As the pandemic continues to unfold, effective, technology‐based solutions are needed to help patients with atrial fibrillation (AF) maintain their health and well‐being during the outbreak of COVID‐19.
This single‐center, pilot study investigated the effects of a 4‐week (eight sessions) virtual AF self‐management program. Questionnaires were completed at baseline and 1 week after the intervention, and assessed AF knowledge, adherence to self‐management behaviors, mental health, physical function, and disease‐specific quality of life in patients with AF. Secondary outcomes included knowledge of COVID‐19, intervention, acceptability, and satisfaction.
Of 68 patients who completed baseline questionnaires, 57 participated in the intervention and were included in the analysis (mean age of 73.4 ± 10.0 years, 60% male). Adherence to AF self‐monitoring behaviors, including monitoring their heart rate (
This pilot study suggests that a virtual patient education program could have beneficial effects on adherence to guideline‐recommend self‐care of AF, emotional wellbeing, physical function, and knowledge of COVID‐19 in patients with AF. Future randomized studies in larger samples are needed to determine the clinical benefits of the intervention.
Atrial fibrillation (AF) is a leading cause of prolonged disability, repeat hospitalizations and premature death in the United States.
The outbreak of novel coronavirus disease 2019 (COVID‐19) and the extraordinary measures taken to reduce the spread of the virus introduced an entirely new set of challenges for patients with AF. Persons with underlying cardiovascular disease have a higher incidence of severe illness, complications and death from COVID‐19.
Initiatives to improve education and self‐management for patients with AF are well established,
A pre‐post design was employed for this pilot study, which was conducted with patients treated at an outpatient electrophysiology clinic at an academic medical center in North Carolina during an acute phase of the pandemic (April 28, 2020 to June 2, 2020) when mandatory shelter‐in‐place orders were issued for all residents in the state of North Carolina. Patients were prospectively screened for eligibility using automated EHR algorithms.
Individuals meeting inclusion criteria were sent an invitation to participate in the intervention through the EHR‐based patient portal. Patients who provided consent to participate were emailed a packet of information about the program and instructions to help troubleshoot technical issues. Patients did not receive compensation for completing study questionnaires or for participating in the intervention. The protocol and procedures were approved by the institutional review board at the University of North Carolina; all participants provided electronic informed consent.
The AF‐At‐Home Program was developed to improve AF management by focusing on self‐monitoring, skill development, and behavioral risk factor modification. The program included eight 1‐h group sessions, occurring 2 days per week over 4 consecutive weeks and was delivered with a secure video‐conferencing platform. Each session included 40 min of didactic instruction followed by 20 min of interactive group discussion and questions answered by the session leader. Sessions were led by a diverse group of health care professionals from cardiovascular electrophysiology, cardiac psychology, endocrinology, clinical pharmacy, and social work. Content was based on guideline‐recommended topics for AF patient education
Baseline demographic and clinical data were collected by medical chart abstraction for all study participants. Standard definitions
Patients completed questionnaires at baseline (pre‐intervention) and again 5 weeks later (1‐week post‐intervention). The primary outcomes for this study were AF‐related health knowledge, adherence to guideline‐recommended self‐management behaviors, mental and physical health outcomes, and general and AF‐specific quality of life. Secondary outcomes included self‐reported knowledge of COVID‐19 and assessment of intervention acceptability and satisfaction with the program.
A series of questions evaluated patients' knowledge and understanding of AF, adherence to self‐management behaviors, and confidence in implementing self‐management skills (details provided in Table
The National Institutes of Health Patient‐Reported Outcomes Measurement Information System (PROMIS)–29 profile, version 2.0, was used to assess global health status and quality of life. The PROMIS‐29 is a self‐administered, extensively validated, quality of life questionnaire with eight domains that assess the following symptoms during the previous 7 days: pain intensity and interference, fatigue, sleep disturbance, physical functioning, depression, anxiety, and ability to participate in social roles and activities.
