Table of Contents  
Year : 2021  |  Volume : 8  |  Issue : 5  |  Page : 60-65

Use and adherence to oral anticoagulants in a tertiary care hospital

1 Department of Pharmacology, St. John's Medical College, Bengaluru, Karnataka, India
2 Department of Vascular and Endovascular Surgery, St. John's Medical College, Bengaluru, Karnataka, India

Date of Submission29-Dec-2020
Date of Acceptance25-Jun-2021
Date of Web Publication30-Aug-2021

Correspondence Address:
Atiya Rehman Faruqui
Department of Pharmacology, St. John's Medical College, Bengaluru, Karnataka
Login to access the Email id

Source of Support: None, Conflict of Interest: None

DOI: 10.4103/ijves.ijves_174_20

Rights and Permissions

Background: In patients at risk of thromboembolism, oral anticoagulants (OAC) are effective, but there are limited data from India on OAC use and adherence. This study in a tertiary care hospital aims to record the pattern of prescription of OAC and medication adherence over 6 months. Materials and Methods: Medications prescribed were recorded, and a follow-up at 3 and 6 months was done to record outcomes and adherence to medication. Modified Morisky medication adherence scale (MMAS-8) was used and compliance to monthly International Normalized Ratio (INR) monitoring was recorded. Results: Of the 140 patients included, mean age was 55.32 (±18.04 years), and 84 (60.0%) were males. Indications for OAC were deep vein thrombosis 64 (45.7%) and atrial fibrillation 42 (30.0%). After initial injectable anticoagulants, majority of patients were shifted to Vitamin K antagonists. Most common OAC was nicoumarol 73 (52.1%), followed by warfarin 43 (30.2%), dabigatran 12 (9.2%), and apixaban 12 (9.2%). Over 6 months, 12 patients were lost to follow-up and 16 (11.4%) deaths were recorded. Monthly INR was done by 23 patients till the 3rd month but only by five patients between 3rd and 6th month. By the 3rd month, 19 (86.3%) and between 3rd and 6th month 4 (13.7%) patients were nonadherent. At 6th month, 78 (86.6%) had a MMAS-8 score of 8 (high adherence). The presence of a caretaker in 37 (41.2%) emerged as the most important factor contributing to adherence. Conclusion: Early identification of patient-specific barriers for INR monitoring and adherence to medications should be sought in patients with thromboembolism.

Keywords: Adherence, deep vein thrombosis, oral anticoagulants, thromboembolism

How to cite this article:
Pradhyumna M, Faruqui AR, Joshi SS. Use and adherence to oral anticoagulants in a tertiary care hospital. Indian J Vasc Endovasc Surg 2021;8, Suppl S1:60-5

How to cite this URL:
Pradhyumna M, Faruqui AR, Joshi SS. Use and adherence to oral anticoagulants in a tertiary care hospital. Indian J Vasc Endovasc Surg [serial online] 2021 [cited 2022 May 28];8, Suppl S1:60-5. Available from:

  Introduction Top

Global burden of disease 2016 reported mortality rates for atrial fibrillation and venous thromboembolism as 1.7 and 9.4–32.3, respectively.[1]

Oral anticoagulants (OACs) are the mainstay in preventing thromboembolism; however, it is important to maintain adherence and persistence to these agents as it determines their efficacy and safety.[2],[3],[4]

For Vitamin K antagonists, the rates of nonadherence ranges between 22% and 58%.[5] Study involving novel oral anticoagulants (NOACs), higher proportion of rivaroxaban (72.7%) patients were found to be adherent compared to dabigatran (67.2%, P < 0.001) and apixaban (69.5%, P < 0.001). Moreover, rivaroxaban was associated with better treatment persistence than patients with warfarin.[6]

Adherence to prescribed anticoagulants predicts outcomes in patients. Several factors such as affordability, polypharmacy, socioeconomic status, and adverse reactions to the drug are known to significantly affect adherence.[7] Data on the use of warfarin and newer anticoagulants and their adherence in patients from India are very limited.

