Abstract

Warfarin remains a widely prescribed oral anticoagulant despite its narrow therapeutic index and its association with life-threatening bleeding events. This systematic review examined the impact of clinical risk factors on international normalized ratio (INR) control in patients receiving Warfarin therapy. Following the PRISMA-P guidelines, we searched four electronic databases (PubMed, SCOPUS, Embase, and Web of Science) using Medical Subject Headings to identify eligible cohort studies published between 2016 and 2020. A total of 15 original research articles met the inclusion criteria. The risk of bias was assessed using the Newcastle-Ottawa scale. This review identified several clinical risk factors influencing INR control, including drug interactions with Warfarin, herbal supplements, dietary intake of vitamin K, and multiple comorbidities (anemia, heart failure, chronic kidney disease, psychiatric disorders, prosthetic valve), kidney transplant recipient, hemodialysis, and hypoalbuminemia. These clinical risk factors adversely affect the time in therapeutic range, thereby increasing the risk of adverse outcomes such as thromboembolism, major bleeding, and mortality. Overall, these clinical risk factors compromise Warfarin efficacy by contributing to subtherapeutic or supratherapeutic INR levels. Identifying and addressing these clinical risk factors prior to initiating Warfarin therapy is essential to achieving optimal INR control.

Keywords

Clinical, Risk Factor, International Normalized Ratio, Warfarin

Share and Cite:

1. Introduction

Warfarin is an anticoagulant that works by competitively inhibiting the vitamin K epoxide reductase complex 1 (VKORC1), resulting in reduced hepatic activation of clotting factors II, VII, IX, X, proteins C and S [1]. It has been widely used for over six decades as the primary treatment to reduce the risk of cardiovascular events such as venous thromboembolism (VTE) and ischemic stroke in patients with atrial fibrillation (AF) [2]. Warfarin is characterized by a narrow therapeutic index, substantial interpatient variability in metabolism, and numerous potential food and drug interactions [3] [4]. Despite the proven benefits of non-vitamin K-dependent oral anticoagulants (NOACs), Warfarin continues to be used, particularly in patients with rheumatic or valvular AF, prosthetic valve replacement, hypercoagulable states or those with limited financial access to newer medications [5]. Regular monitoring of the international normalized ratio (INR) and appropriate dose adjustment are essential to ensure the safety and effectiveness of Warfarin therapy [6].

INR is a standardized and widely accepted measure of anticoagulation intensity, introduced approximately 2 - 3 decades ago. In real-world clinical settings, subtherapeutic and supratherapeutic INR levels are frequently observed in a substantial proportion of patients (approximately 30% - 50%) on routine Warfarin therapy. Poor INR control markedly increases the risk of thromboembolic complications, whereas supratherapeutic INR levels elevate the risk of major bleeding complications [7]. For most indications, maintaining an INR within the target range of 2.0 - 3.0 is considered optimal for ensuring both safety and efficacy of Warfarin therapy [8]. Excessive anticoagulation increases bleeding risk, while subtherapeutic anticoagulation predisposes patients to thrombosis. Furthermore, spontaneous INR elevations in critically ill patients may be associated with increased mortality due to over-anticoagulation [9]. A multi-country study highlighted the challenges of maintaining optimal INR control, even in clinical trial settings [10].

Time in therapeutic range (TTR) quantifies the percentage of time during which a patient’s INR remains within the target therapeutic range and serves as a reliable indicator of INR control quality [11]. A TTR between 60% and 75% is considered adequate, according to current recommendations [10]. The incidence of adverse events, including thromboembolism, major bleeding, and mortality, is significantly higher in patients with a lower TTR group (≤69%) compared to those with a higher TTR (≥70%) [12]. Notably, each 1% increase in TTR has been associated with a 0.46% reduction in recurrent thromboembolic events and a 0.30% reduction in major bleeding annually [13]. Despite the availability of direct oral anticoagulants (DOACs), Warfarin remains widely prescribed due to its accessibility and cost-effectiveness [14]. Additionally, Warfarin continues to be the only recommended anticoagulant for patients with prosthetic heart valves [15].

The therapeutic goal of Warfarin therapy is to administer the lowest effective dose that maintains INR within the target range of 2.0 - 3.0 [16]. However, achieving consistent INR control within the target range is challenging due to the influence of various clinical and non-clinical risk factors. Our previous work has explored how non-clinical risk factors contribute to poor INR control [17]. This review aims to analyze the impact of clinical risk factors on INR control among patients receiving Warfarin therapy. Given the persistent global reliance on Warfarin and the serious implications of INR mismanagement, publication of this review is essential. It may contribute to improved clinical practice by providing an evidence-based resource that supports decision-making, enhances risk mitigation, and ultimately improves patient care in anticoagulation management.