AF‐related quality of life was assessed with the Atrial Fibrillation Effects on Quality of Life questionnaire (AFEQT), a widely used 20‐item measure of patients’ AF symptoms, daily activities, treatment concerns, and satisfaction with treatment during the past month.
Given the rapid spread of inaccurate or misleading medical information about COVID‐19 during the pandemic,
Patients’ provided feedback about their experience and satisfaction with the program at follow‐up. Items were rated on a 5‐point scale (1—Strongly disagree; 2—Disagree; 3—Agree; 4—Strongly agree; 5—N/A, I did not take part in the educational program). Questions included: (1) This educational program helped me understand how to manage my AF during COVID‐19; (2) when I completed the program, I felt more confident about managing my AF during a public health emergency; (3) I am satisfied with this COVID‐19–AF educational program.
We determined that recruitment of at least 49 patients would provide a power of more than 80% to detect an effect of the intervention on PROMIS domain scores from pre to post intervention, with a two‐sided alpha level of 0.05, and a planned drop‐out rate of 25% which is the median dropout rate for studies of internet‐based education and lifestyle interventions.
Study outcomes were analyzed separately among those who participated in at least one session of the intervention (program participants, n = 57) and those who attended zero sessions of the intervention but completed the questionnaire at both study baseline and follow‐up (non‐participants, n = 11) unless otherwise indicated. Categorical data were summarized using frequencies and percentages and continuous data are reported as mean ± SD. The Kolmogorov‐Smirnov test was used to test for data normality, and non‐parametric methods were used where indicated. Baseline characteristics were calculated for the entire sample and compared among program participants and non‐completers using
The Wilcoxon signed rank test was used to compare the primary outcome measures at baseline and study follow‐up. Effect size estimates were calculated for AF‐related quality of life (AFEQT) and mental health and physical function outcomes (PROMIS) using the nonparametric equivalent of the Cohen's
Effects of the atrial fibrillation (AF)‐at‐home program on self‐management skills in patients with AF [Color figure can be viewed at
Sixty‐eight patients consented to participate in the study and completed both baseline and follow up questionnaires. The majority of patients who agreed to participate in the educational program attended at least one session of the intervention (84%) with a mean attendance of 5.4 out of eight sessions (SD = 2.8). Eleven patients completed both the baseline and follow‐up questionnaires but attended zero sessions of the intervention (referred to as non‐participants).
Baseline characteristics of the study population are shown in Table
Baseline characteristics of atrial fibrillation (AF) patients who did and did not participate in the AF at home program
Overall sample (N = 68) | Program participation (n = 57) | Did not participate (n = 11) |