  Materials and Methods Top

Ethics clearance

The study was reviewed and received an approval from the Institutional Ethics Committee.

Study design and duration

This was a prospective observational study with follow-up at 3rd and 6th month. Data were collected for 12 months from February 2016 to January 2017. Follow-up was done at 1st, 3rd, and 6th month after recruitment, by telephone call or personal follow-up, until August 2017.

Eligibility criteria

All adult patients of both genders, who visited the vascular surgery and cardiology departments and were prescribed an OAC, were considered eligible to participate. Patients who would not be able to come back for follow-up were excluded.

Data collection

Cases were identified as per the selection criteria mentioned above from the inpatient wards. A written informed consent from patient and a first-degree relative was taken. A patient interview for collecting relevant medical history and chart review for documenting drugs and comorbidities were done using a structured case report form. A telephonic call at the end of 1 and 3 months was made to record any outcome/complications.

At the end of 6 months, patients were met in person and adherence to medication over 6 months was evaluated using Modified Morisky medication adherence scale (MMAS-8). Patients with a score ≥6 were considered adherent and score <6 as nonadherent to OAC.

Data management was done on a free software “Epi info” and then extracted ont IBM SPSS Statistics for Windows, Version 22.0. Armonk, NY: IBM Corp.

Sample size calculation

Based on search from the literature, Kim et al.,[8] the adherence to OAC was reported as 27.5%. Considering an estimated proportion of 0.28, an alpha error of 5%, and a confidence interval (CI) of 90%, the sample size obtained was 217 patients. Taking a dropout rate of 20%, the final sample size required was 250. Taking into consideration the time frame for the study, the absence of funding for more staff, and the logistics of follow-up at 6 months, we decided to recruit all patients in a 1-year duration and finally 140 patients were included.

Statistical analysis

We summarized baseline data of patients as mean with standard deviation for continuous variable (Age). Moreover, frequencies with percentage, for categorical variables (demographic data, comorbidities, socioeconomic status, habits, presence of AF or deep vein thrombosis [DVT], polypharmacy, and reason for adherence and nonadherence). Categorical variables were assessed using Chi-squared test. To assess the factors affecting nonadherence to medication, we used multivariate logistic regression analysis.

A P < 0.05 was considered significant for all tests. Statistical analysis was done using commercially available software (SPSS version 22).

  Results Top

A total of 140 patients were recruited from the wards of cardiology, vascular surgery, and general surgery [Figure 1].
Figure 1: Study flow diagram

Click here to view

Patient characteristics at baseline

Of the 140 patients, 84 (60.0%) were males with a mean age of 55.32 (±18.04) and 56 (40.0%) were females with a mean age of 56.3 (±17.6). Majority of the participants 82 (58.6%) were from the urban population. The socioeconomic status of the patient was analyzed using the standard of living index (SLI) [Table 1].
Table 1: Baseline characteristics of patients

Click here to view

Patient data pertaining to the associated comorbidities showed that a higher number of patients were hypertensive 66 (47.1%); 52 (37.1%) were diabetic; 32 (22.9%) had dyslipidemia; 21 (15.0%) had a history of ischemic heart disease; 4 (2.9%) were diagnosed with cancer; 5 (3.6%) patients manifested with heart failure; and 6 (4.3%) patients were diagnosed with coagulopathies. The most common coagulopathy noticed was hyperhomocystenemia. Multiple comorbidities, defined as 2 or more of any of the comorbidities in the study, were present in 58 (41.4%) individuals.

Considering the pattern of recruitment of patients, vascular indications compared to cardiac indications predominated the indications for prescription of an anticoagulant. DVT 64 (45.7%) was the major factor under vascular disorder and atrial fibrillation 42 (30.0%) under cardiovascular. Polypharmacy (>4 medications) was noted in 94 (67.1%) patients. Other conditions in our study which required the use of an anticoagulant were pulmonary embolism 13 (19.2%), arterial thrombus 6 (4.3%) rheumatic heart disease with mitral stenosis 5 (3.5%), postcardiac valve replacement 4 (2.8%), stroke 3 (2.1%), and left ventricular clot 3 (2.1%).

