow-up, and screening purposes [3]. During the COVID-19 pandemic, telehealth services have been widely used for screening suspected cases [4,5], as limited physical contact reduces everyone’s exposure to COVID-19. This successful strategy safely provides access to health care services and expands assistance to remote rural areas and underserved communities in situations of shortage of specialized health professionals [3]. Using health information platforms, health care workers can systematically collect qualified data on suspected cases. Such data can be used as a proxy for the spread of infectious diseases for health surveillance [6-11].

The mathematical modeling of epidemics is a well-established field that has been successfully used for tracking the spread of SARS-CoV-2 infection, supporting the decision-making process on diverse aspects of public health response to the COVID-19 pandemic. Such models are usually defined as compartmental models. The population under study is divided into compartments based on qualitative characteristics, with different assumptions about the nature and rate of transfer across compartments. The urgency of the COVID-19 pandemic has motivated the need for more research in this area, with several models for this pandemic outbreak being presented in the last few years [12-18]. The response of the current models depends on the quality of input data, particularly the transmission rate, initial conditions, and other parameters present in compartmental models. Telehealth systems may feed numerical models developed to model virus spread in a specific region.

Here, we report using a high-quality data set obtained from a state-based telehealth service for disease surveillance both for forecasting the geographical spread of new cases of COVID-19 and feeding an advanced computational model that tries to reproduce the virus spread dynamics. We used data from a toll-free telehealth service of the 14.8 million population from the state of Bahia, Brazil [19], to simulate the spatiotemporal spread of COVID-19 using a compartmental model with diffusion in Salvador, the state capital.

 Methods Telehealth Service and Data Collection

Telecoronavirus was a toll-free phone-teletriage service offered to the population of the state of Bahia, Brazil (population of 14.8 million, 417 municipalities, located in the northeast of the country; Multimedia Appendix 1) [19]. Risk screening was provided for cases suspected of COVID-19 from March 24, 2020, to July 31, 2020, during the first epidemic wave, starting 18 days after the first confirmed case of COVID-19 in the state. This toll-free service operated 7 days a week and was publicized statewide by the state government through television, the internet, social media, billboards, radio, and newspapers. In total, 77% (320/417) of all cities in Bahia and all sanitary districts of the state capital, Salvador, accessed the telehealth service.

Structured data were collected during teletriage by a health team of medical students supervised by physicians. Data were registered in a responsive web application for planning and surveillance purposes. The data set was designed to quickly identify users, city, residence neighborhood, date, sex, age, and COVID-19–like symptoms (fever, cough, breathlessness, rhinorrhea, and gustatory or olfactory disorder). For this study, the call reporting at least one of those symptoms was considered a suspected case of COVID-19.

Strategies to improve the service quality were applied, including educational and technical support for the health care team, updating the online application based on clinical protocols, and monitoring the quality of information registered by the health care team, as previously described [19]. Data of telehealth service were categorized according to the phone call date, the type of COVID-19–like symptom reported by the user, and the city of residence. For Salvador, the state capital, calls were categorized according to the COVID-19–like symptoms and the user’s health district. The COVID-19 cases’ notification date and their related geolocation in the state of Bahia were obtained through the Brazilian Ministry of Health. For Salvador, notification data were obtained through the Salvador Municipal Health Department. We further quantified the cities of the state of Bahia and the districts of Salvador, where phone calls reporting symptoms of COVID-19 preceded the first notification of the disease. The lag in days between the first call and the first notification dates was also annotated.

 Ethical Considerations

Anonymized consolidated data were provided by the Bahia State Health Secretary for research purposes. The ethics committee of the School of Medicine of Bahia approved the project (approval number 4459774, on 12/13/2020). The requirement for obtaining informed consent was waived due to the characteristics of the research.

 Data Source and Preparation for Study

We extracted Telecoronavirus data for Salvador inhabitants (population: 2.88 million; surface area: 692.818 km²) for use in the prediction model. The call records were grouped in epidemiological weeks and residential areas informed by Telecoronavirus users. Additionally, they were grouped in sanitary districts using the list of neighborhoods comprising each of the 13 sanitary districts of the city of Salvador. Official numbers of COVID-19 confirmed cases were also obtained by sanitary districts.

 Use of Telehealth Data for Feeding Epidemiological Modeling

We further tested the hypothesis that telehealth systems can help feed numerical models developed to predict a virus spread in a specific region using a spatiotemporal model presented in [12,20-22] and explained as follows.

