The goal of the patient safety movement is to reduce adverse outcomes or injuries resulting from health care processes. It is imperative that these adverse outcomes are avoided, prevented, or minimized [41]. With the improvements in safety standards and policies, more attention is accorded to analyzing safety issues and the sources of these issues [42]. Errors and inefficiencies in inpatient care are frequently the results of conflicting, incomplete, or suboptimal systems in which patients participate and interfere [43]. The report published by the Institute of Medicine at the beginning of the 21st century resulted in the increased and rapid adoption of health information technology in health care settings, especially for patient safety purposes [44]. For example, the wrong site, wrong side, wrong procedure, and wrong individual (WSWP) errors have been mitigated to some degree by electronic health records. However, these errors continue to be quite significant [45]. DT-assisted safety management systems can be implemented within the Safety 4.0 framework to manage complex safety procedures with minimum human error [46]. In addition to providing operators with contextual information about the surroundings, DTs can guide them through safety tasks [47].

Our study found different DT applications (n=3) that contribute to patient and system safety management through system twinning and product twinning. For instance, an opportunity that DT developers seized was the COVID-19 safety crisis within the health care ecosystem [21]. In 2020, COVID-19 disrupted the health care system and caused an ecosystem crisis that harmed public safety, including resource shortages, misinformation, and medical errors. Virtual interventions and technology initiatives became the preferred mode of service. For example, DTs could accurately contribute to vaccination strategy development. Vaccinations can be targeted based on the number of social contacts of each individual, and infections can be restricted to isolated hotspots and delayed by precisely targeted vaccination, inherent immunity, and public health measures that reduce the infection peak. Thus, DT technology supports decision-making to control the spread of the virus [21]. Apart from this, twinning systems contributed toward predicting the COVID-19 spread. Alrashed et al [23] used a DT system to simulate different strategies and scenarios to minimize the impacts of the virus spread and prevent it while continuing to provide necessary services to citizens with no interruptions to ensure their safety with minimal risk. Safety risk for patients was not only caused by the virus itself but also by the inability of the systems to respond to the new challenges. These challenges will continue to impact the system even after the COVID-19 crisis. Talukder [22] suggested a system architecture that ensures safety within an ecosystem disrupted by COVID-19. It mitigates system-related challenges and increases patient safety in post–COVID-19 health care delivery [22]. In conclusion, DTs (systems and products) can ensure safety management in health care systems by identifying potential threats, redesigning the systems to mitigate hazards, and improving the safety strategies implemented. This leads to the right care at the right price and time for everyone and everywhere at any point of care in a safe manner.

Although the applications of DT technology in patient safety were inspired by COVID-19, it is essential to investigate its potential in other settings and crisis situations. In addition, this technology helps address safety issues in health care without interrupting day-to-day work; it can also be used to address other medical safety issues, such as surgical errors, workplace safety issues, and medical bias in diagnosis.

The American Health Information Management Association describes health information management as the process of collecting, analyzing, and securing digital and traditional medical records that are vital to providing quality patient care [48]. Health care organizations seek to analyze patients' information efficiently and quickly, both internally and externally [49]. However, they face many challenges such as privacy, exchange restrictions, and extensive data. Increasing amounts of patient data are forcing health care institutions to replace traditional approaches that cannot cope with increases in data. The United States is taking steps to boost health care information and communication access by leveraging advances in information and communication technologies [50]. Health information management systems have grown rapidly in recent years and are being used to derive important health trends and provide timely preventive care [51]. DT technology can revolutionize clinical research with the changes that it can bring to the basis of health care systems and medical practices. By leveraging this technology, users can better ask questions, get better answers, and gain data-driven actionable insights without compromising the health of real-life subjects. In fact, using a DT, people can gather, aggregate, and represent individual information about their health and well-being [52]. As more data are collected, more DTs will be enabled, leading to more discoveries and better treatment, thus allowing the assembly of more data with less cost, and especially eliminating the risks and consent issues associated with actual human subjects [53]. In our review, we found applications related to information management DTs (n=2).

