Application of Alfalfa App in the management of oral anticoagulation in patients with atrial fibrillation: a multicenter randomized controlled trial

Abstract

Background

In recent years, mobile medical technology has made great progress in chronic disease management, but its application in patients with atrial fibrillation (AF) still needs to be clarified.

Objective

This study aims to determine whether the newly developed smartphone app for patients with AF (Alfalfa App) can improve anticoagulation knowledge, drug treatment compliance, and satisfaction of AF patients.

Methods

Alfalfa App integrates the functions of patient education, remote consultation, and medication reminder through a simple user interface. From June 2020 to December 2020, patients with AF were recruited in five large tertiary hospitals in China. Patients were randomly divided into the Alfalfa App or routine nursing groups. Patients’ knowledge, medication adherence, and satisfaction with anticoagulation were assessed using validated questionnaires at baseline, 1 month, and 3 months.

Results

In this randomized controlled trial, 113 patients with AF were included, 57 patients were randomly assigned to the Alfalfa App group, and 56 patients were randomly assigned to the routine nursing group. Forty-eight patients in the Alfalfa App group completed a three-month follow-up, and 48 patients in the routine nursing group completed a three-month follow-up. Basic demographic data were comparable between the two groups. The average age of AF patients was 61.65 ± 11.01 years old, and 61.5% of them were male. With time (baseline to 3 months), the knowledge scores of the Alfalfa App group (P<.001) and the routine nursing group (P = .002) were significantly improved, the compliance scores of the routine nursing group(P<.001) and Alfalfa App group(P<.001) significantly improved. Compared with the routine nursing group, patients’ knowledge level and medication compliance using the Alfalfa App at 1 month and 3 months were significantly higher (all P < .05). There were significant differences in knowledge and compliance scores between the two groups with time (all P < .05). The satisfaction degree of drug treatment in the Alfalfa App group was significantly better than that in the routine nursing group (all P < .05).

Conclusions

Alfalfa App significantly improved the anticoagulation knowledge, drug treatment compliance, and satisfaction of AF patients. In oral anticoagulation management for AF patients, mobile medical technology that integrates the functions of patient education, remote consultation, and medication reminder may be helpful.

Trial registration

Registration number, ChiCTR1900024455. Registered on July 12, 2019.

Introduction

Atrial fibrillation (AF) is the most common arrhythmia in clinical practice. In 2010, about 33.5 million people worldwide suffered from AF [1], which is estimated to double in 2030 [2]. Oral anticoagulation is recommended for AF patients with stroke risk factors to reduce the risk of stroke or systemic embolism [3]. The 2016 European Atrial Fibrillation Management Guidelines pointed out that better patient knowledge helps enhance self-management and common decision-making, and patients should be educated more comprehensively [3]. However, previous studies have shown that even after receiving oral/written information, AF patients’ knowledge about arrhythmia and its management has little improvement [4,5,6].

When used outside the controlled environment of linical trials, the safety and effectiveness of drugs will be affected by patient compliance, resulting in worse health outcomes and higher health care costs [7, 8]. However, long-term non-adherence to new oral anticoagulants (NOACs) is very serious in China. The results of a cohort study show that only 35% of patients in China’s economically advanced areas have been treated with NOACs for more than one year [9].

Mobile healthcare has been considered to promote the redistribution of medical resources. The World Health Organization has defined mHealth or mobile health as medical and public health practices supported by mobile devices, such as mobile phones, patient monitoring devices, personal digital assistants, and other wireless devices [10]. By improving the communication between patients and doctors and promoting self-monitoring, the appropriate application of information technology may greatly improve the management of oral anticoagulant therapy for AF patients. However, there needs to be more data on implementing mobile medical technology for patients with atrial fibrillation.

Recently, through the use of smartphone applications in various fields, studies have attempted to optimize patient management without increasing the burden on healthcare providers [11,12,13]. As a new tool, Health Buddies app aims to improve the compliance of elderly AF patients by providing a virtual contract with the patient’s grandchildren [14]. The mAF application aims to integrate clinical decision support, education, and patient engagement strategies [15]. Although these studies have had some success, there are still some problems such as low usability of applications and complex user interfaces.

