Using Journey Mapping and Service Blueprinting to Design Digital Health Behavior Change Innovations: A Scoping Review

Abstract

Introduction:

Solutions to support disease self-management and health-related behavior changes require a deep understanding of patient experiences, needs, and challenges across the care journey. Journey mapping (JM) and service blueprinting (SB) are valuable tools for visualizing user experiences and system processes over time. This scoping review explores how JM/SBs have been applied to design digitally enabled interventions targeting health-related behaviors among patients and the public.

Methods:

The JBI reviewer manual was used to guide the review. Studies were sourced from Embase, Psych Info, PubMed, Medline, Web of Science, and Scopus. Inclusion criteria required studies to describe how JM/SBs informed the design of a digitally enabled health innovation that aimed to impact health or health care-related behaviors of patients or the public. A two-level screening process and iterative data extraction were applied.

Results:

A total of 28 studies met the inclusion criteria, with a majority published between 2019 and 2024. The JM/SBs rarely used behavioral science theory and were structured, organized, and presented in diverse ways. Most studies designed their digital health behavior change innovations by using JM/SB to identify relevant innovation touchpoints across the patient journey. Patients frequently participated in the digital health behavior change innovation design process, with JM/SBs often serving as sensemaking tools. Innovations tended to address multifaceted health service problems through multimodal, digitally enabled solutions.

Conclusions:

JM/SBs are emerging as versatile tools to help digital health innovations to conduct user research, engage diverse partners, identify complex problems, and ideate creative solutions. However, limited integration of behavioral science theory indicates an area for future exploration.

Introduction

Over the past decade, the academic literature has seen a substantial increase in studies focused on the development of digital health behavior change interventions. ¹ These innovations are generally designed with behavioral science theories, models, and frameworks, and follow a design process that typically includes: (a) empathizing with users (including understanding the behavior change context), (b) defining key problems and opportunities (including identifying relevant behavioral barriers and facilitators), (c) ideating potential solutions (including creating relevant behavior change strategies), and (d) prototyping and testing the digital behavior change innovation. ² As outlined by Klonoff (2019), the use of behavioral science theory is an important factor for designing any digital health tool that aims to increase patient adherence to treatment, especially for chronic diseases such as diabetes. ³

In recent years, digital health behavior change intervention developers have been adopting tools and methods from the product design and service design fields in order to enhance the usability, accessibility, and effectiveness of their interventions in real-world settings.^4-7 Specifically, journey mapping (JM)^8-10 and service blueprinting (SB)^11-14 have emerged as powerful tools that help to clarify user experiences and system needs. These tools are outlined and defined in Figure 1.

Figure 1.

Journey mapping and service blueprinting.

The JM can be traced back to 1981 when Jan Carlzon developed a framework called moment mapping to identify the stages of a customer’s buying process. Today, JMs are used to map users’ experiences, tracing their journey across different stages, highlighting pain points, and identifying opportunities for tailored support.^8-10 Meanwhile, SB was introduced by G. Lynn Shostack in 1984 in the Harvard Business Review. Today, SBs are used to examine the operational side of interventions, detailing how diverse systems and processes support user activities.^11-14

Over the past several decades, a wide range of methods have been employed to describe patient journeys and develop JMs and SBs. As highlighted by Sijm-Eeken et al ¹⁵ and Davies et al, ¹⁰ there is no standardized approach for patient JM, and they vary significantly in quality and structure. Joseph et al ¹⁶ categorized JM techniques into five distinct but interrelated visual approaches, each serving different analytical and practical functions to examine patient experiences and system interactions.

Given the diversity in JM/SB conceptualization, this article will conceptualize JM/SB structures using a row-and-column format, which will allow for systematic comparisons across studies. Analyzing the presence, content, and organization of rows and columns will provide a structured yet flexible foundation for cross-comparison, facilitating clearer insights into how these mapping approaches are being operationalized in the literature.

Using a row-and-column conceptualization, JMs and SBs tend to share a common structural logic to systematically capture different aspects of a user journey. ¹⁷ The JM/SB columns usually depict longitudinal phases that a service user experiences over time. For example, for a newly diagnosed pre-diabetic, a JM might include columns such as awareness of the issue, consideration of treatment options, decisions about next steps, accessing treatment, and health self-management. The JM columns might also depict user actions around a specific point in time, such as an appointment visit. In this case, columns might include: before visit, arriving to facility, during appointment, departing from facility, after visit.

