Importance
Significant variations in dose requirements of warfarin and its analogues (acenocoumarol and phenprocoumon) make selecting the appropriate dose for an individual patient difficult. Genetic factors account for approximately one-third of the variation in dose requirement. The clinical usefulness of genotype-guided dosing of warfarin has been previously assessed in randomized clinical trials that were limited by lack of power and inconsistent results.
Objective
To compare genotype-guided initial dosing of warfarin and its analogues with clinical dosing protocols.
Data Sources and Study Selection
MEDLINE (inception to December 31, 2013), EMBASE (inception to December 31, 2013), and the Cochrane Library Central Register of Controlled Trials (inception to December 31, 2013) were searched for randomized clinical trials comparing genotype-guided warfarin dosing vs clinical dosing for adults with indications for anticoagulation.
Data Extraction and Synthesis
Two investigators extracted data independently on trial design, baseline characteristics, and outcomes. High-quality studies were considered those that described an appropriate method of randomization, allocation concealment, blinding, and completeness of follow-up.
Main Outcomes and Measures
The outcomes analyzed included the percentage of time that the international normalized ratio (INR) was within the therapeutic range, the percentage of patients with an INR greater than 4, and the incidence of major bleeding and thromboembolic events. Summary standardized differences in means (or Mantel-Haenszel risk ratios) were obtained using a random-effects model.
Results
In 9 trials, 2812 patients were randomized to receive warfarin, acenocoumarol, or phenprocoumon according to a genotype-guided algorithm or a clinical dosing algorithm. Follow-up ranged from 4 weeks to 6 months (median, 12 weeks). The standardized difference in means of the percentage of time that the INR was within the therapeutic range was 0.14 (95% CI, −0.10 to 0.39) in the genotype-guided dosing cohort (P = .25). The risk ratio for an INR greater than 4 was 0.92 (95% CI, 0.82 to 1.05) for genotype-guided dosing vs clinical dosing. The risk ratios for major bleeding and thromboembolic events were 0.60 (95% CI, 0.29 to 1.22) and 0.97 (95% CI, 0.46 to 2.05), respectively, for genotype-guided vs clinical dosing.
Conclusions and Relevance
In this meta-analysis of randomized clinical trials, a genotype-guided dosing strategy did not result in a greater percentage of time that the INR was within the therapeutic range, fewer patients with an INR greater than 4, or a reduction in major bleeding or thromboembolic events compared with clinical dosing algorithms.
Warfarin and its analogues (acenocoumarol and phenprocoumon) are widely prescribed for the prevention of thromboembolic events associated with atrial fibrillation, arterial and venous thrombosis, pulmonary embolism, and prosthetic heart valves. It is very difficult to determine accurate doses for these agents owing to a narrow therapeutic index and up to a 20-fold interindividual variability in dose-response.1 Underdosing increases the risk for thrombotic events, whereas overdosing increases the risk for hemorrhagic events.2,3
Genotypes of the cytochrome P450 isoform CYP2C9 and the vitamin K epoxidereductase complex subunit 1 VKORC1 have been found to influence warfarin dose requirements.4-13 The *2 (R144C) and *3 (1359L) allelic variants of CYP2C9 (GenBank AY702706) cause reductions in enzymatic activity of approximately 30% and 80%, respectively, resulting in an increase in bleeding risk.6 Ten VKORC1 (GenBank AY587020) single-nucleotide polymorphisms determine low-, intermediate-, and high-dose requirements.8,12 These genotypes alone account for approximately 35% of the variability in dosing, and genotype plus clinical characteristics explain approximately one-half of interindividual dose variability,11-15 suggesting that CYP2C9 and VKORC1 genotyping followed by genotype-guided warfarin dosing would be superior to standard clinical dosing algorithms in achieving a therapeutic effect while minimizing thrombotic and hemorrhagic complications.