This study aims to determine whether a smartphone app (Alfalfa App) can improve anticoagulation knowledge, drug treatment compliance, and satisfaction of AF patients. The application integrates the functions of patient education, remote consultation, and medication reminder through a simple user interface.

Methods

Study design and participants

This randomized controlled study was conducted in five large tertiary hospitals in China (Fujian Medical University Union Hospital, Affiliated Fuzhou First Hospital of Fujian Medical University, Xinyang Central Hospital, Affiliated Hospital of Jining Medical University, and First Hospital of Shanxi Medical University), and AF patients were recruited from June 2020 to December 2020. The patients were randomized according to a computer-generated program, using a random numbers table. Patients were randomly assigned to the online management (Alfalfa App) group or the offline management (routine nursing) group at a ratio of 1:1 by researchers who did not participate in the clinic. Inclusion criteria of patients included adult patients aged > 18 years with AF diagnosed with electrocardiogram and 24-hour Holter. Patients unable to sign informed consent, have a cognitive impairment, and have valvular atrial fibrillation are excluded. Medical staff who participated in the study were required to have more than three years of experience in anticoagulation clinics, and they received one month of standardized training before the start of the trial.

Ethics considerations

This study aligns with the Helsinki Declaration and obtained ethical approval from the local ethics committee. All patients provided written informed consent. The research has been registered in China Clinical Trial Registration Center (registration number: ChiCTR1900024455).

Features of the Alfalfa App

Alfalfa is a mobile device application that can provide convenient information exchange channels for medical staff and patients. Its functions include patient education, remote consultation, medication and blood test time reminder, etc. Alfalfa App has designed different versions for patients and doctors, respectively, and its usability has been evaluated in previous studies, and the results show acceptable usability [16].

Patient education

The patient education plan consists of 10 parts. Through a short video of about 4–5 min, patients can learn about AF and learn self-management methods. The education content includes understanding and treating AF, the importance of anticoagulation and self-management of AF, etc. (Supplementary Fig. 1).

Remote consultation

Patients can report the blood test results (International Normalized Ratio(INR) value, prothrombin time, and blood test date), drug dosage, physical and dietary status, etc., to their doctors and then adjust the dosage and take a blood test on time according to the doctor’s advice, thus realizing the remote management of patients (Supplementary Fig. 2).

Medication and blood test time reminder

The Alfalfa app can set reminders for taking medicine and blood tests according to user needs and send reminders automatically at regular intervals through the app (Supplementary Fig. 3).

Auxiliary function

The auxiliary functions of the Alfalfa App include anticoagulant community, INR extreme warning, and advice on blood pressure control. The anticoagulant Community module can serve as a channel for patient-patient and patient-health staff direct dialogue and communication. Patients can share their experiences of taking oral anticoagulants (OAC) in the anticoagulant community. They can also ask questions about how to use the Alfalfa App and how to handle adverse reactions. To reduce the incidence of anticoagulation-related clinical events, Alfalfa App has added the function of warning of INR extreme value and suggesting blood pressure control.

Intervention

Routine care at the hospital includes information provided by the cardiologist during the clinic and an information booklet on AF and OAC treatment given to the patient at the first visit. After baseline inclusion, AF patients in the control group were required to participate in the study in hospitals at a fixed time. Patients randomly assigned to the intervention group will be administered with Alfalfa App. Alfalfa App provides educational materials, remote counseling, medication and blood test time reminders, patient engagement strategies, and structured follow-up to support the comprehensive management of AF patients.

Outcome measures

Structured follow-up was planned at 1, 3, 6, and 12 months after patient enrollment, including assessment of anticoagulant knowledge, medication compliance, satisfaction and clinical outcomes, which are the primary outcomes of our study (Fig. 1). Medication compliance is defined as the patient’s medication treatment according to the doctor’s instructions, and the consistency of the patient’s medication with the doctor’s instructions. Alfalfa App will automatically send reminders to patients before the follow-up time point. When the follow-up time point is reached, patients in the App intervention group will answer the questionnaire embedded in the APP online, and patients in the routine care group will answer the paper questionnaire offline.