On the contrary, JM/SB rows depict different aspects or perspectives of a user’s experience. In JMs, these rows typically involve mapping out users’ actions, thoughts, emotions, pain points, and touchpoints throughout the phases of their journey. In SBs, rows depict a systems perspective to the user experience, and usually involve mapping out users’ actions (ie, actions performed by the patient, such as scheduling a follow up appointment), frontstage actions (ie, actions performed by service providers who directly interact with the user, such as a doctor explaining treatment options), backstage actions (ie, actions performed behind the scenes that aren’t visible to the user, such as doctors coordinating referrals to dieticians), and support processes (ie, processes that allow the service experience to function, such as automated patient notification systems).

Overall, the general column-and-row structure of JM/SBs can help innovation design teams organize key information systematically, with columns representing stages the user progresses through and rows providing different layers of detail about actions, emotions, touchpoints, and supporting processes. For designing digital behavior change innovations, this structure can help break down complex interactions into manageable components, allowing design teams to identify pain points and opportunities for user-centered behavioral interventions.

Despite the potential of using JM and SB to bring behavior change innovation design to a new era, to our knowledge, no review of the literature has been conducted to map how JM and SB have been used to inform digital health behavior change innovation development for patients and the public. Because of the lack of understanding of how these tools have been applied in this context, a scoping review would be beneficial to explore unique contributions, identify gaps in the literature, and highlight opportunities for future research and practice. A scoping review on JM/SB in the context of digital health behavior change innovation development could serve as a foundation for the development of more user-centered innovations in the future.

Aims and Objectives

The overall aim of this scoping review is to map how studies have used JM and SB to develop digital health behavior change innovations. Addressing this aim will provide valuable insights for readers designing digital health solutions aimed at supporting patients in achieving behavior changes, particularly in contexts like diabetes care. To do so, this review will answer the following questions outlined in Table 1.

Table 1.

Main Research Questions.

Question	Description
Question 1	What are the main characteristics of studies that have used JM/SB to design digital health behavior change interventions?
Question 2	How do studies use JM/SB to design digitally enabled behavior change interventions? a. How did studies build their JM/SB? b. How did studies use their JM/SB to innovate?
Question 3	How do studies integrate behavioral science theories, models, and frameworks with their JM/SB to design digitally enabled behavior change interventions?

Methods

Study Design

The JBI reviewer’s manual guided the conduct of this scoping review, with some adaptations made to align with the methods of rapid scoping reviews.^18-20 Adaptations consistent with rapid scoping review methods included expediting the review process (completed in approximately four months), restricting the search scope (eg, limiting language, article type, exclusion of gray literature, and use of narrow search terms), and streamlining the screening process (duplicate screening was conducted only on a pilot subset of sources during level 1 and level 2 screening).^19,20 The reporting of this scoping review follows the PRISMA-ScR (Preferred Reporting Items for Systematic Reviews and Meta-Analyses Extension for Scoping Reviews). ²¹ A scoping review protocol was drafted internally with no deviations noted.

Search Strategy

The search strategy was developed by the lead author and included searches in the databases: Embase, PsychInfo, PubMed, Medline, Web of Science, and Scopus on September 20, 2024. An example of a search is summarized in Table 2. The search terms directly relate to the review objectives to understand how JM/SBs are used to design digital health behavior change innovations. A review of gray literature sources was outside of the scope of this paper, which aimed to focus on the current state of the academic literature.

Table 2.

Search Term Example.

Strategy	Description
Terms	((“service blueprint*” or “journey map*”) and (design* or implement* or innovat* or develop* or creat* or solution*) and (health* or patient* or care* or wellness or wellbeing))
Limits	Language: English Type: Article or Review

Eligibility Criteria

To be included in the study, evidence sources had to clearly describe how JM/SB informed the design of a digitally enabled health innovation that aimed to impact health or health care-related behaviors of patients or the public. If it was not clearly explained how the JM/SB is leading to the development of design features, content, innovation, and improvements for a digitally enabled solution, it was excluded. Reviewers followed a detailed set of inclusion-exclusion criteria outlined in Covidence reference management software (Veritas Health Innovation Ltd) that gave directions related to the study characteristics, the target population, the intervention, and the context. These inclusion criteria are summarized in Table 3.

Table 3.

Summary of Inclusion Criteria.