Injectable anticoagulant most prescribed was enoxaparin. OACs were started following initial anticoagulation with injectable agents. Nicoumarol 73 (52.4%) was the most common OAC prescribed. [Table 2] gives details of anticoagulants used.

Determinants of adherence and nonadherence to medication

Among the factors favoring adherence, the presence of a caretaker influenced the parameter in 37 (41.2%) patients. Severity of the disease 20 (23.3%) and other reasons which could not be ascertained 32 (35.5%) were also noted. The most common reason for nonadherence was self-stoppage of medicines due to patient's perception of feeling better 13 (59.0%); cost 4 (18.1%); and other reasons in 6 (22.9%) (no improvement, complimentary medicine, and others) were seen.

Follow-up data

During a period of 6 months, we had 12 patients who were lost to follow-up, all during the 3rd month. There were 16 (11.4%) deaths recorded during follow-up. Nine of the 16 deaths were in patients with AF, three were due to vascular disorder (2 DVT and 1 splenic vein thrombosis), one case of RHD with MS, one patient had cancer, and two had stroke. All deaths occurred during the 3rd month of follow-up. Twenty-three (20.5%) patients were nonadherent to medication, of them 19 (86.3%) stopped medication within 3 months of initiation, and 4 (13.7%) patients between 3rd and 6th month of follow-up. Data on International Normalized Ratio (INR) monitoring by patients on Vitamin K antagonist showed that only 23 of the patients monitored it at least once a month in the first 3 months of initiation, and between the 3rd and 6th month of initiation 5 monitored it once a month [Table 3].
Table 2: Details of anticoagulants used

Click here to view
Table 3: Clinical outcomes at 6th month

Click here to view

Factors determining nonadherence to anticoagulant medication

Among the several parameters evaluated using a Chi-square analysis, factors with a P < 0.2 were incorporated into the univariate regression model. The factors included were low SLI (P = 0.057), polypharmacy (P = 0.122), and smoking (P = 0.013). On analysis, smoking (odds ratio [OR]: 3.286; 95% CI: 1.248–8.650; P = 0.016) and low SLI (OR: 2.643; 95% CI: 0.950–7.351; P = 0.06) significantly affected adherence to medication. The factors were incorporated into a multivariate regression model, which thereby demonstrated that patients with smoking (OR: 3.94; 95% CI: 1.416–10.972; P = 0.009) and low SLI (OR: 3.36; 95% CI: 1.122–10.070; P = 0.030) were nonadherent [Table 4], [Table 5], [Table 6].
Table 4: Comparison of characteristics that influence baseline factors over adherence and nonadherence

Click here to view
Table 5: Comparison of factors affecting nonadherence* (Univariate regression analysis)

Click here to view
Table 6: Comparison of factors affecting nonadherence**(Multivariate regression analysis)

Click here to view

  Discussion Top

In the present study, we assessed the utilization of OAC and the medication adherence of patients, in a tertiary care hospital. This hospital gets patients mainly from the southern and partly northeastern regions of India. In order to obtain a representative population for all indications warranting the need for OAC, we included patients with all eligible indications from different wards of the hospital (including general and private wards).

