The objectives of this study will be achieved by recruiting 3000 participants from the outpatient clinic of ASST-Fatebenefratelli-Sacco, Buzzi Children’s Hospital, and IRCCS MultiMedica in Milan, Italy. All participants who meet the inclusion criteria and provide consent to participate will be enrolled consecutively. Recruitment will take place from June 2024 to September 2025.

Participants eligible for inclusion in the study must meet the criteria in Textbox 1.

In addition to the above criteria, 70% of the target population should include adult participants meeting at least one of the conditions in Textbox 2. Family history of diabetes or hypertension will also be investigated.

The procedures and evaluations in Textbox 3 will be performed during screening to enroll eligible participants.

Characterizing aspects will be noted, such as scars, piercings, beards, glasses, or the use of facial products, sunscreen cream, or nail polish. Furthermore, current pregnancy will be reported.

The flowchart of the process is schematized in Figure 1.

Comestai is a mobile app that leverages a combination of signal processing and artificial intelligence (AI) technologies, along with a proprietary mathematical backend server, to analyze the skin of the upper cheek region through a smartphone’s front camera (iOS or Android).

The Comestai app captures video images using signal processing AI technologies and analyzes the data using photoplethysmography techniques (Figure 2) [9,14-16]. This approach provides real-time measurements of physiological parameters, including SpO₂, HR, PR, and RR, and biochemical parameters Hb, HbA_1c, and total cholesterol. Unlike other apps, the app uses a novel technology that extracts ambient light RGB signals from face videos in real time, eliminating the need for cloud servers or an internet connection.

For vital signs data collection using the Comestai app, the study staff will use the following mobile phones for remote photoplethysmography “by face” measurements:

In addition to the equipment for vital signs data collection using the Comestai app, reference devices will be used to measure the participant’s vital signs for comparative purposes.

Data from the investigational device will be collected by trained staff members. Vital signs data obtained using the app and reference data will be stored on a secured cloud. The app will be activated only when the cell phone is offline, ensuring that no information is stored in the Comestai cloud. Consequently, the company will not have access to any data collected during the tests.

In addition, the app will be deleted after use to prevent any data from being uploaded into the system. Data collected using the reference devices will be stored without any identifiable participant information and will only be identifiable using the participant identification number provided during screening.

The raw data of participants’ vital signs will be locked, encrypted, and accessible only to the study center team. Raw data can be processed only after unlocking the files to calculate the final measured results.

The coordinator center (Buzzi Children’s Hospital) will be responsible for securely storing the data in a locked and secured location. The Endocrinology Unit of ASST-Fatebenefratelli-Sacco and Buzzi Children’s Hospital will handle data entry and ensure data quality.

This study will be conducted in compliance with the Good Clinical Practice guidelines, including the archiving of essential documents.

No medical decisions will be made based on the data obtained from the Comestai app, reference devices, or blood sample results.

Clinical data for the study will be meticulously organized, easily accessible, and thoroughly cleaned both throughout and after the study concludes (post data lock). Activities in clinical data management encompass data collection and entry, tracking electronic case report form (eCRF) usage, database design, data validation, and database locking.

As previously outlined, the eCRF will serve as the primary tool for capturing clinical data. The data fields within the eCRF are clearly defined and maintain consistency throughout. The eCRF is designed to be succinct, intuitive, and user-friendly.

Database edit checks will be conducted to identify discrepancies in the data, which could stem from inconsistent entries, missing data points, range discrepancies, or deviations from the study protocol. Any incomplete discrepancies will be addressed by site investigators. Continuous quality control measures for data processing will be implemented at regular intervals throughout the study duration.

Temporal interruption of vital signs data collection using the app may occur in the following cases:

Numerical variables will be presented as minimum and maximum values, means, SDs, medians, and quartiles. Categorical variables will be presented as absolute and relative frequencies.

The accuracy of the app measurements compared to the reference devices and standard tests will be assessed for all parameters. CIs will be calculated using the bootstrap method. Participants with abnormal adverse events (defined as 3 SDs or more above the mean) and those with invalid reference device values will be excluded from the analysis.

In order to provide a more comprehensive view of the model’s performance, the proposed approach’s effectiveness was evaluated using various quality criteria, including the mean error, SD, mean absolute error, root mean square error, and mean absolute percentage error. The correlation between measurements obtained using the app and those from reference devices and standard tests will be evaluated using the Pearson correlation coefficient. The agreement between the 2 sets of measurements (app vs reference devices/standard tests) will be represented using Bland-Altman plots. Sensitivity, specificity, positive predictive value, negative predictive value, accuracy, and likelihood ratios will be calculated to determine the ability of the new app to accurately measure vital signs.

Considering the average values of pressure/saturation/HR among a general population with a sample of 3000 participants, there will be a power exceeding 95% for all evaluated parameters in a 2-sided test with a P value of .05 if the discrepancy between the 2 measurement methods is as reported in Table 2.

The study was approved by the ethics committee (Comitato Etico Territoriale Lombardia 1, Milano, Italy, approval code CET 98-2024). The study will be conducted according to the Declaration of Helsinki guidelines, international guidelines, and current clinical trial laws.

All participants will give their written consent after being fully informed about the study. For individuals younger than 18 years, both legal guardians’ written consent and assent by adolescents themselves will be collected.

Data will be collected and stored securely to maintain confidentiality, with only authorized personnel having access to identifiable information.

The study protocol was registered in ClinicalTrials.gov (NCT06427564). If modifications in the original protocol are required, additional amendments will be demanded by the Ethics Committee. In this scenario, the approved related information will be updated on ClinicalTrials.gov.