Category	Inclusion criteria
Characteristics	Article Type: Full text, non-protocol, journal article. Language: English language. Restrictions: Any date/any methodology.
Population	Target Population: Patient- or public-focused (ie, the innovation informed by the JM/SB impacts patient- or public-related behaviors).   • NOTE: The paper can be both clinician AND patient/public focused (ie, as long as the JM/SB impacts SOME patient- or public-related behaviors).
Intervention	Innovation Type: JM/SB must be used in the study methodology to inform the development a digitally enabled innovation, implementation, intervention, or improvement that aims to impact patient/public health/health care behaviors.   • NOTE: The innovation does NOT yet have to be developed, but the paper should clearly outline how the JM/SB is informing design.   • NOTE: “Digitally enabled” means that the innovation must involve SOME digital features, even if it is mainly a non-digital innovation (eg, an in-person process improvement).   • NOTE: The goal of the innovation, intervention, or improvement must involve optimizing patient-public health/health care behaviors.
Context	Setting: Health care or health-related (eg, the intervention could involve changing navigation with a health care service, or it could involve improving nutrition or physical activity behaviors in the public)

Evidence Selection

Evidence sources were handled using Covidence reference management software (Veritas Health Innovation Ltd). Duplicate studies were automatically removed, and the remainder of the studies were imported for a 2-level screening process. During level 1 screening, PV pilot-tested the eligibility criteria with two other reviewers (BI and AB), which involved screening with a selection of 22 titles and abstracts. Discrepancies were tracked, and meetings were conducted with PV, BI, and AB to refine the eligibility criteria and reach consensus on inclusion. The remaining title and abstract screening were completed by PV. During level 2 screening, PV screened a selection of 11 full texts with two new reviewers (NL and JW) using the refined eligibility criteria. Discrepancies were tracked, and meetings were conducted to discuss and confirm inclusion. The remainder of the full-text screening was completed by PV. Studies that met the eligibility criteria were included for data extraction.

Data Extraction and Analysis

A data extraction form was iteratively developed using Microsoft Excel (Redmond, Washington). Data were extracted from the included sources by PV. The data collected related to each study objective and are outlined in Table 4.

Table 4.

Data Collected.

Data collection category	Data collected
What were the main study characteristics?	Author, title, year, location, design, main study problem or opportunity, data collection methods, target population, participant population, main health issue.
How did studies build their JM/SBs?	Choice of JM or SB, perspective of JM/SB, justification for JM/SB, who built the JM/SB, methods used to build JM/SB, use of theories, models, or frameworks to build the JM/SB, and structure of JM/SB [including number of rows, type of rows, number of columns, type of columns, and extra features].
How did studies use JM/SBs to innovate?	Methods used to innovate, who led the innovation, structure of how JM/SB led to innovation, the nature of the innovation, evaluation of the JM/SB innovation process, and reflections on JM/SB innovation process.
How did studies use behavioral science theories, models, and frameworks (TMFs)?	Types of behavioral TMFs used to understand the problem, types of behavioral TMFs used to diagnose behavioral barriers and enablers, and types of behavioral TMFs used to select active behavior change strategies.* *Note, the presence of behavioral science TMFs is assessed by PV, who is a behavioral scientist.

A critical appraisal of the evidence sources was not conducted, as this is not typical in scoping reviews. A descriptive analysis of the included studies was conducted by PV. In order to clearly and meaningfully describe the data, PV inductively created categoric labels and visual summaries to describe and clarify key concepts. ¹⁸

Results

Selection of Studies

Of the 911 evidence sources identified, 651 duplicates were removed, allowing for 260 sources to be screened based on their titles and abstracts. A total of 112 sources were deemed relevant for full-text review, and 28 papers were included in the final review. Figure 2 shows the PRISMA-ScR flow diagram for the evidence selection process.

Figure 2.

PRISMA-ScR flow diagram.

Main Study Characteristics

Key characteristics of the included studies are summarized below and in Appendix A.

Dates and locations of the included studies

All 28 studies except 2^22,23 were published between 2019 and 2024, with 6 between 2019 and 2020,^24-29 7 between 2021 and 2022,^30-36 and 13 between 2023 and 2024.^37-49 In total, 26 different journals were used, with the only repeated journals being the International Journal of Environmental Research in Public Health and BMC Medical Informatics and Decision Making. Studies took place in 13 different countries, including the United States of America (5/28),^{23,24,26,40,49} Australia (4/28),^31,37,41,48 Canada (4/28),^27,35,36,43 Korea (3/28),^38,44,45 Argentina (1/28), ³³ Brazil (1/28), ³⁹ China (1/28), ³⁴ Fiji (1/28), ⁴⁷ Ireland (1/28), ²⁸ Spain (2/28), ²⁵ Sweden (1/28), ²² Thailand (1/28), ³⁰ and the United Kingdom (1/28). ⁴⁶ Two studies^29,32 had an international focus with one of these studies focusing on low- and middle-income countries. ²⁹