2. Materials and Methods

We conducted this systematic review in accordance with the preferred reporting items for the systematic review and meta-analysis (PRISMA) 2020 guidelines [18] (see Checklist A1).

2.1. Search Strategy and Study Selection

We conducted an advanced systematic literature search in four electronic databases (PubMed, SCOPUS, Embase, and Web of Science) to identify eligible cohort studies published between 2016 and 2020, a period during which clinical guidelines for Warfarin use were well established, and routine INR monitoring practices remained consistent [19].

In these four databases, we searched the original research articles using the Medical Subject Headings (MeSH) and keywords according to the PICO concept. The target population was “adult patients”, which was searched under two sub-concepts: population-1 “adult” and population-2 “patient”. The intervention was “Warfarin”. For the outcome, we searched the databases under two concepts: outcome-1 “risk factors” and outcome-2 “International Normalized Ratio”. Although “clinical risk factors” would be the ideal search term, no subject heading specially addressed it; thus, we used “risk factor” as a proxy.

To construct the final search string, we first combined the subject headings and keywords for population-1 “adult” and population-2 “patient” using the “OR” Boolean Operator as they belonged under the same concept. Then, we added the intervention “Warfarin” using the “AND” Boolean Operator. Lastly, we combined the subject heading and keywords for outcome-1 “Risk Factor*” and outcome-2 “International Normalized Ratio” using the “OR” Boolean Operator as they also belonged under the same concept. The complete search strings, both by individual concepts and combined under PICO (with and without filters), are presented in Table A1 in Appendix.

We initially identified 1202 articles across the four databases. After removing duplicates, 989 articles were selected to be screened based on titles and abstracts. The first two authors [TR and NKJ¹] independently screened articles based on titles and abstracts and selected 88 eligible articles for full text review based on the inclusion and exclusion criteria. Finally, we included 13 original research articles for data extraction and analysis (Figure 1).

Figure 1. PRISMA chart used for the selection of articles.

2.2. Inclusion and Exclusion Criteria

We included all original research articles that employed a cohort design, either prospective or retrospective, investigated the effects of clinical risk factors on INR control, involved adult patients aged 18 years and above as study population, patients receiving Warfarin, and were published in English. We excluded studies that focused solely on non-clinical factors, included study population under 18 years of age, were unpublished reports, case reports, or non-cohort observational studies like cross-sectional studies, and reviewed articles, including systematic reviews.

2.3. Data Extraction

The first two authors [TR and NKJ] independently extracted data using NVivo software and entered information into a structured data collection form. The extracted data included: Authors’ names, country of study, year of publication, study design, sample size, primary findings, and conclusion. They further categorized the extracted data under specific themes and subthemes, such as: Drug interactions with Warfarin, herbal supplements, multiple comorbidities (anemia, heart failure, prosthetic valve, malignancy, psychiatric disorders, diabetes, chronic kidney disease, respiratory disease), kidney transplant recipient, hemodialysis and hypoalbuminemia. Finally, the data was included for analysis when both the authors came to a general agreement through discussion. They invited the third author, either NKP or CGY², to join when they could not reach to an agreement.

2.4. Quality Assessment

We assessed the methodological quality of each included cohort study using the Newcastle-Ottawa Scale (NOS). The NOS is a validated tool specifically designed for evaluating the quality of non-randomized studies, including both cohort and case-control designs. We used the version of NOS tool adapted for cohort studies, as our included studies were either prospective or retrospective cohort studies in design. For quality assessment of cohort studies, this measurement tool is based on three broad domains: Selection (maximum of 4 points: ¹representativeness of the exposed cohort, ²selection of the non-exposed cohort, ³ascertainment of exposure, and ⁴demonstration that the outcome measure was not present at the start of the study), Comparability (maximum of 2 points: ¹comparability of cases and controls based on the design or analysis with a major factor, and ²comparability with additional factors) and Outcome (maximum of 3 points: ¹assessment of the outcome, ²follow-up length, and ³adequacy of follow-up of the cohorts) [20].

