Recommendations for sharing cardiovascular implantable electronic device data with patients based on clinician interviews

Abstract

Background

Cardiovascular implantable electronic devices (CIEDs) capture an abundance of data that requires specific medical knowledge to interpret. Currently, there are no clear recommendations for sharing clinically meaningful CIED data with patients from the abundance of data in CIED remote monitoring reports.

Objective

The objectives of this study were to explore clinicians’ perceptions of clinical care gaps related to CIED data that could be improved through patient access to data and to generate recommendations for sharing CIED data with patients.

Methods

In this institutional review board-exempt, 2-phase sequential qualitative study, 42 semistructured phone interviews were conducted with 32 clinicians in the United States who specialize in the care of patients with CIEDs. Phase 1 explored clinical care gaps related to CIED data, and phase 2 involved questions about sharing CIED data with patients in 3 device-specific scenarios.

Results

Clinicians identified mechanisms by which sharing data with patients could close gaps in care, for example, improving timeliness of anticoagulation for patients with atrial fibrillation. However, clinicians expressed concerns about sharing data as well as beliefs that it would not alter patient care.

Conclusion

Based on clinician interviews, patient access to data from CIED reports has potential to improve care. Recommendations synthesized from the findings included displaying battery status and presence/absence of episodes and therapies and providing reassurance that the device is delivering therapies appropriately. Shared decision-making processes before device implantation and during ongoing care should include options for sharing CIED data with patients and discussion of potential benefits to understand patients’ needs and preferences.

Key Findings.

• ▪Data such as battery status and the presence/absence of episodes could be useful for patients with cardiovascular implantable electronic devices (CIEDs) to receive and understand from remote monitoring reports.
• ▪Reasons for sharing data with patients include facilitating next steps in their care plan, improving confidence in care, improving understanding of their device and health condition, preparing for cardiology visits, improving symptom self-monitoring and medication adherence, and responding to patients’ preferences for data.
• ▪Patient use (enrollment and adherence) of remote monitoring was identified as a key clinical care gap in the care of patients with CIEDs.

Introduction

Cardiovascular implantable electronic devices (CIEDs), including implantable loop recorders (ILRs), pacemakers (PMs), and implantable cardioverter-defibrillators (ICDs), monitor cardiac activity, and some also provide therapies for patients who have cardiac conditions.¹ These devices collect a vast amount of data related to device activity, programming, and cardiac functioning. Remote monitoring of CIED data has become the standard of care for patients with CIEDs.^2,3 The 21st Century Cures Act in the United States requires health care systems to allow patient access to their health information without delay or charge.⁴ However, the amount of data that patients receive directly from their devices is unclear. Some CIED vendors, such as Medtronic and Boston Scientific, have mobile applications for remote monitoring that share some data with patients (eg, battery status).⁵ Patient access to data has been shown to improve engagement in health care, including adherence to medical regimens,⁶ and could be useful for patients with CIEDs who are managing cardiac conditions.

Prior research shows that patients have varying levels of interest and needs related to receiving CIED data and feedback from remote monitoring,^7,8 including follow-up phone calls after transmissions are received by the clinic.⁹ One of our studies explored patients’ prioritization of data elements from remote monitoring reports for ICDs and PMs.¹⁰ However, the reasons why these data were important to patients (such as how they were related to their specific cardiac condition) were not fully explored. Among the vast amounts of data collected by devices, patients may not know which data elements are clinically relevant for their situation or how to make meaning of multiple data elements in the context of their medical condition. Also, in our previous research, patients expressed preferences toward having access to simple displays of data that can provide useful, personally relevant information.^7,11 Given their expertise in the clinical context and meaningful use of CIED data, clinicians’ views on care gaps that can be addressed by sharing CIED data with patients, ways to inform or educate patients about their data, and communication about health status and the device are important to integrate with patients’ preferences and needs.

The 2023 expert consensus statement on practical management of the remote device clinic includes techniques for incorporating reports into electronic health records and a mandatory minimal data set for remote monitoring reports for clinicians.² While the document mentions patient access to device information, there are no guidelines or recommendations related specifically to how to share CIED data with patients in a useful way. Research with patients and clinicians can help address this need.

This study aimed to elicit key point-of-care opportunities, or clinical care gaps, that can be addressed through using and sharing CIED data as well as recommendations from clinicians for sharing CIED data with patients. For this study, we defined a clinical care gap as an issue where CIED data are not used most effectively in the care and treatment of patients with CIEDs, potentially hindering the ability to achieve the optimal treatment plan or goals. To this end, this project was based on the following 2 research questions:

• 1.What are key clinical care gaps that can be addressed by sharing CIED data with patients?
• 2.What are clinicians’ recommendations for sharing CIED data with patients?

Methods

Design

This sequential qualitative study included 2 phases of semistructured phone interviews with clinicians in the United States involved in the care of patients with CIEDs and adhered to the COnsolidated criteria for REporting Qualitative research (COREQ) guidelines.¹² The study was reviewed by the Parkview Health Institutional Review Board and was exempt because of minimal risk to participants, as outlined by the criteria in Title 45 of the Code of Federal Regulations 46.104(d)(2)).¹³ The authors are in compliance with the regulations of the institutional review board and have both clinical and nonclinical backgrounds (2 electrophysiologists, 1 nurse/device clinic supervisor, 2 research nurses, 2 user experience researchers, and 4 research scientists with experience in health informatics, human-computer interaction, and psychology). Medtronic, PLC, provided funding for the study and input on the objective and general study design, but did not have a role in data collection, analysis, or manuscript preparation and reviewed the manuscript solely for technical accuracy in aspects related to their devices.

Setting and sample

The study was conducted in a large not-for-profit hospital system in the Midwestern United States. Participants were individuals (clinicians) across the United States who were involved in the care of people (patients) with CIEDs (ICDs, PMs, and ILRs) from various health care settings (eg, public nonprofit, private, and academic health centers). Clinician eligibility requirements were met by those physicians and/or advanced practice providers (APPs) who specialized in cardiology or electrophysiology. Eligible participants also included registered nurses who worked in a cardiac device clinic or specialized in electrophysiology. Clinicians were ineligible if they did not speak English, were CIED representatives, were still in training or less than 1 year outside of training, or were retired.

