Upon successfully securing a sustainable funding source, we initiated consultations with several development teams, leveraging our collaboration with the biomedical engineering team from the Universidad Politécnica de Madrid. Following a public tender process, the project was awarded to Kairos DS [8]. During several collaborative sessions with Kairos DS, we meticulously crafted a statement of purpose for our digital health solution, including a description of its main functionalities and features.

To secure funding for a new medical app, efforts were diversified across private funding contests, collaboration with pharmaceutical companies [9,10], and exploration of public grants focused on health care innovation [11], tapping into multiple sources to maximize potential financial support.

An analysis of the previous remote monitoring experience, the Digireuma study [6], was conducted. For this, semistructured telephone interviews were conducted by an external expert with a randomly selected subset of patients, in order to gain a comprehensive understanding of user experiences, capturing both the strengths and potential areas for improvement of the application and implementation methodology. Clinicians who participated in the previous project completed a questionnaire to evaluate the usability of the medical interface, the usefulness of the solution and its utility in medical practice. Further information on these questionnaires can be found in Text S1 inMultimedia Appendix 1.

We used a user-centered design approach [14], for which by definition must consider the needs and experiences of end users via a continuous series of evaluations [15,16]. Initial user profiles were established based on interviews with clinicians and patients [6], categorizing users into health care providers and patients. For health care providers, functionalities such as patient data monitoring, alert resolution, and telematic contact were developed. For patients, features included ePROs registration, medication management, reminders, and access to educational content covering disease information, management strategies, and relevant hospital information.

As a starting point for designing the architecture of the system, the MAM protocol and the agents that comprise the proposed clinical intervention were considered, combining face-to-face and telematic follow-up protocols [12]. The profiles, needs and functionalities previously defined were then considered. With all of this in mind, the system and components to be developed were then defined: a mobile app for patients, an application for health professionals, and services for communication, data processing and storage.

Throughout the design and development process, it was necessary to evaluate possible technologies and their deployment (Text S2 in Multimedia Appendix 1). The system architecture was designed using cloud-based technologies, specifically Amazon Web Services (AWS), to optimize resource usage and ensure data security [17-19]. To ensure cost-effectiveness, scalability, and ease of maintenance, we selected hybrid and cross-platform technologies [20], allowing IMIDoc to be deployed on both Android and iOS platforms efficiently.

In addition, to carry out the clinical evaluation study, a series of requirements had to be met and specific documentation had to be submitted to the Information Systems Security Office of the Madrid Health Service.

In a dedicated meeting with the multidisciplinary development team, the scope of the digital solution was outlined by conducting a scoping analysis to delineate its potential capabilities and limitations. The name IMIDoc was selected for the digital health solution proposed in this project. The primary purpose of IMIDoc was defined as addressing an unmet need in the care of patients with RA and SpA, which is the providing of continuous monitoring. Conventional clinical practices often allow for discrete, scheduled medical consultations, leaving gaps in patient care when immediate attention is most needed. The IMIDoc app was designed to bridge this gap, ensuring timely medical oversight of patients.

This project was financially supported by a grant from the Fondo de Investigaciones Sanitarias (FIS) of the Instituto de Salud Carlos III (PI22/00777). The grant specifically facilitated the technical development of our digital solution and will allow us to carry out the clinical-technical evaluation in a clinical study.

In the platform ideation phase, relevant aspects and features that the digital solution should incorporate were identified. During this phase we considered the opinion of 41 patients and 4 clinicians about their experience with the previous digital solution (Table S1 in Multimedia Appendix 1). Regarding the patients, usability ranged from 85.3%‐97.5%; most found the digital solution easy to use, and well adapted. Over 90% (38/41) of participants found the application was well adapted to its objective, with an adequate level of detail. Most patients reported considered that it allowed them to record their current health situation and would recommend it. The areas patients most frequently cited as needing improvement were the application’s design and the amount and detail of informational content related to the disease. On the other hand, clinicians rated the overall usability and design (0‐10, from worse to best) of the application with a mean (SD) of 8.50 (1.22), the app utility with 8.50 (1.16), and the feasibility with 7.56 (2.41). They primarily expressed concerns regarding the feasibility of the implementation in clinical practice, particularly in terms of time and logistics, followed by the performance of the solution.

In accordance with the objectives of the clinical study, in which the implementation of the MAM model in clinical practice through IMIDoc will be evaluated, Textbox 1 specifies the clinical and ePROS variables that the solution was designed to capture and that comprise the monitoring and follow-up of patients with RA and SpA.

In addition to the capture of clinical variables, instruments have been implemented to evaluate the usability of the applications, satisfaction, and user experience: System Usability Scale (SUS) and the Net-Promoter Score (NPS) indicator [18,19]. These were evaluated at 10 weeks and 34 weeks, respectively.

