Technology-Based Interventions in Oral Anticoagulation Management: Meta-Analysis of Randomized Controlled Trials

Introduction

Oral Anticoagulation Management

Oral anticoagulants (OACs) have been used for decades, especially warfarin, which has been in use since the 1950s. Until recently warfarin has been a fundamental drug in clinical anticoagulant therapy, and although the results of two previously published meta-analyses of atrial fibrillation suggested that novel OACs are not inferior to warfarin in terms of safety and efficacy [1,2], warfarin has the advantage of being inexpensive and having a wide range of indications. Warfarin is widely used in conditions that are prone to thrombosis, such as atrial fibrillation, irregular heartbeat, myocardial infraction, artificial heart valve replacement, recurrent stroke, deep vein thrombosis, and pulmonary embolism [3]. However, it remains challenging to balance effectiveness and safety in treatment [4]. Statistical data from relevant literature estimated that over 6 million patients in the United States have received anticoagulant therapy [5], leading to increased risks of bleeding, thromboembolism, hospitalization, and mortality. Currently, rough statistics indicate that about 1 in 10 surgical patients in the United States receive OACs [6], and approximately 2 million people start warfarin therapy in the United States every year [7]. Therefore, effective anticoagulation management measures are urgently needed.

Technology-Based Interventions for Oral Anticoagulation Management

With the continuous development of telemedicine health service technology, more and more technological devices have been applied to help in the management of patients with chronic diseases, including those with diabetes, hypertension, heart failure, and others [8]. Telemedicine is defined as a long-distance medical practice that can be characterized by health services through a wide range of technical applications and services [9]. It is a new way of providing high-quality resources to primary medical institutions through remote consultation on the basis of internet convenience. Through this approach, medical services and medical activities are mainly carried out through computers, various remote communications, medical technology, and medical equipment. These remote forms of communication enable contact between patients and medical personnel, medical institutions, and medical equipment. This process, in turn, can assist in diagnoses, treatment, monitoring, and follow-up [10]. As a result, telemedicine has become an increasingly popular model for providing accurate international normalized ratio (INR) monitoring for patients taking warfarin. Other terms used to describe telemedicine include connected health [11], mobile health (mHealth) [12], and electronic health (eHealth) [13], which are collectively called technology-based interventions [14]. There is interest in learning whether these interventions can assist medication management and enhance patient compliance, with results suggesting some benefit [15].

Aim of the Study

Telemedicine technology has the potential to enhance multiple aspects of anticoagulant therapy management [16], including patient education, symptom monitoring, follow-up, and encouragement and tracking of medication adherence, given its accessibility by telephone, internet, voicemail messaging, and apps. However, no systematic review or meta-analysis has been published to summarize what is currently known on this topic. Thus, we aimed to evaluate the effectiveness and safety of technology-based interventions in OAC management by performing a meta-analysis.

Methods

Literature Search

Eligible studies were identified by searching PubMed, EMBASE, Cochrane, Cumulative Index to Nursing and Allied Health Literature, PsycINFO, Scopus, and other relevant databases, and the results were combined using literature traceability methods. Searches were conducted on November 1, 2019. Search keywords included warfarin therapy+, oral anticoagulation management+, telephone, eHealth, apps, and telemedicine+. Search strategies are detailed in Multimedia Appendix 1. Only papers written in English were considered. No restriction regarding publication date was applied.

Inclusion and Exclusion Criteria

Inclusion Criteria

The studies included in the analysis met the following criteria: (1) the study was a randomized controlled trial (RCT); (2) the subjects were taking warfarin; (3) technology-based interventions were used to manage OACs; (4) results included the time in therapeutic range (TTR), bleeding, and thromboembolism events; and (5) the results were reported in English.

Exclusion Criteria

Exclusion criteria included: (1) studies that were retrospective, observational, reviews, model research, literature reviews, or conference summaries; (2) the results of the study did not involve the TTR, bleeding, and thromboembolism events; (3) the study was a duplicate report; and (4) technical intervention was only used as a means to collect data.

Document Screening and Data Extraction

All references were initially screened by title and abstract by 2 reviewers for relevance. Finally, full-text analysis for eligibility was performed independently by 2 authors, HD and CL. Disagreements were discussed and resolved by consensus or third-party arbitration.

The required data were extracted by a researcher using a literature data extraction table, and another researcher confirmed the accuracy and authenticity of the data. The extracted content included (1) basic information of study, such as research topic, author, and date; (2) baseline characteristics of the study subjects, such as sample size, median age, and sex; (3) follow-up time for interventions; (4) efficacy and safety information after interventions, such as TTR, bleeding, and thromboembolism events; and (5) other outcome indicators, including time within expanded target INR range, mortality, and hospitalization events.

Literature Quality Evaluation

Risk of bias assessment of the included RCTs was performed using the Cochrane risk of bias tool based on the Cochrane Handbook for Systematic Reviews of Interventions' literature evaluation criteria [17].