Clinicians were recruited through affiliations and professional networks using convenience and purposeful snowball sampling strategies. For phase 1, 2 team members, including C.D. (female, health informatics researcher, and principal investigator of the study), and a research nurse with a background in cardiology, recruited clinicians using email and posting on social media and professional society forums, networking within the hospital system that employed the research team, and drawing on relationships established in professional organizations. As part of the snowball sampling method, eligible contacts were invited to forward the invitation to colleagues. Participants from phase 1 were invited to complete an interview for phase 2. For phase 2, 2 team members, C.D. and S.P. (female, research nurse), recruited clinicians via email, posting on social media and professional society forums, and disseminating flyers. Once sessions were scheduled, interview activity sheets (Supplemental Appendix 1) were sent to potential participants to review in advance as desired.

Procedure

Participants (in phases 1 and 2) were scheduled for a 45-minute session with researchers. Interview sessions lasted, on average, 46 minutes, ranging from 25 to 69 minutes. The average durations for the review of the study information sheet and interview activities were 11 and 35 minutes, respectively.

The lead interviewer was C.D. for all interviews except 1 in phase 2, led by S.P. Phase 1 interviews also included a research nurse with a background in cardiology for all interviews, and an observer (H.-Y.L., male, user experience researcher) for most interviews. Phase 2 interviews included a research nurse with a background in cardiology (S.P.) for most interviews, and observers for nearly half of the interviews (J.C., female, research nurse; and/or H.-Y.L., male, user experience researcher). Interviews took place using virtual meeting spaces. Interviewees were not asked to disclose their location at the time of the interview. Before any study activities, participants reviewed the study information sheet and provided verbal consent. All clinicians who scheduled a session agreed to participate, and no clinicians dropped out once the interview began. Participants received a $50 electronic gift card for participation (1 person declined). All interviews were recorded and transcribed using Microsoft Teams (Microsoft Corporation, Redmond, WA). Team members debriefed after interviews and consolidated notes. All data were stored on secure servers, and transcripts were anonymized. Participants were not requested to review transcripts for accuracy or clarification.

Interview activities

Phase 1 and phase 2 interviews began with demographic and practice-related questions (Supplemental Appendix 1). The interview activity sheets (Supplemental Appendix 1) were developed with the guidance of an electrophysiologist (author J.S.) and device clinic supervisor (author T.A.). The phase 1 interview activity sheet was developed using the critical incident technique.¹⁴ The purpose was to identify clinical care gaps related to CIED data and explore the factors that contribute to or perpetuate the gaps to understand whether sharing data with patients could help address the issues. The interview guide was tested with a retired cardiologist to assess clarity, flow, and timing of the interview.

Given the variety of gaps that emerged during data collection, the interview guide was modified with more direct prompts to explore clinical scenarios wherein sharing CIED data with patients could be useful, which was the overarching objective of the study. However, most clinical care gaps emerging from the interviews did not involve such scenarios, and the research team ended recruitment for phase 1 once the minimum amount of information needed for phase 2 was collected (14 interviews).

The phase 2 interview activity sheet consisted of 3 scenarios inspired by clinical care gaps in phase 1 that involved ILRs, ICDs, and PMs. Clinicians were asked how they currently share CIED data with patients, if at all, and questions related to sharing CIED data with patients, including what data, if any, would be useful for patients to receive and understand in the given scenario, as well as how and why the data would be useful. We purposely avoided defining “useful” for clinicians and let them explain what it meant to them to avoid leading them in a specific direction. For the ICD and PM scenarios, the 9 highest prioritized data elements selected by patients in the prior study¹⁰ were included as examples of data that could be shared from remote monitoring reports. At the end of the interview, clinicians were asked to reflect on their thoughts about the highest priority data elements that patients chose, and if the selections were surprising or not surprising to them. Participant recruitment for phase 2 ended after 28 interviews, as recurrent patterns had emerged in the data and sufficient information was gathered to meet study objectives. We reached a balance in representation among the positions included in the sample (electrophysiology [EP] registered nurses, electrophysiologists, and EP APPs, cardiologists, and cardiology APPs), which was important since the positions have various types of engagement with patients and CIED data.

Data analysis

Phase 1 thematic analysis consisted of an inductive approach guided by the objective of understanding the nature of clinical care gaps related to CIEDs. Codes were established to characterize the clinical care gaps, including type of device and clinical condition (if applicable), and categories that the gaps were related to (eg, patient understanding and alert settings). The analysis team (author C.D., 2 research nurses [including author S.P.], and 2 user experience analysts [H.-Y.L. and Y.R.T.]) coded 8 transcripts independently, meeting to discuss emergent codes and establish consensus. Once the team reached consensus with minimal discussion, the remaining 6 transcripts were coded by 2 team members independently and reviewed by a third member as a quality check, with discussions as needed. The team members rotated between coding and performing the quality checks. This approach was suitable given the variety of clinical care gaps and was feasible given the number of transcripts. Although the clinical care gaps were explored in detail during the phase 1 interviews, for the purpose of this report, we present a summary of the types of clinical care gaps that emerged from the interviews. Author C.D. grouped the categories that each gap was related to into broader themes. If the gap was characterized by more than 1 category, the gap was grouped according to the main issue expressed in the interview. The research team, including clinical experts T.A. and J.S., reviewed and discussed the types of clinical care gaps in preparation for phase 2 to determine which types of clinical care gaps would be most relevant for exploring clinicians’ recommendations for sharing data with patients in phase 2.

