AF affects around 1–2% of the global population. The prevalence is anticipated to increase twofold in the next three decades due to the ageing population.1 In Malaysia, a 2016 registry showed that the prevalence of AF was one in 185 individuals.2 Individuals with AF have a fivefold risk of stroke, and one in five strokes occurs in this patient group.3
Since its introduction in 1954, warfarin has been the mainstay oral anticoagulant for non-valvular AF treatment, reducing stroke risk by two-thirds, irrespective of the baseline risks.4 Despite its extensive clinical use, it has several limitations, including the need for frequent international normalised ratio (INR) monitoring, a narrow therapeutic index, and frequent drug–drug and food–drug interactions.5
Direct oral anticoagulants (DOACs), such as dabigatran, apixaban, rivaroxaban and edoxaban, are an alternate therapeutic option introduced from 2010 onwards. DOACs have a more predictable pharmacokinetic profile, broader therapeutic index, significantly fewer drug interactions, and no food interactions.5 Landmark trials, such as RELY, ARISTOTLE, ROCKET-AF, and ENGAGE AF-TIMI 48, comparing DOACs to warfarin in patients with non-valvular AF generally demonstrated that DOACs are either superior or non-inferior to warfarin in reducing stroke and systemic embolism.6–8 Therefore, the prescription rate of DOACs increased. As patents for innovative DOACs come to an end, a broader penetration of DOACs can be anticipated, primarily attributed to the reduced costs of generic versions.
Although DOACs have a predictable dose response and are administered in a fixed dose regimen, numerous studies have demonstrated inter-individual variability of DOACs in both the plasma concentration and anti-thrombin or anti-factor Xa activities.9–11 In the Testa et al. study, all 330 patients from four anticoagulant clinics in Italy had highly variable plasma concentrations for apixaban, dabigatran and rivaroxaban. The coefficient of variation (CV) of their peak and trough plasma concentrations was 46% and 63%, respectively.12 That study also showed that non-valvular AF patients with a higher peak anticoagulant concentration were associated with greater bleeding complications, while low trough levels were associated with increased risk of thrombotic complications.13–15
Furthermore, real-world studies have indicated a relationship between plasma concentration and clinical outcomes, in which higher activities correlated with a greater number of bleeding events, and lower activities were associated with a greater number of thrombotic events. Despite using different parameters and instruments, studies have consistently shown this trend.9,16,17 For instance, Bernier et al. showed that patients experiencing bleeding complications had DOAC concentrations exceeding the 95th percentile of the mean, while those having thrombotic events had concentrations below the 5th percentile of the mean.18,19 Patel et al. questioned the prevailing idea of ‘one-size-fits-all’ dosing of DOACs, and their review highlighted the relationship between DOAC plasma concentration and clinical outcome, suggesting selective monitoring in specific clinical scenarios to enhance patient safety and treatment efficacy.20
Furthermore, the patient cohorts in the pivotal trials of DOACs over warfarin in patients with non-valvular AF were heterogeneous, and predominantly involved subjects from Europe and North America. Despite the availability of the assays, including point-of-care (POC) instruments, that can measure the pharmacodynamic response of DOACs, these methods have not yet achieved significant adoption in clinical practice, and routine monitoring of DOAC therapy is not currently recommended.
Therefore, we conducted this systematic scoping review of the published literature regarding POC monitoring in the era of DOACs to better understand the landscape of POC testing, and to identify potential gaps in the feasibility and challenges in implementation of POC testing and monitoring.