According to Lutze [24], eHealth systems can manage knowledge by implementing DTs that are based on artificial intelligence. He proposed a DT that collects daily activity data from smart assistance systems linked to wearable sensors for elderly people and extracts behavioral knowledge for information management. The technology suggested accounts for systems, processes, and group changes to provide unbiased conclusions based on learned, trained data. It has a human-centered design, as it allows the self-determination and autonomy of patients to share or refuse the usage of their data with providers and clinical staff. It also establishes solid robustness by automatically tracing the use of all knowledge sources and verifying conclusions drawn about patients after system changes [24]. Auditability is established by tracing the use of all knowledge sources and recording and verifying conclusions drawn about patients and users after system changes. Robustness is supported by automatically checking the containment of a patient within the designated user group of the system and verifying the continued validity of the assessed acceptance conditions after system changes. Moreover, human oversight is facilitated in all critical situations [24].

Another example of an information management DT was suggested by Pang et al [25]. During pandemics, sharing information among different cities and countries in real time through a shared learning model (federated learning) remains critical while ensuring enhanced privacy protection. Pang et al used DT technology in a novel collaborative paradigm that allows DTs in multiple cities to share the local strategy and status quickly without violating any privacy rules to help manage the COVID-19 pandemic [25]. These 2 examples show that DTs allow information management and knowledge extraction by encoding, storing, retrieving, and sharing data in a secure, smart, and real-time environment.

Health and well-being goals are challenging to achieve for many individuals. In response to this challenge, a growing number of technologies are being developed to improve people's diet, physical activity, sleep, and mental health. By promoting behavior change and controlling health care costs through modern digital health interventions, people can maintain better health and a healthier lifestyle. Today, a variety of sensors are miniaturized and widely used to track basic physiological indicators on the move to help with better health and well-being management. Moreover, because smartphones are extremely easy to access, mobile health apps are currently considered the most beneficial platform for promoting healthy lifestyles and changing behavior [54]. In modern medicine, personal health management services are viewed as electronic, remote, and digitally enabled care that helps individuals manage their own care and reduce the need for in-clinic visits that are typically expensive and time-consuming [55].

In our review, we found that DTs were used for health management and well-being promotion (n=3). One such DT was introduced by Díaz et al [26] in 2021. Their DT application was called DTCoach. It is a user-centered smart coach that serves as a mentor and an accompanying educator to the users. It helps the users make the necessary adjustments in their posture and performance based on measurements taken that characterize their individual capabilities [26].

Another example was ClouDTH, suggested by Liu et al [27]. This cloud-based health care system uses personal data from digitally twinned wearable medical devices to achieve a convergent interaction between the medical and physical spaces, and their virtual twins. Then, it facilitates self-management of health for elderly patients [27]. Patients' needs depend on many factors, and age is one of them. Elderly patients have higher demands for many medical services. Therefore, DTs are used extensively in geriatric care to support health management promotion of elderly patients. Furthermore, another example that we cover in our review was introduced by Tröbinger [28], which is a new DT approach serving as an alternative to telemedicine. It consists of a transparent remote doctor visit and a better-informed and robot-assisted telerehabilitation initiative that allows bidirectional telepresence control [28]. In summary, DTs are used to accompany patients and give them control over their health by promoting well-being and lifestyle activities and supervising them to maximize their performance in a safe environment.

Operating a health care facility on a day-to-day basis impacts patient experiences and organizational goals [56]. Thus, operations management helps in understanding and optimizing the business processes inside medical departments to reduce and alleviate the effects of overcrowding, waiting times, delays, and other problems that facilities are facing [57]. DTs have the potential to contribute to the effective operation of health care units. Our review found that most DT initiatives to improve health services have health care operations management focusing on revolutionizing clinical processes and enhancing medical care. They replicate hospitals or treatment facilities and help improve their performance in a safe manner with less risk. Applications are numerous and range from predicting resource shortage to managing patient flow. Using DT technology, an institution can execute a digital stress test to observe how the technology would fare under extreme conditions like crises. By creating a virtual twin of a hospital, stakeholders can review the operational strategy, capacity, staffing, and care model on the DT to determine what actions to take and mitigate future challenges. Our review identified extensive efforts (9 out of 17studies) to support operational control and account for health care system challenges. The efforts consisted of digitally twinned systems and processes for performance control.