Fig. 1

Flow chart

Patients’ knowledge was assessed by the Jessa Atrial Fibrillation Knowledge Questionnaire (JAKQ) at baseline, 1 month, and 3 months. Drug compliance was assessed by the eight-item Morisky Medication Adherence Scale (MMAS-8) at baseline, 1 month, and 3 months. Medication satisfaction was assessed by the Treatment Satisfaction Questionnaire for Medication version II (TSQM-II) at 3 months. Clinical outcomes (including bleeding and thrombotic events) during follow-up were also recorded.

JAKQ is a short, complete, and effective questionnaire on AF-specific knowledge. It can be used in daily practice to assess patients’ knowledge of their condition [17, 18]. JAKQ consists of 16 questions: 8 about AF, 5 about OAC therapy, and 3 about vitamin K antagonists (VKAs) or new oral anticoagulants (NOACs). JAKQ only contains multiple-choice questions with one correct answer, two distractions, and one “I do not know” option. One point for the correct answer; Incorrect and “I do not know” answers are 0 points. The score on JAKQ is divided by the number of completed questions to get a percentage.

MMAS-8 is a widely used self-report questionnaire [19]. The first seven items are Yes/No, and the last is Likert’s 5-point answer. MMAS-8 has been proven to be quite useful in clinical practice because it captures the root causes of underutilization, such as forgetfulness, while considering the situations related to patients’ persistent behavior. MMAS-8 is reliable (α reliability = 0.83) for evaluating the drug compliance of outpatients with hypertension, and it is significantly related to blood pressure control, indicating that MMAS-8 is sufficient to detect the drug compliance of patients effectively [19].

TSQM-II is a general questionnaire with 11 items used to evaluate the treatment satisfaction of any treatment so that it can be compared among people with different diseases [20]. The questionnaire is divided into five parts: treatment effect (2 items), side effects (3 items), treatment convenience (3 items), and global satisfaction (2 items). On Likert’s five-point or seven-point scale, patients rate their treatment experience between “extremely dissatisfied” and “extremely satisfied.” The score of each scale ranges from 0 to 100. The higher the score, the higher the satisfaction with treatment.

Clinical outcomes included safety and efficacy outcomes. Safety outcomes were total bleeding, major bleeding, and minor bleeding. The efficacy outcome was thromboembolic events. Total bleeding included all types of bleeding events. Major bleeding is defined by the International Society of Thrombosis and Hemostasis (ISTH) as occurring in a critical organ (Intracranial, intraspinal, intraocular, retroperitoneal, intra-articular or intra-pericardial, intramuscular compartment syndrome) or as a drop in hemoglobin level of at least 2 g/dl or transfusion of at least 2 units of red blood cells [21]. Minor bleeding events were defined as failure to meet the criteria for major or clinically significant bleeding. Thromboembolic events include ischemic stroke and systemic embolism. Systemic embolism is defined as acute vascular occlusion of a limb or organ documented by imaging, surgery, or autopsy [22]. Clinical events were adjudicated by an independent, blinded clinical events committee that reviewed hospital records.

Statistical analysis

Considering the importance of patients’ knowledge in the comprehensive management of AF and the fact that various educational interventions that have been tried before have not achieved good results [3,4,5,6], the primary endpoint events of this study is set to patients’ knowledge level. According to the power calculation (power of 80%; alpha of 5%), at least 56 AF patients (28 in each study group) must be included to improve the knowledge level by 25% compared with the baseline during the follow-up period [17]. This estimated effect size was based on previous pilot data showing a 29.4% increase after a few days and a 24.9% increase after one month [17]. Considering the 15% lost follow-up rate, the minimum number of patients finally included is 66.

SPSS 25.0 (IBM, Armonk, USA) was used for statistical analysis. The continuous variables were tested for normality by the Kolmogorov-Smirnov test. Data with a normal distribution are expressed as mean ± standard deviation (SD). Non-normally distributed data are expressed as median and interquartile spacing (IQR). The categorical variables are reported as numbers and percentages. If they conformed to a normal distribution, Student’s t-test was used to compare the differences in continuous variables. If not, non-parametric statistical tests were used. Differences between categorical variables were compared by the chi-square test. The data on patients’ knowledge and compliance at three-time points (baseline, 1 month, and 3 months) are presented in the form of charts. Wilcoxon test or Friedman test is used to study the influence of Alfalfa App management or routine nursing on the knowledge level and compliance. Mann-Whitney U test, independent t-test, chi-square test, or mixed model were used for group comparison. P-value < 0.05 is considered statistically significant.