In 2007, the US Food and Drug Administration issued a labeling change advising physicians to consider the use of genetic tests to improve their initial estimate of warfarin dose. A second label change in 2010 noted that if the genotype of a patient was known, it should be considered in dose selection and follow-up.16 However, prior randomized clinical trials of genotype-guided warfarin dosing have been underpowered with regard to clinical end points and their results have been inconsistent.17 As a result, the biomedical research and clinical communities remain largely undecided as to the usefulness of incorporating genotypic information into warfarin prescribing decisions.17 Given the equipoise surrounding this issue, we performed a meta-analysis of randomized clinical trials that compared genotype-guided dosing of warfarin, acenocoumarol, or phenprocoumon with clinical dosing algorithms in adults with indications for oral anticoagulation.
A systematic search of published studies in any language in MEDLINE, Cochrane, BioMed Central, and PubMed databases from their inception to December 31, 2013, was performed independently by both authors (K.S. and D.L.B.). Search terms included warfarin, coumarin, genotype, pharmacogenetics, and dosage, as well as combinations. A filter for randomized clinical trials was used. In addition, the bibliographies of retrieved articles and prior reviews on the subject were searched for other relevant studies.
For inclusion, studies were required to be prospective randomized trials of genotype-guided dosing algorithms vs clinical dosing algorithms in patients with indications for oral anticoagulation with warfarin, acenocoumarol, or phenprocoumon. Included studies had to report at least 1 of the end points of interest.
Patient characteristics, study design, and outcomes were systematically reviewed and recorded independently by the 2 authors. Disagreements were resolved by consensus.
The methodologic quality of each trial was evaluated for the risk of bias using standard criteria: method of randomization; allocation concealment; patient, investigator, and outcome assessor blinding; selective outcome reporting; incomplete outcome ascertainment; and other potential sources of bias as recommended by the Cochrane Collaboration.18 The Jadad score19 for evaluating randomized clinical trials was also applied, with a score of 3 or greater indicative of high quality.
The primary end point of this analysis was the percentage of time that the international normalized ratio (INR) was within the therapeutic range. The justification for selecting this primary end point was that all 9 eligible trials included this metric as either a primary or secondary end point. In addition, this metric is accepted as a measure of anticoagulation quality and has been shown20 to be a surrogate for bleeding and thromboembolic risk. Secondary end points were the percentage of patients with an INR greater than 4 and the incidence of major bleeding and thromboembolic events. End point definitions were those used in the individual trials.
Data from each trial were entered on an intention-to-treat basis according to the recommendations of the Cochrane Collaboration and the Preferred Reporting Items for Systematic Reviews and Meta-analyses statement.21 Baseline characteristics were summarized, and weighted means and rates according to individual trial sample size were reported. Trials were compared with standardized difference in means and Mantel-Haenszel risk ratios (RRs) as the measures of effect. Risk ratios were used because accurate time-to-event data were not available in all trials. Summary standardized differences in means, RRs, and 95% CIs were calculated using a random-effects model for combining results across studies, which incorporates between- and within-study variance and provides a more conservative summary. A random-effects model was preferred because heterogeneity across patient characteristics and clinical trial design would be unlikely to result in a consistent treatment effect across trials.22 When no events were observed within a treatment group, a 0.5 correction factor was added to all values of that end point for calculation of the RR and its variance.23,24 To determine whether there was heterogeneity between individual trials, we assessed the Q statistic (a weighted index of effect estimate differences across studies assuming a χ2 distribution) and I2 statistic ([Q − df]/Q × 100), where df indicates degrees of freedom. Because the I2 value quantifies heterogeneity on a scale of 0% to 100% and represents the extent of inconsistency among trial results rather than a sampling error independent of the number of studies, an I2 value of 75% or greater was considered representative of a high level of heterogeneity.25 To assess for publication bias risk, funnel plots (precision [inverse of SE] vs effect size) were evaluated. Additional statistical tests for funnel plot asymmetry were not conducted given the limited specificity and power of these tests when fewer than 10 studies are included in a primary meta-analysis.26
Heterogeneity was explored in subgroup analysis by study quality (high vs low), study location (United States vs other countries), and sample size (≤200 patients vs >200 patients). The Q test for heterogeneity was used to evaluate treatment effect among subgroups. Sensitivity analyses were performed for each outcome to determine whether any single study disproportionally influenced the pooled estimate by excluding individual trials one at a time and recalculating the combined standardized difference in means or RR for the remaining studies.
