The successful implementation of digital interventions in health care settings critically depends on their adoption by various stakeholders and health care team members. This process is often hindered by an inadequate consideration of the necessary behavioral changes during the implementation phase. To effectively support these changes, a thorough understanding of the influencing factors is essential. Frameworks, such as the nonadoption, abandonment, scale-up, spread, and systems (NASSS) approach, offer a broad evaluation of the success of digital health solutions [1,2]. Previous reviews have used the NASSS framework to assess contextual factors influencing implementation success of online consultations in primary care [3] and computerized clinical decision support systems in hospitals, identifying adopters as a key determinant [4].

The theoretical domains framework (TDF) provides a theory-informed approach to identify determinants of behavior and potential reasons for implementation issues, focusing particularly on the adopters [5]. Ignoring these behavioral influences can result in poor individual uptake, unsuccessful implementation on both local and systemic levels, or limited growth and innovation of the intervention. This aspect is particularly overlooked when introducing novel remote monitoring technologies [6].

In the neurorehabilitation setting, clinician attitudes and digital literacy, training availability, organizational readiness and support, perceived usefulness, perceived ease of use, and patient engagement have been identified as critical determinants of technology uptake [7,8]. While these studies provide valuable insights into specific clinical contexts, they focus primarily on clinical and organizational aspects of digital health implementation. Our systematic review builds upon this foundation by integrating 2 complementary behavioral frameworks, NASSS and TDF, to comprehensively characterize barriers and facilitators across diverse health care settings. This approach allows for a broader understanding of the complex interplay between technological, organizational, and behavioral factors influencing digital health adoption, offering a more nuanced perspective on how to enhance implementation strategies across different health care environments.

Health care digitization is an increasingly important field and is crucial to meet the evolving demands of health care, as described in the Topol Review by Health Education England [9]. This review forecasts substantially impacts from technological advancements such as telemedicine, smartphone apps, and diagnostic sensors and wearables. The COVID-19 pandemic further accelerated changes in health care delivery, exemplified by the adoption of technology-enabled virtual wards allowing patients to receive care at home instead of in the hospital [10]. The implementation of innovative wearable sensing solutions in the UK National Health Service has encountered obstacles, particularly the lack of consideration for human behavior, which significantly affects the potential of the technology and patient safety [11]. Moreover, engaging key stakeholders is often overlooked during the digital solutions implementation process [12].

The aim of this study was to describe and characterize barriers and facilitators to the adoption and implementation of digital solutions within health care using behavioral frameworks.

This systematic review was conducted in accordance with the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines, which is an evidence-based minimum set of items for reporting systematic reviews and meta-analyses [13].

The review was registered on PROSPERO (CRD42022264937).

This review sought to answer the following question: Which human behaviors influence the adoption and implementation of digital-based interventions in health care?

A systematic search with expert librarian support was performed using the electronic databases MEDLINE, Embase, Health Management Information Consortium, and PsycINFO, for papers published in the English language. The search was conducted using a list of terms relating to human behaviors and digital tools in health care; the complete search strategy is available in Multimedia Appendix 1.

All identified studies were uploaded to Covidence (Veritas Health Innovation), a Cochrane-supported systematic review package tool [14]. Initial title and abstract screening were conducted by 2 investigators (SBP and FMI) to determine if the eligibility criteria were met. Discrepancies were resolved by discussion. Studies meeting the inclusion criteria underwent full-text screening (SBP and FMI); further studies not captured by the search were identified through bibliometric cross-referencing of reference lists.

Eligibility criteria were developed for title and abstract and refined full-text screening, and the final search was performed on July 13 2021. Included studies either reported outcomes relating to a change in behavior or practice of health professionals, or the practicalities of delivering such interventions using technological supports. Full inclusion and exclusion criteria are outlined in Textbox 1.

For all included studies, barriers and facilitators were independently identified, extracted, and classified by 2 investigators (SBP and FMI) using 1 or more of the 7 NASSS framework domains: the condition or illness, the technology, the value proposition, the adopter system, the organization, the wider (institutional and societal) context, and the interaction and mutual adaptation between all these domains over time [1]. All extracted data were then further organized using the NASSS framework subcategories with the exception of the adopter system, which was further classified using the 14 domains of the TDF, allowing for a comprehensive framework for examining implementation factors [5] by SBP and FMI. These included knowledge, skills, social or professional role and identity, belief about the capabilities, optimism, reinforcement, intentions, goals, memory, attention, and decision processes, environmental context and resources, social influences, emotion, and behavioral regulation. Where studies were independently reviewed by SBP and FMI, consensus was achieved through discussion of any discrepancies.

Risk of bias was assessed using the Mixed Methods Appraisal Tool (MMAT) [15]. This is a critical appraisal tool designed for systematic reviews, including qualitative, quantitative, and mixed methods studies. It consists of frameworks for 5 categories of studies; qualitative research, randomized controlled trials, nonrandomized studies, quantitative descriptive studies, and mixed methods studies. Quality assessment was performed by one reviewer (SBP) and validated by a second reviewer (FMI), independently.