Results

One hundred thirteen patients were enrolled in this randomized controlled trial, with 57 patients randomly assigned to Alfalfa App and 56 randomly assigned to routine nursing. Forty-eight patients in the Alfalfa App group and 48 in the routine nursing group completed three months of follow-up. The mean age of AF patients was 61.65 ± 11.01 years old, and 61.5% were male. Hypertension and diabetes were the most common comorbidities in both groups, and patients randomly assigned to Alfalfa App and routine nursing were well matched in different demographic characteristics(Table 1).

Table 1 Baseline characteristics of atrial fibrillation patients

Values are median (interquartile range) except as noted.

Effect on knowledge level

The baseline score of JAKQ was similar in Alfalfa App group(median (IQR): 12.5%(6.3%,37.5%); mean ± SD: 24.5 ± 21.8%) and routine nursing group(median (IQR): 21.9%(6.3%,37.5%); mean ± SD: 26.0 ± 19.6%)(P = .57; Fig. 2). Compared with the routine nursing group, the knowledge level of the Alfalfa App group was significantly higher after 1 month(62.5%(50.0%,75.0%) vs. 25.0%(12.5%,37.5%); P<.001) and 3 months(78.1%(68.8%,92.2%) vs. 25.0%(12.5%,48.4%); P<.001). With time, the knowledge scores of the Alfalfa App group(12.5%(6.3%,37.5%) − 78.1%(68.8%,92.2%); P<.001) and the routine nursing group(median ( 21.9%(6.3%,37.5%) − 25.0%(12.5%,48.4%); P = .002) were significantly improved. There was a significant difference in knowledge scores between the two groups with time (median: 12.5–78.1% vs. 21.9–25.0%; P < .001). These findings were similar when subgroups of NOAC patients were analyzed (Supplementary Fig. 4).

Fig. 2

Score on the JAKQ over time in the routine care group (red triangles) and the Alfalfa App group (green dots). Statistical analysis: Friedman tests for differences over time within each group and mixed model for comparison between groups

Alfalfa App’s educational programs improve the level of knowledge in different aspects of AF patients. Compared to baseline, significantly more patients after three months knew that doctors often prescribed anticoagulants to AF patients to prevent stroke (97.7% vs. 34.8%; P < .001). In addition, 93.2% of OAC patients were aware of bleeding as a possible side effect of anticoagulants after three months, compared to only 11.6% at baseline (P < .001). At the outset, only 11.6% of patients knew they should contact their doctor and continue taking medication when bleeding, compared to 79.5% after three months (P < .001). In the beginning, only 25.5% of patients said they would consult with their doctor if they needed surgery, and 97.7% after three months (P < .001). Finally, 84.1% of patients taking vitamin K antagonists (VKAs) or NOACs knew what to do when they forgot to take their medication; This was only 9.3% at the start of the study (P < .001).

Effect on adherence

Compared with the baseline MMAS-8 score of the routine nursing group (median (IQR): 4.8(1.5,6.0); mean ± SD: 4.26 ± 2.37; P = .10), the score of the Alfalfa App group (median (IQR): 6.0(3.5,7.9); Mean ± SD: 5.28 ± 2.50) was higher (P = .03; Fig. 3). Compared with the routine nursing group, the compliance of the Alfalfa App group was significantly higher after 1 month (7.8(6.8,8.0) vs. 5.3(4.5,6.8); P<.001) and 3 months(8.0(7.0,8.0) vs. 5.5(4.5,6.8); P<.001). With time, the compliance scores of the routine nursing group( 4.8(1.5,6.0) − 5.5(4.5,6.8); P<.001) and Alfalfa App group(6.0(3.5,7.9) − 8.0(7.0,8.0); P<.001) significantly improved. The compliance scores of the two groups were significantly different with time (median: 6.0 to 8.0 vs. 4.8 to 5.5; P < .001). These findings were similar when subgroups of NOAC patients were analyzed(Supplementary Fig. 5).

