SARS-CoV-2, a member of the coronavirus family, is the virus that causes COVID-19, an infectious respiratory disease. Following a flu-like early phase, a very severe respiratory condition linked to the development of bilateral interstitial pneumonia may emerge later in the disease. Breathlessness, dyspnea, difficulty breathing, and elevated heart rate are some of its symptoms. In actuality, until the need for hospitalization becomes urgent, COVID-19 pneumonia causes a drop in blood oxygen saturation without the patient's awareness. Because of this, it is essential to keep an eye on people who are being monitored for COVID-19 infection at home to make sure their saturation level does not drop below the set threshold especially if they have not had any prior respiratory illnesses. Medical personnel can determine whether to hospitalize a patient who is asymptomatic, symptomatic, or pre-symptomatic and receiving home isolation by trending the saturation data. This allows for an early hospitalization before the patient's clinical picture deteriorates. Thus, the creation of an intelligent telemonitoring system that makes use of conventional diagnostic tools like an oximeter and thermometer will enable to satisfy the following constraints: (i) those who are not in immediate danger should be left at home, saving hospital beds; (ii) watch those who could experience a respiratory crisis; (iii) keep an eye on the portion of the population (such as those under precautionary quarantine) that has not been tested for COVID-19 but may be asymptomatic or pre-symptomatic; (iv) assess critical factors to prevent a respiratory crisis in individuals with mild COVID-19; (v) enable general practitioners to monitor the patient round-the-clock, reducing the possibility of infections from prolonged and/or direct contact.

In order to achieve these goals, we designed and developed reCOVeryaID (D'Auria et al., 2023b), an intelligent telemonitoring application for symptomatic, asymptomatic, and pre-symptomatic coronavirus patients; in this paper, starting from the theoretical model shown in D'Auria et al. (2023b), we focus on highlighting its technologies and main functionalities, pointing out that a wide application of this framework in Italian hospitals could be really essential for both the health care system and the population. Specifically, reCOVeryaID is a web application aimed at remotely monitoring patients potentially, currently or no longer infected with COVID-19; it consists of a prototype implementation allowing patients and medical doctors to effectively interact with each other, to cope with the issues listed above. The necessary equipment consists of a saturimeter for measuring oxygen saturation (SpO₂) and heart rate, a thermometer for measuring body temperature the reCOVeryaID web-app, the patient's smartphone (or laptop), and the doctor's smartphone (or laptop). The carried out measurements are the SpO₂ value, the body temperature value, and the heart rate value. Eventually, the framework returns both short-term and long-term alerts, which need to be further investigated by medical doctors. In this way, patients, especially the frail ones, can be constantly monitored by the app, which provides prompt alerts to the physician if abnormal trends in vital parameters are identified in both the short and long term.

The remainder of the paper is organized as follows. In the next section we show the related work in the eHealth context, with particular focus on monitoring vital signs. Then, Section 3 illustrates in detail the used technologies, both from the client side and from the server side, and the overall application behavior. Additionally, Section 4 shows the framework functionalities, and Section 5 exposes the experimental evaluation, which concerns user satisfaction. Eventually, Section 6 concludes the paper and outlines future work.

Based on the current pandemic scenario and events that have transpired since the epidemic's onset, it is evident that telemedicine has not really taken off in Italy or the rest of the world. Aside from the clinical benefit, it might have saved a significant amount of money for a number of health systems, and it might have been very helpful in making primary care physicians' monitoring of COVID-19 patients and not just patients more dependable and manageable right from the start of the pandemic (Charles, 2000). On the other hand, telemedicine has tremendous potential, as this is highlighted by the following papers that precede or coincide with the first part of the pandemic (Ali and Khoja, 2020; Chauhan et al., 2020; Hollander and Carr, 2020; Monaghesh and Hajizadeh, 2020). This is demonstrated by the fragmented trials that are currently taking place in various parts of the world, as well as by the rise of startups, tools, and practical technologies for remote monitoring; additionally, health insurance companies are showing interest in this, as they are now including teleconsultations and far more advanced solutions in their fee-for-service packages; several national health systems could gain from wearable technology to virtual triage, from remote examinations (at least those that are practical) to vital sign monitoring.