Statistical Analysis

The meta-analysis of RCTs was performed using RevMan (version 5.3; Cochrane) software. Heterogeneity was assessed using a chi-square test, and quantitative analysis was performed using I². Values of P≥.05 and I²≤50% were considered to represent no heterogeneity, in which case a fixed-effect model was used. If P<.05 and I²>50%, a random-effects model was used [18].

Results

Search Results

A total of 6784 papers were retrieved from the systematic literature search, and 9 papers were retrieved by other means, totaling 6793 papers. After removing duplicate studies, 2 authors independently reviewed and excluded another 5652 studies that did not meet predetermined selection criteria, based on the title and abstract of each paper. After reading 33 eligible full-text studies, 18 were excluded and 15 were selected for inclusion in the analysis [12,19-31]. The systematic search results are shown in Figure 1.

‎View this figure

Basic Characteristics and Quality Evaluation of the Literature

The 15 RCTs included a total of 2218 patients: 1110 patients in the technology-based intervention group and 1108 patients in the traditional intervention group. General information (such as sample size, median age, sex, and specific intervention measures) from the studies is shown in Table 1. The Cochrane systematic evaluation method was used for quality evaluation, and overall, the included studies had low risk of bias and were high quality, as shown in Figure 2.

‎View this figure

Meta-Analysis

Time Within Target INR Range

Of the 15 included studies, 9 included TTR values, for a total of 1043 patients. A random-effects analysis was used for the meta-analysis of these 9 RCTs. As shown in Figure 3, the TTR of the technology-based intervention group was significantly higher than that of the control group (mean difference [MD]=6.07; 95% CI 0.84-11.30; I²=72%; P=.02).

‎View this figure

Bleeding Events

Of the 15 included studies, 14 included major bleeding values, for a total of 2139 patients. A random-effects model was used for the meta-analysis of these 14 RCTs. There were fewer major bleeding events in the technology-based intervention group than in the control group, but the difference was not statistically significant. (MD=1.02; 95% CI 0.78-1.32; I²=0%; P=.90). There were 5 papers that included minor bleeding values, for a total of 519 patients. A random-effects analysis model was used for the meta-analysis of these 5 RCTs. There were fewer minor bleeding events in the technology-based intervention group than in the control group, but the difference was not statistically significant (MD=1.06; 95% CI 0.77-1.44; I²=41%; P=.73). Major and minor bleeding event analyses are shown in Figure 4.

‎View this figure

Thromboembolism Events

Of the 15 included papers, 11 had thromboembolism values, for a total of 1959 patients. A random-effects model was used for the meta-analysis of these 11 RCTs. As shown in Figure 5, there were fewer thromboembolism events in the technology-based intervention group than in the control group, but the difference was not statistically significant (MD=0.71; 95% CI 0.49-1.01; I²=0%; P=.06).

‎View this figure

Sensitivity Analysis Results

The data were analyzed with fixed- and random-effects models, and the consistency of these results reflects the reliability of the combined results to some extent. The two effect models were used to analyze the combined effect of each risk factor and calculate 95% CIs. The results were similar, indicating that the results of this study are stable.

Other Results

The time within expanded target INR range of the technology-based intervention group was higher than that of the control group, but the difference was not statistically significant (MD=2.13; 95% CI –1.22 to 5.49; I²=35%; P=.21; Figure 6). There were fewer mortalities and hospitalization events in the technology-based intervention group than in the control group, but the difference was not statistically significant (mortality MD=0.61; 95% CI 0.26 to 1.41; I²=0%; P=.25; hospitalization MD=1.02; 95% CI 0.85 to 1.23; I²=23%; P=.84; Figure 7). There were 3 papers [28,30,32] that mentioned that patients' knowledge of anticoagulation was significantly improved through telemedicine intervention. There were 4 studies [22,25,26,28] that involved quality-of-life assessments; quality of life was higher in the intervention group than in the control group and it improved from baseline. There were 2 papers [21,25] that mentioned the higher costs in the technical intervention group than in the control group, and 1 paper [24] found that the costs for the nontechnical intervention group were higher. Patient satisfaction surveys were conducted in 5 studies [24,25,28-30], all of which indicated that the technical intervention group had higher satisfaction. There were 4 papers [25,26,28,32] that included data on INR tests, and the technical intervention group tended to have larger numbers of INR tests performed.

‎View this figure

‎View this figure

Discussion

Principal Findings

We performed a meta-analysis of 15 RCTs to determine whether technology-based interventions were beneficial for patients compared with traditional interventions in oral anticoagulation management. TTR reflects the quality and effectiveness of warfarin and is an important determinant of bleeding and thromboembolism in anticoagulation management. A total of 9 studies were involved in the TTR analysis, and almost all studies reported that technology-based interventions were effective. The number of bleeding (including major and minor) events in the technology-based intervention groups was lower than that in the control groups, but the difference was not statistically significant. Outcomes were similar for the thromboembolism events. Time within expanded target INR range, mortality, and hospitalization were not significantly different between the technology-based management model and traditional management model. However, patients’ anticoagulation knowledge, patient satisfaction, quality of life, and number of INR tests performed were positively affected by technology-based management.

