EHD can be broadly classified into 3 categories, described in depth in subsequent sections of this study. Refer to Table 1 for the taxonomy of key health information systems. Figure 1 visualizes the higher-level categorizations of these systems. The three categories are as follows:

Ensuring data input into health information systems is complete, of high quality, and able to be linked and shared across different systems within and across organizations is a major challenge with the current digital health information systems. In practice, management software data completeness issues are common, particularly on older systems where administrative and patient records are not linked, potentially leading to loss of information when data are transferred between systems [48]. Similarly, PAS used in hospitals have been shown to be inaccurate for reporting care outcomes [51]. The data can have limitations due to varied coding systems across organizations, data quality and completeness, and other issues [52]. In EMRs and EHRs, interoperability can be a major challenge, particularly a lack of integration across care contexts such as primary and acute care [46]. Registries have also had issues with interoperability due to instances of variation in standards used to define common data elements across different registries [44].

Interoperability is also a major challenge for emerging consumer health information repositories such as smartphone apps. Many apps do not include functionality that allows users to easily export data from apps and share it with health professionals [5]. In addition, the breadth of data collected by these apps could prevent integration into clinical informatics systems [71].

Data quality issues are also a significant limitation of different repositories of web-based patient health information, though the quality issues are different from clinical information systems. WPR data may not be representative, are often skewed toward a small number of health professionals, and therefore may not be sufficient for consumers to make an informed decision about a provider [91]. There is little research to show that these data can be used to measure clinical quality for individual health care providers, and the results of research in this area show mixed [91] association [73]. Coupled with this, health care organizations and health professionals have reservations about the validity of WPR data [74] and the ability of a single bad review to reach a wide audience [92]. These concerns may limit health professionals’ uptake of these data to understand the quality of care.

Social media provides a rich source of real-time health data, although it also has many limitations. These include collecting a large amount of irrelevant data, challenges in standardizing and validating data, and potential biases in the demographics and geographic location of social media users and identifying these biases in the data [84,93]. A noted limitation of search engine data is its potential to dilute cultural differences that shape web-based search behaviors or completely exclude data sets from geographic regions where those search engines are unavailable [93]. As with smartphone apps, wearables have significant data quality issues, with noted inconsistencies in data accuracy [85,94]. Wearable devices also encounter quality and completeness issues, compounded by interfaces that may not effectively highlight these gaps [95], leaving users unaware when data are not being collected [96].

Data security and privacy are a major concern with contemporary information systems and are often top of mind for health care organizations. When implementing EMRs and EHRs, ensuring data stored within and shared across organizations are governed in a way that provides private and secure data exchange can be challenging [25]. In consumer-facing systems such as appointment booking engines, data have been misappropriated for financial gain [97,98], suggesting issues to be resolved about the privacy and security of these data and the social license for its secondary use.

Implementing any new technology in existing workflows can be challenging, and digital health information systems are no exception. Resistance to changes in workflow can be a notable barrier to the adoption of clinical information systems [28,56]. In the context of certain types of data collection, such as patient-reported outcome measures, inadequate infrastructure to support the collection of these data [99] is a major challenge. System maintenance after implementation can also be a challenge for health information systems. Limitations associated with clinical registries include the expenses related to maintenance, which often require staff with specialized skills to operate [44], and the necessity for sustained funding to support the registry in the long term [10].

Another challenge of modern information systems such as wearable devices is a lack of guidelines or tools to inform the best practice use of remote monitoring devices in clinical care [61]. Currently, very few wearable devices are subject to regulatory standards that govern other medical equipment [86], making it challenging to know which wearables are safe to use in the context of health care. Integration of these devices and the data they produce into health care remains challenging [100], as they currently do not integrate with clinical workflows or other information systems such as EHRs [86]. This same challenge occurs with health data gathered by apps, as there are currently few pathways for use in the clinical setting by health professionals [101].

One of the major limitations of the current digital health information systems is their impact on the workloads of health professionals. The digitization of health data collection parallels an increased administrative load for health professionals. This issue is particularly pronounced with EHRs, as health professionals typically enter data. This has been shown to increase the workloads of health professionals and contribute to burnout [25].

