Utilization review of concomitant use of potentially interacting drugs in Thai patients using warfarin therapy

PURPOSE: In Thailand, there has been no study determining the concomitant use of medications, known to potentially interact with warfarin, in patients receiving warfarin therapy. This paper examined the frequency of which specific interacting drugs were concomitantly used in warfarin users. METHODS: We retrospectively examined the database of warfarin outpatient medical records from a regional 756-bed hospital located in the north of Thailand. All patients receiving warfarin from 10 June 1999 to 4 August 2004 were reviewed to identify all drugs possessing interaction potential with warfarin. The potential of significant interactions were divided into high, moderate and low, according to the extent of evidence documented in textbooks and literature. RESULTS: Among 1093 patients receiving warfarin therapy, 914 (84%) patients received at least one potentially interacting drug and half of them (457 patients) received at least one drug with high potential for interaction. The most frequently concomitant drug that increased INR was acetaminophen (63%, 316/457). Propylthiouracil was the most frequently concomitant drug that decreased INR response (4%, 19/457), while diclofenac was the most frequently concomitant drug that increased bleeding risk (16%, 73/457). CONCLUSIONS: About a half of patients receiving warfarin therapy was prescribed concomitant drug(s) that has a high potential of interactions with warfarin. These patients should be closely monitored and counselled to watch for signs and symptoms of bleeding and thrombosis to avoid adverse events associated with drug interactions.     

An observational study examining the impact of capecitabine on warfarin antithrombotic activity and bleeding complications

OBJECTIVE: The objectives of this study are to quantify the frequency of concomitant use of capecitabine and warfarin, and to quantify the rate of bleeding events and elevated international normalized ratio (INR) among concomitant users of warfarin and capecitabine. RESEARCH DESIGN AND METHODS: We conducted a retrospective population-based study within the Henry Ford Health System (Detroit, MI) and the Kaiser Permanente Medical Care Program of Northern California (Oakland, CA). The study population included patients prescribed concomitant capecitabine and warfarin from 1 April 1997 through 31 July 2002. Data from the medical records of concurrent users were extracted through 31 August 2002. MAIN OUTCOME MEASURES: Concomitant use of capecitabine and warfarin, bleeding events, and INR laboratory results, collected from computerized databases and medical record review. RESULTS: Overall, 11% of capecitabine users also received warfarin (99 / 883). Among 17 patients who received warfarin for venous access device prophylaxis, one bleeding event occurred during concomitant capecitabine/warfarin use (rate = 35.7 bleeding events per 100 person-years, 95% confidence interval [CI] 0.9-198.9), and no events occurred during use of warfarin alone (95% CI 0.0-136.2) (p = 0.50). Among patients prescribed warfarin for indications other than port prophylaxis, no bleeding events occurred during concomitant use of capecitabine and warfarin (95% CI 0.0-34.6), and one event occurred during warfarin use alone (rate = 9.2 bleeding events per 100 person-years, 95% CI 0.2-51.3) (p = 0.54). We found one INR elevation > 3.0 among concomitant capecitabine/warfarin users receiving warfarin for port prophylaxis (rate = 35.7 per 100 person-years) and no INR elevations > 3.0 during use of warfarin alone (p = 0.46). Among patients using warfarin for indications other than port prophylaxis, the rates of INR > 3.0 were 309.7 per 100 person-years (95% CI 213.2-434.9) during concomitant capecitabine/warfarin use and 193.5 events per 100 person-years (95% CI 119.8-295.8) during use of warfarin alone (p = 0.09). CONCLUSIONS: The results of our study show a low prevalence of capecitabine and warfarin concomitant use. We did not find large differences in the rates of bleeding events and elevated INR in patients receiving concomitant capecitabine and warfarin when compared with use of warfarin alone. While these results do not imply a lack of biologic interaction, our findings indicate that patients appear to be appropriately managed in clinical practice.     