For phase 2, interview data were analyzed in NVivo software using a structural coding approach to capture content within each domain area from the interview guide,¹⁵ followed by thematic analysis of the content within and across the domains. The analysis team (C.D., H.-Y.L., S.P., and J.C.) coded 2 transcripts as a team to establish the codebook and then coded a third transcript independently and met to refine the codebook. Because of the volume of transcripts in phase 2, the team measured interrater reliability with the goal of dividing the transcripts among individual team members for coding. Team members individually coded transcripts until Light’s κ, the average of all possible pairwise Cohen κ values between pairs of raters, reached 0.80 (from −1 to 1, where 1 is total agreement) after 7 transcripts, with regular meetings to reach consensus and refine the codebook if needed. The remaining 19 transcripts were then divided for coding (3 team members coded 5 transcripts and 1 coded 4). The team members met regularly to discuss any issues or questions that emerged during the coding process. 2 team members (C.D. and H.-Y.L.) synthesized the content within each of the codes into themes across the 3 scenarios. Author C.D. aggregated the responses coded for how data are currently shared into counts. Transcripts were not returned to participants for review, nor were participants asked to provide feedback on the findings for either phase 1 or phase 2.

The interview findings were derived from the phase 2 interviews and synthesized into recommendations for sharing CIED data with patients. The recommendations were drafted by the principal investigator of the study (C.D.) and reviewed for relevance and accuracy by the research team, including clinician consultants T.A. and J.S. The recommendations were based on phase 2 interview findings across the 3 scenarios and were structured around the areas of inquiry from the phase 2 interviews, including what data elements and/or qualitative or descriptive data would be useful for patients to receive and understand, how often patients should receive the data and in what modality, supporting information that could help patients understand the data, and how to tailor the messaging on the basis of medical history or other factors.

Results

Participant characteristics

42 interviews with 32 participants were conducted across the 2 phases (Table 1). The 32 clinician participants were mostly employed full-time (87.5%) and had an average of 16.7 years of experience with patients with CIEDs, ranging from 2 to 35 years of experience.

Table 1.

Participant characteristics in phase 1, phase 2, and total

Characteristic	Phase 1 (N1 = 14)	Phase 2 (N2 = 28)	Total (N = 32)
Age range
26–35 y	3 (21.4)	4 (14.3)	5 (15.6)
36–45 y	2 (14.3)	7 (25.0)	8 (25.0)
46–55 y	3 (21.4)	9 (32.1)	10 (31.3)
56–65 y	5 (35.7)	7 (25.0)	8 (25.0)
66–75 y	1 (7.1)	1 (3.6)	1 (3.1)
Gender
Female	8 (57.1)	16 (57.1)	18 (56.2)
Male	6 (42.9)	12 (42.9)	14 (43.8)
Race
White	11 (78.6)	24 (85.7)	27 (84.4)
Asian	3 (21.4)	3 (10.7)	4 (12.5)
Prefer not to disclose	0 (0.0)	1 (3.6)	1 (3.1)
Position
EP RN∗	8 (57.1)	8 (28.6)	11 (34.4)
Electrophysiologist	3 (21.4)	7 (25.0)	8 (25.0)
EP APP	1 (7.1)	6 (21.4)	6 (18.8)
Cardiologist	1 (7.1)	4 (14.3)	4 (12.5)
Cardiology APP	1 (7.1)	3 (10.7)	3 (9.4)
US region
Midwest	10 (71.4)	20 (71.4)	22 (6.9)
Southeast	3 (21.4)	5 (17.9)	7 (21.9)
South	0 (0.0)	2 (7.1)	2 (6.3)
Northeast	1 (7.1)	1 (3.6)	1 (3.1)

Values are presented as n (%).

APP = advanced practice provider; EP APP = electrophysiology advanced practice provider; EP RN = electrophysiology registered nurse.

^∗EP RN includes cardiac device clinic nurses and electrophysiology nurses. The majority (all but 1) of the EP RNs were device clinic nurses.

Phase 1: Clinical care gaps (N₁ = 14)

From phase 1, 47 clinical care gaps were identified (Supplemental Appendix 2). The majority (46) were related to (1) how CIED data are shared, retrieved, used, and managed in the health system (18 care gaps); (2) patients’ understanding of their device and remote monitoring (7 care gaps); (3) lack of use of remote monitoring (8 care gaps); and (4) capabilities of the device and programmer (related to programming and engineering) (13 care gaps).

Notably, clinicians (n = 8) were concerned about patients’ enrollment and adherence to remote monitoring. 1 clinician indicated that 20% of their patients were not enrolled in remote monitoring. Clinicians suggested that patients’ disinterest in remote monitoring, residence in rural areas without sufficient Internet service, or decisions to leave monitors at home during travel were reasons for nonadherence. Nonadherence was identified as a key clinical care gap because reports for clinicians to review cannot be generated without data capture.

While many of the clinical care gaps could not be improved by sharing data with patients, clinicians identified other gaps for which sharing data with patients may be involved. These were related to improving data access between providers and departments, improving timeliness of notifications of atrial fibrillation (AF), and improving the quality of information for clinical decision making by collecting patient symptoms experienced at the time of episodes.

Phase 2: Scenarios and usefulness of CIED data (N₂ = 28)

To examine opportunities for sharing data with patients, clinicians in phase 2 were asked to discuss what data would be clinically relevant and useful to share with patients depending on 3 scenarios and how they recommend sharing data in terms of frequency, timing, and modality, supporting information, and other considerations for content and presentation.

Current data sharing with patients

First, clinicians were asked whether patients currently receive data from their device, and if so, how data are shared and what types or amount of data are included. Clinicians reported that clinics reach out to patients, typically by phone call, when the findings (data) in the report require follow-up. In terms of sharing CIED data with patients in general, 28.6% of participants (n = 8) expressed that patients receive their data from transmissions through a patient portal (Table 2). 1 clinician specified that patients receive a cover sheet with their report that summarizes the data. 10 clinicians stated that they share data with patients during in-person visits or through follow-up calls, suggesting that select data elements may be brought up for discussion and that patients do not necessarily receive the data from their transmission reports unless requested. Some clinicians indicated that they did not typically share data and were not aware of what may or may not be available to patients through the patient portal. For others, generic letters or messages included high level information, such as confirmation that their data have been reviewed and everything is normal, the device is functioning properly, battery status, and timing for the next scheduled transmission.