The sample consisted of a larger proportion of vascular cases (DVT + pulmonary embolism + arterial thrombus) 83 (59.2%), compared to cardiac indications, most of whom were cases of atrial fibrillation. Warfarin is still the mainstay of OAC treatment in India and also the western world, 60 years after its introduction. More than 2 million North Americans are on Vitamin K antagonist, and the number continues to grow with the aging population. Our findings are consistent with these data, but with a replacement of warfarin by nicoumarol 73 (52.1%) in the group of Vitamin K antagonist. This form of prescription could be due to factors such as cost and lesser drug interactions of nicoumarol, as compared to warfarin.[9] In Asians, the average initiation dose requirement was found to be 3.4 mg versus 5 mg in whites, as per a study of warfarin management within the clinic for 3 or more months, to quantitate influence of ethnicity on warfarin dosing. The adjusted mean (95% CI) weekly warfarin doses for patients with an INR goal of 2–3 for Asian Americans was 24 mg (21–27).[10] This coincides with the initiation dose in our study, wherein 3 mg of nicoumarol was initiated in 53 (42.4%) patients, thereby mounting to an average dose of 24 mg/week. The frequency of INR testing is also variable over time and is dictated by dose response.[11] Data pertaining to frequency of INR monitoring demonstrated a decline with each follow-up, with a drop from 23 patients monitoring it at least once a month in the first 3 months of initiation to only five patients, between the 3rd and 6th month of initiation. The frequency of monitoring between patients varied in our study and was seen dependent on their follow-up INR values, frequency of visit, and costs of monitoring.

In chronic disease, adherence decreases dramatically after the first 3 months of therapy. The rate of nonadherence to medication for chronic conditions usually varies from 22% to 55%, including those treated with OAC.[7] In our study, we noted that 23 (20.5%) patients were nonadherent to medication at the end of 6 months, with a highest rate of nonadherence seen within the 3rd month of initiation. There were 12 (9.4%) patients on dabigatran and 11 (8.6%) patients on apixaban. Rivaroxaban was not prescribed to the patients visiting our hospital, though evidence from the EINSTEIN – DVT study (Primary Outcome: Hazard ratio [HR]: 0.68; 95% CI: 0.44–1.04; P < 0.001) and the ROCKET– AF study (Primary Outcome: 0.88; 95% CI: 0.74–1.03; P < 0.001 for noninferiority) as well as once daily dosing benefit therapy with this agent. Affordability from the patients' perspective could be a reason for avoiding prescription of rivaroxaban, though interestingly we noticed that none of the patients on other NOACs had discontinued medication.

A cross-sectional study by Mayet, in an outpatient anticoagulant clinic with 192 patients, demonstrated that 89 (46.4%) patients according to MMAS-8 had high adherence score (score = 8) and 103 (53.6%) had a low score (score = 6). In our study, 78 (69.0%) had a score of 8 (high adherence) and 12 (9.7%) had a score of 6–8. The mean age of the study population in the cross-sectional study was 53.8 ± 15.2, and this was consistent with our population age group of 55.32 ± 18.04.[12] Although the study did demonstrate that unemployment was a significant independent factor for poor INR control (OR: 2.91:95% CI: 1.10–4.92), our study demonstrated low SLI was an independent factor (OR: 3.36: 95% CI: 1.12–10.07). Several factors have been associated with drug adherence such as acceptance of illness, trust in the therapist, belief in the therapy, higher level of education, and stability of family background. In a questionnaire-based study involving 264 patients in Italy, 96% patients had stated that doctor–patient relationship plays a role despite a monitored and computer-assisted method of dose adjustment.[13] From our study, we noticed that the presence of a caretaker 37 (41.2%) benefitted adherence to therapy.

This study shows that INR monitoring was done in very few patients and therefore implicates the need for better strategies to improve INR monitoring, such as availability of monitors, at affordable prices and portable INR devices in the nearest health-care facilities. This will help improve adherence by regular monitoring and maintenance of INR in the therapeutic range. This study also involves patients with diverse indications, thereby providing scope for further evaluation of each, to enhance compliance to respective therapies.


This study has fewer number of patients on newer anticoagulants, thereby the adherence in these subsets of patients could not be adequately compared., Lack of rivaroxaban prescription also limits the comparability with the most novel NOAC. We relied primarily on self-reported metrics which may be subject to bias.