The analysis and design phase of the project made it possible to identify the user profiles and the functionalities and characteristics of the applications included in the IMIDoc digital health solution. Figure 1 shows a schematic of the main agents, components, and services of the system architecture using the MAM applied to RA and SpA. The figure shows the users (patients and health care professionals) accessing their respective applications and communicating with the service infrastructure deployed in the cloud.

The first steps in the development of the IMIDoc applications (the mobile app for patients and the web interface for professionals) were focused on addressing the interface design. For this purpose, we used the Figma platform, which is a cloud prototyping platform that facilitates collaborative work.

The IMIDoc mobile app (Figure 2) was designed to incorporate functionalities for recording disease activity, as well as some other domains potentially affected by RA or SpA.

In addition, the IMIDoc app included scenarios for medication management. To achieve this, the professional team carefully consulted the latest clinical practice guidelines for RA and SpA. This comprehensive review of current guidelines informed the creation of tools for monitoring disease activity and managing medication within the app.

A section with educational and informative content to help patients understand their disease and treatment options was also included; specifically, 26 educational articles were generated by clinicians with multimedia components, images and 12 educational videos. Specialists in various fields, including immunology and obstetrics, were engaged to contribute to the development of educational resources. The material produced included illustrated resources designed to improve patient understanding of their condition and the treatment options available. In addition, a diverse range of formats was used to adapt to different preferences. This included the creation of infographics, interactive texts, and videos. Each format was carefully selected to best suit the content being presented, with the aim of enhancing user engagement and understanding.

Another feature of the app involves task and medical appointment reminders in a personalized questionnaire that is complemented by push notifications.

The IMIDoc clinician web platform is a web-based application for health care professionals, primarily clinical rheumatologists [21]. The IMIDoc clinician web platform includes scenarios for monitoring ePROs in a prespecified time frame and generates events or alerts. These features will enable clinicians to respond in near real-time to significant changes in disease activity. Figure 3 shows an image of the screen displaying the data recorded by a patient.

The top section displays basic patient information, including patient code, sex, and last application access. Below, the left panel shows recent incidents (flares, infections, medication issues, or others), the number of swollen joints (NAT, Spanish acronym for “Número de Articulaciones Tumefactas”), the number of tender joints (NAD, Spanish acronym for “Número de Articulaciones Dolorosas”), enthesitis, and access to a homunculus visualization, along with options to view patient messages. The right panel below features bar graphs tracking patient progress, including pain scale (EVA, Spanish acronym for “Escala Visual Analógica”), patient global assessment (VGP, Spanish acronym for “Valoración Global del Paciente”), erythrocyte sedimentation rate (VSG, Spanish acronym for “Velocidad de Sedimentación Globular”), C-reactive protein level (PCR, Spanish acronym for “Proteína C Reactiva”), Bath Ankylosing Spondylitis Disease Activity Index score (BASDAI), and Axial Spondyloarthritis Disease Activity Score (ASDAS).

Through IMIDoc, clinicians could quickly identify the health status of patients and their level of adherence based on the automatic processing of data. The IMIDoc classification algorithm assigns 6 different levels to patients as shown in Figure 4.

As mentioned above, there are several platforms available from technology providers such as Amazon Web Services (Amazon), Azure (Microsoft), Google Cloud (Google) and IBM Cloud Computing (IBM), among others. For the implementation of IMIDoc, we selected Amazon Web Services (AWS) to deploy the backend components. In addition, AWS offered a number of features of interest for our implementation [15,16]. We used AWS Fargate which is a serverless computing engine for containers, in which a Node.js-based application has been deployed that features a RESTful API to respond to requests using GET, POST, PUT, DELETE methods via HTTPS. For the storage of ePROs data and clinical actions, a relational database PostgreSQL with Amazon RDS was implemented and deployed in the European region (Spain). An important aspect to consider was security, for which the Security Hub service was enabled. This integrates functionalities to detect and manage security events, in addition to other components such as a Firewall and distributed denial of serviceprotection. Moreover, AWS complies with various certifications, accreditations, European laws and regulations such as ENS (high category), DESC CSP, CISPE and the General Data Protection Regulation (GDPR). It should be noted that these characteristics, in terms of security and data processing, are requirements mandated by the Information Systems Security Office for the approved use of the application in health care. Finally, AWS allows us to use a continuous integration and delivery service and, in this way, allows us to maintain and prepare system updates for subsequent release in production.

The system has modules that run automatically, monitoring the data, establishing the patient’s health status, and generating notifications that are sent to the patients’ app through the Google Firebase service. Table 1 specifies the types of notifications included in IMIDoc.

Google Analytics was used to track and process dimensions and metrics related to the use of both applications: country, device, operating system, app version and user interactions with the app (number of sessions, duration of sessions, screen views, and interaction times). Google Analytics service and mechanisms were integrated to record the usage metrics of both applications.