Fig. 3

Score on the MMAS-8 over time in the routine care group (red triangles) and the Alfalfa App group (green dots)

Effect on satisfaction

At 3 months, the four dimensions of the TSQM-II questionnaire were significantly different between the two study groups (Fig. 4). Compared with the routine nursing group, the scores of effectiveness (66.7(50.0,83.3) vs. 50.0(50.0,58.3); P<.001), side effects(100.0(91.7,100.0) vs. 91.7(83.3,100); P = .006), convenience(66.7(55.6,81.9) vs. 55.6(50.0,66.7); P<.001) and global satisfaction(66.7(58.3,83.3) vs. 54.2(50.0,66.7); P<.001) of patients in Alfalfa App group were significantly higher. When subgroup analysis was performed among patients using NOAC, we found that patients in the standard of care and Alfalfa App groups reported similar satisfaction with side effects (100(91.7,100) vs. 100(91.7,100); P = .42), while the rest of the outcomes were consistent(Supplementary Fig. 6).

Fig. 4

Score on the TSQM-II in the routine care group and the Alfalfa App group

Sensitivity analysis

We performed sensitivity analyses for excluded patients in Alfalfa App intervention group and routine nursing group, filling in missing values of data for missed patients by multiple interpolation with SPSS. The sensitivity analysis results, showed that the between-group comparison outcomes between Alfalfa App intervention group of 57 patients who included lost patients and routine nursing group of 56 patients who included lost patients, and within-group comparison outcomes in each group were consistent with the main findings of the analyses in this paper, suggesting that the current conclusions with the exclusion of lost patients are robust (Supplemental Table 1 and Supplemental Table 2).

Effect on clinical outcome

During the one-year follow-up period, there were 18 (18.8%) bleeding events, 3 (3.1%) major bleeding events, 15 (15.6%) minor bleeding events, and 1 (1.0%) thrombotic event. Figure 5 shows the specific bleeding and thrombosis events of AF patients in different groups. Overall, there was no significant difference in the risk of bleeding (P = .60, OR = 1.32,95% CI: 0.47–3.69) or thrombosis (P>.99, OR = 0.98,95% CI: 0.94–1.02) between the Alfalfa App group and the routine care group. These findings were similar when subgroups of NOAC patients were analyzed.

Fig. 5

Clinical outcomes of patients in Alfalfa App group and routine care group

Discussion

In this multicenter randomized controlled trial, we investigated the effect of the smartphone application Alfalfa on oral anticoagulation management in AF patients. The main findings are as follows: (1) Compared with the routine nursing group, patients who use Alfalfa App have significantly higher knowledge levels at 1 month and 3 months. There was a significant difference in knowledge scores between the two groups with time. (2) Compared with the routine nursing group, patients who use Alfalfa App have significantly better medication compliance at 1 month and 3 months. There were significant differences in compliance scores between the two groups with time. (3) Compared with the routine nursing group, the satisfaction score of the Alfalfa app group was significantly higher.

Patients’ knowledge of AF and OAC treatment is essential to their overall management [23]. However, due to the time constraints of doctors and the lack of effective educational interventions, this aspect needs to be addressed and systematically solved. A cross-sectional survey in China shows that the management of AF patients in community health service centers is not standardized, 75.8% of primary care physicians have insufficient knowledge of OAC treatment for patients with non-valvular AF, and most primary care physicians do not provide sufficient education for AF patients [24]. Previous studies have shown that short educational intervention through information pamphlets helps improve patients’ awareness of AF anticoagulant therapy [25]. Similarly, our research also found that routine nursing (educational intervention based on an information manual) can significantly improve patients’ knowledge levels. However, the Alfalfa app can not only significantly improve patients’ knowledge level but also the change in knowledge score with time is more significant than routine nursing, which shows that compared with traditional educational interventions, educational measures based on mobile medical care may be more effective. The alfalfa app can be a good educational tool for AF patients.