Methods
Arksey and O’Malley’s Scoping Review Framework
The methodology for this scoping review was based on the methodological framework established by Arksey and O’Malley and further refined by recent advances in scoping review techniques.21–23 Searches were conducted on 25 January 2024, using combinations of the following keywords: ‘DOAC’, ‘direct oral anticoagulant’, ‘NOAC’, ‘new oral anticoagulants’, ‘non-vitamin K anticoagulants’, ‘rivaroxaban’, ‘dabigatran’, ‘apixaban’, ‘edoxaban’, ‘POCT’, ‘point of care’, ‘point of care testing’, ‘rapid testing’, ‘bedside testing’, ‘laboratory-independent’ and ‘near patient testing’.24,25 These searches in three electronic databases (PubMed, Cochrane, and Scopus) yielded 409 records. After eliminating duplicates in accordance with the PRISMA guidelines (Figure 1), 342 records were evaluated for relevance to POC monitoring of DOACs. Following the removal of non-English-language articles and those not directly related to POC in DOAC monitoring, a total of 89 studies were included in the qualitative synthesis of the review.
Results
Synthesis of Evidence on POC Testing for DOACs
Publication Landscape
This review of literature spanning from 2009 to 2024 showed consistent interest in POC testing for monitoring of DOACs (Table 1). Analysis of the 89 publications showed that the highest number of studies, 53.93%, were conducted between 2015 and 2020, reflecting a rising focus as the usage of DOACs increased. Publications from 2009 to 2014 accounted for 20.22%, while those from 2021 to 2024, covering the COVID-19 pandemic and post-pandemic phase, comprised 25.84%. This suggests a slightly reduced but ongoing research. There was a total of 17 published clinical trial articles (Supplementary Table 1) and 29 published non-clinical trial articles (Supplementary Table 2). In general, the studies on POC testing for DOACs demonstrated substantial progress from 2009 to 2024.
Between 2009 and 2014, efforts were focused on developing assays to quantitatively measure DOAC therapy. This foundational work established the groundwork for a surge in research activity from 2015 to 2020. During this time, research expanded beyond assay development to investigate inter-individual variability and the correlation between the pharmacokinetics and pharmacodynamics of DOACs and the incidence of thrombotic and bleeding events in patients with non-valvular AF. The increase in studies was driven by the increased usage of DOACs and the need for potential monitoring of the anticoagulation therapy across various clinical settings, including emergency scenarios in high-risk non-valvular AF patients. However, the onset of the COVID-19 pandemic redirected global research priorities towards urgent public health concerns, which subsequently slowed the momentum in DOAC monitoring research. As a result, research output on POC testing decreased in the post-pandemic period. Therefore, it is crucial to refocus research efforts on POC testing technologies for DOAC monitoring in this patient group.
Geographically, Europe has the highest proportion of first authors with 68.54%, followed by North America at 21.35%. The contributions from Oceania, Asia and South America are comparatively lower. The DOAC landmark trials were initiated by European research teams and the majority of the assay’s development was conducted in Europe. Most of these studies were conducted by academic institutions, with 70.79% of first authors being affiliated with universities. This highlights the strong academic focus on enhancing DOAC therapy monitoring through innovative POC testing methods.
Role of POC Testing in DOAC Therapy Monitoring
The use of DOACs led to a significant advance in the development of assays to quantify DOAC drug level (pharmacokinetic) and activity (pharmacodynamic). In 2013, the Scientific Committee on Control of Anticoagulation of the Scientific and Standardization Committee of the International Society on Thrombosis and Haemostasis recommended essentially two types of tests to monitor pharmacokinetics and pharmacodynamics of DOACs: quantitative laboratory measurement of plasma levels by chromatography–mass spectrometry (LC-MS/MS); and laboratory measurement of coagulation tests.26 High-performance liquid chromatography (HPLC) is the gold standard to quantify DOAC level in subjects’ plasma. However, it could not measure the pharmacodynamic properties of DOACs. Therefore, multiple assays from bench to bedside had been developed to more comprehensively assess DOAC activity.