P < .05 was considered statistically significant, and all tests were 2-sided. Statistical analyses were performed with Comprehensive Meta-analysis, version 2 (Biostat Inc).
The electronic search yielded 660 citations that were screened by reviewing the title or abstract of each. Of these, 39 publications were reviewed in full and 9 trials27-36 were included in the meta-analysis (Figure 1). The 9 trials enrolled patients between 2001 and 2013 from the United States, Israel, and Europe (Table 1). Funding was derived from governmental, foundation, hospital, health care system, university, and industry sources. The genotypic testing and dosing algorithms are presented in Table 1. Six studies28,31-35 used CYP2C9 and VKORC1 testing, 2 studies27,29 used only CYP2C9, and 1 study30 used CYP2C9, VKORC1, and CYP4F2. The randomization to genotype guidance or clinical dosing pertained to the period of warfarin initiation. Subsequent dosing decisions were generally determined by clinical protocols. Of the 2812 patients randomized, 1411 patients were assigned to genotype guidance and 1401 individuals were randomized to various clinical dosing algorithms alone. Baseline characteristics of the study populations are provided in the eTable in the Supplement. Most patients enrolled in the studies were white and were evenly divided between males and females. The mean or median ages of patients randomized ranged from 45 to 70.5 years in the 9 studies. Atrial fibrillation or atrial flutter, deep venous thrombosis, and pulmonary embolism were the most common indications for anticoagulation. The range of follow-up was 4 weeks to 6 months, with a median follow-up duration of 12 weeks.
Study quality is summarized in Table 2. Although only one study was double-blind,33 all studies were randomized and reported on patient attrition. A Jadad score of 3 or higher, indicating high quality, was achieved by 7 of 9 studies.
The standardized difference in means of the percentage of time that the INR was within the therapeutic range was 0.14 (95% CI, −0.10 to 0.39; P = .25; I2 = 88%) greater for the 1413 patients in the genotype-guided treatment arms than the 1399 patients in the group with clinically based doses (Figure 2A). An INR greater than 4 was reported in 340 of 1316 (25.8%) of the patients with genotype-guided dosing compared with 365 of 1305 (28.0%) of those with clinically based doses. The RR for an INR greater than 4 in the genotype-guided cohort was 0.92 (95% CI, 0.82 to 1.05; P = .21; I2 = 0%) (Figure 2B). Major bleeding was reported in 12 of 1297 (0.9%) patients in the genotype-guided dosing group compared with 21 of 1289 (1.6%) patients in the clinical dosing group. The RR for major bleeding was 0.60 (95% CI, 0.29 to 1.22; P = .16; I2 = 0%) (Figure 2C). Among the patients randomized to genotype-guided dosing, 14 of 1299 individuals (1.1%) had thromboembolic events compared with 16 of 1287 patients (1.2%) randomized to clinical dosing (RR, 0.97; 95% CI, 0.46 to 2.05; P = .93; I2 = 0%) (Figure 2D).
Subgroup Analyses, Sensitivity Analyses, and Publication Bias
Subgroup analysis showed no significant differences in outcomes when results were analyzed by study quality, study location, or sample size (results not shown). Sensitivity analyses to assess the potential effect of qualitative differences on study design and patient selection showed that exclusion of any single trial from the analyses for the percentage of time of INR within the therapeutic range, INR greater than 4, major bleeding episodes, and thromboembolic events did not alter the overall findings of the analysis (data not shown). The exclusion of data from the study of Verhoef et al,35 which was the only study to use the warfarin analogues (acenocoumarol and phenprocoumon) did not change the overall results for any end point. Visual inspection of the funnel plots did not suggest publication bias (eFigure in the Supplement).