Study designs of the included studies

Different study designs were adopted to explore the use of JM/SB in designing digital health behavior change innovations, with the most common designs being co-design/co-creation (7/28)^{25,31,35,37,42,43,48} and qualitative research (9/28).^{22,27,31-33,36,41,44,47} The remainder of studies described their study designs as multi-methods/mixed methods (4/28),^24,26,39,45 case study (3/28),^28,30,46 SB (2/28),^34,38 narrative review (1/28), ²⁹ quality study (1/28), ²³ design thinking study (1/28), ⁴⁰ iterative design study (1/28). ⁴⁹

Participant involvement within the included studies

Most studies involved the primary patient target population of the digital health behavior change intervention as participants (24/28). The number of participants involved from the target population varied significantly, with 4 studies involving 0 patients,^23,29,34,46 4 studies involving 1 to 10 patients,^32,36,37,40 7 studies involving 10 to 20 patients,^{22,33,38,42,44,47,48} 5 studies involving 20 to 30 patients,^{25,31,41,43,49} 5 studies involving 30 to 100 patients,^26-28,30,35 and 3 studies involving an unclear amount.^24,39,45 Among the studies that involved the primary patient target population, a minority (7/24) only involved the primary patient target population, while a majority (17/24) involved multiple types of participants, including health care professionals (eg, managers, doctors, nurses, assistants, administrators, pharmacists, front desk staff, physiotherapists, nutritionists, community health agents, public health representatives, medical topic experts, social support advisors, health education experts), caregivers, care companions, family members, community leaders, and digital design experts. The studies that did not involve the primary patient target population only involved health care professionals (2/28) or did not do primary research (ie, a narrative review and a hypothetical case scenario) (2/28).

Target population of the digital health innovations

The types of target populations that the digital health behavior change interventions were designed for were diverse in nature. Studies’ innovations were designed for: (1) populations with specific health-related conditions (ie, patients with cancer, dental disease, heart disease, COVID-19, opioid addiction, speed disorder, cognitive impairment, and generalized non-communicable chronic disease), (2) populations who needed specific health services support (eg, patients needing surgery, patients accessing lab results, patients transitioning to home, patients requiring medication, patients needing outpatient services, and post-discharge stroke patients), and (3) populations who represented a demographic in need of health-related support (eg, families with infants diagnosed with severe combined immune deficiency, youth wellbeing support, elderly care support, caregivers in need of support, expectant mothers in need of support, and support for those using Chinese medicine).

Main problem or opportunity the digital health innovation addressed

Although all studies designed an innovation that required some behavioral action, not all studies described the main problem or opportunity they were solving as behavioral in nature. Main problems or opportunities spanned two general categories: (1) behavioral problems or opportunities (eg, adequate treatment preparation, self-management strategies, treatment adherence, consumer decision-making, wellbeing behavior changes) and (2) user experience problems or opportunities (eg, more supportive care, service accessibility, experience predictability, informational needs, systems navigation, and treatment process efficiency).

How Did Studies Build Their Journey Mapping/Service Blueprinting?

Choice of journey mapping or service blueprinting?

Most studies employed JM (25/28), with SB being the less common technique (3/28).

Justification for journey mapping/service blueprinting?

Studies gave a diverse range of justifications for using JM/SB, which generally included: (1) understanding user perspectives; (2) visualizing complex processes; (3) identifying gaps, barriers, and enablers; (4) defining and prioritizing key problems or opportunities; (5) informing ideation of target solutions; (6) informing system optimization; (7) enhancing communication and collaboration; (8) supporting iterative design; and (9) educating relevant partners.

Methods used to build journey mapping/service blueprinting?

Most studies (22/28) constructed their JM/SB using researcher-led synthesis methods. In these studies, researchers used the data they collected during their research process to construct the JM/SB. A minority of studies (6/28) constructed their JM/SB using participatory methods, where patients and/or health care providers used their own personal experiences to build the JM/SB themselves. Table 5 provides a detailed breakdown of the different methods studies used to build their JM/SB, with several studies using multiple methods.

Table 5.

Methods Used to Build the JM/SB.