Each item was scored as 1 point if the criterion was clearly satisfied or 0 points if it was not satisfied or unclear. We used numerical scores (0 or 1) rather than stars for clarity and consistency in reporting and to calculate the score. A study could therefore achieve a maximum score of 9 (Selection = 4, Comparability = 2, Outcome = 3). Higher scores indicated better methodological quality. Based on this scoring system, we assessed 15 articles. The NOS scores ranged from 4 to 8 out of 9, indicating no studies of poor or low-quality articles (defined as NOS scores between 1and 3). Five studies were classified as medium or moderate quality (NOS scores of 4 - 6), while the remaining 10 were considered as high quality (NOS scores of 7 - 9) (see Table 1).

Table 1. Detailed Newcastle-Ottawa Scale for each included cohort study.

3. Results

We analyzed and extracted data from 15 cohort studies involving a total of 173,804 patients receiving Warfarin therapy. This systematic review identified multiple clinical risk factors that influence INR control in these patients. A summary of the included studies is presented in Table 2.

Table 2. List of selected articles that are included in the review with their main findings and conclusion.

3.1. Drug Interaction

We found significant interactions between Warfarin and various drugs that negatively impact INR control. Al-Momany et al. reported that concurrent medication use, approximately 46.9% of their study population, has interrupted INR stability. Notably, they identified salicylates and amiodarone as contributors to poor INR control and reported that the simultaneous use of Warfarin with aspirin plus NSAID/clopidogrel and levofloxacin resulted in severe bleeding, including bleeding from the upper gastrointestinal tract, surgical site, eye, and nose [21]. Botton et al. identified significant interactions between Warfarin and amiodarone, verapamil, and acetaminophen, all of which contributed to decreasing TTR values [22]. Similarly, Bernaitis et al. found that concurrent use of platelet inhibitors was significantly associated with lower TTR in a cohort from Singapore [23].

Our review also confirmed the strong interaction between antibiotics and Warfarin. Jaakkola et al. found that the use of antibiotics increased the risk of severe over-anticoagulation by 4.6 times when used concurrently with Warfarin [24]. Al-Momany et al. highlighted that antibiotics, especially ciprofloxacin, azithromycin, and clarithromycin, significantly contributing to poor INR control [21]. Interestingly, Botton et al. found that the concurrent use of glibenclamide with Warfarin had a protective effect and was associated with a significantly higher TTR [22]. Bertram et al. demonstrated that patients using PPI alongside Warfarin had significantly lower mean TTR compared to those on Warfarin alone (84.4% vs. 89.4%, p < 0.0001) [25].

3.2. Herbal Supplements

In our review, Al-Momany et al. described that herbal supplements contributed to supratherapeutic INR, accounting for 15.02% of non-therapeutic INR cases in their study. Hawthorn, a widely used traditional cardiovascular remedy in Jordan and Ginseng, a traditional Chinese medicine, were both associated with elevated INR levels and increased bleeding risk [21].

3.3. Comorbidities

Several studies in this review explored the role of comorbidities in INR control. Jaakkola et al. described that 57.3% of patients with excessive anticoagulation had chronic medical illnesses, including active cancer, heart failure, and mechanical heart valves [24]. In addition, Liang et al. reported that comorbidities such as diabetes mellitus, congestive heart disease (CHD), and peripheral arterial disease (PAD) are significantly associated with INR instability, likely due to the effects of the underlying diseases [26].

Björck et al. reported a higher risk of poor INR control among the anaemic patients on Warfarin therapy [iTTR < 60% by 36% (Odds Ratio, OR: 1.36, 95% Confidence Interval, CI: 1.22 - 1.51)], chronic obstructive pulmonary disease (COPD) [iTTR of <60% (OR: 1.45, 95% CI: 1.33 - 1.59)] and dementia [iTTR (OR: 1.47, 95% CI: 1.02 - 2.12)] [27]. Prochaska et al. further confirmed that diabetes is a statistically significant predictor of subtherapeutic INR control (OR:1.4, 95% CI: 1.0 - 2.0), with patients experiencing recurrent episodes of subtherapeutic INR [28].

3.4. Chronic Kidney Disease (CKD)

Our review identified that chronic kidney disease (CKD) significantly influences INR control. Yang et al. reported that 56.3% of AF patients with CKD were on Warfarin, with a slightly higher proportion (62.3%) among those on dialysis. The TTR value was lower than the targeted value across all CKD patients with different eGFR. This study reported that supratherapeutic INR was more common than subtherapeutic INR in CKD patients, and this reduction in TTR in CKD patients was not mediated by poor monitoring of INR [29].