Randomized, placebo-controlled trial of oral phytonadione for excessive anticoagulation

STUDY OBJECTIVE: To compare the efficacy of managing excessive anticoagulation in the absence of bleeding by either omitting warfarin therapy alone or administering oral phytonadione in addition to omitting warfarin therapy. DESIGN: Randomized, double-blind, placebo-controlled study. SETTING: Clinical pharmacy anticoagulation service in a group model health maintenance organization. SUBJECTS: Thirty nonbleeding patients with international normalized ratios (INRs) ranging from 6.0-10.0. INTERVENTIONS: Patients were randomized to receive either a single oral dose of phytonadione 2.5 mg or placebo. Both groups omitted warfarin doses until the INR became less than or equal to 4.0. MEASUREMENTS AND RESULTS: The mean calculated time to reach an INR of 4.0 was significantly greater in the placebo than the phytonadione group (2.6 vs 1.4 days, p=0.006). Overcorrection of anticoagulation was significantly more common in patients receiving phytonadione. Overt warfarin resistance was not observed in either group after reinitiating warfarin therapy. No major bleeding or thromboembolic complications occurred, and minor bleeding episodes were similar in both groups. CONCLUSION: The addition of oral phytonadione 2.5 mg reduced the time to achieve an INR of 4.0 by approximately 1 day compared with omitting warfarin therapy alone. Adverse events did not differ between the two groups. Both strategies were effective in managing asymptomatic patients with INRs of 6.0-10.0. Oral phytonadione may be most appropriate for patients at high risk for bleeding in whom the benefit of prompt INR reduction would outweigh the thromboembolic risk associated with INR overcorrection.     

Effect of levofloxacin coadministration on the international normalized ratios during warfarin therapy

STUDY OBJECTIVE: To evaluate the effect of levofloxacin coadministration on the international normalized ratio (INR) in patients receiving warfarin therapy. DESIGN: Prospective analysis. SETTING: Outpatient clinic at a Veterans Affairs medical center. PATIENTS: Eighteen adult patients receiving warfarin. INTERVENTION: On the basis of clinical diagnosis and judgment, levofloxacin was prescribed to the 18 patients for treatment of various types of infection. The INR was measured before and at 2-8-day intervals after the coadministration of levofloxacin therapy, and once after completing therapy. Warfarin dosages were adjusted when necessary. MEASUREMENTS AND MAIN RESULTS: Warfarin dosages were changed in seven patients as a result of the first nontherapeutic INR values obtained after start of levofloxacin therapy. Owing to a concern regarding noncompliance and the adverse effect of bleeding, warfarin dosage was adjusted in one patient even though his first INR value was in the high end of the therapeutic range (2.98, therapeutic range 2-3). One patient withdrew from the study after the first INR measurement after levofloxacin coadministration. Because of a concern about the possible bleeding complication, warfarin dosage was also adjusted in this patient after obtaining his first INR value. Therefore, only the INR values obtained before and the first INR values obtained after levofloxacin administration were compared to evaluate the effect of levofloxacin on INR determination of warfarin therapy. The INR values obtained before levofloxacin administration did not differ significantly from the first INR values obtained after levofloxacin coadministration (mean +/- SD 2.61 +/- 0.44 vs 2.74 +/- 0.83, 95% confidence interval -0.449-0.196, p=0.419). CONCLUSION: The INR values measured before and after concomitant levofloxacin therapy were not significantly different. However, the ability to detect a significant difference may be affected by the small number of patients studied. Further studies with a larger sample are required to better determine the effect of levofloxacin coadministration on INR monitoring during warfarin therapy     

Retrospective evaluation of a possible interaction between warfarin and levofloxacin

STUDY OBJECTIVES: In order to clarify the clinical significance of a suspected drug interaction, we sought to determine if the international normalized ratio (INR) is affected when levofloxacin is administered in patients receiving long-term warfarin therapy. DESIGN: Retrospective cohort study using pharmacy and medical records. SETTING: Outpatient clinic. PATIENTS: Forty-three patients receiving long-term warfarin therapy who subsequently were prescribed either levofloxacin (22 patients) or felodipine (21 controls); felodipine was chosen as it has been shown not to interact with warfarin. Patients in both groups were required to have a documented INR before the start of levofloxacin or felodipine (pre-INR) and either during levofloxacin or felodipine therapy or within 15 days after the drug had been started (post-INR). MEASUREMENTS AND MAIN RESULTS: Patient demographics were similar between the two treatment groups. The mean +/- SD age of the patients in the levofloxacin and control groups was 59.5 +/- 8.7 and 65.3 +/- 11.5 years, respectively (p=0.07). The mean change between the pre- and post-INR (primary outcome measure) was 0.31 +/- 0.82 (pre-INR 2.46, post-INR 2.76) and 0.21 +/- 0.54 (pre-INR 2.46, post-INR 2.67) in the levofloxacin and felodipine groups, respectively (p=0.65). A post hoc power analysis, based on a sample-derived, weighted standard deviation of 0.68, revealed that the study had 66% power to detect a change of 0.5 in the INR value. The percentage of patients who required a warfarin dosage adjustment based on the post-INR (secondary outcome measure) was 41% (9 of 22 patients) in the levofloxacin group and 33% (7 of 21 patients) in the felodipine group. CONCLUSION: Although our primary analysis did not detect a warfarin-levofloxacin interaction, the potential for such an interaction, especially in idiosyncratic cases, cannot be ruled out. Clinicians should closely monitor INR values when levofloxacin is administered jointly with warfarin.     