Method used to build JM/SB	References	Count
Patient-Led JM/SB Design
Built by Patients (At Patients and Provider Workshop)	25,37,43	3
Built by Patients (At Home)	25,26	2
Built by Patients (At Patients-Only Workshop)	31,42	2
Health care Provider-Led JM/SB Design
Built by Providers (In a Focus Group)	25	1
Researcher-Led JM/SB Design
Built using Interview Data (Patients and Providers)	24,33,38,44,47,48	6
Built using Observation Data (Patients and Providers)	24,33,35,38,39,44	6
Built using a Predefined Theory, Model or Framework	22,29,34,38,46,47	6
Built using Literature Review	29,39,40,45	4
Built using Patient Feedback and Validation	22,25,26,30	4
Built using Interview Data (Provider Only)	28,34,45,49	4
Built using Interview Data (Patients Only)	26,27,41	3
Built using Survey Data (Provider Only)	23,34,39	3
Built using Informal Discussions with Providers	30,39	2
Built using Provider Feedback and Validation	28,44	2
Built using Researcher’s Own Expertise	23,40	2
Built using Focus Group Data (Patients Only)	22	1
Built using Focus Group Data (Provider Only)	25	1
Built using Observation Data (Patients Only)	28	1
Built using Observation Data (Provider Only)	34	1
Built using Video Analysis Data	32	1
Built using a Patient Case Data	36	1
Built using a Systems Review Data	38	1
Built using an Application Design Review Data	45	1
Built using a User Personas	45	1
Built using a Document Review Data	28	1
Built using a Hypothetical Scenario (Patients Only)	46	1
Built using Conversation Circles (Provider and Patients)	39	1
Built using a Listening Labs (Patients Only)	25	1

Use of theories, models, or frameworks to build journey mapping/service blueprinting?

As shown in Table 5, a minority of studies used a predefined theory, model, or framework to build their JM/SB (6/28). The theories, models, or frameworks that were used to support the development of JM/SBs were Healthcare Access Framework, ⁵⁰ the Customer Purchase Decision Making Model, ⁵¹ the McKinsey Customer Decision Journey Model, ⁵² the “RACE” model, ⁵³ the Theory of Inventive Problem Solving,^54,55 the Nurturing Care Framework for Early Childhood Development, ⁵⁶ the Chronic Illness Trajectory Framework, ⁵⁷ and the Donabedian Framework of Structure-Process-Outcome (SPO). ⁵⁸

Formats of journey mapping/service blueprinting

The formats of the JM/SBs were extremely diverse, with no JM/SB repeating the same column and row arrangement. Figure 3 summarizes these JM/SB formats, and Appendix 1 lists the different column and row categories used by each of the included studies.

Figure 3.

Structures of JM/SBs.

The number of columns (ie, the journey phases) ranged from 0 to 10. The JM/SB columns could generally be described by (1) their structure and (2) their categorization. The column structures included three main types: (a) structured (where the columns were clearly defined from one another), (b) unstructured (where the columns overlapped and were not clearly differentiated), and (c) adaptable (where the columns shifted from before vs. after the innovation was proposed). Most of the studies used structured columns (22/28), while the remainder of studies used unstructured columns (3/28)^22,34,47 or adaptable columns (3/28).^23,38,49 Column categories were extremely diverse but could generally be classified into three main groups: (a) patient-centered categories (where the phases reflected thoughts or behaviors experienced by the patient), (b) service-related categories (where the phases reflected operational flows from the organization perspective), and (c) transformation-related (where the phases represented a transformation from an undesired state to a desired one).

The number of rows (ie, levels of an experience or system) ranged from 0 to 9. The JM/SB rows could also generally be described by (1) their structure and (2) their categorization. The row structures similarly included three main types (a) structured, (b) unstructured, and (c) adaptable. The majority of studies used structured rows (21/28), while the remainder of studies used unstructured rows (5/28)^{22,24,41,45,47} and adaptable rows (2/28).^38,49 Row categories were also extremely diverse but could generally be classified into four main groups: (a) patient-centered rows (where the rows represented patients’ physical, emotions, cognitive, and behavioral journeys), (b) interaction-related rows (where the rows represented some sort of touchpoint between a patient and a person or service), (c) operational-related rows (where the rows represented system inputs or operations), and (d) SB-related rows (where the rows directly followed the guidance for rows from traditional SB).

There were several unique features from many of the JM/SB formats across the studies, but some main notable features were the addition of multiple personas within a JM/SB, additional labeling or color coding within the JM/SB, and connecting JM/SBs to defined key performance indictors or desired outcomes.

How Did Studies Use Their Journey Mapping/Service Blueprinting to Innovate?

Types of journey mapping/service blueprinting innovation approaches?