In this meta-analysis of 2812 patients undergoing anticoagulation with warfarin or a warfarin analogue, an initial dosing algorithm based on patient genotype was not superior to clinical dosing algorithms alone with regard to the percentage of time the INR was within the therapeutic range, percentage of time the INR was greater than 4, major bleeding episodes, or thromboembolic events. Only one prior meta-analysis42 of 3 randomized studies has been performed. The inclusion of more contemporary studies randomizing approximately 4-fold more patients to either genotype-guided vs clinical dosing for warfarin initiation in the current meta-analysis makes the current findings unique.
Warfarin and its analogues have been used as oral anticoagulants for more than 50 years.43 In 2004, 31 million warfarin prescriptions were filled in the United States.44 Although that number has likely declined with the introduction of newer anticoagulants,45 warfarin remains a commonly used medication with significant morbidity and mortality associated with its use. Warfarin is a leading cause of drug-related emergency department visits44 and is the most cited reason for drug-related mortality.46 Most of the adverse events associated with warfarin use are thought to occur within the initiation period, which can last several weeks.47
By targeting VKOR, warfarin interferes with the posttranslational modification of the vitamin K–dependent blood coagulation proteins (factors II, VII, IX, and X). The effect of warfarin on coagulation is monitored in the clinical laboratory with the prothrombin time, which is adjusted to reflect the potency of the tissue factor used in the assay, resulting in the INR. The warfarin dose is adjusted to maintain the INR in a predetermined therapeutic range for the specific clinical indication for anticoagulation. Patients with a subtherapeutic INR are at increased risk of thrombosis, whereas those with a supratherapeutic INR are at risk of hemorrhage. Because of interindividual variability in response to a dose of warfarin, determining and maintaining the therapeutic dose is very difficult.43
The variable response to warfarin doses reflects individual differences in pharmacokinetic and pharmacodynamic factors.48 Pharmacokinetics involves drug absorption, distribution, metabolism, and elimination. For warfarin, variation in metabolism by genetic variation of a cytochrome P450 molecule can lead to a large variability in drug concentrations, a situation referred to by Roden et al48 as “high-risk pharmacokinetics.” Coding or regulatory variation in VKORC1, encoding the warfarin target, is a key pharmacodynamic contributor to variability in warfarin response. Augmenting clinical warfarin dosing protocols with pharmacogenomic data increases the predictability of warfarin dose response from as low as 20% to 60%.48
Customization of clinical decision making based on the genetic information of individual patients is the hallmark of the movement toward what is referred to as personalized medicine. Given that genetic differences in CYP2C9 and VKORC1 account for the largest single source of individual variation in warfarin dose requirements, genotype-based dosing would seem to be an ideal application of personalized medicine. However, the present meta-analysis indicates that genotype-guided warfarin dosing does not result in laboratory or clinical evidence of improved outcomes. Specifically, the percentage of time that the INR was within the therapeutic range, the incidence of supratherapeutic INRs, and the occurrence of major bleeding episodes or thromboembolic events were not significantly different in the genotype-guided and clinical dosing groups.