| |
---|---|---|---|---|
Demographics | ||||
Age (years) | 73.4 ± 10.0 | 74.1 ± 9.2 | 69.5 ± 13.5 | .166 |
Sex | .383 | |||
Male | 39 (57.4%) | 34 (59.6%) | 5 (45.5%) | |
Female | 29 (42.6%) | 23 (40.4%) | 6 (54.5%) | |
Race/ethnicity | .820 | |||
White | 66 (97.1%) | 55 (96.5%) | 11 (100%) | |
African American | 1 (1.5%) | 1 (1.8%) | 0 (0%) | |
Other | 1 (1.5%) | 1 (1.8%) | 0 (0%) | |
Marital status | .828 | |||
Married | 54 (79.4%) | 45 (78.9%) | 9 (81.8%) | |
Divorced | 5 (79.4%) | 4 (7.0%) | 1 (9.1%) | |
Single | 5 (7.4%) | 4 (7.0%) | 1 (9.1%) | |
Widowed | 4 (5.9%) | 4 (7.0%) | 0 (0%) | |
Employment status | .121 | |||
Employed | 12 (17.6%) | 9 (15.8%) | 3 (27.3%) | |
Not employed | 43 (63.2%) | 39 (68.4%) | 4 (36.4%) | |
Unknown | 13 (19.1%) | 9 (15.8%) | 4 (36.4%) | |
AF history | ||||
AF Type | .044 | |||
Paroxysmal | 52 (76.5%) | 41 (71.9%) | 11 (100.0%) | |
Persistent or permanent | 16 (23.5%) | 16 (28.1%) | 0 (0.0%) | |
Time since AF diagnosis (months) ‡ | 60.7 ± 52.8 | 57.5 ± 47.2 | 78.0 ± 78.3 | .289 |
Prior procedures | ||||
Ablation | 29 (42.6%) | 24 (42.1%) | 5 (45.5%) | .837 |
LAA occlusion | 2 (2.9%) | 2 (3.5%) | 0 (0.0%) | .528 |
Prior cardioversion | 30 (44.1%) | 26 (45.6%) | 4 (36.4%) | .572 |
PM/ICD implant | 19 (27.9%) | 17 (29.8%) | 2 (18.2%) | .431 |
Cardiovascular comorbidities | ||||
Hypertension | 38 (55.9%) | 32 (56.1%) | 6 (54.5%) | .922 |
Previous MI | 12 (17.6%) | 12 (21.1%) | 0 (0.0%) | .094 |
Coronary heart disease | 20 (29.4%) | 20 (35.1%) | 0 (0.0%) |
|
Hyperlipidemia | 39 (57.4%) | 34 (59.6%) | 5 (45.5%) | .383 |
Heart failure | 17 (25.0%) | 15 (26.3%) | 2 (18.2%) | .568 |
TIA/CVA | 8 (11.8%) | 8 (14.0%) | 0 (0.0%) | .186 |
Diabetes mellitus | 5 (7.4%) | 4 (7.0%) | 1 (9.1%) | .809 |
Obstructive sleep apnea | 23 (33.8%) | 18 (31.6%) | 5 (45.5%) | .373 |
Thyroid disease | 14 (20.9%) | 11 (19.6%) | 3 (27.3%) | .569 |
Chronic lung disease | 15 (22.1%) | 11 (19.3%) | 4 (36.4%) | .211 |
Chronic kidney disease | 7 (10.3%) | 6 (10.5%) | 1 (9.1%) | .886 |
Anxiety | 12 (17.9%) | 10 (17.9%) | 2 (18.2%) | .980 |
Depression | 6 (8.8%) | 6 (10.5%) | 0 (0.0%) | .260 |
CHA2DS2‐VASc | ||||
0 | 2 (2.9%) | 1 (1.8%) | 1 (9.1%) | .187 |
1 | 12 (17.6%) | 9 (15.8%) | 3 (27.3%) | .360 |
≥2 | 54 (79.4%) | 47 (82.5%) | 7 (63.6%) | .158 |
Medications | ||||
Aspirin | 14 (20.6%) | 13 (22.8%) | 1 (9.1%) | .303 |
P2Y12 | 3 (4.4%) | 3 (5.3%) | 0 (0%) | .436 |
Anticoagulation therapy | 52 (76.5%) | 46 (80.7%) | 6 (54.5%) | .061 |
Warfarin | 5 (7.4%) | 4 (7.0%) | 1 (9.1%) | .809 |
DOAC | 47 (69.1%) | 42 (73.7%) | 5 (45.5%) | .064 |
Beta blocker | 47 (69.1%) | 41 (71.9%) | 6 (54.5%) | .447 |
Calcium channel blocker | 4 (6.0%) | 3 (5.3%) | 1 (9.1%) | .633 |
Antiarrhythmics | 22 (32.8%) | 18 (31.6%) | 4 (36.4%) | .601 |
Lifestyle factors | ||||
BMI‡ | 28.1 ± 6.8 | 27.1 ± 5.7 | 33.1 ± 10.1 |
|
Alcohol consumption | 46 (67.6%) | 37 (64.9%) | 9 (81.8%) | .272 |
Smoking status | .523 | |||
Current | 1 (1.5%) | 1 (1.8%) | 0 (0%) | |
Never | 33 (48.5%) | 26 (45.6%) | 7 (63.6%) | |
Former | 34 (50.0%) | 30 (52.6%) | 4 (36.4%) |
aData are presented as means ± SD.