Studies used their JM/SB to create digitally enabled health innovations using many different methods. Overall, studies used eight major ways to transition from JM/SB to innovation creation. Figure 4 outlines the different types of approaches studies used to facilitate this transition. Appendix 1 describes the approach taken by each included study. Although some studies integrated multiple types of approaches, Type #3: Inspiration-Based Innovation (9/28)^{22,25,30-33,37,40,48} and Type #1: Phased Based Innovation (9/28)^{22,27-29,31,36,44,46,47} were the most common approaches. Type #6: Multi-User Innovation Mapping (4/28),^34,35,41,47 Type #8: Before vs. After Innovation Embedding (4/28),^23,34,38,49 Type #2: Innovation Content Mapping (3/28),^25,26,42 Type #4: Innovation Prioritization Mapping (3/28),^24,43,44 Type #5: Level Based Innovation Mapping (1/28), ³⁹ and Type #7: Innovation Structure Mapping (1/28) ⁴⁵ were also utilized.

Figure 4.

Different ways of using JM/SB to innovate.

Who was involved in the transition from journey mapping/service blueprinting to innovation?

In half of studies (14/28),^{25,26,30,31,35,37,40-43,45,47-49} patients were directly involved in the digital health behavior change innovation design and improvement process. In the other half of studies (14/28),^{22-24,27-29,32-34,36,38,39,44,46} the study team led the process of creating innovations using previously collected data, but without the direct involvement of patients. Patients were involved in the innovation development in diverse ways, including commenting on innovation ideas in an advisory group, recommending innovations in interviews or focus groups, prioritizing core needs based on their ability to be assisted by technology, brainstorming solutions to core needs from the JM/SB, co-creating the innovation recommendations, iteratively co-designing the final digital health innovation, evaluating innovation ideas based on feasibility, impact and relevance; prioritizing solutions using voting, based on aesthetic, flow, and content, and giving feedback during usability testing that led to improvements in the innovation.

What was the nature of the innovations?

Studies designed either (1) fully digital innovations (ie, innovations that were one digital product or multiple digital solutions) or (2) digitally enabled innovations (ie, innovations that included digital components but involved other multi-level and multi-model inputs). Of the fully digital innovations (12/28), the majority (9/12) described the development of one digital product innovation,^{25,31,33,36,42,43,45,48,49} while the minority (3/12) described multiple digital product innovations.^22,32,46 Singular digital products included digital health apps,^31,45,49 patient portal innovations,^33,36 an online course, ⁴² a web based self-management program, ⁴³ a virtual community of practice, ²⁵ and an eHealth innovation. ⁴⁸

Of the digitally enabled innovations (16/28),^{23,24,26-30,34,35,37-41,44,47} all innovations involved multiple inputs and actions, both digital and non-digital. The digital innovations proposed in these studies were numerous and diverse, and included innovations such as websites, apps, text messages, patient portal innovations, dashboards, automated alerts, virtual care coordinators, online forms, communications platforms, online guidelines, remote monitoring systems, novel algorithms for identification, online scheduling, educational videos, online games, online questionnaires, digital navigators, telemedicine, online support groups, and more. Only 1/28 of the studies suggested innovations involving AI or ML. ³²

Was the journey mapping/service blueprinting process evaluated?

Only two studies evaluated the JM/SB-driven innovation process itself.^28,42 In Álvarez-Pérez et al, ⁴² participants involved in the co-creation process gave feedback on their overall experience being involved in this co-design. Participants felt their experiences were incorporated into the innovation. Emotionally, participants felt they were a part of something meaningful and liked expressing and sharing experiences with others going through similar situations. Face-to-face sessions were preferred.

How Did Studies Use Behavioral Science Theories, Models, and Frameworks?

As assessed by PV, only two studies appeared to use a theory, model, or framework from behavioral science to support the design of their digital health behavior change innovation.^43,46 Howell et al ⁴³ used Social Cognitive Learning Theory, ⁵⁹ which is a theory grounded in behavioral science, representing an evolution of traditional behavioral theories by including cognitive and social dimensions. Howell et al used social cognitive learning theory alongside data from their qualitative interviews, focus groups, personas, and JM to ideate and design their digital health platform, I-Can Manage. Specifically, they describe that, in their ideation phase, the I-Can Manage program was theoretically underpinned by social cognitive learning theory and the construct of self-efficacy. The authors achieved this through incorporating action-oriented information and behavioral exercises as a core feature of the I-Can Manage program to promote self-management. Their program featured four sources of self-efficacy information including mastery learning, vicarious experiences, social persuasion, and emotional and physiological control. Modha ⁴⁶ used McKinsey’s Consumer Decision Journey Model, ⁵² which has connections to behavioral science due to its interpretations of consumer behavior but is not explicitly framed as a behavioral science model. This cyclical model outlines five stages of consumer decision-making behavior including (1) initial consideration; (2) active evaluation; (3) moment of purchase; (4) post-purchase experience; and (5) loyalty loop. In contrast to Howell et al’s ⁴³ use of Social Cognitive Learning Theory for ideation and operationalization of their digital health program alongside their JM activity, Modha et al ⁴⁶ used McKinsey’s Consumer Decision Journey Model to build their patient journey map, as outlined in Table 5.

