We used a descriptive qualitative design guided by the COM-B model. Qualitative data were collected through semistructured interviews focused on identifying barriers and facilitators to tracking symptoms and signs relevant to cardiovascular health (eg, fatigue, chest pain, palpitations, and blood pressure) in the context of cancer treatment and survivorship. Interviews also explored participants’ perspectives on desirable features of digital health solutions to support cardiovascular symptom monitoring, with these features subsequently mapped to COM-B domains.

The target behavior in this study was tracking cardiovascular symptoms and signs between clinical appointments. This behavior involves complementary roles for patients with cancer and medical team participants. Patients engage in tracking using available tools and strategies and report this information during or between visits. The medical team often inquires about cardiovascular health during appointments and may ask patients to monitor specific signs or symptoms between visits to support clinical care discussions and decisions. Because tracking involves actions and expectations on both sides of the clinical relationship, both patients with cancer and medical team participants were interviewed.

The COM-B framework guided both the development of the interview guides and the analysis of qualitative data. The model conceptualizes behavior as the interaction of capability (physical and psychological), opportunity (physical and social), and motivation (reflective and automatic). In this study, the framework was used to classify participant-reported barriers and facilitators related to tracking behaviors.

Medical team participants (n=12) and patients with cancer (n=12) who were currently undergoing or had previously undergone cardiotoxic anticancer therapy as defined by established cardio-oncology guidelines [12] (eg, tyrosine kinase inhibitors, chemotherapy, immune checkpoint inhibitors, and radiation) participated in one-on-one semistructured interviews to discuss their experiences and perspectives related to tracking symptoms and signs relevant to cardiovascular health. Eligible medical team participants were (1) aged 18 years or older, (2) English-speaking, (3) currently practicing as a licensed health care professional or allied health professional involved in cancer or cardio-oncology care (eg, physician, advanced practice provider, nurse, or allied health discipline such as physical therapy, exercise physiology, or nutrition), (4) employed within an oncology or cardio-oncology care setting, and (5) involved in the care of patients who were receiving or had received cardiotoxic anticancer therapy.

Eligible patients were (1) aged 18 years or older; (2) English-speaking; (3) diagnosed with cancer; (4) currently undergoing or had previously undergone a cancer regimen that includes a targeted, biologic, radiation, or immune-based therapy (eg, tyrosine kinase inhibitors or immune checkpoint inhibitors); (5) able to engage in physical activity (self-report); and (6) interested in symptom or sign tracking or increasing physical activity levels to improve health (self-report). Although physical activity was discussed during interviews, this analysis focuses on tracking symptoms and signs relevant to cardiovascular health; physical activity–related findings are reported separately.

Medical team participants were recruited through word of mouth and professional referrals. Potential patients were recruited via medical team participant-based referrals within the cardio-oncology clinic. Medical team participants introduced the study to eligible patients during routine medical appointments. If patients expressed interest, members of the research team contacted them via telephone or email to provide additional study information and confirm eligibility. The telephone was the primary mode of contact unless the participant preferred email. If a potential participant could not be reached by phone, a voicemail was left. Research staff followed a 3-call limit protocol when attempting contact. Eligible participants who provided informed consent were enrolled and scheduled for a semistructured interview with a trained member or members of the research team.

Of the 32 patients with cancer screened through the EMR, 7 were ineligible. The remaining 25 preliminarily eligible patients were approached by the clinical study team during medical appointments in the cardio-oncology clinic and asked whether they were interested in speaking with the research study team. The research team attempted to contact all 25 patients by telephone or email. Of the total, 5 patients were unreachable despite repeated contact attempts, and 4 were not contacted after the recruitment target had been met. Moreover, 16 patients were successfully screened by the research team; of these, 2 were ineligible and 2 declined to participate. In total, 12 patients with cancer were enrolled and completed interviews.

Semistructured interviews were conducted remotely via videoconferencing (eg, Zoom [Zoom Communications] and Microsoft Teams) between December 12, 2023, and July 29, 2024. Interviews were facilitated by the lead author (RMB), one of the research team members (MKT), or both, who were trained in qualitative interviewing techniques and followed a semistructured guide designed to promote open discussion.

