Restricted Access

Reason: Access restricted by the author. A copy can be requested for private research and study by contacting your institution's library service. This copy cannot be republished

Case-control study of risk factors contributing to bleeding risk in community-based patients receiving warfarin therapy

posted on 09.02.2017, 03:27 authored by Diug, Basia
Warfarin is well recognised as the mainstay anticoagulant therapy for prophylaxis and prevention of stroke. Atrial fibrillation (AF) is a common primary indication for warfarin therapy and an increase in incidence and prevalence of AF has resulted in a 9% rise per annum in warfarin prescription. Nevertheless, the literature suggests that warfarin is under-prescribed, possibly due to its recognised complications which may lead to morbidity or mortality. Warfarin’s narrow therapeutic index necessitates patients adhere strictly to a daily regimen and regular monitoring through blood tests of their International Normalised Ratio (INR). AF patients who are below their therapeutic INR range <2.0 are at a 5-fold increased risk of stroke whilst patients with elevated INRs >6.0 have been associated with significantly increased bleeding risk. The aim of this thesis was to identify factors contributing to bleeding risk in patients receiving long-term warfarin therapy in the community. A case-control study was conducted between March 2008 and June 2009 with community-based patients residing in metropolitan Melbourne. Cases were patients previously stabilised on warfarin therapy who had recorded an INR ≥ 6.0 on routine testing. Controls were patients stabilised on warfarin for at least three months after initial titration and without recording INR levels that were out of their therapeutic range. Inclusion criteria included being aged ≥18 years; residing within a 50km radius of the central business district, and stabilized on warfarin for a minimum of 3 months. Patients were excluded if they were hospitalised at the time of their INR measurement, were living in a nursing care facility, or were unable to be interviewed within a 30 day period after registering the INR level of interest. A structured patient questionnaire was developed and data was collected on the following characteristics: demographics, health status, medications and psychosocial factors. The patients’ general practitioner was contacted, to obtain a current list of their medications and to confirm specific patient details. Reasons and demographic factors of patients non-recruited were also collected. Logistic regression models were used to analyse the data. In this case-control study, a total of 487 patients: 157 cases and 329 controls were recruited. Several important risk factors for bleeding were identified. Cases were more likely than controls to have multiple co-morbidities, be maintained at a higher INR target range according to their primary indication, and have poor visual acuity. Cases were also more likely to be employed but have an income less than $28,079 per annum. Patients taking aspirin and prednisolone were at an increased risk, and those currently taking antibiotics had a 4.3 fold increased risk of bleeding, with 20% of those taking antibiotics receiving contra-indicated macrolides and metronidazoles. Cases were less likely to use a tablet dispenser or dosette for medication management, and more likely to self-report poor medication adherence. A high proportion of cases (41%) reported taking 10 or more medications. In addition, several unrecognised risk factors were identified in the form of psychosocial deficits. The following psychosocial factors; cognition, depression, social isolation, functional independence and functional health literacy were examined using validated quantitative measures. Psychosocial deficits showed a strong association with increased bleeding risk, both in the univariate and adjusted models. In addition, when multiple psychosocial deficits were present there was a strong association, with poor health literacy displaying the strongest relationship (adjusted OR=4.8, 95% CI, 2.9-7.8), followed by a depressed mood (adjusted OR=3.1, 95% CI, 2.9-7.8). A sub-study showed that the two genetic polymorphisms CYP2C9 and VKORC1, linked to warfarin metabolism or its action respectively, were not associated with an elevated INR, therefore by inference, with an increased risk of bleeding in this cohort. However, there was a strong association with the two polymorphisms and dosage. The average daily dose declined from 6.2mg/day, for patients who were rapid metabolisers and had a low sensitivity; to 2.2mg/day, for patients who were intermediate metabolisers with a high sensitivity. This is clinically significant and could influence dosing in the future. The significance of this thesis is the identification of several novel risk factors for bleeding risk as well as confirmation of well established risk factors in a large Australian population. With an increasingly ageing population and foreseeable continued usage of warfarin, these psychosocial characteristics should be regularly assessed. When such psychosocial deficits are identified, appropriate management strategies should be implemented to optimise patient safety.


Principal supervisor

John McNeil

Year of Award


Department, School or Centre

Public Health and Preventive Medicine

Campus location



Doctor of Philosophy

Degree Type



Faculty of Medicine Nursing and Health Sciences