The Susceptible, Exposed, Infected, Recovered, Deceased Model for COVID-19

The COVID-19 dynamics may be modeled as compartmental models, in which the population under study is divided into compartments and has assumptions about the nature and time rate of transfer from one compartment to another [23]. These models have been used extensively in biological, ecological, and chemical applications [24-26]. They allow for an understanding of the processes at work and predict the dynamics of the epidemic. One of the simplest compartmental models is the “susceptible, infected, removed” model proposed in 1927 by Kermack and McKendrick [27], in which the population is divided into susceptible, infected, and recovered compartments. This basic “susceptible, infected, removed” model can be extended in several ways by enriching the number of compartments, as the “susceptible, exposed, infectious, recovered, deceased” model. Here, we work with a spatiotemporal “susceptible, exposed, infectious, recovered, deceased” model, presented in [20-22], given by,

Where s(x, t), e(x, t), i(x, t), r(x, t), and d(x, t) denote the densities of the susceptible, exposed, infected, recovered, and deceased populations, respectively. The sum of all the compartments, except for d(x, t), is represented by npop, which is the total living population. βi and βe denote the transmission rates between symptomatic and susceptible individuals and asymptomatic and susceptible individuals, respectively (units 1/days), a denotes the incubation period (units 1/days), γe corresponds to the asymptomatic recovery rate (units 1/days), γi the symptomatic recovery rate (units 1/days), δ represents the mortality rate (units 1/days), and νs, νe, νi, νr are the diffusion parameters of the different population groups as denoted by the subscripted letters: units km²/(persons.days).

To use this model, we need to define all parameters that govern the system of equations and the initial population of each compartment. Note that all these parameters can be considered time- and space-dependent. We need to infer several hypotheses about the initial conditions, especially those related to the exposed compartment, consisting of asymptomatic cases, which are more challenging to estimate.

 Model Construction for Salvador, Bahia

We define the beginning of the simulation as April 1, 2020, and simulate 180 days. The initial infected population is set according to the 7 days moving average data provided by the Brazilian Ministry of Health, following the procedure in [12] and [28]. The susceptible population is based on the estimation of the population of each Sanitary district, given by the Brazilian Institute for Geography and Statistics. Recovered and deceased populations start with zero assigned cases (ie, we consider that nobody died or recovered from COVID-19 at the beginning of the simulation).

The exposed compartment is the most nontrivial compartment to assign an initial condition. In previous works, the estimation was based on the amount of the infected population. It was considered that the exposed population could be about 10 times the number of the infected [29]. Here, we consider the 7-day moving average data from telehealth on April 1, 2020, as the initial exposed population. The telehealth data indicated rises in cases approximately 2 weeks before the notification system. Therefore, the calls may indicate a better estimation than the relation with the infected, especially at the epidemic’s beginning.

Each compartment’s population is divided by the area of each sanitary district and distributed in the 12 areas as people/km². Table 1 shows the differences between how the initial exposed population is considered in this work (related to the telehealth system) and how it would be represented when using the old approach (by multiplying the number of infected individuals by 10). Without the need for any calculation, it is possible to see that there is no correlation between the numbers of each approach.

The biological parameters of the simulation are defined based on the literature, as α = 1/7 day–1, γi = 1/24 day–1, γe = 1/6 day–1, δ = 1/160 day–1 [21]. On the other hand, the contact rate and the diffusion coefficient have to be estimated. They are based on the social distancing estimation, representing the homestay rate for Bahia.

Initial exposed population (people/km2) based on 10 times the number of infected and the telehealth calls.

Sanitary district Ten times the number of infected Telehealth
Barra Rio Vermelho 27.14 53.00
Brotas 8.57 34.57
Centro Histórico 2.86 13.42
Liberdade 4.29 17.70
Boca do Rio 5.71 23.42
Cabula/Beiru 8.57 43.57
Itapuã 10.00 36.14
Pau da Lima 4.29 25.85
Subúrbio Ferroviário 7.14 30.28
Itapagipe 2.86 18.42
São Caetano or Valéria 7.14 26.57
Cajazeiras 2.86 10.28
Results

The telehealth service received a total of 111,795 calls, 83,175 (74%) of which reporting at least one COVID-19 symptom (fever, cough, breathlessness, rhinorrhea, and gustatory or olfactory disorder) during the first 4 months of the COVID-19 epidemic in the state of Bahia (Figure 1). Olfactory or gustatory dysfunction (smell and taste change), considered highly specific for COVID-19 during the first wave of the pandemic [30-32], was reported in 47% of the calls, suggesting a high frequency of COVID-19 cases among users of the telehealth service.