The interview guides were developed based on the COM-B model domains, including motivation, opportunity, and capability [11]. Open-ended questions were structured to elicit participants’ perspectives related to tracking symptoms and signs relevant to cardiovascular health. Patients were asked about their experiences tracking their own symptoms and signs, whereas medical team participants were asked about their perspectives on patients’ tracking behaviors. Topics included current tracking practices, contextual factors influencing tracking (eg, available tools and time constraints), motivational factors, and perceptions of potential digital tools to support tracking. Example prompts included: “How do you currently keep track of your own/your patients’ signs and symptoms?” (capability), “Can you describe any opportunities or external factors that help you track signs and symptoms?” (opportunity), and “Can you describe your reasons for tracking signs and symptoms?” (motivation). The full interview guides are provided in Multimedia Appendix 1.

Interviews lasted approximately 60‐90 minutes. All interviews were audio- and video-recorded with the participant’s permission, transcribed using the auto-generated transcript from Zoom or Teams, and reviewed and corrected, when necessary, by a research team member (MKT). Participants were given the option to turn their camera on or off. Participants were assigned unique identifiers not associated with personal identifying information. Participants received an electronic US $25 gift card delivered via email upon completion of the interview.

Sociodemographic information from patients (age, sex, race, ethnicity, education, employment, income, height, and weight) was collected via REDCap surveys. Similar demographic information was collected verbally from medical team participants before interviews, including age, sex, race, ethnicity, job position, and years of experience in their current role.

Demographic and medical characteristics were analyzed using descriptive statistics. Rapid qualitative analysis (RQA) was applied to analyze the one-on-one interviews [13-15]. RQA is a team-based qualitative approach in which research team members review audio recordings and transcripts, summarize key points, and identify patterns across interviews to generate themes.

A total of 5 study team members (RMB, MKT, SMSG, VPA, and AG) were involved in transcript review, data summarization, and were adequately trained in RQA. Research assistants first reviewed and corrected auto-generated transcripts for accuracy. Following this step, a researcher (RMB) reviewed each transcript and interview recording and identified segments corresponding to the COM-B domains and subdomains. Relevant excerpts were grouped according to these domains and further categorized as facilitators or barriers.

Each transcript was then independently reviewed by a second researcher (MKT), who repeated the process while having access to the initial coding for comparison. After double review, identified themes and supporting quotes were entered into a structured database to allow for comparison across participants. The research team met regularly to reconcile interpretations and ensure consistency.

Themes were finalized by examining patterns across participants and grouping similar concepts within COM-B subdomains. Because each subdomain is inherently aligned with a COM-B domain, themes were organized accordingly within capability, opportunity, or motivation and illustrated with representative quotes, with subcomponents (eg, physical or psychological capability; physical or social opportunity; and automatic or reflective motivation) used for classification.

This study was approved by The Ohio State University Institutional Review Board (IRB #2021C0018). All participants provided informed consent before participation. Interviews were audio-recorded with participant permission. Identifying information was removed and replaced with unique identifiers to maintain confidentiality.

Participant demographic and clinical characteristics are summarized in Table 1.

In total, 12 patients with cancer participated in interviews. The mean age was 65.3 (SD 12.6) years, and participants were evenly distributed by sex (6/12, 50% female). Most identified as White and non-Hispanic (10/12, 83.3%). The mean BMI was 26.9 kg/m² (SD 4.53), with 41.7% (5/12) classified as overweight and 33.3% (4/12) classified as obese. The majority had completed a graduate or professional degree (7/12, 58.3%) and were retired or disabled (10/12, 83.3%). The most frequently reported annual household income category was US $100,000-US $199,999, representing 41.7% (5/12) of participants. Cancer diagnoses included acute myeloid leukemia (3/12, 25%), breast cancer (2/12, 16.7%), and a range of other malignancies represented in smaller frequencies. Participants reported receiving multiple cancer-directed therapies, including chemotherapy (6/12, 50%), radiation (5/12, 41.7%), tyrosine kinase inhibitors (4/12, 33.3%), stem cell transplant (4/12, 33.3%), hormonal therapy (3/12, 25%), surgery (3/12, 25%), immune checkpoint inhibitors (1/12, 8.3%), and other therapies (4/12, 33.3%).

A total of 12 medical team members were interviewed. Most participants were female (7/12, 58.3%). The mean age was 37.8 (SD 6.4) years, and participants had been in their current position for a mean of 4.5 (SD 2.78) years. Participants identified as White and non-Hispanic (6/12, 50%), Asian (3/12, 25%), White and Hispanic (1/12, 8.3%), Black or African American (1/12, 8.3%), or multiple races or ethnicities (1/12, 8.3%). Professional roles included cardio-oncologists (4/12, 33.3%), oncologists (3/12, 25%), radiation oncologists (1/12, 8.3%), and allied health professionals, including a nurse practitioner, physical therapist, exercise physiologist, and dietitian (each 1/12, 8.3%).