However, 3 caveats should be considered in interpreting these data. First, the studies included in the present meta-analysis addressed only the process of dose initiation and not the subsequent modifications throughout the duration of anticoagulant therapy. Second, the control groups in most studies received doses according to established clinical algorithms and were aggressively monitored with frequent INR testing; this scenario is not equivalent to the current standard of care.17 In fact, the largest study34 using a standard-of-care control group found a significant benefit of genotype-guided dosing in terms of the percentage of time that the INR was within the therapeutic range and the incidence of excessive anticoagulation (INR >4). Like the other studies included in this meta-analysis, this trial34 was underpowered to assess clinical outcomes. As noted by Zineh and colleagues,17 it is unlikely that the addition of 1 or 2 explanatory covariates affecting a relatively small proportion of the trial population would substantially affect the performance of a multivariable model in a setting in which frequent INR monitoring is conducted. Finally, the percentage of time that the INR was within the therapeutic range varied across the studies, resulting in significant heterogeneity for that end point and raising concern that the methods of ascertainment of this outcome are likely to have differed between studies. This outcome measure, a surrogate for safety, has been criticized as a primary outcome measure for warfarin pharmacogenetic trials and should instead be considered as a piece of a “totality-of-evidence” approach to pharmacogenetics.17
The threshold at which sufficient evidence exists that allows for the translation of pharmacogenetic data into tools that physicians can use to improve health outcomes for individual patients has not been established. In an era of cost containment and patient-centered care, setting a threshold should require the establishment of improved outcomes, reduction in adverse events, and reduction in cost, as well as improvements in quality of life before the widespread use of genotype-guided testing for each patient starting warfarin therapy. The US Food and Drug Administration16 relabeling of the warfarin package insert in 2010 came as a surprise because none of the above measures had been demonstrated. The package insert16 of warfarin advises that “lower initiation doses should be considered for patients with certain genetic variations in CYP2C9 and VKORC1 enzymes.” This change was based on an accumulation of data demonstrating that the allelic variants CYP2C9 and VKORC1 are associated with increased warfarin levels, INRs that are out of the therapeutic range, and increased bleeding risk.17 Moreover, cost-effectiveness analyses of a genotype-guided strategy for warfarin dosing have been inconclusive and suggest that cost savings are sensitive to the upfront costs of genetic testing, overall effectiveness, and the individual patient’s risk of bleeding.36 It therefore appears unlikely that genotype-guided initiation of warfarin therapy will become the standard of care in the near future.
Our analysis should be interpreted within the context of several limitations. First, the studies in this meta-analysis used different genotype-based and clinical warfarin dosing algorithms, different outcome definitions, and different durations of warfarin dose initiation and follow-up. Second, the number of clinical events was small, raising the possibility of type II error. Third, we did not evaluate genotype-specific outcomes because these data were not readily available for individual outcome measures. Finally, data were extracted only from randomized clinical trials and therefore may not be representative of patients seen in daily practice.
The strengths of the present study include efforts to identify and systematically review all randomized clinical trials of genotype-guided warfarin dosing since the inception of major biomedical literature databases, thereby limiting the likelihood of publication bias and risk of confounding from nonrandomized studies. In addition, we performed sensitivity analyses that revealed no suggestion of inconsistency among trial results or missing data.
Although there is great public appetite for the concept of personalized medicine, this approach does not appear to offer incremental benefit over clinical dosing of warfarin alone in patients with indications for oral anticoagulation. At this time, it would appear prudent to allocate increasingly scarce financial resources to establishing better infrastructure for INR testing, communication between patient and provider, implementation of validated clinical anticoagulation protocols, and promoting patient adherence rather than to testing for genotype and incorporating that information into genotype-based dosing algorithms.
Accepted for Publication: March 27, 2014.
Corresponding Author: David L. Brown, MD, Cardiovascular Division, Washington University School of Medicine, 660 S Euclid Ave, Campus Box 8086, St Louis, MO 63110 (dbrown@dom.wustl.edu).
Published Online: June 16, 2014. doi:10.1001/jamainternmed.2014.2368.
Author Contributions: Dr Brown had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Brown.
Acquisition, analysis, or interpretation of data: All authors.
Drafting of the manuscript: All authors.
Critical revision of the manuscript for important intellectual content: All authors.
Statistical analysis: All authors.
Conflict of Interest Disclosures: None reported.
2.Hylek
EM, Skates
SJ, Sheehan
MA, Singer
DE. An analysis of the lowest effective intensity of prophylactic anticoagulation for patients with nonrheumatic atrial fibrillation.