Calls were registered by 320 (77%) out of 417 cities of Bahia State. The majority of these users of the telehealth service were female (48,873, 60%), with a median age of 38 (IQR 28-49) years. The demand for the service progressively increased between the first and ninth week, with a peak in the 23rd week. In the 4th month of operation, we observed a reduction in the number of daily calls to values close to those at the beginning of the service (Table 2).

During the first month of operation (April 2020), calls reporting COVID-19–like symptoms were registered in 205 cities, including areas from the north and east of the state. By this time, COVID-19 cases were notified in 136 cities (Figure 2). The first call to the telehealth service preceded the first COVID-19 notification in 181 (43%) out of 417 municipalities of Bahia. In these municipalities, the call occurred on average 30 (IQR 11-42) days before the notification. Additionally, for 68 (38%) of these 181 municipalities, the symptoms registered in the telehealth service included olfactory and gustatory disorders. These symptoms specific for COVID-19 were reported on average 14 (IQR 6-31) days before the first notified COVID-19 case (Multimedia Appendix 2).

Flowchart of calls for Telecoronavirus service according to local and presence of COVID-19–like symptoms.

Symptoms reported by users and the number of monthly calls.

Characteristic Salvador (n=61,797) Other cities (n=21,378) State (n=83,175)
Age (years), median (IQR) 39 (29, 50) 37 (27, 48) 38 (28, 49)
Month, n (%)
March 178 (0.3) 49 (0.2) 227 (0.3)
April 10,009 (16) 3464 (16) 13,473 (16)
May 23,227 (38) 4822 (23) 28,049 (34)
June 20,743 (34) 8634 (40) 29,377 (35)
July 7640 (12) 4409 (21) 12,049 (14)
Sex: female, n (%) 36,315 (60) 12,558 (60) 48,873 (60)
Fever, n (%) 30,501 (49) 10,468 (49) 40,969 (49)
Cough, n (%) 30,307 (49) 11,287 (53) 41,594 (50)
Shortness of breath, n (%) 15,518 (25) 6054 (28) 21,572 (26)
Smell or taste change, n (%) 28,741 (47) 9248 (44) 37,989 (47)

We also analyzed the spatial-temporal distribution of calls and notified cases in the districts of Salvador, the state capital. In all districts, an increase in the number of calls reporting COVID-19–like symptoms preceded an elevation in the number of notifications of COVID-19 (Figure 3).

Next, we evaluated if the data obtained with the telehealth service could be helpful in feeding a mathematical model to predict disease spread in Salvador. We simulated 180 days of the epidemic. To validate the results of our model with the available data, we compared the number of notified COVID-19 cases. First, we showed the values obtained for the whole city (Figure 4). Then, we integrated the values of the sanitary districts and plotted them in time. The simulation provides curves of accumulated infection similar to the data of notified cases.

Spatial distribution of calls related to COVID-19–like symptoms (orange) and notifications (green) by months of operation of the Telecoronavirus in Bahia state.

Seven-day rolling average of number of calls reporting COVID-19–like symptoms (blue line) and number of confirmed cases of COVID-19 (red line) in the period of operation of the telehealth service in Salvador, Bahia state, Brazil. (A) All cities. (B) Stratified by sanitary district.

Comparison between simulation and real data of cases at Salvador (total). Dashed blue lines: real data. Pink line: simulation with exposed population based on telehealth calls.

 Discussion Principal Findings

A critical challenge in predicting the spread of cases, especially in locals with a low test rate, is to timely estimate the number of suspected cases. Here, we demonstrated that curated data obtained from a telehealth service could be helpful in anticipating areas with new cases of COVID-19. A rise in calls referring to COVID-19–like illness preceded, in about two weeks, the increase in the notification of COVID-19 in the majority of Bahia state cities demonstrating the potential utility of this instrument for syndromic surveillance in the early phase of the epidemic curve. Additionally, it contains relevant information about the disease under investigation, such as the profile of the affected population, type, and frequency of symptoms, which can help characterize susceptible populations, which is important in the context of epidemics caused by new infectious agents. Calls captured in the telehealth service were also valuable in feeding a numerical disease spread model. Data obtained by a toll-free telehealth service can achieve the goals required for surveillance purposes, such as the following: (1) to be collected early after disease onset, (2) to represent the majority of the population in the covered area, and (3) to be rapidly accessed by the decision-making officials [5,33].