Themes were organized by COM-B domain and categorized as facilitators or barriers. Table 2 summarizes the themes identified among patients and medical team participants, along with the number of participants who endorsed each theme. Illustrative quotes can be found in Multimedia Appendix 2.

Within the capability domain, 3 facilitator themes and 4 barrier themes were identified among patients, compared with 2 facilitator themes and 1 barrier theme among medical team participants. Opportunity included 3 facilitator themes and 2 barrier themes among patients, and 4 facilitator themes and 3 barrier themes among medical team participants. Within the motivation domain, 4 facilitator themes and 3 barrier themes were identified among patients, while 4 facilitator themes and 2 barrier themes were identified among medical team participants.

Digital health solution features were also described and mapped to COM-B domains. Among patients, 5 motivation-related, 3 opportunity-related, and 3 capability-related features were identified. Among medical team participants, 2 motivation-related features and 1 feature related to opportunity and capability were identified.

Both patients and medical team participants described specific features that could support cardiovascular symptom tracking.

Among medical team participants, the most frequently endorsed feature was reminders, alerts, or triaging capabilities (7/12). Medical team participants described the value of automated prompts to encourage patient logging, as well as alert systems that would highlight abnormal values and help prioritize review. One participant noted:

Another medical team participant emphasized the importance of triaging information efficiently:

Of the 12 medical team participants, 6 highlighted the importance of tracking functionality that allows visualization of trends over time. Medical team participants described wanting both patients and medical team participants to see symptom progression and recovery patterns:

Moreover, 4 medical team participants emphasized the potential benefit of incorporating reward systems or positive reinforcement to promote long-term tracking behaviors. One participant explained:

Finally, 5 medical team participants stressed that any digital solution must be simple and user-friendly, with clear and actionable outputs:

This qualitative study examined barriers and facilitators to cardiovascular symptom and sign tracking from both patient and medical team perspectives, in a cardio-oncology setting, using the COM-B framework. We found that tracking behaviors related to cardiovascular health were shaped by interrelated factors across capability, opportunity, and motivation domains. Patients described variability in knowledge regarding cardiovascular health during and after cancer treatment, including uncertainty about which symptoms or signs to monitor and when to communicate changes. Medical team participants similarly identified gaps in standardized processes for cardiovascular symptom tracking and expressed concerns about time constraints and information burden. At the same time, both groups identified various motivational drivers, including perceived value of monitoring for maintaining cardiovascular health, reassurance, and improved clinical decision-making. Social opportunity, particularly patient–medical team participant communication and family involvement, also emerged as central to supporting tracking behaviors.

These findings suggest that cardiovascular symptom tracking in cancer care is not limited by awareness of cardiotoxicity alone but is influenced by clarity of expectations, workflow integration, and alignment between patient and medical team participant roles.

Randomized trials in oncology have demonstrated that structured electronic symptom monitoring can improve some clinical and PROs. For example, in the PRO-TECT cluster-randomized trial, weekly electronic symptom reporting reduced visits to the emergency department and delayed deterioration in physical function and HRQoL among patients with metastatic cancer [7]. Similarly, the use of an EMR-integrated symptom and needs monitoring in an ambulatory, racially, ethnically, and linguistically diverse oncology population (My Wellness Check [American Heart Association]) has been associated with reductions in emergency department usage and hospitalization [8]. These findings are consistent with broader evidence that PRO monitoring enhances symptom detection and clinical communication, supporting improved system-level outcomes in oncology care [16].

These studies establish the potential clinical impact of systematic symptom monitoring but do not examine the behavioral mechanisms influencing long-term engagement in symptom tracking, particularly for cardiovascular risk in cardio-oncology contexts.

Our findings extend this literature by identifying capability-, opportunity-, and motivation-related determinants influencing the adoption and implementation of cardiovascular symptom tracking. Whereas previous trials primarily evaluate downstream clinical outcomes, our qualitative data illuminate upstream behavioral and contextual factors (including knowledge gaps, workflow constraints, uncertainty about symptom attribution, and perceived role boundaries) that may influence the feasibility of structured cardiovascular monitoring.

Our findings align with the qualitative work on cardiotoxicity-focused mHealth tools, where both patients and medical team participants expressed interest in digital cardiovascular symptom tracking but stressed the importance of clear reporting guidance for urgent symptoms, integration with existing clinical systems, and minimal workflow disruption [10]. Additionally, a recent systematic review of mHealth tools for cardiovascular health in patients with cancer found a fragmented array of technologies and measures, limited integration with electronic health records (EHRs), and few studies specifically addressing structured cardiovascular symptom collection [9]. More broadly, mHealth technologies enable continuous monitoring of both physiological and patient-reported data, but their implementation is often constrained by variability in engagement and limited integration into clinical care workflows [9], aligning with recent implementation-focused reviews demonstrating that the effectiveness of remote symptom monitoring depends on patient engagement, workflow integration, and actionable use of reported data [16].