Discussion

This scoping review mapped how studies used JM and SB to develop digital health behavior change innovations. Specifically, this review mapped the main characteristics of these studies, the methods studies used to build their JM/SB, how studies used their JM/SB to innovate, and how studies integrated behavioral science theory models, and frameworks.

Overall, 14 key takeaways were elicited from the results of this scoping review, which have been outlined in Table 6.

Table 6.

Key Takeaways.

Key takeaway	Takeaway description	Implications for future work
1. Using JM/SB for innovation is an emerging trend.	The results suggest that there is a growing use of JM/SB for digital health innovation.	Given the emerging nature of JM/SB use, researchers should prioritize continual evaluations on this topic.
2. JM/SB supports innovation for a range of populations and locations.	The results suggest that JM/SB is adaptable to support various populations, locations, and methodologies.	Given the adaptability of JM/SB, researchers should review how JM/SB could be used to support innovation for equity-deserving groups. Researchers could also explore culturally, linguistically, and contextually tailored JM/SBs to enhance inclusivity, accessibility, and global applicability.
3. JM/SB shifts the focus of innovations on multi-user needs and systems issues.	The results suggest that JM/SB can be helpful for broadening the scope of innovation design, focusing on end-to-end user experiences, and integrating innovations for multiple users and systems.	Given a broadened focus, JM/SB can help teams develop more effective and engaging solutions that integrate insights from design thinking, service, design, and behavioral science. Future studies could examine how JM/SB helps innovators integrate feedback loops from users at different stages to refine the user experience continually.
4. JM is a more popular tool than SB.	The results suggest that JM is a commonly used tool than SB for digital health behavior change design. Despite this, teams included diverse rows in their JM that often mimicked common SB rows.	Future studies might benefit from contributing to the literature by providing clearer definitions, educational resources, and describing the differentiating and complementary uses of JM and SB. Researchers may also consider whether distinguishing the “front stage,” where users interact with services, from the “backstage,” where providers and system-level processes operate, is helpful to ensure alignment between user experiences and the workflows that support them.60,61
5. JM/SB is useful for supporting tasks across the entire design process.	The results suggest that JM/SB can be used throughout the entire digital health behavior change design process, including empathizing with users, defining problems, ideating solutions, and prototyping and testing innovations.	Future studies might benefit from contributing to the literature by developing frameworks and methodologies that clarify the strategic application of JM/SB at different stages of the design process.
6. There is a lack of use of behavioral science theory within JM/SBs.	The results suggest that JM/SBs are often not grounded in behavioral science theories, models, and frameworks.	Future studies should consider how behavioral science theories, models, and frameworks could support the use of JM/SB to design digital health behavior change innovations. Researchers should investigate the barriers preventing the integration of theory into JM/SB and provide solutions to address them. Established tools like COM-B 62 or the BCT Taxonomy 63 may provide a valuable starting point for researchers exploring how to better integrate behavioral insights with JM/SB.
7. Patient partners are involved in JM/SB construction and usage.	The results suggest that both research teams and patient partners are actively involved in JM/SB construction and usage.	Future studies should consider evaluating different participatory approaches in JM/SB construction and use for innovation design. Specifically, reflection about how JM/SB can be used as “boundary objects”, to diverse partners together for creative collaboration would be adventageous.64-67
8. There is extreme diversity in JM/SB formats.	The results suggest that there is vast diversity in how studies are constructing their JM/SBs.	Although the range in structure promotes adaptability, it also challenges cross-study comparisons. Future studies should review best practices for building JM/SB (as suggested in Figure 3). Expert panels could be convened to develop a standardized taxonomy for JM/SB formats.
9. There is extreme diversity in JM/SB innovation approaches.	The results suggest that there is vast diversity in how studies use JM/SBs to create actionable innovations.	Future studies should consider reviewing best practices for translating insights from JM/SBs into actionable innovations using structures (as suggested in Figure 4). Research should focus on creating adaptive toolkits to help teams leverage diverse JM/SB approaches while maintaining strategic alignment with innovation goals.
10. There is a focus on digitally enabled innovation.	The results suggest that JM/SB may promote the development of more digitally enabled innovations, where multimodal, solutions are proposed outside of just digital technology.	Future studies might benefit from exploring how JM/SB guide the development of both digital and non-digital innovations to address complex, health care needs.
11. There is a lack of process evaluation regarding the use of JM/SB.	The results suggest that the process of using JM/SB itself is rarely evaluated.	Future studies should consider evaluating the process of using JM/SB to understand its effectiveness and improve its implementation. Process evaluations could include longitudinal studies to measure the sustainability and scalability of JM/SB-driven solutions over time.
12. Studies tend to use JM/SB for exploration, not innovation.	The results suggest that many studies had to be excluded from this paper because they didn’t clearly describe how their JM/SB was used to create a digitally enabled innovation. Studies that were excluded often used JM/SB for exploratory purposes but lacked clarity in translating insights into actionable innovation.	Future studies that use JM/SB should explicitly detail how insights from JM/SB are transformed into specific, implementable design solutions for digitally enabled health innovations. Improved methodological descriptions could help other studies bridge the gap between exploration and innovation.
13. Studies tend to describe more linear approaches to innovation.	The results suggest that there may be a lack of iterative approaches during innovation development (ie, moving unidirectionally through the design process, rather than going back and forth). Although this may be true, this result may also be due to the nature of academic studies and paper construction.	Future studies should address the constraints that academic traditions impose on iterative JM/SB-based innovation. Exploring creative approaches to reimagining ethics approval processes, funding structures, and journal publication practices could prove valuable.
14. There is a notable absence of AI/ML-based innovation.	The results suggest that there is a notable absence of AI/ML-based innovation, despite the potential for JM/SB-based design to support these types of innovations.	Future studies should explore how AI/ML innovations could leverage JM/SB insights, supporting just-in-time tailoring to user needs.68-70 With AI/ML, researchers should consider how digital health behavior change innovations could achieve greater personalization and scalability, aligning with the health care sector’s shift toward intelligent, data-driven solutions.71,72 Researcher could additionally consider how platforms like ChatGPT could potentially enhance each phase of the JM/SB innovation design process.73,74