First, in an outbreak caused by a pathogen of airborne transmission, such as COVID-19, the fear of contamination reduces seeking face-to-face medical assistance. The majority of individuals with mild symptoms avoid being exposed in clinical care facilities, reducing the number of individuals who get tested, and therefore confirmed infections may be underestimated. From the patient perspective, telehealth assistance becomes more attractive. Accordingly, a substantial increase in telemedicine service was detected in the first weeks of the pandemic [5,34]. Consequently, data from telehealth services become more suitable for surveillance purposes than laboratory-confirmed infection or electronic medical records obtained from health care units.

Second, telehealth services can cover a population from a wide geographic area. The service offered via telephone dismisses reading or writing skills, digital literacy, or internet access and can be used by all age groups and social strata. If the service is toll-free, most of the population may be covered, which is especially important for low- to middle-income countries, as observed in this case.

Third, data sets containing time-sensitive information collected from telehealth services can be accessed simultaneously by public health authorities for syndromic surveillance. Automating data extraction and real-time interpretation of information in the context of relevant public health emergencies is still challenging. Structured surveys designed for telescreening assistance, as described here, can serve this purpose and offer daily reports for the stakeholders. Although several aspects, such as technical or operational, ethical, and legal points, should be addressed to use these data for disease surveillance, analysis of real-time data obtained from telehealth services is feasible. It has already been performed for contact tracing, screening, and monitoring clinical conditions in emergency response to epidemics [35] and can support decision-making on public health policy.

Our result suggests that the data set from a telehealth service may also be helpful as input to a mathematical model to predict COVID-19 spread. The initial conditions of the exposed and infected populations are essential to short-term predictions. The estimate of exposure may be challenging since this population has no symptoms. Therefore, the hypothesis of assuming the number of exposed as the number of calls might be better than an estimation based on the number of infected, since the telehealth data indicated rises in cases approximately 2 weeks before the notification system.

The simulation for a city of about 2.9 million inhabitants showed that most sanitary districts had a good agreement between the actual data and the simulation. The simulation overestimated the cases in 1 district area and underestimated them in 2 areas. Other districts had an excellent agreement until day 100 and lost accuracy afterward. A loss of accuracy in a fast-changing infectious disease is expected since it is challenging to reproduce long-term predictions. The simulation may still be improved due to several aspects needing to be considered. For example, we used estimates for parameters such as transmission rate and diffusion, and it would be possible to use a machine learning mechanism to find the best ones for this case. Moreover, we define that all sanitary districts would receive the same input parameters, and we know that different regions might have performed differently regarding the population’s behavior and restriction policies. Besides, we used only spatial spread mechanisms due to diffusion and further use of different approaches, as convection and source terms could have offered better estimates. However, the central idea of this report is to show that telehealth data impact a robust algorithm that worked well to simulate the COVID-19 behavior of different regions (Italy, Brazil, and the United States [12]).

Collecting real-time data may help predictors forecast new surges and prepare the population, authorities, and health systems. Moreover, we may infer that our hypothesis works since our simulation with the telehealth data have provided promising results.

 Limitations

Our study has some limitations. We used COVID-19–like symptoms as a proxy for confirmed infection. Although this is not accurate, the number of phone calls and the trend in service assessment may suggest the direction and amount of disease spread early on. It shall also be stressed that confirmatory tests may not be available in the early stages of a new epidemic infection. In such cases, the use of the syndromic approach, as employed here, is a critical element for predicting infection spreading. Another significant limitation is that the telehealth service was not equally publicized in all cities, which may have implications for interpreting surveillance data in some areas.

 Conclusion

In conclusion, data from telehealth services help model COVID-19 spread and may be helpful in other health situations. Telehealth data and digital health technologies for monitoring disease spread may be especially useful considering the resurgence of new SARS-CoV-2 variants [36]. Data collected from primary health care systems can be used for monitoring the dynamics of COVID-19 cases and the geographic localization of cases. Considering the continuous expansion of telehealth and telemedicine tools in the health care system, the availability of such data may prove a critical tool for modern epidemiology.

Brazil’s map showing the state of Bahia and the capital Salvador.