While considerable time and effort is spent inputting data into contemporary digital health information systems and given the vast amount of data collected, extracting and accessing these data for secondary purposes is notably difficult. Even when data can be extracted, secondary use of data from health information systems has several limitations. In the context of EMRs and EHRs, it can be challenging to obtain longitudinal insights from data as information is captured at each time point a patient interacts with the health system, which for many individuals occurs as sporadic or occasional visits [102]. Similarly, data from consumer complaints and incident reporting have limitations because they capture individual incidents not holistic care [103]. This can make it challenging to identify a strategy for improving quality and can also be an issue for determining whether a complaint is attributed to an individual or a system issue.

Another challenge supporting the secondary use of many digital health information systems is scaffolding the information or supporting end users to interpret it in different contexts from those in which it was collected. For example, registries are often primarily designed for research purposes, and they are not intended to make data readily accessible for use by health services [37]. Patient-reported outcome measure data have a similar limitation. The data are not routinely used by many health care organizations [56]. Health professionals may not use patient-reported measures due to not feeling like they have the capacity to use them, not seeing their value [56], or a lack of knowledge regarding how to meaningfully interpret the data [99]. When using these data for quality improvement, barriers include a lack of timeliness of the data, limitations in determining the cause of a poor outcome, and implementing change based on a poor outcome [104]. It has also been noted that patient-reported outcome measures are only 1 piece of performance data that health professionals have access to, and the data may be ignored if individual providers do not see themselves as outliers in the data set [104].

As health information systems increase the amount of EHD collected and move toward better interoperability, patient consent for secondary use of data becomes an increasing concern. In the context of health data, consumers freely share about themselves, such as that on social media and web-based communities, which is another challenge related to consent exists. Although individuals choose to share this content publicly, it remains unclear if users are comfortable having their data used for research and other secondary applications [84].

Another major challenge with contemporary digital health information systems is data ownership. While all the data collected belongs to patients, the data are housed in information systems that consumers do not have access to. The value of these data is being increasingly recognized, and this is acting as a barrier to access. For example, the device manufacturer often controls access to data from remote monitoring devices, which can raise complex legal and ethical concerns [62]. These technologies collect large amounts of data that may be vulnerable to privacy breaches or be coopted for commercial purposes [105], and custodianship of these data is widely contested [18]. Data collected by bespoke databases in devices has value as a commercial asset [17,18], and vendors may not want to make it available to end users. Furthermore, the data collection architecture in these systems is not governed by any standards, so linking and comparing data, even if it can be accessed, is difficult.

EHD has great potential to transform many aspects of the health sector. There is a growing body of research exploring how data can be used to predict patient outcomes [64,68,106], understand care quality [51,55,107], and personalize treatments [21]. Approaches such as those described in learning health system processes can increase the actionability of the EHD to ensure it improves care processes and outcomes [24]. There are also a growing number of ICTs designed to increase the actionability of EHD from different sources. Much research has been undertaken into visualizing EHD from clinical sources such as PAS and EMRs to uncover insights for end users of the tools [66,108] and how approaches such as learning health system processes can increase the actionability of the EHD. Electronic patient portals use digital technology, such as a website or a smartphone application, to provide patients access to personal health information [109]. These platforms are often linked to EHRs to increase accessibility to data, help involve patients and caregivers in clinical decision-making, and improve communication between patients and their health care team [110,111]. Similarly, initiatives such as OpenNotes, which makes clinical notes available to patients and caregivers, improve transparency and patient-centered care [112,113].

In the following subsections, 4 use cases describe how ICT supports existing care models and how the data collected might be harnessed to transform service provision, health, and well-being. Some of these use cases describe scenarios increasingly realized in the health sector, partly due to the rapid adoption of digital platforms in response to the COVID-19 pandemic. Other use cases are informed by the research but describe hypothetical scenarios requiring evidence to be translated into practice before they can be fully realized by the health sector. These scenarios are designed to illustrate and prompt reflection on how EHD can be harnessed to generate actionable insights and transform health care in the future. Figure 2 presents a visualization of the 4 different use cases.

Digital technologies and the EHD they collect can be used to strengthen patient consultations with health professionals. The value of technology to support patient interactions both within and outside of the consultation is increasingly being recognized as virtual care. Virtual care describes non–face-to-face clinical care professionally enabled through digital mechanisms [27]. In the current clinical practice, health professionals have limited access to clinical decision support tools that harness EHD and support shared decision-making [114]. Genomics data are increasingly available to health professionals to support precision medicine [57] and are likely to be more prominent in future clinical decisions. EHD from patient applications and wearable devices can be challenging for health professionals to access in consultations. When data are available, there is limited guidance on how to incorporate these data [115,116]. In addition to data access challenges, there is some evidence that the use of digital technologies during the consultation may impact patient-provider communication and rapport [117,118].