A Swedish study had shown that CKD was associated with poor TTR in AF patients on Warfarin. The decrease in TTR depends on CKD severity, and they are at a higher risk of INR instability [30]. Inoue et al. reported that creatinine clearance (CrCl) < 30 mL/min and TTR < 65% are independently associated with an increased rate of mortality and adverse events. Patients with lower CrCl had lower TTR, which showed that CrCl < 30 mL/min is independently associated with a decrease in TTR < 65% [31].

3.5. Hemodialysis

Yang et al. identified a history of deep vein thrombosis and pulmonary embolism among dialysis patients. They reported that the proportion of dialysis patients that achieved TTR ≥ 60% was only 21.1% and had a significantly poor INR control [29]. Patients on dialysis had lower TTR and higher time out of therapeutic range, although the frequency of INR monitoring was similar among the cohort. The possible reasons were the reversibility of Warfarin, the clinical familiarity of physicians with Warfarin, and all approved DOACs being at least partially eliminated by kidneys, which may not be ideal for dialysis patients [29].

3.6. Kidney Transplant Recipient (KTR)

In this review, Yanik et al. demonstrated that kidney transplant recipients (KTRs) had 18.9% lower Warfarin requirements after accounting for genetic, clinical, and demographic factors. This reduction in Warfarin dose in KTRs is independent of allograft function. The overall findings of this study showed that lower Warfarin doses should be used when initiating Warfarin therapy for KTRs, and close INR monitoring is required to achieve and maintain INR within the therapeutic range [32].

3.7. Hypoalbuminemia

We found a study by Kawai et al., who reported that during a 1-year follow-up period, low serum albumin level was associated with an increased risk of major bleeding. Patients with a basal serum albumin level of <3.6 g/dl had a higher risk of bleeding compared to those with ≥3.6 g/dl serum albumin. This study also demonstrated that the lower the serum albumin, the higher the risk of supratherapeutic INR control (INR > 3) among patients on Warfarin therapy [33].

4. Discussion

This comprehensive review explores the multifaceted clinical risk factors that influence INR control in patients undergoing Warfarin therapy. Our findings highlight the significant impact of drug interactions, comorbidities and other clinical factors. We found that numerous drugs affect the pharmacodynamics and pharmacokinetics of Warfarin primarily via cytochrome P450 pathways, leading to an increased risk of bleeding, even without altering the INR value [34]. Warfarin is primarily metabolized by hepatic cytochrome P450 enzymes, particularly CYP2C9, which metabolizes the more potent S-enantiomer, and CYP1A2 and CYP3A4, which are involved in R-enantiomer metabolism [35] [36]. Inhibitors of CYP2C9, such as amiodarone and fluconazole, reduce Warfarin clearance, resulting in elevated plasma levels and increased INR [4] [37]. Conversely, inducers such as rifampin and carbamazepine accelerate Warfarin metabolism, potentially lowering INR and increasing thromboembolic risk [38]. The magnitude of these interactions varies depending on the degree of enzyme inhibition or induction and the pharmacokinetic properties of the co-administered drug. Therefore, identifying specific isoenzyme-mediated interactions is essential for tailoring Warfarin therapy and minimizing adverse events.

Martín-Pérez et al. supported our study findings by showing that patients consuming ≥ 10 concurrent medications with Warfarin had a higher risk of INR elevation ≥ 4 [39]. Polypharmacy is prevalent among patients on Warfarin therapy [40], with 48% - 81% of them consuming multiple interacting drugs [41]. As a result, these drug interactions often result in either subtherapeutic or supratherapeutic INR levels, increasing the risk of complications. We found that salicylates, including aspirin, compromise poor INR control and elevate bleeding risk. This finding is supported by Boyce et al., who reported that patients on concurrent aspirin have significantly lower TTR and an increase in minor bleeding [42]. In contrast, Okumura et al. and Mueller et al. found no significant impact of concurrent use of antiplatelet agents or aspirin therapy on TTR [43] [44]. Similarly, drugs like Amiodarone reduce the total body clearance of both (R)- and (S)-Warfarin, impairing INR stability [45]. Our review finding is supported by Saleh (2016), who reported that patients on Amiodarone required lower Warfarin doses [46].