Poor OAC compliance is associated with an increased risk of clinical events [26,27,28]. It is worth noting that long-term non-adherence to OAC treatment is also very serious in China. A study conducted a telephone follow-up of 281 patients with nonvalvular AF who took NOACs in Urumqi, the western city of China. Among them, 48.4% of the patients stopped using NOACs during 1-year follow-up [29]. The results of another cohort study showed that compared with patients taking warfarin, patients taking NOACs in China had lower persistence. Other inventions were needed to improve the persistence of NOACs in patients in China [9]. Many studies have evaluated the methods to improve the compliance of AF patients, including patient education, reminding tools, and consultation, but the results are mixed. FACILITA research shows that the mixed intervention consisting of patient education and simple calendar reminder of drug intake is an effective strategy to improve the treatment compliance of patients with AF with dabigatran. One year later, the average compliance of dabigatran in the intervention group was 89.2%, while that in the control group was 63.2% [30]. The test results of MISOAC-AF show that among patients with non-valvular AF who receive OAC treatment, multi-level incentive interventions (including motivational interviews and targeted drug compliance counseling) significantly improve drug compliance and treatment duration [31]. AEGEAN trial explored the influence of education (using pamphlets and reminder tools) and telephone follow-up on compliance with apixaban. The results showed no difference in compliance between the routine nursing group and the intervention group [32]. However, these long-term proactive interventions may put a greater burden on healthcare providers. The alfalfa app, which integrates the functions of patient participation, education, consultation, and reminder, may improve compliance with drug treatment without increasing the burden.

By referring to similar studies [15, 33, 34], we set the time points for setting questionnaires at baseline, 1 month and 3 months. Our results showed that although the knowledge level of patients in the intervention group continued to increase significantly from 1 month to 3 months (62.5%(50.0%,75.0%) vs. 78.1%(68.8%,92.2%); P = .029), but the growth rate has slowed down. In addition, patient compliance levels appeared to have peaked at one month, with no significant difference between 1 and 3 months (7.8(6.8,8.0)vs. 8.0(7.0,8.0); P = .95), which might indicate that patients had basically completed medication education at 1 month. Future research designs may consider designing more observation points in 1 month to better describe the detailed change process of parameters related to medication education.

Besides improving patients’ knowledge and compliance through education and reminding, remote consultation is also one of the Alfalfa app’s main functions, allowing patients to monitor themselves and give timely feedback. Our research also found that self-monitoring and online feedback based on the Alfalfa app improved drug treatment compliance and anticoagulation satisfaction. Through the remote consultation function, clinicians/pharmacists can quickly know the latest physical condition of patients to re-evaluate the clinical risks of patients and adjust the dosage and use of various drugs. As a public platform for patient-patient and patient-medical staff direct dialogue and communication, the anticoagulant community is another feature of the Alfalfa app. In the anticoagulant community, patients can solve all kinds of problems about the disease or app. At the same time, patients often share relevant experiences and encourage each other. This virtuous circle can encourage patients to participate in self-management and greatly benefit from their treatment.

Hospital anticoagulation clinics have been established in China to optimize the use effect of OAC and reduce the risk of OAC. However, the medical resources of anticoagulation clinics are mostly concentrated in developed central cities. Still, a larger proportion of patients who need anticoagulation live in rural areas and areas with insufficient medical resources. Unfortunately, China’s medical and health resources are insufficient and unevenly distributed geographically [35]. Rural areas and areas with insufficient medical resources usually lack professional anticoagulation doctors or pharmacists [24], resulting in patients living in these areas facing significant risks of bleeding and thrombosis. Mobile medical technology breaks the limitation of time and space and provides new opportunities for anticoagulation therapy. If the Alfalfa app becomes widely available, it will enable most patients living in rural areas and areas with insufficient medical resources to enjoy high-quality online anticoagulation management services without leaving home.

This study has several limitations. First of all, Alfalfa App only showed an improvement in patient knowledge level, compliance, and satisfaction for 3 months. Whether it can improve patients’ long-term knowledge level, compliance and satisfaction need further research. Secondly, due to the small sample size, the effect of this trial on clinical outcome should be interpreted with caution. Finally, there is no cost-utility analysis of Alfalfa app remote management and traditional offline management, and the economic differences between them still need to be clarified.

Conclusion

Alfalfa App significantly improved the anticoagulation knowledge, drug treatment compliance, and satisfaction of AF patients. In oral anticoagulation management for AF patients, mobile medical technology that integrates the functions of patient education, remote consultation, and medication reminder may be helpful.