The various Rheumatology Societies have also recently raised the alarm, or rather pushed for the use of telemedicine systems, citing the need for constant monitoring and prompt diagnosis and treatment for patients with lupus, rheumatoid arthritis, vasculitis, and other similar chronic and frequently autoimmune diseases; consequently, the papers (Danhieux et al., 2020; Dimitroulas and Bertsias., 2020; Mason et al., 2020; Wright et al., 2020; Coupet et al., 2021; Hacker et al., 2021) point out that the use of telemedicine is extremely useful not only to monitor the COVID-19 itself, but also to control the evolution of other types of pathologies, such as chronic diseases. They businesses actually plan to make significant investments in telemedicine, and several of them have already created online resources specifically for patients with rheumatology, but they are limited in scope. All diseases should be treated using a comparable model that is structured and applied globally.

We realize that the road to digitizing public health is still long, given that, for example, in Italy, where we had one of the largest pandemic spikes in the world at the beginning of the first pandemic wave, even electronic medical records, a relatively simple tool for coordinating all the services provided to citizens, never took off. Even with the mandated acceleration of the pandemic, only dematerialized prescriptions have become widely used thus far, limiting telemedicine mostly to general practitioners' (often busy) phones.

Various attempts have been made to mitigate the deficiency of sufficient telemedicine networks by employing artificial intelligence (AI) to forecast illness or diagnose patients (Li et al., 2020). Specifically, the goal of using AI combined with telemedicine is to provide the added value regarding future prediction in the context of numerous diseases. In fact, it has been suggested that artificial intelligence be used to evaluate chest CT scans (Han et al., 2020) and diagnose COVID-19 early. It has also been suggested to identify predictive markers for cytokine storm (Caricchio et al., 2021) or evaluate particular laboratory parameters (like lymphocyte count or hsCRP) to forecast the patient's worsening clinical condition. The primary issue with this study, aside from the challenge of validating new criteria, is that patients must undergo intrusive “screening” procedures, which include blood samples and CT scans, requiring hospitals and laboratories, and therefore increasing the expense of the diagnosis process. In order to avoid intervening until it is too late and the patient has already required hospitalization, other authors have proposed algorithms to predict the risk of death in hospitalized patients (Gao et al., 2020) and the necessity for invasive ventilation (Burdick et al., 2020).

In addition to the aforementioned information and in line with earlier research from multiple studies, it is evident that using an oximeter to monitor saturation level to follow COVID-19 patients is essential, as highlighted by the following papers (Taguchi et al., 1994; Solé et al., 2009; Scott and McDougall, 2017; Elliott and Baird, 2019; Takei et al., 2020); specifically, they need to be closely watched to ensure that their saturation level does not drop below a certain point, particularly if they have not previously experienced a respiratory illness. Medical personnel can determine whether to admit a patient who is asymptomatic, symptomatic, or pre-symptomatic and is receiving home isolation by analyzing trending saturation data. This allows them to make an early admission before the patient's clinical condition deteriorates.

As a result of the discussion conducted so far, we point out that all of the listed issues can be resolved by an integrated home care/telemedicine system exploiting many devices, including a saturation device and a thermometer; specifically, an algorithm examines and filters the data, presenting any alarms to the referring medical doctor in real time. Alternatively, the patient could supply the data required for analysis on their own. The benefits will include avoiding invasive testing like CT scans and blood draws, as well as a decrease in hospital stays for less severe symptoms that do not need for extensive care. The following is a list of other relevant and recent telemonitoring systems that use vital parameter monitoring: (Wiffen et al., 2023) facilitates the improvement of data collection and processing in order to create a reliable app that can be used on a daily basis in clinical settings, while (Ko et al., 2023) performed in-person home visits once a day, with nursing visits up to three times a day for intravenous therapy; patients were also released from the program when they satisfied standard inpatient discharge criteria. Murali (2023) provide a more efficient method of monitoring the patient's healthcare system. Their primary areas of interest are cloud services, IoT, and machine learning (ML) in the context of patient monitoring. Furthermore, their technology may be utilized as an advanced Internet of Things-enabled real-time patient monitoring system to monitor the patient's vital indicators, including blood pressure, heart rate, oxygen saturation, and body temperature. Subsequently, Totuk et al. (2023) examined Value Stream Mapping (VSM) measures using smartphones, pulse oximetry probes, and blood group antigens (BGA). The VSM, smartphone, and BGA all shown very good agreement in the Integrated Comprehensive Care (ICC) of the oxygen saturation of arterial blood (SaO₂) readings. Similar to this, there was excellent agreement between the VSM and smartphone in terms of the ICC value of the Heart Rate (HR) values. Moreover, according to Smith et al. (2021), the success of wearable technologies in the future depends on developing clinical confidence in the accuracy of the data measured and the appropriate interpretation of that data in relation to the person, the environment, and the activity carried out. In fact, wearable physiological monitoring may soon enhance point-of-care diagnostic accuracy and provide crucial information for medical decisions.