Similarly, Macedo et al. found that pain medications such as NSAIDs, acetaminophen, and opioids are usually associated with poor INR control [47]. Our review findings also found that various antibiotics use, especially for respiratory infections, increases the risk of INR elevation, consistent with findings by Martín-Pérez et al. [39]. It is noted that proton pump inhibitors (PPIs), especially esomeprazole and omeprazole, may inhibit CYP2C19, altering Warfarin metabolism and affecting TTR [48]. Additionally, an evidence-based study by Ge et al. supported our review finding that herbal supplements can significantly increase INR due to interactions with Warfarin [49]. Therefore, clinicians should encourage patients to disclose any herbal supplement use.

Our review identified that multiple comorbidities contribute to subtherapeutic or supratherapeutic INR control. A study conducted in France by Rouaud et al. supports our review findings, demonstrating that patients’ comorbidities increase the risk of poor INR control [50]. DeRemer et al. reinforced our review finding, showing that anemia is associated with poor INR control in patients on Warfarin therapy [51], and Martín-Pérez et al. reported a significant association of anemia with supratherapeutic INR control (INR ≥ 4) [39].

Comorbidities also emerged as key contributors to poor INR control. Our review findings on COPD, which causes poor iTTR, are supported by Cohen et al., who also associated COPD with supratherapeutic INR [52]. Martín-Pérez et al. reported that asthma and COPD are linked to INR elevation (INR ≥ 4), proposing that β2 agonists in patients with chronic respiratory disease may interact with Warfarin [39]. Heart failure was another risk factor identified in our review. Similarly, a cross-sectional study reported that heart failure patients are more likely to have poor INR control [46] [53]. This may be due to hepatic congestion and impaired synthesis of clotting factor [54]. Warfarin remains the only anticoagulant option for patients with mechanical valve prosthesis [55]. Our review findings demonstrated that many of these patients have supratherapeutic INR. In contrast, Mvondo et al. reported good outcomes in such patients with 78.9 ± 3.7% and 93.1 ± 2.1 % of patients with mechanical valves had freedom from anticoagulation-related bleeding and neurological events, respectively [56], though other studies indicate a significantly increased bleeding risk [57] [58]. Our review found that cancer increases the risk of supratherapeutic INR, potentially due to polypharmacy, malnutrition, or liver dysfunction [59].

Michal et al. have demonstrated that depression is associated with lower TTR, indicating that psychological conditions such as depression and anxiety may negatively affect anticoagulation outcomes[60] [61]. Our findings also linked diabetes is associated with subtherapeutic INR; both non-insulin and insulin-dependent diabetes strongly predict poor INR control [28] [62].

CKD increases the risk of AF and atrial flutter [63], as well as bleeding, stroke, and increased mortality [64]. Our review findings on poor INR control in CKD patients are aligned with Limdi et al., who reported frequent supratherapeutic INR (INR ≥ 4) in non-dialysis-dependent CKD patients [65]-[67]. Reduced albumin binding in CKD increases free Warfarin activity [68]. Kleinow et al. supported our review finding and found that CKD patients required lower Warfarin doses to maintain the therapeutic INR range compared to those without CKD [69]. Chan et al. demonstrated that AF patients with hemodialysis receiving Warfarin therapy have a higher risk of new stroke compared to those not on Warfarin therapy [70]. In contrast, Kai et al. reported that hemodialysis patients on Warfarin therapy have a lower risk of ischemic stroke and all-cause death with no significant elevation in bleeding [71]. Our review findings have shown that hemodialysis patients on Warfarin therapy have lower TTR. The bleeding risk in hemodialysis patients also could be higher due to the frequent utilization of heparin [72].

Our review findings established that KTR patients have poor INR control and require a low Warfarin dose. A study has shown that the incidence of AF at three years of post-kidney transplantation is 7.3% [73]. KTR with AF have an increased risk of poor clinical consequences, including a high risk of allograft loss and mortality [74]. Finally, we found the association between hypoalbuminemia and poor INR control and increased bleeding risk; this study finding is also supported by Abdelhafiz et al., who demonstrated that one of the predictive factors of minor and major bleeding in younger patients (<75 years) is hypoalbuminemia [75].

Clinical Implications and Recommendations: This review highlights the need for personalized Warfarin therapy and vigilant INR monitoring, particularly for patients with multiple comorbidities or complex drug regimens. Based on our findings, clinicians managing patients on Warfarin therapy are advised to adopt several practical strategies to optimize INR control and reduce adverse outcomes.