Table 2.

Type/amount of data and modality of data that are currently shared with patients from the clinic (N₂ = 28)

Type/amount of data	Modality	n
Data from the report as relevant or per patient request	Follow-up call or in-person visits	10
Entire transmission report	Patient portal	8
Generic letter or message	Patient portal or postal mail	5
Unsure if/what patients receive	Patient portal	5

Clinician-recommended data elements for patients

All clinician participants (n = 28) proposed data elements that could be useful to share with patients across the 3 scenarios. Most clinicians (n = 19) mentioned battery status as a valuable data element, and almost half of the participants (n = 13) indicated that they believed a summary or interpretation of the data would be useful to notify patients that everything is “normal” and that the device is working as it is supposed to, using terms such as “essentially normal function,” “routine information,” or “no concern.”

Monitored atrial tachycardia/atrial fibrillation (AT/AF) episodes were frequently mentioned as useful for PMs (n = 12), as was the presence/absence of AT/AF episodes for ILRs and PMs (n = 19). Table 3 presents the highest priority data elements identified by patients in a prior study¹⁰ and clinicians’ feedback as to whether they perceive these data elements as useful or not useful for patients to receive and understand.

Table 3.

Clinicians’ feedback on highest priority data elements identified by patients (N₂ = 28)

Highest priority data elements for patients (based on the previous study10)		Clinicians’ feedback (useful or not useful)
Data element	Device(s)	Useful (n)	Device(s)	Not useful (n)	Device(s)
Battery longevity	PM	19	ICD, ILR, PM	0	N/A
Monitored AT/AF episodes	ICD, PM	12	ICD, PM	2	ICD
Presence/absence of therapies, including pace-terminated episodes (ATP) and ventricular pacing	ICD, PM	11	ICD	0	N/A
Atrial pacing	PM	11	PM	9	PM
Ventricular pacing	ICD	9	ICD, PM	4	ICD
Observations section of the report∗	PM	5	PM	7	PM
Atrial sensing	PM	4	PM	9	PM
Aborted charges for AT/AF	ICD	3	ICD	6	ICD
Treated AT/AF episodes	PM	3	PM	8	PM
Atrial pacing impedance	ICD	3	PM	11	PM
Pace-terminated AT/AF episodes	PM	2	PM	10	PM
Lead information (impedances and thresholds)	ICD, PM	1	ICD	10	ICD
Pace-terminated (VT/VF) episodes	ICD	0	N/A	3	ICD

AT/AF = atrial tachycardia/atrial fibrillation; ATP = antitachycardia pacing; ICD = implantable cardioverter-defibrillator; ILR = implantable loop recorder; N/A = not applicable; PM = pacemaker; VT/VF = ventricular tachycardia/ventricular fibrillation.

^∗Observations is a section on the vendor report (before clinician review) and does not incorporate clinical interpretations or summaries.

Clinicians also suggested data elements not present in the patient list that they believed would be useful, including presence/absence of AT/AF episodes for ILRs and PMs (n = 19), VT/VF episodes for ICDs (n = 16), presence/absence of general arrythmias and events for ILRs and PMs (n = 11), and AF episode information for ILRs and PMs (n = 8) (Table 4). One clinician stated that they believed the most important information for patients to know, although not a data element on the report, is whether their home monitor is not connected or transmitting.

Table 4.

Useful data elements suggested by clinicians (N₂ = 28)

Useful data element	n	Scenario(s)
Presence/absence of AT/AF episodes (particularly if new)	19	ILR, PM
VT/VF episodes	16	ICD
Presence/absence of general arrhythmias and events	11	ILR, PM
AF burden, frequency, and overall percentage	8	ILR, PM
AF duration	8	ILR, PM
Pacing (general)	8	ICD
Number of AF episodes	6	ILR, PM
Heart failure data (including activity, thoracic impedance, and biventricular pacing)	5	PM, ICD
Aborted shocks or charges (general)	3	ICD
Dates and times of AF events	3	ILR
Device information (including brand name, date of implantation, and MRI capability)	3	PM, ICD
Heart rate	3	ILR, PM
Nonspecified arrythmia burden, duration, frequency, and number of episodes	3	ILR
Presence/absence of tachycardia- or bradycardia-tachycardia–related arrythmias	3	ILR
Presence/absence of ventricular arrythmias (specifically true ventricular rhythms for ICDs)	3	PM, ICD
ECG strips (showing normal rhythm)	2	ILR
Graphical representation and tracking of episodes, even if no AF	2	ILR, PM, ICD
Histograms	2	ILR, PM, ICD
Pauses	2	ILR
All monitored episodes	1	ICD
Any tachycardia therapies	1	ICD
Episodes in general	1	ICD
How much pacing in the upper and lower chambers	1	PM
Number of AT episodes	1	PM
Pacemaker dependent	1	ICD
Presence/absence of bradycardia pacing	1	ICD
Presence/absence of parameters, such as an indicator after a certain increase in AF episodes	1	PM
Programming elements (detection rates—how high the heart rate gets before shock and how low the heart rate gets before pacing)	1	ICD
When their next check is	1	ICD
Whether noise is detected	1	ICD

ECG = electrocardiogram; MRI = magnetic resonance imaging; other abbreviations as in Table 3.

Reasons for sharing data with patients

When asked about reasons for sharing data with patients, clinicians stated that they believed it could help patients facilitate the next steps in their care plan (n = 23), improve confidence, reassurance, and trust in care (n = 15), improve understanding of device status and their health condition (n = 14), prepare for cardiology visits (n = 12), and improve symptom self-monitoring and medication adherence (n = 10). Patients’ preferences were also mentioned as another reason to provide data (n = 16). 4 clinicians expressed that there would be no difference in communication or quality of care by sharing the data. Table 5 outlines the themes and exemplary quotes. Quotes may have been edited slightly for clarity and succinctness.

Table 5.