Thromboembolic consequences of subtherapeutic anticoagulation in patients stabilized on warfarin therapy: the low INR study

STUDY OBJECTIVE: To quantify the absolute risk of thromboembolism associated with a significant subtherapeutic international normalized ratio (INR) in patients with previously stable anticoagulation while receiving warfarin. DESIGN: Retrospective, matched cohort analysis. SETTING: Centralized anticoagulation service in an integrated health care delivery system. PATIENTS: A total of 2597 adult patients receiving warfarin from January 1998-December 2005; 1080 patients were in the low INR cohort and were matched to 1517 patients in the therapeutic INR cohort based on index INR date, indication for warfarin, and age. MEASUREMENTS AND MAIN RESULTS: Stable, therapeutic anticoagulation was defined as two INR values, measured at least 2 weeks apart, within or above the therapeutic range. The low INR cohort included patients with a third INR value of 0.5 or more units below their therapeutic range. The therapeutic INR cohort included patients with a third therapeutic INR value and no INR value 0.2 or more units below their target INR range in the ensuing 90 days. The primary outcome was anticoagulation-related thromboembolism during the 90 days after the index INR. Secondary outcomes were times to the first occurrence of anticoagulation-related complications (bleeding, thromboembolism, or death) in the 90 days after the index INR. Four thromboembolic events (0.4%) occurred in the low INR cohort and one event (0.1%) in the therapeutic INR cohort (p=0.214). The differences in the proportions of thromboembolism, bleeding, or death were not significant between the cohorts (p>0.05). No significant differences were noted in the hazard of thromboembolism, bleeding, or death between the cohorts (p>0.05). CONCLUSION: Patients with stable INRs while receiving warfarin who experience a significant subtherapeutic INR value have a low risk of thromboembolism in the ensuing 90 days. The risk was similar to that observed in a matched control population in whom therapeutic anticoagulation was maintained. These findings do not support the practice of anticoagulant bridge therapy for patients stabilized on warfarin therapy to reduce their risk for thromboembolism during isolated periods of subtherapeutic anticoagulation.     

Effect of age on international normalized ratio at the time of major bleeding in patients treated with warfarin

STUDY OBJECTIVES: Because the risk of major bleeding associated with warfarin increases with increasing international normalized ratio (INR) as well as with advanced age, we evaluated the association between age and INR in patients with major bleeding events related to anticoagulation with warfarin. DESIGN: Retrospective record review. SETTING: Two university-affiliated anticoagulation clinics. PATIENTS: Sixty-six patients (mean age 61.2 yrs, range 21-90 yrs) receiving warfarin therapy who experienced major bleeding, defined as bleeding requiring hospitalization, during a 20-month index period. MEASUREMENTS AND MAIN RESULTS: In patients aged 65 years or older, the mean INR at the time of a major bleeding event was significantly lower than that in patients younger than 65 years (INR 3.1 vs 4.2, respectively; p=0.01). For every 1-year increase in age, mean INR at the time of a major bleeding event decreased by 0.03 (p=0.02). CONCLUSION: Patients aged 65 years or older experience warfarin-related major bleeding events at a mean INR 1.1 units lower (95% confidence interval -1.9 to -0.27) than patients younger than 65 years. Older patients may require more aggressive management of overanticoagulation to minimize the risk of major bleeding.     

Anticoagulation-related outcomes in patients receiving warfarin after starting levofloxacin or gatifloxacin