Evidence Highlights
Evaluation of Existing Methods
During the transition from warfarin to DOACs, methods adopted from warfarin monitoring, that is, POC INR, and prothrombin time (PT) test with an INR (PT/INR), were used. Additionally, various studies in non-valvular AF patients prescribed DOACs identified several disadvantages of these methods. Building upon this foundation, clinical research has adapted assays originally developed for heparin products. Measurement of anti-thrombin activity and anti-factor-Xa activity were key parameters used to gauge the pharmacodynamic effect of dabigatran, rivaroxaban, apixaban and edoxaban. Laboratory-based assays of anti-thrombin activity and anti-factor-Xa activity were developed to measure pharmacodynamic properties of DOACs. Some of these assays are now commercially available for assessing DOAC activity. Comparative studies of the HPLC method and these assays demonstrated a moderate to high degree of correlation, indicating the potential of the adapted assays in measuring DOAC activity.18,27–29
The advance in POC testing for these parameters is a significant step forward in the potential monitoring of DOAC therapy. It could enable timely, informed decisions regarding anticoagulation management, particularly in acute or urgent care settings in which rapid assessment is crucial for optimal outcomes.30 Nevertheless, the potential benefits of such monitoring, either routine or one-off, particularly in high-risk patient groups, could be explored to optimise the treatment efficacy and safety, tailoring anticoagulation therapy to individual patient needs.
Development of New Methods
From activated partial thromboplastin time and prothrombin time to chromogenic assays, the development of recent methods such as viscoelastometry and urine dipsticks has overcome several limitations associated with previous tests. Notably, the DOAC dipstick test has shown high accuracy, with a 100% detection rate of apixaban, rivaroxaban and dabigatran in urine, without interference from other anticoagulants, such as heparin, nadroparin, fondaparinux or warfarin.31 A study by Papageorgiou et al. confirmed the efficacy of the dipstick at a plasma threshold of ≥30 ng/ml. They recommended further research to explore its use in other anticoagulants, to enhance its clinical applicability.32
With regard to viscoelastometry tests, one of them, ROTEM (rotational thromboelastometry), and particularly the clotting time and the maximum clot firmness variables (when measured using the assay of tissue factor activation [i.e. extrinsic pathway of coagulation, or Extem], and assay of contact activation [i.e. intrinsic pathway of coagulation, or Intem]), showed a strong correlation with plasma dabigatran level, suggesting its potential to quantify DOAC activity. However, these results still need validation in broader patient groups.33 A review study on viscoelastic tests showed that the DOAC-specific assays were more sensitive than the non-specific assays.24 Strong correlations were observed between plasma concentration of rivaroxaban and dabigatran and the clotting times measured using several assays. However, normal clotting and reaction times do not reliably exclude significant residual DOAC levels, limiting the clinical utility of these tests. The authors concluded that although viscoelastic assays provide critical, rapid POC data on DOAC activity, they fall short compared with standard anti-Xa activity laboratory tests in detecting residual DOAC levels.24 This limitation suggested a need for continuous improvement of these assays for clinical application.
Improvement of Existing Methods
Subsequently, Schäfer et al. (2021) investigated the integration of machine learning with POC testing in the monitoring of DOAC therapy. This method used artificial intelligence (AI) to improve the predictive accuracy of test outcomes. By incorporating AI into the evaluation of data from existing assays, models could be developed to provide more accurate predictions of patients’ responses to DOACs. This advance increased the reliability of the monitoring of therapeutic levels and the use of tailored and efficient DOAC therapy in non-valvular AF patients.34
Discussion
The rapid POC monitoring of DOACs prescribed in non-valvular AF could be incorporated into the anticoagulant treatment algorithm, tailoring DOAC selection and dosage to the individual’s needs and risks, particularly in those at elevated risk of bleeding or thrombosis as measured using scores, such as HAS-BLED and CHA2DS2-VASc. It is also essential that the POC monitoring of DOAC therapy be cost-effective. Although the upfront costs for POC instruments and training (although involving minimal laboratory work) may be considerable, the long-term benefits of incorporating POC monitoring into clinical practice aligns with the goal of personalised medicine, from the choice of DOACs to the dose.
In addition, patient adherence is an essential component of anticoagulation therapy. While POC testing can assess the steady-state plasma level of the drug at a specific time point, it can only reflect the true plasma level or activity when the patient is consistently adherent to their regimen. Variability in adherence can lead to misleading assessments. Adherence assessment and counselling play an important role in patients’ clinic visits to achieve optimal therapeutic outcomes.