Detecting anomalies in processes is essential to prevent hazards and predict the corrective actions that need to be implemented. With some interventions, such as the solution suggested by Croatti et al [37], DTs can be used to support physical processes through their digital representations and monitor their changes. In this study, by mirroring the real system by building an agent-based smart DT, they aimed to digitalize and support the process of severe trauma management [37]. This DT represents the operative phase of trauma management and starts when the trauma is marked as severe in the previous phase. The fact that this DT starts before the patient's arrival to the unit is very important for this case study [37]. In this way, the trauma team is prealerted about the incoming patient and starts collecting and receiving information directly from the accident site. Its internal state changes when the patient is delivered to the emergency department, where the trauma team starts taking care of the patient. A very preliminary version of a system prototype has been developed according to the designed conceptual model [37].

Another example is Health@Hand, suggested by Nonnemann et al [29]. They twinned the processes of an intensive care unit (ICU) station and integrated them into a digital system (Health@Hand) to allow remote monitoring of the processes. With this intervention, hospital managers can detect anomalies and faults immediately and intervene in an early stage. Moreover, while improving the productivity and efficiency of processes, some digital interventions forget to address the problem of cybersecurity, which may harm the systems. As a solution to this problem, Mylrea et al [33] propose BioSecure. It is a process twinning that allows managers to monitor every step in the supply chain process to ensure good productivity and cybersecurity to secure the system and data by applying Cyber-Informed Engineering.

One of the challenges in health care is providing an optimized scheduling strategy that can effectively use the hospital's resources and prevent delays, errors, and long lengths of stay. An application in the same settings (ICU and process twins) was developed by Chase et al [30] that aimed to optimize patient care by clinician staff at the enterprise level to improve the productivity of the staff and the quality of care delivered to patients. In addition, using resources effectively has always been a challenging decision for managers in all industries. In health care, resource allocation needs to be regulated by providing efficient services on time. Rodríguez-Aguilar et al [36] suggested a DT for a hospital that supports resource allocation (financial and human) called the Digital Health Public Emergency System (DHPES). DHPES provides efficient health services [36]. The DT design seeks to generate virtual instances that emulate the real operation of the provision of highly specialized public services, including the supply of medications, supplies, devices, and equipment as well as the management of human resources and financial resources in the event of a contingency [36].

Another DT initiative called HospiT'Win was developed to manage the patients' pathways inside the hospitals [34,35]. This DT can help hospitals equilibrate demand and supply and control the growing workload while reducing waiting times, lengths of stay, and delays. It also provides key information to decision-makers in real time, controlling the real-time flow of data. As high demand can disturb health care systems, such DT systems might be useful in times of crisis. Hospit'Win performed well during the COVID-19 pandemic by managing the demand [34,35].

Some operational strategies need to be implemented first to evaluate their efficiency. In hospitals, evaluating interventions would disrupt the daily services and activities of the units. This is where DTs could be most useful. Karakra et al [31] proposed a decision support system that employs real-time services data drawn from real systems and devices to enable evaluation of the impact of changes in services without disrupting the daily activities of the hospital. This idea allows foreseeing the effectiveness of changes in the models before they are applied in reality [31].

Providing patients safe and high-quality care is a demanding process. A DT framework for a system is proposed by Augusto et al [32] to optimize a patient’s care pathway in health care units. The system accounts for various arrival processes and simulates different scenarios to determine the best available leverages to optimize the operations of the system even under high demand and variability to account for uncertainties [32]. The framework has been conceived and tested in close collaboration with health care professionals to be as close to the real system as possible. Furthermore, the framework is intended to be used regularly by the head of the emergency unit [32]. Data collection was performed using the hospital information system for the following parameters: patient arrivals, the total length of stay; type of patient including less critical (fast and normal track), moderately critical, and life-threatening emergencies; and number of requested paraclinical examinations per patient. On the other hand, processing times were recorded by interviewing doctors, nurses, caregivers, and interns because the related data in the hospital information system were not reliable enough [32]. The model was validated and shown to reduce the waiting time and length of stay in different scenarios.