Together, these studies suggest that although digital monitoring tools are increasingly available, implementation may depend less on technological capability alone and more on alignment with behavioral determinants and system-level structures. This aligns with findings from previous work on EHR-integrated interventions, which highlight that successful implementation depends on alignment with clinical workflows, defined roles, and manageable data flows rather than technology alone [17,18]. By applying the COM-B framework, our study provides a behavioral lens to inform the design and integration of cardiovascular symptom-tracking approaches in oncology settings. Notably, cardiovascular symptom tracking may differ from general oncology symptom monitoring in several respects: the overlap between cardiotoxicity-related and cancer- or treatment-related symptoms, the evolving nature of cardiovascular risk across the cancer care continuum, and the need for coordination across oncology and cardiovascular specialties.

Cardiovascular symptom tracking in cancer care emerged as a shared behavior shaped by capability, opportunity, and motivation among both patients and medical team participants. Patients were often unsure which cardiovascular symptoms or signs should be tracked and when changes should prompt communication, suggesting that education alone may be insufficient without clear expectations for monitoring and reporting. Although patients generally recognized the relevance of cardiovascular health, many described ambiguity about expectations for monitoring and reporting. Medical team participants similarly described variability in how cardiovascular monitoring was addressed and incorporated into routine care. Such variability likely reflects differences in clinical context, patient complexity, and workflow demands, underscoring the importance of clarifying expectations and responsibilities when considering strategies to support consistent tracking practices. These challenges highlight the importance of clearly defining which cardiovascular symptoms should be tracked and when reporting is warranted, particularly in contexts where symptom attribution is uncertain. They also suggest that effective tracking approaches may need to be integrated into existing workflows and support coordination across oncology and cardiovascular care teams, agreeing with broader systematic reviews identifying knowledge gaps, workflow constraints, and variability in perceived value as central barriers to PRO implementation [19].

Opportunity constraints were prominent. Time limitations, competing priorities, and concerns about information burden shape whether tracking can be integrated into routine workflows. Previous implementation research in oncology has shown that monitoring systems are more likely to be adopted when embedded within existing clinical processes, supported by defined triage pathways, and aligned with established roles rather than added as an additional task [7-9]. EMR-integrated monitoring programs that include automated alerts and predefined response workflows have been described as feasible in ambulatory oncology settings when these structural supports are in place [20]. Observational analyses have also reported associations between participation in such programs and health care usage outcomes [8]. More broadly, implementation science literature emphasizes that long-term adoption of digital monitoring systems depends on minimizing disruption to clinical workflows and reducing cognitive and time burden on medical team participants [21], along with recent umbrella reviews highlighting workflow disruption, limited integration with EHRs, and time and resource constraints as key barriers to implementing patient-generated health data systems in oncology care [19]. The findings of this study add to this literature by identifying behavioral factors that may influence uptake of cardiovascular-specific tracking within cardio-oncology settings.

Motivational influences were equally apparent. Patients described reassurance and a sense of control as reasons for engaging in monitoring, whereas uncertainty about symptom meaning sometimes contributed to inconsistent reporting. Medical team participants were more inclined to engage with monitoring data when it was perceived as clinically meaningful but expressed concern when data volume was high or when response expectations were unclear. This aligns with previous real-world implementation studies demonstrating that long-term engagement with electronic PRO systems may vary over time and is influenced by reminder systems, medical team participant reinforcement, perceived relevance to care, and integration into clinical workflows [22]. These findings are also consistent with previous research showing that perceived usefulness and ease of use are central determinants of engagement with digital health and self-monitoring tools in chronic disease management [23]. These perspectives indicate that perceived usefulness and manageable data flow may be important for long-term engagement over time.

Taken together, these findings indicate that knowledge clarity, workflow integration, and perceived value operate in combination. Introducing a digital tool without addressing expectations or integrating tracking into existing care processes may offer limited benefit.

Our findings inform the design of digital tools intended to support cardiovascular symptom tracking in cancer care. For patients, tools may benefit from providing clear guidance on which symptoms and signs to monitor and when changes should prompt communication with the medical team. Educational components embedded within the tracking process may help address gaps in psychological capability while avoiding additional burden during clinical encounters. Tools should also account for the overlap between cardiovascular, cancer-related, and treatment-related symptoms to avoid excessive data collection that increases patient burden or generates information that is not clinically actionable.