N Engl J Med. 1996;335(8):540-546.
PubMedGoogle ScholarCrossref 3.Odén
A, Fahlén
M, Hart
RG. Optimal INR for prevention of stroke and death in atrial fibrillation: a critical appraisal.
Thromb Res. 2006;117(5):493-499.
PubMedGoogle ScholarCrossref 4.Crespi
CL, Miller
VP. The R144C change in the
CYP2C9*2 allele alters interaction of the cytochrome P450 with NADPH:cytochrome P450 oxidoreductase.
Pharmacogenetics. 1997;7(3):203-210.
PubMedGoogle ScholarCrossref 5.Takanashi
K, Tainaka
H, Kobayashi
K, Yasumori
T, Hosakawa
M, Chiba
K. CYP2C9 Ile359 and Leu359 variants: enzyme kinetic study with seven substrates.
Pharmacogenetics. 2000;10(2):95-104.
PubMedGoogle ScholarCrossref 6.Aithal
GP, Day
CP, Kesteven
PJ, Daly
AK. Association of polymorphisms in the cytochrome P450 CYP2C9 with warfarin dose requirement and risk of bleeding complications.
Lancet. 1999;353(9154):717-719.
PubMedGoogle ScholarCrossref 7.Higashi
MK, Veenstra
DL, Kondo
LM,
et al. Association between CYP2C9 genetic variants and anticoagulation-related outcomes during warfarin therapy.
JAMA. 2002;287(13):1690-1698.
PubMedGoogle ScholarCrossref 8.D’Andrea
G, D’Ambrosio
RL, Di Perna
P,
et al. A polymorphism in the
VKORC1 gene is associated with an interindividual variability in the dose-anticoagulant effect of warfarin.
Blood. 2005;105(2):645-649.
PubMedGoogle ScholarCrossref 9.Hillman
MA, Wilke
RA, Caldwell
MD, Berg
RL, Glurich
I, Burmester
JK. Relative impact of covariates in prescribing warfarin according to
CYP2C9 genotype.
Pharmacogenetics. 2004;14(8):539-547.
PubMedGoogle ScholarCrossref 10.Wadelius
M, Sörlin
K, Wallerman
O,
et al. Warfarin sensitivity related to CYP2C9, CYP3A5, ABCB1 (MDR1) and other factors.
Pharmacogenomics J. 2004;4(1):40-48.
PubMedGoogle ScholarCrossref 11.Wadelius
M, Chen
LY, Downes
K,
et al. Common VKORC1 and GGCX polymorphisms associated with warfarin dose.
Pharmacogenomics J. 2005;5(4):262-270.
PubMedGoogle ScholarCrossref 12.Rieder
MJ, Reiner
AP, Gage
BF,
et al. Effect of
VKORC1 haplotypes on transcriptional regulation and warfarin dose.
N Engl J Med. 2005;352(22):2285-2293.
PubMedGoogle ScholarCrossref 13.Carlquist
JF, Horne
BD, Muhlestein
JB,
et al. Genotypes of the cytochrome p450 isoform, CYP2C9, and the vitamin K epoxide reductase complex subunit 1 conjointly determine stable warfarin dose: a prospective study.
J Thromb Thrombolysis. 2006;22(3):191-197.
PubMedGoogle ScholarCrossref 14.Sconce
EA, Khan
TI, Wynne
HA,
et al. The impact of
CYP2C9 and
VKORC1 genetic polymorphism and patient characteristics upon warfarin dose requirements: proposal for a new dosing regimen.
Blood. 2005;106(7):2329-2333.
PubMedGoogle ScholarCrossref 15.Aquilante
CL, Langaee
TY, Lopez
LM,
et al. Influence of coagulation factor, vitamin K epoxide reductase complex subunit 1, and cytochrome P450 2C9 gene polymorphisms on warfarin dose requirements.