Date of first confirmed case, first call, and first call reporting smell or taste disorder for each municipality of Bahia state, Brazil.

The authors thank all TeleCoronaVirus volunteers, the Government of the state of Bahia, and all participating institutions.

This study is part of the Alert-Early System of Outbreaks with Pandemic Potential (AESOP) program funded by Fundação Oswaldo Cruz and the Rockefeller Foundation. During the period of analyses, TC-S was a PhD student at the Post-Graduation Program in Health Sciences, Universidade Federal da Bahia, which is supported by the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior-Brasil, finance code 001. MBN, VSB, and AC are research fellows from the National Council for Scientific and Technological Development (CNPq).

 Data Availability

Our agreement with the Bahia State Health Secretary for accessing the database denies authorization of access to a third party. Request for accessing the database should be addressed to the state government, Secretaria de Saúde do Estado da Bahia.

None declared.

 Rogers DG Santamaria K Seng EK Grinberg AS Behavioral Health, Telemedicine, and Opportunities for Improving Access Curr Pain Headache Rep 2022 12 26 12 919 926 10.1007/s11916-022-01096-w 36418847 10.1007/s11916-022-01096-w PMC9684808 Karamagi HC Muneene D Droti B Jepchumba V Okeibunor JC Nabyonga J Asamani JA Traore M Kipruto H eHealth or e-Chaos: The use of Digital Health Interventions for Health Systems Strengthening in sub-Saharan Africa over the last 10 years: A scoping review J Glob Health 2022 12 03 12 04090 10.7189/jogh.12.04090 36462201 PMC9718445 Mbunge E Batani J Gaobotse G Muchemwa B Virtual healthcare services and digital health technologies deployed during coronavirus disease 2019 (COVID-19) pandemic in South Africa: a systematic review Glob Health J 2022 06 6 2 102 113 10.1016/j.glohj.2022.03.001 35282399 S2414-6447(22)00023-9 PMC8897959 Khoshrounejad F Hamednia M Mehrjerd A Pichaghsaz S Jamalirad H Sargolzaei M Hoseini B Aalaei S Telehealth-Based Services During the COVID-19 Pandemic: A Systematic Review of Features and Challenges Front Public Health 2021 9 711762 10.3389/fpubh.2021.711762 34350154 PMC8326459 Koonin LM Hoots B Tsang CA Leroy Z Farris K Jolly T Antall P McCabe B Zelis CBR Tong I Harris AM Trends in the Use of Telehealth During the Emergence of the COVID-19 Pandemic - United States, January-March 2020 MMWR Morb Mortal Wkly Rep 2020 10 30 69 43 1595 1599 10.15585/mmwr.mm6943a3 33119561 PMC7641006 Katayama Y Kiyohara K Komukai S Kitamura T Ishida K Hirose T Matsuyama T Kiguchi T Hirayama A Shimazu T The relationship between seasonal influenza and telephone triage for fever: A population-based study in Osaka, Japan PLoS One 2020 15 8 e0236560 10.1371/journal.pone.0236560 32760164 PONE-D-19-28279 PMC7410252 Hughes SL Morbey RA Elliot AJ McEwen SA Greer AL Young I Papadopoulos A Monitoring telehealth vomiting calls as a potential public health early warning system for seasonal norovirus activity in Ontario, Canada Epidemiol Infect 2019 01 147 e112 10.1017/S0950268818003357 30869009 S0950268818003357 PMC6518795 Pivette M Mueller JE Crépey P Bar-Hen A Drug sales data analysis for outbreak detection of infectious diseases: a systematic literature review BMC Infect Dis 2014 11 18 14 604 10.1186/s12879-014-0604-2 25403237 s12879-014-0604-2 PMC4240820 Rodrigues DLG Belber GS Padilha FVDQ Spinel LF Moreira FR Maeyama MA Pinho APNM Júnior ÁA Impact of Teleconsultation on Patients With Type 2 Diabetes in the Brazilian Public Health System: Protocol for a Randomized Controlled Trial (TELEconsulta Diabetes Trial) JMIR Res Protoc 2021 