Medical team participants emphasized that digital tracking must be integrated into existing EMR systems and supported by clear triage processes. Previous implementation work in oncology indicates that feasibility is strengthened when monitoring systems include defined response workflows and designated roles for follow-up [20]. Broader implementation-oriented scholarship in digital oncology similarly emphasizes the importance of integrating patient-generated health data into existing clinical infrastructure and aligning digital tools with established workflows [21].

Broader mHealth research in populations with cancer has documented limited integration of patient-generated data into clinical systems and substantial variability in measures captured across platforms [9], echoing findings from recent reviews that emphasize the need for integrating patient-generated health data into clinical workflows and ensuring data are actionable at the point of care [19]. Participants emphasized that tracking is useful only when incoming information can be interpreted and acted upon efficiently. However, actionability may depend not only on data integration, but also on the availability of clinical resources, referral pathways, and support services that allow medical teams to respond to identified needs. For example, symptom or risk data may be more actionable in settings where patients can be referred to embedded supportive care, rehabilitation, behavioral health, nutrition, or cardiovascular services. In contrast, when follow-up services are unavailable, poorly coordinated, or associated with additional patient costs, tracking may generate information that is difficult to act on and may increase the burden for patients and medical team participants. This reflects broader challenges in translating patient-generated data into actionable clinical insights, particularly when data streams are not standardized or integrated with existing decision-support systems [9]. From an informatics perspective, this underscores the importance of systems that support interpretation within clinical workflows rather than treating data as isolated data entries [21].

In summary, digital cardiovascular symptom tracking systems should prioritize clarity, workflow alignment, and actionable outputs rather than simply increasing data volume. Participants’ perspectives also indicate that minimizing cognitive burden and time constraints for both patients and medical team participants may be central to feasibility. Clear expectations regarding symptom relevance and defined response processes may support more consistent engagement over time.

This study has several strengths. First, it included perspectives from both patients and medical team participants, allowing examination of cardiovascular symptom tracking as an interdependent behavior rather than a unidirectional patient responsibility. Second, use of the COM-B framework provided a structured approach to organizing barriers and facilitators across psychological, social, and environmental domains while maintaining analytic transparency. Third, interviews focused specifically on cardiovascular symptoms and signs tracking in the cardio-oncology context, contributing to a growing body of research that has largely examined symptom monitoring more broadly in oncology.

Several limitations should be considered. Participants were recruited from a single, specialized cardio-oncology clinic and were predominantly highly educated and White, which may limit transferability to other institutions with different workflow structures or to more diverse and underserved populations. These findings may not reflect the experiences of individuals with lower health or digital literacy, limited access to technology, or those receiving care outside of specialized clinic settings. As with all qualitative studies, findings reflect self-reported perceptions and experiences rather than directly observed behaviors. We did not measure actual tracking adherence or clinical outcomes and therefore cannot draw conclusions regarding the effectiveness of specific tracking approaches. Additionally, because all interviews were conducted with participants who had access to videoconferencing capability, they may have been more interested in cardiovascular health or digital tools than nonparticipants, potentially introducing selection bias. Finally, while the COM-B framework provided a useful organizing structure, alternative theoretical approaches may have highlighted different aspects of tracking behavior. Future studies should include more diverse and community-based populations to better understand how cardiovascular symptom tracking can be implemented across different health care settings.

Cardiovascular symptom tracking in cancer care is a shared behavior requiring coordinated engagement between patients and medical team participants. The use of the COM-B framework allowed identification of capability, opportunity, and motivation factors that shape how cardiovascular symptoms and signs are monitored and communicated in cardio-oncology settings. Participants described ambiguity regarding expectations across both groups. Patients emphasized uncertainty about which symptoms warranted tracking and about expectations for communicating symptom information, whereas medical team participants described time pressures, concerns about information burden, and limited standardized processes for reviewing and responding to symptom information. Medical team participants expressed concern about receiving large volumes of symptom information without clear mechanisms for prioritization or response, while patients described reassurance and a greater sense of control as benefits of tracking.

Efforts to strengthen cardiovascular symptom tracking may benefit from addressing behavioral determinants alongside technological considerations. Digital systems that provide clear guidance, align with clinical workflows, and support actionable interpretation of symptom reports may facilitate engagement and perceived value. Future studies should examine how these behavioral insights inform implementation strategies and evaluation of cardiovascular monitoring approaches in routine cancer care.