Once the data had been extracted into the individual NASSS and TDF headings, 2 authors (SBP and FMI) reviewed and discussed each heading and subheading and synthesized the findings from the included studies to answer the review question. All studies with an MMAT score of ≥3 stars were included in the discussion and with specific risk of bias scores being considered in the discussion and synthesis.

A total of 2704 unique citations were retrieved through literature searches. A full-text review was performed for 81 articles, with 12 meeting the inclusion criteria for analysis. A PRISMA flow diagram is provided in Figure 1. Table 1 summarizes the characteristics of the included studies. Of the 12 included studies, there were 4 (33%) qualitative studies, 2 (17%) quantitative randomized controlled trials, 3 (25%) quantitative nonrandomized studies, and 3 (25%) mixed methods studies. Eight studies were carried out in the United States, and one each in India, Australia, the Netherlands, and Tanzania.

Table 2 displays a summary of included study designs, limitations, and conclusions. These were heterogeneous in terms of the intervention being assessed (eg, electronic health care records, alert systems, clinical decision support tools, mobile health platforms) and the adopters using the technology (general practitioners, specialist doctors, physician assistants, pharmacists, nurses, midwives, and administrative staff).

This systematic review aimed to characterize barriers and facilitators influencing the adoption of digital interventions in health care using 2 established behavioral frameworks: the NASSS and TDF. Twelve studies met the inclusion criteria; analysis identified 5 NASSS domains (“condition or illness,” “technology,” “value proposition,” “adopter system,” and “organization”) and 8 TDF domains (“knowledge,” “skills,” “social or professional role,” “beliefs about capabilities,” “beliefs about consequences,” “reinforcement,” “memory, attention, or decision processes,” and “environmental context/resources”). Our findings highlight that successful implementation critically depends on aligning interventions with adopters’ capabilities and workflows, ensuring intuitive technology design, clearly communicating value propositions, providing adequate training resources, and addressing organizational readiness.

Our findings showed the digital intervention itself is a pivotal factor in the success of adoption, particularly when it enhances the clinician’s experience and aligns with existing workflows. Intrusive interventions can disrupt these workflows, leading to reduced engagement by users. The risks associated with software and hardware failures, identified as significant barriers to adoption, must be carefully considered during the design and delivery phases. One study [28] highlighted technical factors, including network issues and a lack of infrastructure, as common barriers, while formal instructions or guidance can mitigate some of these challenges [29]. Where users are unable to recognize the value proposition or experience negative effects from the technology, they may lose interest in adoption of the technology.

Several psychological and social factors were underreported and are critical in understanding the adoption of digital health care interventions, in keeping with previous findings identifying motivation to change and willingness to adopt novel approaches as important but underexplored psychological factors [8]. TDF domains such as optimism, intentions, goals, and behavioral regulation were not reported in included studies or the wider literature, which could be likely due to digital health implementation studies focusing on final evaluation. Emotions such as fear have been described as a barrier; however, it was also not identified in the included studies [28].

This review identified the capabilities, beliefs, and workflows of adopters in the clinical environment as crucial for the successful implementation of digital interventions. Factors, such as knowledge, skills, social and professional roles, and identity are paramount. For example, ensuring adopters have the necessary knowledge and literacy, both basic and digital, to use the digital intervention effectively is vital [30]. Personal and psychological barriers, including age and education levels, have been described as multifaceted [28,30]. In the wider literature, user-friendly design and intuitive navigation are also often cited as facilitators for ease of use [7,8,31].

Our findings indicated that adopters’ beliefs about their capabilities and the consequences of using digital interventions, such as concerns over patient information security, play a significant role. The integration of digital interventions into existing workflows requires a balance between capabilities and cognitive load. Cognitive burdens are a recognized barrier to the implementation of digital health interventions [31], specifically regarding situational awareness and decision-making impairment [32]. The literature suggests a need to adapt consultation skills to accommodate digital interventions, especially in the context of increased remote consultations following the COVID-19 pandemic [33,34]. However, there is limited research into communication skills required while adopting digital health interventions in person. Disruptions, such as alert fatigue from excessive notifications, are recognized barriers and contribute to resistance to engagement with digital interventions [35-37].

Adoption and implementation are significantly influenced by the nature of the intervention. Indeed, factors facilitating or hindering the implementation of an electronic health record system likely differ from those affecting an integrated alert system. As illustrated in one study [25], most facilitators and barriers identified during the initial implementation of Electronic Health Record Systems relate more to the broader concept of health care digitization than to the specifics of an individual digital intervention [38]. Moreover, the scope of digital interventions examined in the included studies was limited, often excluding technologies such as automated laboratory or radiology reports, communication systems, and cloud services. Furthermore, the existing literature tends to inadequately detail the implementation strategies for these digital interventions, with a focus that shifts from the design process to the achieved outcomes [28].