The Future of Behavior Change Design for Diabetes Technology

In addition to the key takeaways outlined in Table 6, it is important to note that this scoping review did not find any studies that described how they used JM/SB to create a diabetes-specific digital health behavior change innovation. Although the use of behavioral science theories to design diabetes-specific digital health innovations appears to be increasing, ¹ the lack of JM/SB use highlights a potential area for future study, especially given the multifaceted challenges of diabetes management. Diabetes requires patients to navigate complex daily routines involving blood glucose monitoring, medication adherence, dietary adjustments, and physical activity, all while balancing emotional and social factors. ⁷⁵ The JMs/SBs could be particularly valuable for mapping these interconnected behaviors and identifying systemic pain points, such as gaps in care coordination, barriers to accessing education resources, or emotional burdens like fear of hypoglycemia. For example, a JM might reveal how patients struggle to transition from receiving initial dietary advice to consistently integrating meal planning into their routines, while an SB could map out the backstage processes needed to provide timely support, such as proactive alerts or dietitian follow-ups.

Limitations

This scoping review adhered to evidence-informed guidance; however, several limitations must be acknowledged. Restricting included studies to English-language publications may have excluded valuable research from non–English-speaking regions, limiting the global applicability of the findings. In addition, restricting the search by type of article may have led to the exclusion of potentially relevant sources that were coded incorrectly in the databases used. The omission of gray literature likely excluded practical insights from industry reports and conference proceedings, where real-world applications of JM/SB may be more extensively documented. In addition, only partial duplicate screening was conducted during study selection, and data extraction was performed by a single reviewer, which could introduce bias or errors. These constraints suggest that future reviews should incorporate multilingual searches, include gray literature, and adopt more rigorous screening and extraction processes to ensure comprehensive and reliable results.

Conclusion

This scoping review underscores the growing importance of JM and SB in driving digital health behavior change innovation. The JM and SB emerged as versatile tools for engaging diverse partners, addressing user experience challenges, and supporting co-creation. However, gaps like the underuse of behavioral science theories and lack of standardization highlight areas for growth. Addressing these gaps can enhance the impact of JM/SB in paving the way for more effective, user-centered, and scalable digital health solutions.

Appendix A.

Summary of Included Studies.