The authors followed the SPIRIT (Standard Protocol Items: Recommendations for Interventional Trials) guidelines for study protocol reporting (Multimedia Appendix 1) to ensure greater scientific rigor. Written informed consent will be obtained from all participants, and data collection will be performed by a trained research team.

Consent will be requested for the use of the blood sample results (complete blood count, HbA_1c, lipid profile, glucose, and creatinine). Consenting to view the biochemical data is optional. A participant who chooses not to provide their biochemical results will still be eligible to participate and complete the recordings.

The project aims to report whether this user-friendly mobile app supports a high level of accuracy in monitoring SpO₂, RR, BP, and PR, demonstrating the same reliability in detecting Hb, HbA_1c, and total cholesterol. Enhancing algorithm performance and achieving a high level of accuracy will be crucial in assessing whether the app can be adopted to improve self-monitoring of health status. This, in turn, could enhance the screening for early abnormalities in vital signs and biochemical parameters.

Today’s health care industry embraces technological innovations to provide better personalized health care services to the public [17]. Mobile-based solutions and apps are one such revolutionary and trending innovation that facilitates enhanced patient care management and efficient diagnosis. They are increasingly reliant on wireless technologies as they evolve [17]. Applications of mHealth typically include the use of mobile devices in the collection of health and clinical data, the provision of health information to practitioners, researchers, and patients, the real-time monitoring of patients; vital signs, and the direct delivery of care (via mobile telemedicine).

In an mHealth context, the advancement of noncontact methods to monitor human vital signs holds substantial promise for enhancing patient care across various environments [6].

A recently proposed RGB camera–based method for noncontact vital sign measurement using photoplethysmography has been identified as an interesting tool for recording vital signs. Photoplethysmography is also used by the Comestai app to assess various vital functions, including SpO₂, RR, BP, and PR [10]. Initial research into photoplethysmography shows promising outcomes; however, scientific evidence regarding the accuracy and benefits of digital health apps that measure vital signs remains limited and somewhat controversial. Many photoplethysmography studies are still susceptible to errors [18-20], with the majority involving only small sample sizes.

In our protocol, we will recruit a large population (3000 participants) to ensure accuracy across various demographics and health conditions, including also pathological conditions. The use of a large population will also help reduce errors in the results, improving the reliability of the tool. Thorough analysis and classification of these signals are essential for the development of affordable and user-friendly PPG-based tools across a wide range of applications [10].

While there is an increasing number of articles and research on SpO₂, RR, BP, and PR monitoring using PPG, a limited number of available publications on advancements related to biochemical parameter estimation are reported [10]. Comestai offers the possibility to monitor Hb, HbA_1c, and cholesterol concurrently with other parameters, thereby broadening its application possibilities in health care services.

Although the protocol included participants with diseases such as hypertension, prediabetes and diabetes, dyslipidemia, and atrial fibrillation to assess the tool’s accuracy even in the presence of pathological values, conducting the study in a controlled setting can be considered a limitation. Its application in real-life scenarios may present challenges that were not accounted for.

As suggested by the World Health Organization (WHO), it is important to disseminate medical procedures and health programs based on mHealth [21]. Mobile apps have become increasingly popular among professionals and individuals involved in the health care sector, with a variety of mHealth apps developed to improve health care facilities for all kinds of patients [22,23]

Specifically, mobile apps can play a vital role in screening, leading to the early detection of vital sign alterations, which is useful for alerting individuals and their caregivers to access medical care services and design and implement treatment plans. Screening is considered a crucial player in addressing chronic noncommunicable diseases (NCDs), which are responsible for nearly 70% of all deaths worldwide [12,24,25].

According to the WHO [12], NCDs cause 41 million deaths each year. This burden is expected to grow over the next decade, with NCD deaths projected to reach 52 million by 2030. Cardiovascular diseases are the leading cause of NCD deaths, accounting for 17.9 million fatalities annually. They are followed by chronic respiratory diseases (4.1 million) and diabetes (2.0 million, including kidney disease deaths linked to diabetes). Early detection and screening are vital strategies in combating NCDs. Disease screening detects diseases early, often before symptoms appear, through tests, examinations, or other rapid procedures [26]. This early detection enables effective prevention or treatment of identified diseases through timely medical intervention for those at risk [27].

Despite screening being recognized as the gateway to care, routine NCD screening at community and health facilities remains challenging due to limited capacity within health systems. A screening tool must be simple, valid, reliable, quick to administer, cost-effective, and user-friendly [23]. According to reports, photoplethysmography wearables have significant potential for estimating the risk of cardiovascular diseases [28], predicting suspected sleep apnea [29,30], and helping in the clinical diagnosis of hypertension [31] and diabetes. This potential can be harnessed through app solutions.

Recent reviews have assessed the role of apps in aiding self-management and transition among chronically ill young people, enhancing the lifestyles of chronic patients, and improving lifestyle in NCDs [32]. As reported by Moses et al [27], regardless of a country’s economic status, there has been a rise in mobile cellular subscriptions per hundred people in developing nations, mobile app solutions may provide equitable health care solutions without barriers.

Our study will enhance and support the accuracy of data on vital sign detection through PPG, also introducing measurements of biochemical risk indicators. The evaluation of a large population will allow for continuous improvement in the performance and accuracy of AI algorithms, reducing errors. Expanding research opportunities to evaluate mHealth solutions as effective and dependable screening tools is useful for preventive strategies, reducing the burden of NCDs [27]. App solutions, such as Comestai, may present new opportunities and challenges related to technology use in the health care sector, offering insights and guiding future research directions.