Despite these variations, there are common threads, such as the critical importance of adopter engagement and training. However, it is essential to note the inherent limitations of the TDF and NASSS frameworks used in these studies. The TDF, while comprehensive, can be complex and cumbersome to apply, potentially overlooking the dynamic and context-specific nature of behavior. The NASSS framework may oversimplify the system-level interactions and changes over time, potentially failing to capture the nuanced and evolving nature of technology implementation in health care settings. Our use of the NASSS framework does not incorporate the associated complexity categorization tools, which would offer a useful scope for future work. Recognizing these limitations is crucial in understanding the context and implications of the review’s findings and in guiding future research and implementation strategies.

Further research to answer several important questions is required. First, the included studies overlooked several domains, including those within the NASSS framework: “the wider system” and “embedding and adaptation over time,” as well as within the TDF: “optimism,” “intentions,” “goals,” “social influences,” “emotion,” and “behavioral regulation.” While not commonly reported in scientific literature unless directly related to study outcomes, these factors are likely critical in the implementation of digital health care interventions and warrant further investigation through qualitative methods that explore these elements to deepen our understanding of the value proposition and other overlooked areas.

Second, academic research and quality improvement in health care often emphasize predetermined protocols and outcomes, neglecting the iterative nature of development and implementation. This could potentially be driven by limited funding opportunities for long-term funding continuing beyond the evaluation phase [4]. Adopting a more flexible approach, similar to start-up methodologies, could enhance learning and integration within health care settings. This approach would include a focus on adopter customer discovery, rapid hypothesis testing, iterative design, and continuous maintenance of digital health solutions, important when integrating novel digital solutions in complex organizations [39,40]. Use of an implementation model, for example, the exploration, preparation, implementation, sustainment framework [41], which describes 4 well-defined phases that align with and guide the implementation process, may offer a useful next step for digital innovators in the implementation stage. Another important consideration beyond implementation is considering service design to help teams execute more successful digital health solution adoption projects [40].

Policy makers should integrate human factors and behavioral evaluations into the implementation of digital solutions to modify workflows effectively. Such integration is crucial for comprehending the financial and technical prerequisites for successful adoption. Collaboration among all key stakeholders, including patients, health care professionals, policy makers, and industry, is essential.

In addition, there should be a strategic allocation of resources, encompassing dedicated training and support for health care staff. This approach is vital to ensure that variations in digital literacy are considered, aiming to cultivate a digitally proficient workforce. NHS England’s commitment of GDP £200 (USD $265) million in funding, with an additional GDP £250 (USD $331) million in 2023 and 2024, underscores the importance of these initiatives [42,43]. It is imperative that these funds are strategically allocated to address the issues highlighted, fostering a more effective and responsive digital health care system.

Several limitations must be acknowledged in interpreting our findings. Considerable heterogeneity was observed among included studies regarding intervention types, settings, and methodologies. This limits direct comparisons and generalizability of specific barriers and facilitators across different technological interventions or clinical contexts. This study was limited to scientific articles and did not include white papers, governmental or organizational reports, or blog posts; these articles could be considered in the future and may provide details relating to NASSS and TDF domains, particularly in the development process. Furthermore, limited qualitative data available restricted a deeper exploration into psychological or social influences on adopter behavior. Finally, this review highlighted the paucity of published literature as multiple factors within the frameworks were devoid of any literature; as a result, this affected the number of meaningful conclusions that could be drawn, and our findings show the starting points when it comes to implementing digital solutions in health care.

Despite these limitations, this review has several key strengths. By integrating 2 established behavioral frameworks, NASSS and TDF, we provided a comprehensive characterization of behaviors influencing digital health adoption that complements existing literature. Unlike previous reviews that primarily focused on clinical or organizational factors alone [7], our approach incorporates adopter-level psychological determinants, such as beliefs about capabilities and consequences, memory processes, attention demands, and environmental contexts. Finally, by explicitly identifying gaps in current reporting practices regarding overlooked behavioral domains (eg, optimism or behavioral regulation), we offer clear directions for future research to deepen understanding of human behaviors critical for successful digital intervention implementation across health care settings.

In conclusion, our systematic review identified critical behavioral domains influencing digital health intervention adoption across diverse health care settings using the NASSS framework and TDF. Key facilitators include intuitive technology design aligned with existing workflows, clear communication of value propositions to adopters, adequate provision of training resources tailored to adopters’ knowledge levels, and ensuring organizational readiness for technological change. Conversely, significant barriers include disruptions to clinical workflow, inadequate adopter training or confidence levels, unclear value propositions leading to disengagement, insufficient consideration of cognitive load impacts, such as alert fatigue, and limited organizational preparedness. Stakeholders aiming for successful implementation must strategically address these identified behavioral factors through targeted interventions such as user-centered design processes, tailored educational programs for clinicians with varying digital literacy levels, proactive management of cognitive burdens associated with new technologies, and fostering organizational cultures supportive of technological innovation. Recognizing and addressing factors that are not commonly reported in academic literature is vital for developing robust frameworks that can guide the successful implementation of digital interventions in future health care settings.