01 21 10 1 e23679 10.2196/23679 33475516 v10i1e23679 PMC7861994 Der-Martirosian C Wyte-Lake T Balut M Chu K Heyworth L Leung L Ziaeian B Tubbesing S Mullur R Dobalian A Implementation of Telehealth Services at the US Department of Veterans Affairs During the COVID-19 Pandemic: Mixed Methods Study JMIR Form Res 2021 09 23 5 9 e29429 10.2196/29429 34477554 v5i9e29429 PMC8462492 King K Ford D Haschker M Harvey J Kruis R McElligott J Clinical and Technical Considerations of an Open Access Telehealth Network in South Carolina: Definition and Deployment J Med Internet Res 2020 05 22 22 5 e17348 10.2196/17348 32441657 v22i5e17348 PMC7275254 Grave M Viguerie A Barros GF Reali A Coutinho ALGA Assessing the Spatio-temporal Spread of COVID-19 via Compartmental Models with Diffusion in Italy, USA, and Brazil Arch Comput Methods Eng 2021 28 6 4205 4223 10.1007/s11831-021-09627-1 34335018 9627 PMC8315263 Kucharski AJ Klepac P Conlan AJK Kissler SM Tang ML Fry H Gog JR Edmunds WJ CMMID COVID-19 working group Effectiveness of isolation, testing, contact tracing, and physical distancing on reducing transmission of SARS-CoV-2 in different settings: a mathematical modelling study Lancet Infect Dis 2020 10 20 10 1151 1160 10.1016/S1473-3099(20)30457-6 32559451 S1473-3099(20)30457-6 PMC7511527 Ilyin SO A Recursive Model of the Spread of COVID-19: Modelling Study JMIR Public Health Surveill 2021 04 19 7 4 e21468 10.2196/21468 33871381 v7i4e21468 PMC8057198 McMahon A Robb NC Reinfection with SARS-CoV-2: Discrete SIR (Susceptible, Infected, Recovered) Modeling Using Empirical Infection Data JMIR Public Health Surveill 2020 11 16 6 4 e21168 10.2196/21168 33052872 v6i4e21168 PMC7674142 Deng Q Dynamics and Development of the COVID-19 Epidemic in the United States: A Compartmental Model Enhanced With Deep Learning Techniques J Med Internet Res 2020 08 21 22 8 e21173 10.2196/21173 32763892 v22i8e21173 PMC7451112 Utamura M Koizumi M Kirikami S An Epidemiological Model Considering Isolation to Predict COVID-19 Trends in Tokyo, Japan: Numerical Analysis JMIR Public Health Surveill 2020 12 16 6 4 e23624 10.2196/23624 33259325 v6i4e23624 PMC7746226 Shapiro MB Karim F Muscioni G Augustine AS Adaptive Susceptible-Infectious-Removed Model for Continuous Estimation of the COVID-19 Infection Rate and Reproduction Number in the United States: Modeling Study J Med Internet Res 2021 04 07 23 4 e24389 10.2196/24389 33755577 v23i4e24389 PMC8030656 Cerqueira-Silva T Carreiro R Nunes V Passos L Canedo BF Andrade S Ramos PIP Khouri R Santos CBS Nascimento JDS Paste AA Paiva Filho IDM Santini-Oliveira M Cruz Barral-Netto M Boaventura V Bridging Learning in Medicine and Citizenship During the COVID-19 Pandemic: A Telehealth-Based Case Study JMIR Public Health Surveill 2021 03 04 7 3 e24795 10.2196/24795 33630746 v7i3e24795 PMC7935247 Viguerie A Veneziani A Lorenzo G Baroli D Aretz-Nellesen N Patton A Yankeelov TE Reali A Hughes TJR Auricchio F Diffusion-reaction compartmental models formulated in a continuum mechanics framework: application to COVID-19, mathematical analysis, and numerical study Comput Mech 2020 66 5 1131 1152 10.1007/s00466-020-01888-0 32836602 1888 PMC7426072 Viguerie A Lorenzo G Auricchio F Baroli D Hughes TJR Patton A Reali A Yankeelov TE Veneziani A Simulating the spread of COVID-19 a spatially-resolved susceptible-exposed-infected-recovered-deceased (SEIRD) model with heterogeneous diffusion Appl Math Lett 2021 01 111 106617 10.1016/j.aml.2020.106617 32834475 S0893-9659(20)30301-3 PMC7361091 Grave M Coutinho ALGA Adaptive mesh refinement and coarsening for diffusion-reaction epidemiological models Comput Mech 2021 67 4 1177 