Reasons to share data with patients with representative quotes

Facilitate next steps in the patient care plan (N2 = 23)

Patients can be proactive: But it's [referring to monitored AT/AF episodes] something that would prompt them potentially to reach out to your provider. Cause if, you know, if they are not on treatment for a-fibrillation, such as anticoagulation for stroke prophylaxis. Then perhaps that could be a prompt if, if that message didn't get to the provider, to consider that therapy. Protect them from stroke. (P34, cardiology physician, scenario 2, PM) Well, I think it would, you know, they could see their own progress. They could compare to the last check and if they had concerns, then, you know, perhaps they could take more ownership of their own I think right now patients are totally reliant on health care providers to pick up on everything and there's not a lot of ownership and part of that's our fault, for the patient to be responsible for, hey, you know what? I've seen my heart rate gradually going up when I'm in A-fib, I'm gonna call them and see if they're concerned about this. (P18, EP APP, scenario 2, PM) Patients can be prepared for what comes next: They would have an ability to know whether or not the reason for which they got a device is actually detected. Because when we talked to the patients prior to loop recorder placement, they know what the plan of care is. If this occurs, namely atrial fibrillation detected, these are the next steps. And so, the fact that they are already versed in the plan of care helps them participate to know what comes next in a timely fashion. (P33, EP physician, scenario 1, ILR) It may lead to a conversation on possible medication changes. For instance, if they’re having a lot of pace-terminated VT/VF episodes that may lead to a change in their medication. Or whether or not they may require an ablation. (P41, EP physician, scenario 3, ICD) Ventricular pacing is also important…if they are pacing a significant amount of time through their defibrillator, then there may be implications about whether or not they would develop pacing-induced cardiomyopathy…they want to be aware if it’s pacing or not. And the implications I guess they would not fully understand unless they ask the question to the clinician. (P26, EP physician, scenario 3, ICD) Patient access to data may improve the efficiency and timeliness of treatment: When they would come to see us in the clinic in-person, that would be probably more efficient…the first half of the clinic visit is talking about what we’ve found in their remote checks and then the discussion of the plan after that. So, if they kind of had that information up front, it would probably allow for a better discussion as to treatment options and therapy options. (P29, EP physician, scenario 1, ILR) If for some reason, you know, I mean things happen. And it doesn't get to the provider. And, you know, somehow it slipped through the cracks….So, they’re advocating for themselves in that case. Perhaps. You know if it wasn't caught or you know. That can be the difference between having another, you know, TI or stroke. You know, getting on anticoagulation even a couple days earlier. If that were the case where, you know, there was a delay. So. (P32, cardiology APP, scenario 1, ILR) Patients could share their data with other providers outside the cardiology clinic (suggested by a few clinicians): I had another patient that she comes in and goes, “thank goodness for the loop recorder.” And I was like, “well, what happened?” And she goes, “We were in a car accident. And that marked the date and time and recorded everything for the ER. we could take it right to ‘em.’” (P17, EP RN, scenario 1, ILR)

Improve confidence, reassurance, and trust in care (N2 = 15)

Yeah, just being aware of what's happening. And what a lot of patients get thrown off about is they just hear radio silence basically, if we don't see anything. So, if there is some sort of way to update them like, “Oh, this month, you look good.” That would probably give them some peace of mind. (P31, EP APP, scenario 1, ILR) I think that the patient would feel that they’re getting a better quality of care to actually discuss the data at that detailed level. Versus telling the patient “You’ve got A-fib” and not going into the details. And I think the patient would interpret that as a higher quality level of care that the doctor is spending that time giving them the details. (P41, EP physician, scenario 1, ILR) It helps the patient understand what is going on…this is why we put the device in. And then just kind of building trust with the patient, I feel like, if the patient understands what is going on, and it helps them to trust you, and maybe they’ll be more open to recommendations, and asking questions if they have them. (P21, EP RN, scenario 3, ICD) If I received a primary prevention ICD and I was seeing that I was having significant non-sustained episodes, that would help me reaffirm my decision to have the device in the first place. Because I can see that there was a risk. (P33, EP physician, scenario 3, ICD)

Support patients’ understanding of device status and their health condition (N2 = 14)

I think it would be useful to patients who have no symptoms that they said to help them understand that even though you feel well, you still have this abnormal rhythm that needs to be treated. (P15, EP RN, scenario 1, ILR) I think it just makes them more enlightened. You know if they’re not certain of the impact of A-Fib or why we’re monitoring…it makes them a little bit more aware of kind of how useful this particular device is. (P25, EP physician, scenario 1, ILR) There are some times patients don’t fully understand the difference between just the pacemaker versus ICD and why they have it. So probably maybe giving more information would either prompt them to ask, “Ok, I hadn’t thought of this. Why is it? Do I really have an ICD?” or for us to even just explain to them more clearly. (P22, cardiology APP, scenario 3, ICD)

Help patients to prepare for cardiology visits (N2 = 12)

Sometimes we look at stuff and we don’t know what it is and we just kind of jump right in and try to kind of diagnose things. It might be helpful for them so that when they come in, we can say “Oh yeah we are going to follow this” or “your battery life estimated was this, at the last check. We can compare it to today.” So it might be kind of a helpful tool in the office, to have that information ahead of time for them. (P25, EP physician, scenario 2, PM) They’ll either know, coming in. Like “Oh, there was a change. I should ask the doctor about it.“ Or when the doctor says, “I need to talk to you about something we saw in your pacemaker.“ The patient hopefully has seen it or had access to look at it, so it won’t be a complete shock. (P39, EP APP, scenario 2, PM) They’ll be able to respond in the portal. Ask any questions up front. And then kind of chew on the information to be better prepared for their office visits. (P31, EP APP, scenario 3, ICD)

Improve symptom self-monitoring and medication adherence (N2 = 10)