First, comprehensive medication reconciliation should be routinely performed to identify potential drug-drug interactions, particularly involving antibiotics, NSAIDs, PPIs, amiodarone, and herbal supplements. Second, individualized Warfarin dose adjustments are warranted in patients with comorbidities such as CKD, liver dysfunction, or heart failure, and lower initiation doses should be considered for those with hypoalbuminemia or undergoing hemodialysis. Third, clinicians should implement more frequent INR monitoring in high-risk groups, including patients with polypharmacy, mechanical valves, cancer, or psychiatric conditions such as depression. Finally, patient education on the importance of reporting herbal or over-the-counter drug use, maintaining consistent dietary vitamin K intake, and adhering to follow-up INR testing is critical for maintaining therapeutic anticoagulation. Incorporating these evidence-informed recommendations and strategies may improve time in therapeutic range (TTR), reduce the risk of bleeding or thromboembolic complications, and enhance the overall Warfarin safety and efficacy in real-world clinical settings.

5. Limitation

There are several potential limitations inherent in our study that should be acknowledged. First, this is an observational review primarily based on cohort studies. We did not identify any randomized controlled trials specifically evaluating clinical factors influencing INR control in Warfarin-treated patients. Therefore, some degree of bias is unavoidable, and establishing causal relationships remains challenging. Second, there was considerable heterogeneity in patient populations, including differences in hemorrhagic and thromboembolic risk profiles. Due to the lack of standardized quantitative data across studies, we were unable to perform formal heterogeneity analysis (e.g. using the I² statistic) for Warfarin indications or outcome measures. Therefore, findings should be interpreted with appropriate caution. Third, we acknowledge the exclusion of genetic factors, particularly VKORC1 and CYP2C9 polymorphisms, which are well-established determinants of warfarin sensitivity and metabolism. These factors were excluded because genetic testing is not universally available or routinely implemented in clinical practice, limiting their practical relevance to this review.

6. Conclusion

Warfarin remains a widely prescribed oral anticoagulant for patients with cardiovascular diseases such as AF, prosthetic heart valves, and hypercoagulable conditions. This review demonstrates that multiple clinical risk factors affect INR control, including drug interactions and comorbidities such as diabetes, CKD, malignancy, prosthetic valve, heart failure, and anemia. Patients on hemodialysis or with hypoalbuminemia are particularly vulnerable to supratherapeutic INR levels and bleeding complications. Moreover, kidney transplant recipients may require lower Warfarin doses and are also at increased risk of poor INR control. In real-world clinical settings, many patients on Warfarin experience subtherapeutic or supratherapeutic INR values. Identifying and managing these clinical risk factors before prescribing Warfarin is essential for achieving optimal anticoagulation. Treating physicians should make prescribing decisions by carefully weighing the cardiovascular benefits of Warfarin against its bleeding risks and tailor treatment to individual patient profiles.

Funding

The authors did not receive any funds to conduct this review.

Ethics Approval

The authors did not need to apply for ethical approval to conduct this review.

Consent to Participate

This is a systematic review where the authors did not need to interact with any study respondents.

Consent for Publication

The authors are giving consent to publish their data, which are presented in this manuscript.

Availability of Data and Material

All the data are presented in the manuscript.

Authors’ Contributions

All of the authors [Tharani Ramasamy (TM), Nowrozy Kamar Jahan (NKJ), Naganathan Pillai (NP), and Christina Gertrude Yap (CGY)] conceived the study. TM and NKJ conducted the screening and selection of the relevant articles. All authors participated in the data extraction under different themes, sub-themes, and risk of bias assessment. TM and NKJ drafted the manuscript, which the remaining two authors reviewed before submitting it for publication.

Acknowledgments

The authors would like to acknowledge that Monash University provides permission to use Monash University Malaysia library’s extensive resources, mainly to access all required full-text articles. In addition, Monash University also allowed the authors to use EndNote software to remove duplicate articles and NVivo software to extract the data for analysis.

Conflicts of Interest

The authors declare no conflicts of interest.

Appendix

Table A1. Detailed search strategy and search string in four electronic databases.

*The asterisk symbol, commonly known as a wildcard, is used in keyword searches to expand results by matching different variations of a root word.

Checklist A1. PRISMA 2020 checklist.

NA = Not Applicable.

List of Abbreviations

NOTES

¹TR: Tharani Ramasamy; NKJ: Nowrozy Kamar Jahan.

²NKP: Naganathan Kathiresan Pillai; CGY: Christina Gertrude Yap.

Conflicts of Interest

The authors declare no conflicts of interest.

References