In the future, access to EHD during consultations may enable shared decision-making between patients and health professionals. This could be achieved by visualizing cohort data within an organization to help patients understand their likely outcomes for a given procedure at that specific care center [66]. It can also support more personalized decision-making in the consultation, such as by harnessing visual interfaces presenting EHD to help patients assess the risk of undergoing a specific procedure [43], supporting more informed consent processes. Finally, patients could consent to sharing health application and wearable device data with their health professional so that the data can be reviewed and discussed in the consultation to understand patient compliance with treatment plans and tailor care to align best with the patient’s lifestyle. Alongside this, the increasing application of disruptive technologies such as AI to EHD could streamline administrative processes in health care. Although not directly beneficial in the consultation, the actionability of EHD for such applications should reduce the administrative burden on health professionals and allow more time for service delivery.

Health care teams currently underuse EHD. While digital technologies such as videoconferencing have been successfully adopted as tools to support team meetings [119], the uptake of ICTs for visualizing data has been slower. There may be a need to modify physical meeting spaces to better incorporate technologies for presenting these data [120].

In the future, there will undoubtedly be greater access to EHD by individual health professionals and teams. Given the considerable time health professionals spend on data entry [25], there will likely be an increasing expectation that they will have access to these data in return for the workload burden. Access to technologies to visualize and scaffold EHD will provide health teams with a holistic view of each patient’s care journey, transforming all aspects of team interactions, including clinical decision-making, quality improvement, research, and learning. Furthermore, ready access to data by health care teams will likely have an important role in ensuring quality control as health professionals incorporate visualization tools into team meetings to review processes and outcomes of care. As part of these team reviews or as individuals, health professionals will likely use digital technologies presenting these data to review the quality of information, annotate points of interest, and edit and improve data in real time. As the application of AI in health data becomes more reliable, there will also likely be an increase in prompts and scaffolds provided by advanced analytics tools that can support decision-making within health professional teams.

Health professionals regularly dedicate considerable time to engaging in education and training activities to stay up to date on the latest evidence [121,122]. Health professionals have limited access to EHD to reflect on their practice and inform decisions about professional development. Despite this, regulatory bodies expect that EHD will be used in training and professional development to maintain registration to practice [123,124]. Research suggests that health professionals would like greater access to electronic data for educational purposes [125].

In the future, health professionals, teams, and organizations will be able to leverage the plethora of EHD to understand care delivery and outcomes and enable personalized reflective practice and learning. A new range of educational technologies will be available to engage health professionals in their data to develop a rich understanding of how their behavior can change care quality. This learning approach could create authentic clinical narratives and link it with evidence-based approaches to technology enabled learning. This would focus efforts in health professional knowledge and skill development and enable accreditation bodies to recognize quality in learning experiences rather than quantity of time spent engaging in learning.

Patients have limited access to data collected by health care organizations about their treatment pathway. While technologies such as patient portals [109] and open notes [113] are making these data more available to patients and their caregivers, uptake of these solutions is limited. However, there may be an increase in patient portal adoption as regulation requires health care providers to increase the accessibility of EHD for individual patients [57]. Coupled with this, vendors from outside of health care are increasingly entering the sector to provide direct-to-consumer care by harnessing AI to enable services to be delivered when and where the consumer desires [88,89].

The health care transparency movement has also explored the value of making cost and quality information about health services available to the public [126,127], but these data are not consistently published for all organizations. There is also a plethora of consumer-centric apps and wearable devices collecting extensive data on individual activities, both general and health-specific, that enable the self-management of health and wellness [6,71,72]. Some of these innovations are beginning to be used to improve patient self-efficacy in managing chronic conditions. The increasing availability of diverse data sets also opens up exciting research opportunities. It will undoubtedly unlock new knowledge about human health and support the discovery of more personalized and adaptive approaches to maintaining health and delivering care. In the future, EHD has the potential to enable personalized preventative health and well-being solutions for members of the public. Digital technologies will increasingly make health care available outside of clinical settings, at home, in the workplace, or in whatever location suits the patient.