Another important aspect is the genetic variation in the proteins responsible for the metabolism of DOACs, which plays a role in the inter-individual variability of the responses. There is a gastrointestinal re-secretion over a P-glycoprotein (P-gp), coded by the ABCB1 gene following absorption of DOACs in the gut. Strong inducers of P-gp will reduce DOAC bioavailability, thereby affecting the plasma concentration, and subsequently affecting the bleeding and thrombotic risks; and vice versa.35 In the Paré et al. (2013) and Sychev et al. (2018) studies, genetic variants of the esterase gene CES1, involved in dabigatran etexilate metabolism, and ABCB1 were found to be associated with decreased trough dabigatran plasma concentration, and also higher peak plasma concentration, respectively.36,37 This is a relatively new field of research, and in Malaysia data are scarce. An association was demonstrated between DOAC drug level and genetic variations. However, there are limited data linking genetic variants with anti-factor Xa or anti-thrombin activities and outcomes, either bleeding or thromboembolic events. In contrast, these POC assays measure the phenotypes, i.e. the DOAC drug level or drug response, which are partly attributed to the drug transporter ABCB1 and drug metabolism CES1 gene.
Practices in Southeast Asian Countries and Malaysia
In Southeast Asia, countries such as Singapore, Thailand and Vietnam have prescribing patterns similar to those in Europe, primarily using a one-size-fits-all dosing strategy, with adjustments based only on factors such as creatinine level. However, there is a notable gap in the literature regarding the monitoring strategy for DOACs. This is consistent with our review findings. We identified only two articles relevant to Asia, including one from Malaysia.
In Malaysia, warfarin is predominantly prescribed in non-valvular AF patients due to its lower cost compared with DOACs. The monthly treatment cost of warfarin is approximately US$1, whereas DOAC cost is approximately US$50 per month, highlighting a significant difference in treatment expense. DOAC prescription has an allocation limit (quota item), with prescriptions restricted to a specific number of available slots and based on whether patients meet established prescribing criteria. Despite a gradual shift towards prescription of DOACs for non-valvular AF patients, warfarin clinics with routine INR monitoring remain highly active. However, given the advantages of DOACs over warfarin and the availability of generic versions of DOACs following expiration of the original patents, there is potential for a further shift in the prescribing pattern towards DOACs. Nevertheless, this shift involves considerations beyond treatment cost: the necessity for POC monitoring to tailor treatment to the individual patient’s needs and achieve net clinical benefit is also crucial. Subsequently, a cost-effectiveness study is required on the dynamics of DOACs replacing warfarin, including consideration of their respective monitoring strategies.
We referenced a local ongoing study conducted by our team at the Sarawak Heart Centre in Samarahan, Malaysia.38 This centre serves a population of 2.45 million within a radius of 100 km2. At Sarawak Heart Centre, approximately 36.5% of patients requiring anticoagulants were prescribed DOACs, primarily dabigatran. Warfarin remains the mainstay treatment due to cost constraints. Given the constraints of limited resources and quota-based prescriptions at Sarawak Heart Centre, which affect the anticoagulant prescribing pattern in this region, along with existing literature on the inter-individual variability of DOACs, we initiated a study on the diverse multiethnic population of our non-valvular AF patients prescribed with dabigatran.38 The preliminary findings of this study were shared at The National Heart Association Malaysia Annual Scientific Meeting 2023, Malaysia.38 A total of 304 non-valvular AF patients on dabigatran were recruited and their whole blood samples were taken. The POC instrument, Clotpro, was used to measure the pharmacodynamic activity of dabigatran, that is, the clotting time. The drug level of recruited patients was assessed using the gold standard, LC-MS/MS. High inter-individual variability was observed in both drug level (CV = 112.5%) and clotting time (CV = 52.3%). The drug level and clotting time were log-transformed and a correlation was observed between the log-transformed drug level and log-transformed clotting time (Pearson correlation, r=0.55, p<0.0001; Figure 2). A total of 12.5% of the subjects were in the lowest quartile for both log-transformed drug level and clotting time. At 1-year follow-up, the overall rate of major adverse cardiovascular and cerebrovascular events was 13.5%, and were more prevalent in subjects in the left lower quadrant compared to the rest of the groups (18.4% versus 12.8%; p=0.341). Six subjects (2%) had ischaemic stroke. Patients in the left lower quadrant (i.e. both log clotting time and log drug level less than or equal to the first quartile) had a ninefold increased risk of ischaemic stroke (HRadjusted 8.7; p=0.024) despite being on dabigatran therapy.38,39 This highlights the significant variability and increased stroke risk in certain groups in which the drug activity and plasma level are suboptimal. These findings underscore the potential of using the instruments available to monitor anticoagulation therapy.