1199 10.1007/s00466-021-01986-7 33649692 1986 PMC7905202 Brauer F Castillo-Chavez C Feng Z Mathematical models in epidemiology 2019 New York, US Springer Brauer F Castillo-Chavez C Castillo-Chavez C Mathematical models in population biology and epidemiology 2012 New York, US Springer Keller JP Gerardo-Giorda L Veneziani A Numerical simulation of a susceptible-exposed-infectious space-continuous model for the spread of rabies in raccoons across a realistic landscape J Biol Dyn 2013 7 Suppl 1 Suppl 1 31 46 10.1080/17513758.2012.742578 23157180 PMC3957468 Jourdan N Neveux T Potier O Kanniche M Wicks J Nopens I Rehman U Le Moullec Y Compartmental Modelling in chemical engineering: A critical review Chemical Engineering Science 2019 12 210 115196 10.1016/j.ces.2019.115196 Kermack WO McKendrick AG A contribution to the mathematical theory of epidemics Proc. R. Soc. Lond. A 1997 01 115 772 700 721 10.1098/rspa.1927.0118 Grave M Viguerie A Barros GF Reali A Andrade RFS Coutinho ALGA Modeling nonlocal behavior in epidemics via a reaction-diffusion system incorporating population movement along a network Comput Methods Appl Mech Eng 2022 11 01 401 115541 10.1016/j.cma.2022.115541 36124053 S0045-7825(22)00534-5 PMC9475403 Reis RF de Melo Quintela B de Oliveira Campos J Gomes JM Rocha BM Lobosco M Weber Dos Santos R Characterization of the COVID-19 pandemic and the impact of uncertainties, mitigation strategies, and underreporting of cases in South Korea, Italy, and Brazil Chaos Solitons Fractals 2020 07 136 109888 10.1016/j.chaos.2020.109888 32412556 S0960-0779(20)30288-5 PMC7221372 Xydakis MS Dehgani-Mobaraki P Holbrook EH Geisthoff UW Bauer C Hautefort C Herman P Manley GT Lyon DM Hopkins C Smell and taste dysfunction in patients with COVID-19 Lancet Infect Dis 2020 09 20 9 1015 1016 10.1016/S1473-3099(20)30293-0 32304629 S1473-3099(20)30293-0 PMC7159875 Menni C Valdes AM Freidin MB Sudre CH Nguyen LH Drew DA Ganesh S Varsavsky T Cardoso MJ El-Sayed Moustafa JS Visconti A Hysi P Bowyer RCE Mangino M Falchi M Wolf J Ourselin S Chan AT Steves CJ Spector TD Real-time tracking of self-reported symptoms to predict potential COVID-19 Nat Med 2020 07 26 7 1037 1040 10.1038/s41591-020-0916-2 32393804 10.1038/s41591-020-0916-2 PMC7751267 Altwairqi RG Aljuaid SM Alqahtani AS Effect of tonsillectomy on humeral and cellular immunity: a systematic review of published studies from 2009 to 2019 Eur Arch Otorhinolaryngol 2020 01 277 1 1 7 10.1007/s00405-019-05672-6 31664514 10.1007/s00405-019-05672-6 Gunasekeran DV Tseng RMWW Tham Y Wong TY Applications of digital health for public health responses to COVID-19: a systematic scoping review of artificial intelligence, telehealth and related technologies NPJ Digit Med 2021 02 26 4 1 40 10.1038/s41746-021-00412-9 33637833 10.1038/s41746-021-00412-9 PMC7910557 Roy B Nowak RJ Roda R Khokhar B Patwa HS Lloyd T Rutkove SB Teleneurology during the COVID-19 pandemic: A step forward in modernizing medical care J Neurol Sci 2020 07 15 414 116930 10.1016/j.jns.2020.116930 32460041 S0022-510X(20)30267-7 PMC7241381 Gao Y Liu R Zhou Q Wang X Huang L Shi Q Wang Z Lu S Li W Ma Y Luo X Fukuoka T Ahn H Lee M Luo Z Liu E Chen Y Shu C Tian D COVID-19 EvidenceRecommendations Working Group Application of telemedicine during the coronavirus disease epidemics: a rapid review and meta-analysis Ann Transl Med 2020 05 8 10 626 10.21037/atm-20-3315 32566563 atm-08-10-626 PMC7290625 Newcomb K Bilal S Michael E Combining predictive models with future change scenarios can produce credible forecasts of COVID-19 futures PLoS One 2022 17 11 e0277521 10.1371/journal.pone.0277521 36378674 PONE-D-22-16941 PMC9665358