STUDY OBJECTIVE: To compare anticoagulation-related outcomes in patients receiving stable dosages of warfarin who started levofloxacin or gatifloxacin therapy. DESIGN: Retrospective medical record review. SETTING: Veterans Affairs medical center. PATIENTS: Of 92 patients receiving the same dosages of warfarin for at least 4 weeks before starting antibiotic therapy, 54 received levofloxacin between January and September 2003, and 38 received gatifloxacin between January and September 2004. MEASUREMENTS AND MAIN RESULTS: Data were obtained through the hospital's pharmacy, laboratory, and general patient databases and through electronic medical records. The INRs evaluated were prefluoroquinolone use, defined as the last INR measured before the start of antibiotic therapy (up to 4 wks earlier), and postfluoroquinolone use, defined as any INR measured during antibiotic therapy through 1 week after discontinuation of the antibiotic. Analyzed outcomes included the percentage of patients with postfluoroquinolone INRs that were above 4, that exceeded the therapeutic goal, or that exceeded the goal by more than 1 point; INR changes of more than 0.5, 1, or 1.5 points above the INR before fluoroquinolone use; major or minor bleeding events; requirement for vitamin K administration; warfarin dosage reduction or withholding doses; and warfarin-related hospital, emergency, or urgent care admissions or visits. No significant differences were noted in baseline characteristics with regard to age, sex, prefluoroquinolone INR, or anticoagulation indications between the two groups. The percentage of patients with a postfluoroquinolone INR above 4 was 2% (1 of 54 patients) in the levofloxacin group versus 21% (8 of 38 patients) in the gatifloxacin group (p=0.003). The percentage of patients receiving vitamin K in the levofloxacin and gatifloxacin groups was 0% (0 of 54 patients) and 11% (4 of 38, p=0.026), respectively. For the other anticoagulation-related outcomes, no significant differences were noted between the groups. CONCLUSION: Patients receiving warfarin who take gatifloxacin may be at higher risk for an INR above 4 compared with those taking levofloxacin. Close monitoring of warfarin therapy while concomitantly receiving gatifloxacin is warranted.     

Guidelines for stroke prevention in patients with atrial fibrillation

Atrial fibrillation (AF) is a major independent risk factor for stroke. AF is most commonly associated with nonvalvular cardiovascular disease and is especially frequent among the elderly. The annual risk for stroke in patients with AF is approximately 5% with a wide range depending on the presence of additional risk factors. For patients who cannot successfully be converted and maintained in normal sinus rhythm (NSR), antithrombotic therapy is an effective method for preventing stroke. The 2 drugs which are indicated for stroke prophylaxis in patients with AF are warfarin and aspirin. For primary prevention, warfarin reduces the risk of stroke approximately 68%. Aspirin therapy is less effective, resulting in a 20 to 30% risk reduction. Combination therapy with aspirin and low intensity warfarin adjusted to an International Normalised Ratio (INR) of 1.2 to 1.5 has not been shown to be superior to standard intensity warfarin with a target INR of 2.0 to 3.0. In patients with AF and a prior history of stroke or transient ischaemic attack (TIA), the absolute risk reduction with warfarin is even greater because of the high risk of stroke in this population. In contrast, aspirin has not been shown to significantly reduce the risk of stroke in patients with AF when used for secondary prevention. When appropriately managed, warfarin is associated with a low risk of major bleeding. In controlled trials of highly selected patients, the annual rate of intracranial haemorrhage (ICH) with warfarin was approximately 0.3%. Studies have shown that specialty anticoagulation clinics can achieve similar low rates of major bleeding. However, these results cannot be extrapolated to the general population. Factors which have been identified as predictors of bleeding include advanced age, number of medications and most importantly, the intensity of anticoagulation. INR values above 4.0 have been associated with an increased risk of major bleeding while values below 2.0 have been associated with thrombosis. Slow careful dosage titration, regular laboratory monitoring and patient education can substantially reduce the risk of complications. In patients with AF, antithrombotic therapy has been shown to be cost effective. For high risk patients, warfarin is the most cost-effective therapy, provided the risks for bleeding are minimised. In contrast, aspirin is the most cost-effective agent for low risk patients. Current practice guidelines for stroke prophylaxis recommend warfarin (target INR 2.5: range 2.0 to 3.0) for AF patients at high risk for stroke including those over 75 years of age or younger patients with additional risk factors. Aspirin should be reserved for low risk patients or those unable to take warfarin. Although these recommendations are strongly supported by the clinical trial evidence, studies show that many patients are not receiving appropriate antithrombotic therapy. In particular, warfarin is underutilised in high risk elderly patients. Additional studies are needed to identify barriers that prevent implementation of the clinical trial findings into clinical practice.     