Clin Pharmacol Ther. 2006;79(4):291-302.
PubMedGoogle ScholarCrossref 17.Zineh
I, Pacanowski
M, Woodcock
J. Pharmacogenetics and coumarin dosing—recalibrating expectations.
N Engl J Med. 2013;369(24):2273-2275.
PubMedGoogle ScholarCrossref 18.Higgins
JP, Altman
DG, Gøtzsche
PC,
et al; Cochrane Bias Methods Group; Cochrane Statistical Methods Group. The Cochrane Collaboration’s tool for assessing risk of bias in randomised trials.
BMJ. 2011;343:d5928. doi:10.1136/bmj.d5928.
Google ScholarCrossref 19.Jadad
AR, Moore
RA, Carroll
D,
et al. Assessing the quality of reports of randomized clinical trials: is blinding necessary?
Control Clin Trials. 1996;17(1):1-12.
PubMedGoogle ScholarCrossref 20.Veeger
NJ, Piersma-Wichers
M, Tijssen
JG, Hillege
HL, van der Meer
J. Individual time within target range in patients treated with vitamin K antagonists: main determinant of quality of anticoagulation and predictor of clinical outcome: a retrospective study of 2300 consecutive patients with venous thromboembolism.
Br J Haematol. 2005;128(4):513-519.
PubMedGoogle ScholarCrossref 21.Moher
D, Liberati
A, Tetzlaff
J, Altman
DG; PRISMA Group. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement.
BMJ. 2009;339:b2535. doi:10.1136/bmj.b2535.
Google ScholarCrossref 24.Friedrich
JO, Adhikari
NK, Beyene
J. Inclusion of zero total event trials in meta-analyses maintains analytic consistency and incorporates all available data.
BMC Med Res Methodol. 2007;7:5. doi:10.1186/1471-2288-7-5.
PubMedGoogle ScholarCrossref 26.Sterne
JA, Sutton
AJ, Ioannidis
JP,
et al. Recommendations for examining and interpreting funnel plot asymmetry in meta-analyses of randomised controlled trials.
BMJ. 2011;343:d4002. doi:10.1136/bmj.d4002.
PubMedGoogle ScholarCrossref 27.Hillman
MA, Wilke
RA, Yale
SH,
et al. A prospective, randomized pilot trial of model-based warfarin dose initiation using CYP2C9 genotype and clinical data.
Clin Med Res. 2005;3(3):137-145.
PubMedGoogle ScholarCrossref 28.Anderson
JL, Horne
BD, Stevens
SM,
et al; Couma-Gen Investigators. Randomized trial of genotype-guided versus standard warfarin dosing in patients initiating oral anticoagulation.
Circulation. 2007;116(22):2563-2570.
PubMedGoogle ScholarCrossref 29.Caraco
Y, Blotnick
S, Muszkat
M. CYP2C9 genotype-guided warfarin prescribing enhances the efficacy and safety of anticoagulation: a prospective randomized controlled study.
Clin Pharmacol Ther. 2008;83(3):460-470.
PubMedGoogle ScholarCrossref 30.Burmester
JK, Berg
RL, Yale
SH,
et al. A randomized controlled trial of genotype-based Coumadin initiation.
Genet Med. 2011;13(6):509-518.
PubMedGoogle ScholarCrossref 31.Borgman
MP, Pendleton
RC, McMillin
GA
et al. Prospective pilot trial of PerMIT vs standard anticoagulation service management of patients initiating oral anticoagulation.
Thromb Haemost. 2012;108(3):561-569.
PubMedGoogle ScholarCrossref 32.Jonas
DE, Evans
JP, McLeod
HL,
et al. Impact of genotype-guided dosing on anticoagulation visits for adults starting warfarin: a randomized controlled trial.
Pharmacogenomics. 2013;14(13):1593-1603.
PubMedGoogle ScholarCrossref 33.Kimmel
SE, French
B, Kasner
SE,
et al; COAG Investigators. A pharmacogenetic versus a clinical algorithm for warfarin dosing.