Let’s say I was a patient, and I thought I was having an arrythmia. Then, I could send my transmission, but then add an addendum or note and say “Hey, at this time when I sent in the transmission, I was feeling chest pain, shortness of breath, and irregular heartbeat.” Then that would include the provider in, to what was going on, and then, there could be a message that’s sent back, that “Yes. It was reviewed I’ll be calling you in 10 minutes.” (P17, EP RN, scenario 1, ILR) I think probably the most important thing is that if they do have an episode if you can kind of pinpoint with them and see if they had symptoms at the time then they would know what to look for, you know, in the future. (P23, EP RN, scenario 1, ILR) If they could access, this could save a lot of ER visits for people that have kind of have phantoms, phantom shocks. (P29, EP physician, scenario 3, ICD) I think that a lot of people who don’t have symptoms of A-fib…they’re lax on taking some certain medications, and if they can see this information in real time, it may keep them more compliant in taking medications if they know even though I feel well and I don’t know it, I’m still having these episodes and I need to take my medication. (P15, EP RN, scenario 1, ILR) If may help with compliance, right? If you’re not taking your medications, and you see, you know, you’re having a lot of VT, then it may prompt people to you know, take their medicine. (P40, EP APP, scenario 3, ICD)

Respond to patients’ preference to see their data (N2 = 16)

For the pacemakers, most of them are FYI type things…I don’t know that the patient would be able to change anything by themselves because they now have had more atrial fib… Some patients would be interested if they have more of that. (P37, EP APP, scenario 2, PM) I think it can kind of be confusing at times if we share too much medical knowledge in terms of, you know, confuse them with the strips and things like that. But there are some patients who really want to see it. (P22, cardiology APP, scenario 3, ICD) I think patients may wanna know how much their device works like how often they pace. Some patients wanna know that information, some don’t. Some are more concerned about that. I find that patients are more concerned about their overall battery longevity. (P24, EP RN, scenario 2, PM) The percent pacing I think again goes along with that patient that’s just curious… “Am I using it?” they want to just know “Is this device really doing anything?” or that sort of thing. (P25, EP physician, scenario 2, PM)

Abbreviations as in Tables 1 and 3.

Concerns about sharing data

Nearly all clinicians (n = 27) had concerns related to patients misunderstanding or misinterpreting data, leading to confusion and anxiety. Specifically, for scenarios 1 (ILRs) and 2 (PMs), clinicians were concerned about patients’ understanding of data about AF, as well as challenges with interpretation:

Physicians themselves have a hard time determining, “Is this true A-fib or not?’ And I think a potential drawback is if the patient were to receive an alert that’s not been filtered by some type of overriding physician to get rid of false abnormal findings, that could lead to potential problems and increase anxiety and stress. (P41, EP physician, scenario 1, ILR)

Clinicians also expressed that data such as aborted charges for ICDs or clinically irrelevant data could cause anxiety for patients.

Let’s just say they’ve had a bunch of non-sustained VT episodes that the physician doesn’t consider clinically relevant. But the patient sees something like that and… it’s a real stressor – really produces a lot of anxiety. (P23, EP RN, scenario 3, ICD)

Several clinicians (n = 15) described other reasons not to share data, including that the data are irrelevant, patients do not want the data, or the extra work required to share data may provide little benefit.

Supporting information to improve patient understanding of their data

Suggestions for educational information included what to expect from the data (eg, what will happen next in their care plan) (n = 9), descriptions of their condition (eg, AF), including causes, guidelines, and how AF is treated (n = 8), discernment of what would be considered normal results for each individual (n = 6), and indication of the proper functioning of the device (n = 6). Other types of supporting information that came up less frequently included what each data point means, a glossary or reference guide, lifestyle suggestions, and what to do with the information in the report (even if that means just knowing that the clinic received the report and there is nothing to worry about).

Nearly half of the clinicians (n = 13) also specified that supporting information should be short, simple, and broken into smaller chunks of information.

For pacing information, maybe just an explanation of why (and) this is your atrium pacing this much, this is your ventricle pacing this much and then letting them know the atrium’s the top chamber of your heart, the ventricles the bottom chamber of your heart. (P21, EP RN, scenario 2, PM)

In contrast, 13 clinicians indicated that additional information provided directly to the patient is not necessary. In general, their preference would be to explain critical results during office visits, as there is too much information to explain clearly in a message along with the data. Furthermore, some of these participants added that data in the reports should already be self-explanatory and sufficient on their own as long as supplemented by continued education at various time points surrounding the implantation and during visits.

Considerations for sharing data with patients

Most clinicians (n = 20) expressed their opinion that patients should receive the data at the same time as their scheduled device checks or interrogations, about 30 days for ILRs and 90 days for PMs and ICDs. Clinicians suggested sending data electronically, specifically through the patient portal (n = 23); however, clinicians also expressed their perception that not all patients have access to or want to use a patient portal. Alternative suggestions included phone calls or sharing during office visits (n = 12) and by postal mail (n = 4).

Half of the clinicians (n = 14) recommended that patients should not receive data before it is reviewed or signed by the clinic. Some data may not be accurate, or there could be noise detected by the device. A few clinicians suggested that the data should be discussed during an office visit or over the phone before anything is released. As 1 EP APP explained for scenario 1, ILR:

Sometimes with loop recorders, it will trigger that it’s atrial fibrillation. But if you really dig into the tracings, it may not be atrial fibrillation. It may be normal due to sinus rhythm with a lot of PACs [premature atrial complexes]…triggered events need to be reviewed to assure we have the right interpretation before we start sending out information. (P37, EP APP, scenario 1, ILR)

Also, a few clinicians expressed their belief in the importance of focusing on reassurance rather than creating stress or worry for patients. For example, 1 cardiology physician described:

It’s really important (for monitored atrial tachycardia or fibrillation) that we stress it is not life-threatening, but we wanted to have a discussion with you regarding what we found on the pacemaker. …if we were to say, “call the office immediately, we found a bad rhythm,” that’s going to be a problem, that would scare them. (P20, cardiology physician, scenario 2, PM)

Half of the clinicians (n = 14) suggested that patients should also receive data when there is an alert or additional findings, and several emphasized the importance of a follow-up phone call from the clinic as the means of relaying information, particularly for scenario 2, PM.