In Sarawak, where access to healthcare is limited by geographical challenges, POC testing could be especially beneficial in public health clinics. There are more than 270 public health clinics scattered across the state, facing significant logistical hurdles in providing timely and effective care. Furthermore, the adoption of POC testing, potentially in the medication therapy adherence clinic for anticoagulation, could streamline the workflow in clinics, reduce the need for multiple patient visits, and minimise the dependency on centralised laboratory services, which are often delayed in rural settings. This would not only improve patient outcomes by ensuring that personalised therapeutic decisions are made by clinicians and pharmacists based on the latest real-time clinical data, but also enhance the overall efficiency of healthcare delivery in Sarawak.
Future Perspectives and Challenges
Integrating POC testing of DOAC therapy into clinical practice could offer real-time, on-site assessment of coagulation status, enhancing patient management, particularly in urgent care settings. However, ensuring the accuracy and reliability of the assays remains a big challenge. One significant gap is the establishment of an optimal therapeutic window for DOACs with available instruments. Although DOACs have a broader therapeutic index than warfarin, the cut-offs for the optimal balance between thrombotic and bleeding risks of these different assays have yet to be defined. Furthermore, as anticoagulation therapy evolves with the increased use of DOACs, developing cost-effective monitoring strategies becomes crucial. While the initial costs for DOAC monitoring technologies such as LC-MS/MS and laboratory-based or POC coagulation tests are higher than those for traditional INR monitoring for warfarin, the long-term potential for cost savings could be significant. Additionally, integrating AI, for example machine learning, could help to develop more affordable and precise assays.
To effectively implement POC testing, it is important to establish cost-effective tests with clearly defined cut-offs for the therapeutic window and ensure that patient counselling and adherence are addressed. Once these elements are in place, POC testing of DOAC therapy could potentially reduce hospitalisations and overall healthcare costs, while tailoring anticoagulant therapy to the individual patient.
Conclusion
With the anticipated emergence of generic DOACs in the near future, a shift towards more widespread prescribing of DOACs in non-valvular AF patients is expected. Monitoring of DOACs is particularly important in high-risk patient groups, as highlighted by the literature on the inter-individual variability of DOAC level with the associated bleeding and thrombotic risks with supraoptimal and suboptimal levels of DOACs. This review highlights significant advances in POC testing for DOACs over the past decade, which are crucial for refining anticoagulant treatment algorithms and enhancing the management of non-valvular AF patients on DOAC therapy in the future.
Our scoping findings have identified a substantial gap in the literature in Asia, particularly regarding the application and studies of POC monitoring for DOACs, and highlight a need for further research and implementation. In Malaysia, with its diverse population and geographical challenges, POC monitoring is potentially useful. This is especially relevant in the state of Sarawak, where remote and rural areas could benefit significantly from precise and timely assessment of anticoagulation status enabled by POC testing in the anticoagulation medication therapy adherence clinic. Addressing the identified literature gaps, which include validation of these technologies and determination of the optimal DOAC therapeutic window, is a crucial step towards improving healthcare outcomes for Malaysia’s non-valvular AF patient population.