Oral anticoagulation management can be challenging when ensuring safety of the patient; as long as warfarin is used, regular monitoring is necessary [33]. The technology-based interventions used to manage oral anticoagulation typically include multiple methods, including telephone, internet, voicemail messaging, and apps. These tools are readily available in this era of technological advancement, and they reduce requirements for capacity and waiting time in outpatient clinics. Through this process, mHealth technologies have systematically demonstrated that they can help improve early diagnosis and treatment of atrial fibrillation and significantly reduce the cost of illness associated with atrial fibrillation [34]. One such example of mHealth technology is SintromacWeb, a management system of internet-based telecontrol tools that has been recognized to have better effectiveness and safety at oral anticoagulation therapy management than a conventional approach does [35]. Technology-based interventions can effectively and professionally teach patients about anticoagulation, provide interpretation of indicators, and provide a timely dose adjustment program.

Comment on Results

It is worth mentioning that the mean age of the patients included in each study was over 60 years, so these patients may have had a relatively poor grasp on learning how to use novel websites and apps. The telephone is an ultra-portable electronic device, and it was chosen as the intervention tool in most of the 15 RCT studies included. Telephone intervention is especially suitable for older patients who are not accustomed to using computers or online services, or who may have difficulties in vision, finger dexterity, and mental state.

Our overall results showed that technology-based interventions significantly improved the effectiveness of oral anticoagulation management but did not significantly increase the safety or other results of oral anticoagulation management. These results may be because of the different follow-up times, which ranged from 1 month to 36 months. The greatest difference between the intervention and control groups in the frequency of major bleeding events occurred in the first week (when anticoagulation therapy was started in the hospital) and after the first month of therapy (when outpatient therapy was presumably stabilized). The 2 groups differed little during the second through fourth weeks of therapy, when patients were generally discharged and warfarin was first monitored in the outpatient setting [19]. That is to say, most patients might attach great importance to any type of intervention early on and then slowly relax later.

Limitations

We acknowledge that our research has some limitations. The included studies had relatively small sample sizes and varying follow-up times, with some as short as 1 month. Warfarin is the oldest and best known OAC, and as long as warfarin is used, monitoring is necessary. Therefore, in order to more comprehensively assess the safety of OACs, longer follow-up times are needed. Furthermore, patients’ adherence to technology-based interventions may be lacking. It would be ideal if intervention compliance studies were added during follow-up. Despite these limitations, this study is the first to comprehensively analyze the effect of technology-based interventions on anticoagulation management. Therefore, larger sample sizes and longer clinical RCTs are needed to further evaluate the impact of technology-based interventions on anticoagulation management.

Comparison With Prior Work

To the best of our knowledge, this is the first meta-analysis of RCTs that evaluates the effectiveness and safety of technology-based interventions for oral anticoagulation management. In published systematic meta-analyses and reviews, most of the literature evaluating the effects of technology-based interventions has focused on atrial fibrillation [34] and cancer [36]. Most studies regarding the administration of OACs have focused on the effects of managers on the effectiveness and safety of anticoagulation management, such as the differences between pharmacist management and physician management or between patient self-management and anticoagulation clinic management [37-41]. The intervention method we chose is a technology-based approach that might have also been included in other studies, such as pharmacist-managed or self-managed interventions [37,41].

Although the study of telemedicine is not specific to OACs and is primarily qualitative, previous studies have highlighted similar challenges [42-44]. Clinical pharmaceutical care provided by Niznik et al [43] through telemedicine (mainly by telephone) in inpatient or outpatient settings was found to have an overall positive impact on outcomes related to clinical disease management, patient self-management, and adherence to the management of various chronic diseases. The common ground in the studies that had a positive impact on outcomes was the use of continuous or regular patterns of care, including telephone interaction and frequent monitoring and intervention [43]. Xia et al’s [44] study was conducted to explore the effects of online and offline management of anticoagulants on therapeutic efficacy and adverse reactions. Considering the convenience and economy of technology-based interventions, online anticoagulation management is more suitable for patients with stable conditions and for whom transportation may be difficult, such as those with disabilities or who live far from the hospital. Lee et al’s study [42] suggested that most outcomes for telemedicine were similar to those for conventional medical care, but the incidence of major thromboembolism events was significantly lower in the telemedicine group. However, the papers they included had a higher risk of bias and were of a lower quality study design, and the level of evidence supporting this conclusion was very low.

Conclusions

This meta-analysis explored the effects of technology-based interventions on oral anticoagulation management. The results demonstrate that the technology-based intervention group had significantly improved TTR compared with the traditional intervention group and that there were no significant differences between the 2 intervention models in terms of time within expanded target INR range and the incidence of major bleeding events, minor bleeding events, thromboembolic events, mortality, and hospitalization. Of the different management options, telephone intervention was found to be the most widely used and most convenient means of technology-based interventions; it enables patients to get a professional reply quickly and is not restricted by the patient’s ability to use the internet. OAC management through technology-based intervention appears to be superior to OAC management through traditional intervention and may provide more convenient and higher quality anticoagulation services for patients. Further research is needed to explore more optimal technology-based interventions in oral anticoagulation management in a wider array of health care settings.