When asked about tailoring reports for patients based on certain characteristics or individual factors, 24 clinicians discussed considerations such as level of experience with the device and understanding of the information and ensuring that reports are in the patients’ language (Table 6). While several clinicians (n = 10) believed that reports should be tailored on the basis of discussion with patients regarding what information they want to receive and how they want to receive it, several clinicians (n = 10) expressed that standardized reports could (or should) be shared with all patients and caregivers.

Table 6.

Considerations for sharing data with patients (N₂ = 28)^∗

Factors that may influence tailoring of reports	Scenario(s)	n
Level of experience, education, or understanding of the information	ILR, PM, ICD	10
Accessibility related to language, literacy, learning modality, and technology proficiency	ILR, PM, ICD	8
Medical condition (cardiac and noncardiac, including stroke, heart failure, and dementia)	ILR PM, ICD	7
Reason for the device	ILR, ICD	2
Level of worry	ILR, ICD	2
Who is to be contacted (eg, nursing home, caseworker, and home care nurse)	PM	1

Abbreviations as in Table 3.

^∗Clinicians may have responses in more than 1 category.

Discussion

The present study explored gaps that clinicians experience in the care of patients with CIEDs and clinician recommendations for sharing CIED data with patients. To our knowledge, this is the first study to explore clinicians’ perspectives on sharing CIED data with patients based on specific clinical scenarios. This work is novel in that it provides information on specific data elements that clinicians believe could be useful for patients with ICDs, PMs, and ILRs as well as detailed descriptions of reasons for sharing data and concerns. The insights gathered from clinician interviews have been abstracted into recommendations for sharing CIED data with patients, contributing the clinicians’ perspectives to the existing literature on sharing CIED data with patients. Given the increasing transparency of digital health data, it is important to understand providers’ and patients’ perspectives on how to share data in clinically meaningful and accessible ways, avoid unnecessary concerns among patients, and meet patients’ needs and preferences.

Clinical care gaps

Clinicians reported a variety of clinical care gaps related to workflow and communication of patient device information across the health system and capabilities of devices related to engineering and programming. Although not the direct focus of this study, an important primary concern among clinicians was related to patient enrollment and adherence to remote monitoring, which is a global issue.¹⁶ Prior studies suggest that receiving some amount of data or feedback from the remote monitoring report could help with patient adherence to remote monitoring.^7,10,17 However, to do this, connectivity must be established. Efforts such as engaging patients with smartphone-based remote monitoring,¹⁸ sending postcards in the mail,¹⁹ and establishing clinic infrastructure to support follow-up processes and education of patients and families²⁰ have positively affected adherence to remote monitoring. Variability in the use of remote monitoring is also due to factors such as limitations in technical infrastructure and recurring fees for patients. These must be addressed to make remote monitoring accessible to patients.^9,15 Future studies should build on existing research examining approaches to successful remote monitoring adherence.

Sharing data with patients

Most clinicians agreed that some data should be shared or reviewed with patients; however, there were varying opinions as to how much data sharing could help patients increase engagement in care. While CIED data show increasing potential to provide valuable indicators about health status, the data have the potential to be misunderstood by patients. Moreover, many clinicians believed that most of the data should not be shared until data are interpreted by a trained medical professional. Indeed, a large study of patients with CIEDs from 67 device clinics across the United States showed that 57% of transmissions were dismissed by clinicians, as they did not provide new information or provided inaccurate data (eg, noise detected by the device); of the 43% of transmissions that were passed on to clinicians to review, only 13% merited clinical alerts.²¹ While some clinicians noted how sharing data with patients could improve timeliness of treatment, such as anticoagulation after detection of AF with ILRs, more exploration is needed to understand how the alerts could be filtered so as not to overburden patients with data or create unnecessary anxiety.²²

Patients have varying preferences for receiving CIED data, suggesting that reports require tailoring to individual needs.^7,23,24 Clinicians also expressed the importance of understanding patients’ preferences for what CIED data they want to receive.²⁵ One way to integrate tailoring in the patient portal is to provide layering so that users can choose the level of complexity and avoid being overwhelmed.^26,27 Clinicians can assist patients in understanding which data elements may or may not be clinically relevant on the basis of the patients’ health history. Furthermore, the discussion should include that treatment plans may be updated on the basis of CIED data.

Patients’ understanding of their device

Clinicians expressed the importance of patients understanding the purpose of the device. A survey study in more than 300 patients with CIEDs showed that patients overall do not understand the reason for their CIED, functions of the device, and battery life.²⁸ Consistently, some participating clinicians proposed that sharing device data with patients could help strengthen patients’ basic understanding of their device and improve engagement in their care (ie, ask questions during visits and understand when to call the clinic about symptoms). At the very least, patients could receive information about their device and its operability, such as the device identifier, the plans and process for tracking the device, and how they will be notified if there is an issue with the device.²⁹

Clinically relevant data elements

The majority of the clinicians thought that it would be useful for patients to receive battery status in all scenarios, and some clinicians also thought that monitored AT/AF episodes could be useful for both ICDs and PMs. Clinicians also proposed arrhythmia episodes, ventricular episodes, and AF data, which are data elements that clinicians report using frequently for clinical decision making.³⁰ Sharing these data with patients could potentially increase the timeliness of follow-up and improve follow-up conversations, thereby increasing timeliness of treatment, such as anticoagulation. Similarly, patients with CIEDs could review the data before a visit, respond with information about symptoms, and facilitate conversations about treatment. The more patients are aware of what is happening with their device and how that relates to their health, the more prepared they may be for discussions about their treatment and care, similar to the way in which accessing clinical notes can increase shared decision making.³¹

Clinicians described different types of educational information that could be helpful to patients, such as information about treatment options and care plans, descriptions of their condition, and explanation of “normal” results. Although some clinicians described including educational material, such as a glossary or reference guide, along with the report itself, clinicians also emphasized the importance of education along various time points in the care trajectory that would enhance understanding of the reports. Although education modalities were not explored in these interviews, educational videos with information about their device, remote monitoring, and device data could be a potential avenue to explore with patients with CIEDs.³² Video materials to support shared decision making for ICD implantation and deactivation show promise for improving patient understanding^33,34 and could also be explored for continuing education to support understanding of the data in context.