In this context, the proposed framework reCOVeryaID leverages an innovative combination of (wearable) devices to generate short-term and long-term alerts related to COVID-19; additionally, the reCOVeryaID prototype's flexibility and generality allow it to be combined with other systems that address various diseases, which is another strength of the system; indeed, it might be combined with complex event processing and intelligent agent-based telemonitoring systems (Persia et al., 2021; De Lauretis et al., 2023), Italian clinical notes (D'Auria et al., 2023a), and a multi-agent system for epilepsy prediction and detection in neuropediatrics (Bertoncelli et al., 2023).

In order to improve browsing of related works, Table 1 summarizes them, listing first authors, year, and main topic.

Figure 1 shows the architecture of the proposed framework. Specifically, the technologies used to implement the reCOVeryaID cross-platform application can be classified into client-side technologies and server-side technologies.

The functionalities provided by the multi-platform reCOVeryaID application are: system registration and user account creation (Section 4.1), user account activation (Section 4.2), password retrieval (Section 4.3), login (Section 4.4), log off (Section 4.5), homepage (Section 4.6), user profile (Section 4.7), new measurement (Section 4.8), measurement history (Section 4.9), long-term alert history (Section 4.10), settings (Section 4.11), and information (Section 4.12).

In order to validate the developed framework, we evaluated the impact of our system on different categories of users (i.e., patients and medical doctors) engaged in performing the tasks mentioned in Section 4. To evaluate the results, we used the TLX (NASA Task Load Index factor)³; more specifically, we asked users to express their opinions about the capability of our system to provide an effective user experience in completing all the tasks described in Section 4, based on the TLX evaluation protocol. In fact, TLX is a multidimensional rating procedure that provides a score from 1 to 5 on six subscales: mental demand, physical demand, temporal demand, low performance, effort, and frustration. Lower TLX scores are better.

Figure 5 shows the scores achieved by each of the three categories of the 20 patient involved (expert, medium expert, and nonexpert in medical apps); Figure 6 does the same for the three medical doctors involved.

In summary, nonexpert users evaluate our system better, as they are likely to appreciate the feedback provided by the app more; on the contrary, expert users might already know what to expect. The best results are obtained with the subscales mental demand and effort for patients, and with mental demand, physical demand and effort for medical doctors.

In this paper, we have shown the technologies and the main functionalities used for developing reCOVeryaID, an intelligent telemonitoring application for symptomatic, asymptomatic and pre-symptomatic COVID-19 patients. The main goal is to make it operational in Italian hospitals very soon, after final validation.

In order to make the developed prototype even more useful in an actual situation, further effort will be focused on refining its current iteration. In particular, we hope to improve through ongoing communication with the medical personnel engaged in the COVID-19 emergency as well as General Practitioners.

Furthermore, because of its ease of use and adaptability, reCOVeryaID can also be seamlessly applied to other domains, not limited to the COVID-19 emergency. In fact, it can be utilized to track patients with conditions like diabetes or hypertension, which broadens its scope and reinforces physicians' acceptance of it. It is obvious that adding more vital signs to the system like blood pressure and combining the various metrics in a novel way to identify potentially dangerous situations early on or even prevent them involves further utilizing artificial intelligence techniques like logical inference and knowledge representation.

The original contributions presented in the study are included in the article/Supplementary material, further inquiries can be directed to the corresponding author.

DD'A: Writing – original draft, Writing – review & editing. FB: Writing – original draft, Writing – review & editing. ST: Writing – original draft, Writing – review & editing. DC: Writing – original draft, Writing – review & editing. AM: Writing – original draft, Writing – review & editing.