  Conclusion Top

To the best of our knowledge, this is the first study in our hospital to monitor adherence to OAC. Patient-specific barriers for INR monitoring and adherence to medications have been identified in patients with thromboembolism. Studies such as this one point out the importance of early identification of barriers impacting continuation of prescribed treatment in patients in order to improve their outcomes.


We would like to thank the faculty and residents from Departments of Pharmacology, Vascular Surgery and Cardiology for their Guidance and Support.

Financial support and sponsorship


Conflicts of interest

There are no conflicts of interest.

  References Top

Wendelboe AM, Raskob GE. Global burden of thrombosis: Epidemiologic aspects. Circ Res 2016;118:1340-7.  Back to cited text no. 1
Ho PM, Magid DJ, Shetterly SM, Olson KL, Maddox TM, Peterson PN, et al. Medication nonadherence is associated with a broad range of adverse outcomes in patients with coronary artery disease. Am Heart J 2008;155:772-9.  Back to cited text no. 2
Rhee MK, Slocum W, Ziemer DC, Culler SD, Cook CB, El-Kebbi IM, et al. Patient adherence improves glycemic control. Diabetes Educ 2005;31:240-50.  Back to cited text no. 3
Böhm M, Schumacher H, Laufs U, Sleight P, Schmieder R, Unger T, et al. Effects of nonpersistence with medication on outcomes in high-risk patients with cardiovascular disease. Am Heart J 2013;166:306-14.e7.  Back to cited text no. 4
Prandoni P. Adherence to the novel oral anticoagulants: An unmet need. Curr Med Res Opin 2015;31:2175-7.  Back to cited text no. 5
McHorney CA, Crivera C, Laliberté F, Nelson WW, Germain G, Bookhart B, et al. Adherence to non-vitamin-K-antagonist oral anticoagulant medications based on the Pharmacy Quality Alliance measure. Curr Med Res Opin 2015;31:2167-73.  Back to cited text no. 6
Fischer MA, Stedman MR, Lii J, Vogeli C, Shrank WH, Brookhart MA, et al. Primary medication non-adherence: Analysis of 195,930 electronic prescriptions. J Gen Intern Med 2010;25:284-90.  Back to cited text no. 7
Kim JH, Kim GS, Kim EJ, Park S, Chung N, Chu SH. Factors affecting medication adherence and anticoagulation control in Korean patients taking warfarin. J Cardiovasc Nurs 2011;26:466-74.  Back to cited text no. 8
Gschwind L, Rollason V, Lovis C, Boehlen F, Bonnabry P, Dayer P, et al. Identification and weighting of the most critical “real-life” drug-drug interactions with acenocoumarol in a tertiary care hospital. Eur J Clin Pharmacol 2013;69:617-27.  Back to cited text no. 9
Dang MT, Hambleton J, Kayser SR. The influence of ethnicity on warfarin dosage requirement. Ann Pharmacother 2005;39:1008-12.  Back to cited text no. 10
Gopalakrishnan S, Narayanan S. Oral anticoagulants: Current Indian scenario. In: Medicine Update. Bombay Association of Physicians: The Association of Physicians of India; 2013. p. 410-3.  Back to cited text no. 11
Mayet AY. Patient adherence to warfarin therapy and its impact on anticoagulation control. Saudi Pharm J 2016;24:29-34.  Back to cited text no. 12
Barcellona D, Contu P, Sorano GG, Pengo V, Marongiu F. The management of oral anticoagulant therapy: The patient's point of view. Thromb Haemost 2000;83:49-53.  Back to cited text no. 13


  [Figure 1]

  [Table 1], [Table 2], [Table 3], [Table 4], [Table 5], [Table 6]


Similar in PUBMED
   Search Pubmed for
   Search in Google Scholar for
 Related articles
Access Statistics
Email Alert *
Add to My List *
* Registration required (free)

  In this article
   Materials and Me...
   Article Figures
   Article Tables

 Article Access Statistics
    PDF Downloaded63    
    Comments [Add]    

Recommend this journal