Adherence to guideline‐recommended self‐monitoring behaviors, including monitoring their heart rate (2.4 ± 1.0 vs. 3.0 ± 1.0;
Mental health and physical function were improved at the end of the intervention compared with baseline among program participants (Table
Differences in primary outcomes at baseline and study follow‐up
Program participation | N | Baseline | Follow‐up |
| Effect size | |
---|---|---|---|---|---|---|
AFEQT total score | ||||||
Yes | 57 | 76.7 ± 17.9 | 79.2 ± 16.1 | .252 | 0.15 | |
No | 11 | 80.9 ± 11.2 | 81.7 ± 15.0 | .824 | ||
AFEQT symptom subscale | ||||||
Yes | 57 | 81.9 ± 19.4 | 84.1 ± 15.6 | .476 | 0.10 | |
No | 11 | 79.2 ± 16.9 | 79.2 ± 19.6 | .819 | ||
AFEQT daily activity | ||||||
Yes | 57 | 73.4 ± 24.6 | 75.5 ± 23.9 | .260 | 0.15 | |
No | 11 | 82.8 ± 15.5 | 85.2 ± 16.4 | .266 | ||
AFEQT treatment concern | ||||||
Yes | 57 | 77.6 ± 17.0 | 81.0 ± 16.1 | .139 | 0.20 | |
No | 11 | 79.3 ± 11.5 | 77.6 ± 25.3 | .964 | ||
AFEQT current control | ||||||
Yes | 54 | 79.0 ± 19.2 | 80.9 ± 21.1 | .251 | 0.015 | |
No | 10 | 70.0 ± 30.2 | 71.7 ± 30.5 | .659 | ||
AFEQT treatment relieved | ||||||
Yes | 51 | 78.8 ± 21.1 | 79.7 ± 22.7 | .624 | 0.06 | |
No | 10 | 70.0 ± 32.2 | 70.0 ± 24.6 | .826 | ||
PROMIS‐physical function | ||||||
Yes | 57 | 47.7 ± 8.8 | 49.3 ± 7.9 |
| 0.36 | |
No | 11 | 51.8 ± 6.3 | 51.9 ± 6.1 | .645 | ||
PROMIS‐anxiety | ||||||
Yes | 56 | 51.8 ± 9.4 | 49.5 ± 8.4 |
| 0.33 | |
No | 11 | 53.0 ± 10.8 | 54.3 ± 10.8 | .646 | ||
PROMIS‐depression | ||||||
Yes | 55 | 48.5 ± 7.4 | 46.6 ± 7.3 |
| 0.26 | |
No | 11 | 46.7 ± 9.2 | 48.4 ± 10.7 | .346 | ||
PROMIS‐fatigue | ||||||
Yes | 56 | 47.4 ± 10.7 | 46.9 ± 10.0 | .919 | 0 | |
No | 11 | 45.8 ± 6.9 | 46.0 ± 10.2 | .718 | ||
PROMIS‐sleep disturbance | ||||||
Yes | 57 | 56.0 ± 2.6 | 45.4 ± 8.2 |
| 0.80 | |
No | 11 | 57.8 ± 2.7 | 48.8 ± 10.7 |
| ||
PROMIS‐social activities | ||||||
Yes | 56 | 51.4 ± 10.6 | 52.1 ± 10.5 | .626 | 0.06 | |
No | 10 | 55.5 ± 10.0 | 55.8 ± 9.1 | 1.000 | ||
PROMIS‐pain | ||||||
Yes | 57 | 48.0 ± 8.0 | 48.1 ± 7.5 | .852 | 0.03 | |
No | 11 | 50.0 ± 7.5 | 49.8 ± 6.9 | .789 |
a:Data are presented as means ± SD. For PROMIS domains, a positive value represents worsening pain, pain interference, fatigue, sleep disturbance, depression, and anxiety and an improvement in physical functioning and ability to participate in social roles and activities.