Comparison of Two Methods for INR Determination in a Pharmacist-Based Oral Anticoagulation Clinic

Warfarin is a commonly used oral anticoagulant that is usually initiated after the definitive diagnosis of a certain thromboembolic disorder or disease. Warfarin therapy will usually be prescribed for 6–12 weeks or more, and some patients may continue therapy throughout life, depending on the type of thromboembolic disorder. Major problems associated with warfarin therapy include adverse effects such as bleeding complications and drug-drug or drug-food interactions. In addition, thromboembolic complications may occur due to subtherapeutic dosages of warfarin. The laboratory reference standards for monitoring warfarin therapy are the prothrombin time (PT) and the International Normalized Ratio (INR). While both the PT or INR will reflect the clinical response in the patient, results reported as INR values have been shown to be more accurate than those reported as PT values. Thirty-two patients were enrolled in this study. Our objectives were to compare INR values measured by both the Coumatrak and conventional laboratory method, and to demonstrate the effects of pharmacist intervention on managing patients receiving warfarin therapy. Results from our study reveal that INR monitoring by Coumatrak is similar to the conventional laboratory method. In addition, our study indicates that patients receiving warfarin therapy can be monitored and managed effectively by pharmacists.     

Prediction of the international normalized ratio and maintenance dose during the initiation of warfarin therapy

AIMS: A pharmacokinetic/pharmacodynamic model, with Bayesian parameter estimation, was used to retrospectively predict the daily International Normalized Ratios (INRs) and the maintenance doses during the initiation of warfarin therapy in 74 inpatients. METHODS: INRs and maintenance doses predicted by the model were compared with the actual INRs and the eventual maintenance dose. Cases with drugs or medical conditions interacting with warfarin or receiving concurrent heparin therapy were not excluded. As the study was retrospective, model predictions of the maintenance dose were not those that were administered. Mean prediction error (MPE) and percentage absolute prediction errors (PAPE) were used to assess the model predictions. RESULTS: INR MPE ranged from -0.07 to 0.06 and median PAPE from 10% to 20%. Dose MPE ranged from -0.7 to 0.17 mg and median PAPE from 16.7% to 37.5%. Accurate and precise dose predictions were obtained after 3 or more INR feedback's. CONCLUSIONS: This study shows that the model can accurately predict daily INRs and the maintenance dose in this sample of cases. The model can be incorporated into computer decision-support systems for warfarin therapy and may lead to improvement in the initiation of warfarin therapy.     

Unreliability of international normalized ratio for monitoring warfarin therapy in patients with lupus anticoagulant

STUDY OBJECTIVE: To compare the international normalized ratios (INRs) of patients positive for lupus anticoagulant and the INRs of control patients receiving warfarin therapy with equivalent therapeutic chromogenic factor X levels. DESIGN: Prospective case series. SETTING: A 625-bed, adult, private, tertiary care teaching hospital. PATIENTS: Sixty-eight outpatients positive for lupus anticoagulant and 57 control patients receiving long-term warfarin therapy. MEASUREMENTS AND MAIN RESULTS: Concomitant INR and chromogenic factor X activity were measured in all patients. In 44 control patients (77%) and 46 patients with lupus anticoagulant (68%), chromogenic factor X activity was 22-40% of normal, which is therapeutic. Of the 44 control patients, 4 (9%) had an INR above 3.0, and none had an INR above 4.0. In contrast, 18 (39%) of the 46 patients with lupus anticoagulant had an INR above 3.0, and 5 (11%) had an INR above 4.0. CONCLUSION: At least 10% of patients with lupus anticoagulant receiving long-term warfarin therapy may have falsely high INR values, which could lead to inappropriate warfarin dosage reduction. Monitoring warfarin therapy by chromogenic factor X activity in patients with lupus anticoagulant avoids this INR artifact.     

Low-molecular-weight heparin may alter point-of-care assay for international normalized ratio