N Engl J Med. 2013;369(24):2283-2293.
PubMedGoogle ScholarCrossref 34.Pirmohamed
M, Burnside
G, Eriksson
N,
et al; EU-PACT Group. A randomized trial of genotype-guided dosing of warfarin.
N Engl J Med. 2013;369(24):2294-2303.
PubMedGoogle ScholarCrossref 35.Verhoef
TI, Ragia
G, de Boer
A,
et al; EU-PACT Group. A randomized trial of genotype-guided dosing of acenocoumarol and phenprocoumon.
N Engl J Med. 2013;369(24):2304-2312.
PubMedGoogle ScholarCrossref 36.Eckman
MH, Rosand
J, Greenberg
SM, Gage
BF. Cost-effectiveness of using pharmacogenetic information in warfarin dosing for patients with nonvalvular atrial fibrillation.
Ann Intern Med. 2009;150(2):73-83.
PubMedGoogle ScholarCrossref 37.Kovacs
MJ, Rodger
M, Anderson
DR,
et al. Comparison of 10-mg and 5-mg warfarin initiation nomograms together with low-molecular-weight heparin for outpatient treatment of acute venous thromboembolism: a randomized, double-blind, controlled trial.
Ann Intern Med. 2003;138(9):714-719.
PubMedGoogle ScholarCrossref 38.Ageno
W, Johnson
J, Nowacki
B, Turpie
AG. A computer generated induction system for hospitalized patients starting on oral anticoagulant therapy.
Thromb Haemost. 2000;83(6):849-852.
PubMedGoogle Scholar 39.Ansell
J, Hirsh
J, Hylek
E, Jacobson
A, Crowther
M, Palareti
G. Pharmacology and management of the vitamin K antagonists: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines (8th edition).
Chest. 2008;133(suppl 6):160S-198S.
PubMedGoogle ScholarCrossref 40.Gage
BF, Eby
C, Johnson
JA,
et al. Use of pharmacogenetic and clinical factors to predict the therapeutic dose of warfarin.
Clin Pharmacol Ther. 2008;84(3):326-331.
PubMedGoogle ScholarCrossref 41.Klein
TE, Altman
RB, Eriksson
N,
et al; International Warfarin Pharmacogenetics Consortium. Estimation of the warfarin dose with clinical and pharmacogenetic data.
N Engl J Med. 2009;360(8):753-764.
PubMedGoogle ScholarCrossref 42.Kangelaris
KN, Bent
S, Nussbaum
RL, Garcia
DA, Tice
JA. Genetic testing before anticoagulation? a systematic review of pharmacogenetic dosing of warfarin.
J Gen Intern Med. 2009;24(5):656-664.
PubMedGoogle ScholarCrossref 44.Wysowski
DK, Nourjah
P, Swartz
L. Bleeding complications with warfarin use: a prevalent adverse effect resulting in regulatory action.
Arch Intern Med. 2007;167(13):1414-1419.
PubMedGoogle ScholarCrossref 45.Ruff
CT, Giugliano
RP, Braunwald
E,
et al. Comparison of the efficacy and safety of new oral anticoagulants with warfarin in patients with atrial fibrillation: a meta-analysis of randomised trials.
Lancet. 2014;383(9921):955-962.
PubMedGoogle ScholarCrossref 46.Budnitz
DS, Pollock
DA, Weidenbach
KN, Mendelsohn
AB, Schroeder
TJ, Annest
JL. National surveillance of emergency department visits for outpatient adverse drug events.
JAMA. 2006;296(15):1858-1866.
PubMedGoogle ScholarCrossref 47.Landefeld
CS, Goldman
L. Major bleeding in outpatients treated with warfarin: incidence and prediction by factors known at the start of outpatient therapy.
Am J Med. 1989;87(2):144-152.
PubMedGoogle ScholarCrossref