Recommendations for sharing CIED data with patients

Table 7 summarizes the findings related to the main areas of inquiry in the phase 2 interviews. The rationale based on the interview findings is included next to each recommendation. There are several assumptions underlying these recommendations, including the following:

• •Patients are connected (linked to the monitoring clinic) and adherent to remote monitoring.
• •Patients have access to different modalities of communication (eg, patient portal, phone call, and postal mail).
• •Clinics have the resources and structure to provide different options based on patients’ information needs and preferences.
• •Education and discussions with patients will be available in the patients’ language.
• •Patients with abnormal findings that require follow-up will be contacted via phone.

Table 7.

Recommendations for sharing CIED data with patients

Recommendations	Rationale
Data that would be useful for patients to receive and understand
1. Battery status and basic information indicating that the device is working properly with simple language such as “device working normally” or “no concerns.”	Most clinicians (n = 19) indicated that including battery status would be useful for patients regardless of the device. Battery status was also one of the highest-prioritized data elements for patients with PMs in the prior study. Nearly half of the clinicians (n = 13) proposed a summary of the data with a statement confirming that the device is working properly. This could also reassure patients.
2. Confirmation of whether episodes or therapies have occurred, in a simple (yes/no) format.	19 clinicians indicated that the presence/absence of AT/AF episodes would be useful, and 11 clinicians indicated that presence/absence of general arrythmias and events would be useful, with the binary yes/no format being potentially more meaningful than the number of episodes or events.
Frequency, timing, and modality
3. Share data with each device check or interrogation. Take care if data are shared after alerts (in between scheduled checks). If patients receive data at the same time as clinicians, include an explanation that the data have not been reviewed for accuracy (eg, “true” atrial fibrillation”). Phone calls may be more appropriate for sharing alert data.	Most clinicians (n = 20) expressed that patients should receive the data at the same time as their interrogations, and half of the clinicians (n = 14) expressed that patients should also receive data when there is an alert. However, half of the clinicians (n = 14) indicated that patients should not receive data before they are reviewed or signed by the clinic, and several emphasized the importance of a follow-up phone call for alerts.
4. Provide options for how patients receive the data, including electronic communication, such as EHR messaging/patient portal, by phone, or letter in the mail, to accommodate patients’ preferences.	Most clinicians (n = 23) expressed that the patient portal is optimal while adding that the patient portal is not used by all patients. Several clinicians suggested relaying information by phone call or office visits (n = 12) or by mail (n = 14).
5. Provide mechanisms for patients to associate symptoms or events that they have experienced with dates and times of episodes and for clinicians to review these associations. These mechanisms could include a follow-up phone call or on a report where patients send information back to the clinic.	10 clinicians described symptom self-monitoring and medication adherence as a reason for sharing data, including a suggestion for patients to be able to send information back to the clinic after receiving a report. Facilitating the care plan was one of the top reasons for sharing data (n = 23), and this bidirectional communication could support that.
Supporting information
6. Provide educational opportunities before implantation, when possible, and at regular intervals to support patients’ understanding of their device and data from their device. Education should focus on why the device was implanted, what may happen next on the basis of the data, descriptions of the condition, functioning of the device, and discernment about what normal results mean, and what is normal for them.	Nearly half of the clinicians (n = 13) expressed that supporting information is important before implantation and throughout their care and office visits. Categories for educational information included what happens next on the basis of the data (n = 9), descriptions of the condition (n = 8), indications of proper functioning of the device (n = 6), and what would be considered normal results per the individual (n = 6).
7. When sending a message with the data, include a statement that none of the data are life-threatening.	Nearly all clinicians (n = 27) expressed concerns related to patients having unnecessary anxiety about their data, and a few clinicians stressed the importance of providing reassurance, for example, stating that the data are not life-threatening.
Tailoring to patients’ individual needs and preferences
8. Assess patients’ desire for information, that is, if they prefer “no news is good news,”8 if they would rather have more detailed information, or something in between.	One of the reasons for sharing data was to respond to patients’ preferences (n = 16). However, 15 clinicians expressed reasons not to share data, including that patients may not want to receive it. Clinicians also had mixed opinions on tailoring reports individually based on discussion with patients and providing a standard report. As a start, assessing patients’ interest in receiving data could help clinicians meet patients’ preferences.

AT/AF = atrial tachycardia/atrial fibrillation; EHR = electronic health record.

Limitations

Although the study included a sample of clinicians across the United States, the sample included predominantly white clinicians working in the Midwest. Information about the types of health systems and populations served was not considered. Furthermore, the study is vulnerable to sampling bias because of the use of convenience and snowball sampling methods through professional societies and networks of the authors, potentially limiting the breadth of perspectives on the topic. The themes and recommendations were not reviewed with participants for accuracy and to check researcher bias. Furthermore, the recommendations are derived from the themes that emerged from the aggregated interview data by the principal investigator of the study and reviewed by the research team. As such, the recommendations do not necessarily represent the views or consensus among all the clinicians who participated in the study. In addition, this study only included interviews with clinicians; therefore, the findings cannot be assumed to represent patients’ preferences or experiences related to receiving their CIED data.

Conclusion

Based on clinician expertise, when sharing CIED data with patients, it is important to keep information simple and, when appropriate, to provide reassurance regarding device activity to avoid triggering anxiety for patients. With access and proper understanding of the device and data, sharing specific data from CIEDs has the potential to improve patient care. These recommendations, based on clinicians’ perspectives, should be taken into consideration with patients’ values, beliefs, and priorities. Shared decision-making processes, from implantation and throughout patient device follow-up, should include ongoing education and options for sharing data with patients and discussion of concerns and potential benefits.

Disclosures

The authors have no conflicts of interest to disclose.