After identifying relevant articles, each article was assessed thematically with data extracted and tabulated in Table 1. The findings from these articles can be subsumed into 2 overarching themes: (1) the need to evolve service delivery models in health care, which was expressed in 5 articles; and (2) the potential value of using patient journey data in an LHS, which was described in 3 articles. These 2 thematic categories will be examined in the subsequent sections.
Theme 1: The Need to Evolve Service Delivery Models in Health Care
With 5 articles stressing the urgency to evolve service delivery models in health care settings, it was the most prominent theme of the literature sample. In the article “The Indiana university center for healthcare innovation and implementation science: bridging healthcare research and delivery to build a learning healthcare system” [13], Azar et al detail that an “estimated 75,000 deaths every year could be prevented if high-quality care was more efficiently and effectively implemented” [13]. The authors quote the United States National Institute of Health, in that this considerable problem is not due to a paucity of knowledge, but rather poor incorporation of health care discoveries into daily practice [13]. Azar et al [13] clarify that over the past 3 decades, medical knowledge has increased, with 11 systematic reviews and 74 clinical trials being published every day, yet only 14% of these new findings are actually implemented in health care settings and translated into practice [13]. Therefore, to mitigate the risks to human health, the authors propose a paradigm shift in how health systems and service delivery should be conceptualized. They present 2 contrasting perspectives: (1) a traditional model of service delivery and (2) an innovative and adaptive model of health care service delivery. In the traditional model, organizations are viewed as machines that perform predictable, repeated tasks with replaceable parts that operate in stable and nondynamic settings [13]. In the adaptive model, health care systems are viewed as complex, dynamic, adaptive, and evolving systems comprised of a network of semiautonomous individuals (ie, health care professionals) who interact in nonlinear ways [13]. As health care needs and interactions are interdependent, interconnect, and changing over time [13], the authors insinuate that it is vital to design health care services to support the fluidity of systemic evolution. Thus, their article expressed the criticality of designing and developing an adaptable agile learning system that integrates hospital systems, population health, individual patients, and health care personnel [13].
Fung et al [14] present a systems approach to redesigning care in their article “Regional process redesign of lung cancer care: a learning health system pilot project.” Their novel approach enables timely access to cancer treatment for patients with lung cancer to a centralized specialty service that addresses clinical and operational challenges [14]. However, the authors caution that, despite its potential value, there is limited evidence of successful implementation of the LHS vision [14]. Thus, to streamline and operationalize the LHS concept, they developed the Ottawa Health Transformation Model as a regional approach to guide service delivery change and to integrate the nuances of the patient journey with best practices [14]. Further, the article laments that all facets of care need to be examined to address the complexity of health systems and to improve patient experiences holistically rather than just isolated parts [14]. The article concludes with the caveat that the value of the LHS approach in relation to service delivery is the creation of a system that can facilitate best practice adoption and fluid innovation [14]. Similarly, in their concept analysis and systematic review, Gartner et al [15] detail that a performant health care system is crucial for every country and that the current siloed health care business practices must be evaluated and challenged [15]. The authors suggest that fragmented health care services can compromise patient care, inhibit sustainable service delivery, and result in suboptimal use of financial and human resources [15]. Further, the authors state that repeated calls to improve the overall performance and quality of global health care delivery have occurred since 2001 [15]. The calls for transformational change in health care have been made by well-established national and international organizations such as the Institute of Medicine [15,20,21]; The National Academies of Sciences, Engineering, and Medicine [15,22]; and The World Health Organization [15,23,24]. Gartner et al [15] suggest that understanding the patient journey through an LHS view can facilitate the improvement of health care service delivery through a feedback loop in which data can be used to identify problem areas to support continuous improvement [15]. Lastly, in a similar yet contrasting view, Sun et al [16] express in their paper “Health management via telemedicine: learning from the COVID-19 experience” that telemedicine provides numerous opportunities to improve care efficiency, accessibility, and patient outcomes [16]. However, they state that many challenges exist, such as the digital divide, usability, and technology interoperability [16]. Further, the authors detail that the delivery of telemedicine services must evolve to support continuity of care throughout the patient journey [16]; specifically, by including the seamless integration of data from the clinical workflow of multidisciplinary care teams to support the LHS [16]. Nonetheless, they clarified that the implementation of a telemedicine business model must be supported by rigorous evidence-based practices, including clinical trials [16]. They warned that such precautionary measures are necessary to facilitate the seamless integration of telemedicine into routine care, ensuring the quality and safety of virtual care delivery [16]. Lastly, Yu et al [17] recount that data are only important and useful when they can be transformed into knowledge. In a health care context, the importance of data is realized when data sets of individual patients can be aggregated with similar patient data to inform patient populations [17]. Further, the value of clinical data lies in its interpretation in a clinical context among continuing care providers and when it is shared with the patient or their caregivers [17]. Additionally, the data set of a citizen (ie, patient) becomes of greater importance when it is combined with that of other citizens and when it can be aggregated for comparative statistical analysis to inform the health system on the health status of a population or subset [17].
Theme 2: The Potential Value of Using Patient Journey Data in an LHS
The potential value of using patient journey data in an LHS was expressed in 3 articles. In the article “Patient journey mapping: current practices, challenges and future opportunities in healthcare,” Joseph et al [1] describe how the data gleaned from patient journey maps could improve the health system by identifying varying patient experiences. Additionally, Joseph et al [1] detail that journey mapping approaches hold a significant value in improving complex health care processes for patients and providers alike. Further, the authors express that closely integrating patient journey mapping techniques into the health care system could create an LHS [1]. In their study “Learning health system for breast cancer: pilot project experience,” Levine et al [18] report that clinicians need accurate and timely information on patient outcomes associated with various treatment modalities. Moreover, the authors describe that electronic health records are perceived to be helpful technologies, but access to patient data is often difficult [18]. However, despite the data accessibility challenges expressed in their study, the researchers were able to combine, read, and extract electronic health records data to view the patient journey [18]. Specifically, Levine et al [18] developed a prototype leveraging IBM Watson technology, with capabilities to validate artificial intelligence using natural language processing and to denote the clinical course of patients (ie, patient journey) in support of an LHS platform [18]. Their study findings illustrated a means by which the vision of an LHS could potentially be achieved by using artificial intelligence [18]. Despite the preliminary nature of their study, the authors were able to demonstrate that the hospital had the necessary data to formulate a view of the patient journey, which could be extracted and used in ways to support clinical decision-making [18]. Lastly, in their observational study, Sharma et al [19] used an incremental and iterative approach, engaging administrative and clinical domain experts to demonstrate that human actors, rather than IT, are the central focus of data movement [19]. The authors evaluated a kidney transplant referral pathway and established the relationship between human actors, organizations, the complexity of data administration, and data flow bottlenecks [19]. Their study illustrated the manual and often cumbersome tasks that clinical staff must perform to access and visualize health data from fragmented IT systems [19]. The authors express broadly that IT systems that are not interoperable can lead to data access challenges and complicate the clinical workflow and health care providers’ ability to effectively and efficiently perform their job functions [19]. They further reveal that in a kidney transplant referral context the lack of centralized and timely access to patient data can delay patients’ registration on the transplant list, as the time and effort to complete referral forms are greatly increased [19]. Sharma et al [19] propose that an LHS with linked patient data can improve population health outcomes and inform interventions by providing timely and intuitive access to health information.