As shown in Table
Knowledge, beliefs, and behaviors related to COVID‐19
AF‐At‐Home Program Participants | |||
---|---|---|---|
Baseline | Follow‐up |
| |
NSAIDs increase the risk of COVID‐19 infection and worse outcomes |
| ||
True | 9 (16.1%) | 13 (23.2%) | |
False | 27 (48.2%) | 33 (58.9%) | |
Unsure | 20 (35.7%) | 10 (17.9%) | |
Hydroxychloroquine can prevent or treat COVID‐19 |
| ||
True | 5 (8.8%) | 4 (7.1%) | |
False | 35 (61.4%) | 46 (82.1%) | |
Unsure | 17 (29.8%) | 6 (10.5%) | |
ACE‐I and ARBs increase the risk of COVID‐19 infection and worse outcomes |
| ||
True | 7 (12.3%) | 11 (19.6%) | |
False | 13 (22.8%) | 22 (39.3%) | |
Unsure | 37 (64.9%) | 23 (41.1%) | |
Discontinue taking ACE‐I and ARBs immediately if infected by COVID‐19 |
| ||
True | 2 (3.5%) | 2 (3.5%) | |
False | 25 (43.9%) | 43 (75.4%) | |
Unsure | 30 (52.6%) | 12 (21.1%) | |
Delay or avoid seeking medical attention for symptoms of a heart attack or stroke due to fears of contracting COVID‐19a | ‐ | ||
Agree | ‐ | 4 (7%) | ‐ |
Disagree | ‐ | 53 (93%) | ‐ |
If I had to go to the hospital for worsening cardiac symptoms, I would not get the medical care I need because of COVID‐19a | |||
Agree | ‐ | 13 (22.8%) | ‐ |
Disagree | ‐ | 44 (77.2%) | ‐ |
a
Items were not assessed in the baseline questionnaire–data presented are for the follow‐up questionnaire.
Very few patients reported that they would delay seeking medical care for acute cardiovascular symptoms due to fears of COVID‐19 infection (7.0%). However, nearly one in five reported concerns about receiving sub‐optimal care if they were hospitalized for AF during the pandemic. In addition, Figure
There was no correlation between the number of intervention sessions attended and the primary or secondary outcomes (data not shown). A majority of program participants (75%) reported improvements in AF knowledge and self‐management skills after completing the AF‐At Home‐Program. Participants also reported feeling more confident in their ability to recognize symptoms and manage AF exacerbations (58.8% strongly agreed, 39.2% agreed) and more than two‐thirds reported they were highly satisfied with the program.
COVID‐19 has revealed the clear and pressing need for technology‐based approaches to delivering continuous education and support to patients with AF during a public health emergency. In this pilot study, we demonstrated the feasibility of developing and rapidly deploying a tailored AF‐self management intervention delivered by a broad range of health care professionals during a period of mandatory quarantine. The results of this study show that the AF‐At‐Home program was effective at increasing self‐confidence in disease management and adherence to guideline‐recommended AF self‐management behaviors, including self‐ monitoring (heart rate, heart rhythm, and blood pressure), symptom identification and management and may have broader applications in routine care outside beyond this pandemic. The program was also effective in reducing sleep disturbance, anxiety and depression. Additionally, we observed a high prevalence of misinformation and inaccurate beliefs about COVID‐19 in this sample of AF patients at baseline. Preliminary findings from this investigation suggest that the AF‐At‐Home program and other technology‐based, direct‐to‐consumer, communication strategies may be effective at reducing uncertainty and inaccurate beliefs about COVID‐19 in vulnerable persons.
Previous studies have suggested that clinic‐based interventions are effective in increasing disease‐specific knowledge, long‐term adherence to anticoagulation therapy, reducing symptom burden and improving quality of life.
We extend this work by demonstrating the acceptability and preliminary efficacy of a structured AF self‐management program that was adapted for rapid delivery to address the secondary health impacts of an ongoing public health emergency. We further demonstrate that such an intervention can be delivered 100% remotely, thereby minimizing the risk of COVID‐19 exposure among health care providers and patients. The high percentage of women (40%) and older adults who participated in this study also suggests that the technology‐based programs may facilitate access to underserved populations by overcoming traditional barriers to nonattendance (e.g., inadequate transportation, lack of insurance, and work obligations, and caregiver responsibilities).