STUDY OBJECTIVE: To compare the international normalized ratio (INR) measured by a point-of-care testing device with that measured by a reference laboratory method for patients receiving either warfarin only or warfarin plus low-molecular-weight heparin (LMWH). DESIGN: Retrospective study. SETTING: Outpatient anticoagulation clinic. SUBJECTS: Ninety-one patients receiving warfarin for various indications; 59 of them receiving only warfarin and 32 receiving warfarin plus LMWH. INTERVENTION: Capillary blood was obtained for INR determination by a point-of-care device, and venous blood was obtained for INR determination in a standard reference laboratory. MEASUREMENTS AND MAIN RESULTS: Ninety-one patients had INR pairs run on a point-of-care device and by the laboratory. In both the patients receiving only warfarin and in those receiving warfarin plus LMWH, the mean INR as determined by the point-of-care testing device was statistically significantly higher than the mean INR determined by the laboratory. Although the differences were statistically significant in both groups, the clinical significance of this difference was accentuated in the patients receiving warfarin plus LMWH. The measure of divergence between the point-of-care and laboratory methods was greater in the group receiving warfarin plus LMWH than in the warfarin-only group, with a mean +/- SD percent change between the INR values of 24.19 +/- 27.54% in the warfarin plus LMWH group and 7.21 +/- 17.73% in the warfarin-only group. In assessing the clinical impact of such variability, a greater degree of discordance in dosing adjustment decisions was noted for patients receiving warfarin plus LMWH. In this group, a 25% rate of discordance was noted compared with 8% in the warfarin-only group. Such discrepancy in dosing decisions based on the point-of-care INR would have resulted in discontinuation of LMWH therapy before the patient acquired a true therapeutic INR, with use of the laboratory measurement. CONCLUSION: The INR measured with the point-of-care device in patients receiving concurrent LMWH and warfarin therapy may be inaccurate. Patients receiving LMWH in addition to warfarin should have INRs checked by means of the standard reference laboratory method.     

不同抗凝强度华法林治疗非瓣膜性房颤的安全性及缺血性脑卒中发生的危险因素评估

目的 评估华法林不同抗凝强度治疗非瓣膜性房颤的安全性,以及缺血性脑卒中发生的危险因素。方法 纳入2012年1月—2013年12月收治的130例非瓣膜性房颤患者,根据华法林抗凝治疗的强度分为中强度组：华法林中等强度抗凝治疗,国际标准化比率(international normalized ratio,INR)控制在2.0＜INR≤3.0;低强度组：华法林低等强度抗凝,INR控制在1.6≤INR≤2.0,记录2组患者治疗和随访期间缺血性脑卒中、出血栓塞等不良反应的发生率,ROC曲线法分析INR诊断抗凝出血风险,多因素Logistic回归分析患者缺血性脑卒中的危险因素。结果 中强度组缺血性脑卒中、短暂性脑缺血发作和外周动脉栓塞发生率分别为6.70%,3.45%和1.72%,与低强度组的8.33%,4.17%和4.17%比较,无统计学差异(P>0.05);中强度组华法林用量(3.13±0.45)mg·d^-1,INR值2.61±0.32,出血发生率为24.14%;低强度组华法林用量(2.63±0.32)mg·d^-1,INR值 1.84±0.30,出血发生率为9.72%。采用INR诊断患者出血风险,ROC曲线下面积为0.858(95%CI：0.791~0.924),INR的cut-off值2.85,该值下判断出血敏感性为81.1%,特异性为67.2%;多因素logistic回归分析发现年龄、合并高血压、糖尿病、心力衰竭、脑卒中病史、INR、治疗窗内时间、卒中危险评分是非瓣膜性房颤患者缺血性脑卒中发生的独立危险因素(P<0.05)。结论 中、低强度华法林抗凝治疗均有较好的抗凝效果,非瓣膜性心房颤动患者伴有血栓栓塞危险因素应尽早应用华法林抗凝治疗,严密监测INR,INR值控制在1.6≤INR≤2.0,降低和避免出血并发症。     

Is INR between 2.0 and 3.0 the optimal level for Chinese patients on warfarin therapy for moderate-intensity anticoagulation?

AIM: To examine the optimal range of International Normalized Ratio (INR) for Chinese patients receiving warfarin for moderate-intensity anticoagulation. METHODS: This was a retrospective cohort study conducted at the ambulatory setting of a 1400-bed public teaching hospital in Hong Kong. The INR measurements and occurrence of serious or life-threatening haemorrhagic and thromboembolic events among patients newly started on warfarin from 1 January 1999 to 30 June 2001 for indications with target INR 2-3 were analysed. The INR-specific incidence of bleeding and thromboembolism were calculated. RESULTS: A total of 491 patients were included, contributing to 453 patient-years of observation period. Forty-seven of the 491 patients experienced 25 haemorrhagic events (5.5 per 100 patient-years) and 27 thromboembolic events (6.0 per 100 patient-years). The percentage of patient-time spent within therapeutic INR range (2-3), INR <2 and INR >3 were 50, 44 and 6%, respectively. The incidence of either haemorrhagic or thromboembolic events was lowest (< or =4 events per 100 patient-years) at INR values between 1.8 and 2.4. CONCLUSIONS: An INR of 1.8-2.4 appeared to be associated with the lowest incidence rate of major bleeding or thromboembolic events in a cohort of Hong Kong Chinese patients receiving warfarin therapy for moderate-intensity anticoagulation.