The intervention did not have an effect on AF‐related quality of life. The absence of an effect on AF‐related quality of life could have been due to the relatively brief duration of the study which was selected for its practicability and implementation in a wide variety of health care settings and is consistent with the American Heart Association's goals for integrating telehealth solutions into existing care delivery systems.
Perhaps one of the most concerning findings from this study was the high rate of inaccurate knowledge, beliefs and behavioral responses to COVID‐19 among persons with AF. Fear and uncertainty have fueled the rapid spread of false information about the virus which is then repeated and “re‐tweeted” on television programs, the internet and social media platforms.
This intervention was specifically developed to address disruptions in medical services during the most acute phase of the COVID‐19 pandemic, however, the need for safe, socially‐distant interventions will remain during more “chronic” phases of the pandemic, and possibly for months even after a vaccine becomes available. It also remains to be seen whether these services can be adopted and implemented in routine care, and whether the AF‐At‐Home program is effective in populations with less access to technology, low health literacy, and persons living in rural communities, as these populations may be disproportionately affected by digital inequalities and disruption to routine care during a public health emergency. Further randomized clinical studies are also needed to quantify the impact of the AF‐At‐Home program on clinically meaningful outcomes such as reductions in thromboembolic events, and utilization of inpatient, outpatient, or emergency medical services, and to identify patients who would benefit from this type of technology‐based intervention.
This study is not without limitations. First, this study was conducted at an academic medical center with predominantly older, white patients with varying levels of AF knowledge, adherence and health literacy at baseline. Thus, the generalizability of these results to other patients with AF and to other types of organizations may be limited. The lack of randomization is another limitation of this study. Second, while we were able to examine within‐subject changes in the primary outcomes among persons who did and did not participate in the intervention, the study did not include a pre‐specified comparison condition (e.g., usual care) and was not powered to examine between‐group differences in outcomes. Third, in response to the disruptions to routine care that occurred while stay‐at‐home orders were in place, the recruitment period was intentionally short (8 days). We recognize that this limits generalizability, but we felt that rapid, direct communication with patients was the priority. Nevertheless, this may have introduced a selection bias, as persons who regularly use technology may have responded to the study invitation in that timeframe. The reliance on self‐report measures that are prone to social desirability bias may also have affected our results. Similarly, while specific changes in AF knowledge were evaluated, additional dimensions of disease knowledge and health behavior change should be examined in future studies (e.g., knowledge of anticoagulation, procedures, and preventative health behaviors). Fourth, as with any observational study, it is possible that residual confounding may have affected our results. Finally, this pilot study was too small and too short in duration to assess the effect of the intervention on meaningful clinical outcomes or adjust for important clinical covariates in our analyses. Well‐designed trials are needed to clarify these issues and to determine the optimal frequency and timing of intervention sessions to maximize program value in routine clinical care.
As the pandemic continues to unfold, effective, technology‐based solutions are needed to help patients with AF maintain the health and wellbeing. Findings from this proof‐of‐concept study indicate that a virtual self‐management program for persons with AF may improve disease self‐management, mental health, and physical function during the pandemic and may have broader applications in routine care outside beyond the pandemic. The program was also effective in reducing misinformation and inaccurate beliefs about COVID‐19. Larger studies with longer follow‐up are needed to determine the efficacy of this intervention in reducing complications of AF and improving important quality of life outcomes.
Anil K. Gehi, MD: Research Grant: Bristol‐Myers Squib Foundation, Honoraria/Consulting Fees: Biosense‐Webster, Abbott, Biotronik, Zoll Medical. Jennifer Walker, MSN, ANP has received salary support from the Bristol‐Myers Squib Foundation. Sriram Machineni, MD: Research funding: Novo Nordisk, Boeringher Ingelheim, Consulting Fees: Novo Nordisk, Rhythm Pharmaceuticals.
The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.
Supplementary Material
Click here for additional data file.
We want to thank Tanya Lulla, Lindsay Mosteller and Brittany Becker for their contribution to this study.