Meta-Analysis to Assess the Quality of International Normalized Ratio Control and Associated Outcomes in Venous Thromboembolism Patients Elizabeth S. Mearns, Christine G. Kohn, Ju-Sung Song, Jessica Hawthorne, Joy Meng, C. Michael White, Monika K. Raut, Jeff R. Schein, Craig I. Coleman PII: DOI: Reference:
S0049-3848(14)00301-6 doi: 10.1016/j.thromres.2014.05.035 TR 5546
To appear in:
Thrombosis Research
Received date: Revised date: Accepted date:
13 March 2014 24 May 2014 27 May 2014
Please cite this article as: Mearns Elizabeth S., Kohn Christine G., Song Ju-Sung, Hawthorne Jessica, Meng Joy, White C. Michael, Raut Monika K., Schein Jeff R., Coleman Craig I., Meta-Analysis to Assess the Quality of International Normalized Ratio Control and Associated Outcomes in Venous Thromboembolism Patients, Thrombosis Research (2014), doi: 10.1016/j.thromres.2014.05.035
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
Meta-Analysis to Assess the Quality of International Normalized Ratio Control and Associated Outcomes in Venous Thromboembolism Patients
PT
Short Title: Quality of INR Control and Outcomes in VTE
SC
RI
Elizabeth S. Mearns, PharmDa,b (
[email protected]); Christine G. Kohn, PharmDa,b (
[email protected]); Ju-Sung Song, BSb (
[email protected]); Jessica Hawthorne, BSb (
[email protected]); Joy Meng, BSb (
[email protected]); C. Michael White, PharmDa,b; Monika K. Raut, PhD, MSc (
[email protected]); Jeff R. Schein, DrPH, MPHc (
[email protected]); Craig I. Coleman, PharmDa,b (
[email protected]) a
University of Connecticut School of Pharmacy, Department of Pharmacy Practice, 69 N Eagleville
NU
Road, Storrs, CT, 06269-3092, US; bHartford Hospital Division of Cardiology, 80 Seymour Street,
MA
Hartford, CT, 06102-5037, USA; cJanssen Scientific Affairs, LLC, Raritan, NJ, USA
Craig I. Coleman, PharmD
TE
Professor, University of Connecticut
D
Corresponding Author/Requests for Reprints:
AC CE P
University of Connecticut/ Hartford Hospital Evidence-Based Practice Center 80 Seymour Street
Hartford, CT 06102-5037, USA +1 860-545-2096
+1 860-545-2277 (fax)
[email protected] Abstract word count: 250 Total word count: 3,202 Tables: 3 Figures: 4 Target journal: Thrombosis Research
1
ACCEPTED MANUSCRIPT ABSTRACT INTRODUCTION: Patients with venous thromboembolism (VTE) frequently require vitamin K
PT
antagonists (VKAs) to prevent recurrent events, but their use increases hemorrhage risk. We performed a meta-analysis to assess the quality of international normalized ratio (INR) control, identify
RI
study-level predictors of poor control and to examine the relationship between INR control and adverse
SC
outcomes in VTE patients.
NU
MATERIALS AND METHODS: We searched bibliographic databases (1990-June 2013) for studies of
MA
VTE patients receiving adjusted-dose VKAs that reported time in range (2.0-3.0) or proportion of INRs in range and/or reported INR measurements coinciding with thromboembolic or hemorrhagic events.
D
Meta-analysis and meta-regression analysis was performed.
TE
RESULTS: Upon meta-analysis, studies found 59% (95%CI: 54-64%) of INRs measured and 61%
AC CE P
(95%CI: 59-63%) of the time patients were treated were spent outside the target range of 2.0-3.0; with a tendency for under- versus over-anticoagulation. Moreover, this poor INR control resulted in a greater chance of recurrent VTE (beta-coefficient=-0.46, p=0.01) and major bleeding (beta-coefficient=0.30, p=0.02). Patients with an INR3.0 made up 48% (95%CI: 34-61%) of major hemorrhage cases. Upon meta-regression, being VKA-naïve (-14%, p=0.04) and treated in the community (-7%, p50% signifying an important degree of statistical heterogeneity). Publication bias
6
ACCEPTED MANUSCRIPT was assessed using the Egger’s weighted regression statistic, with a p-value 3 months). No hierarchy was used in the model for
To evaluate the relationship between INR control and recurrent VTE and major bleeding events, two additional and distinct analyses were undertaken. A weighted least squares linear regression was performed; with major adverse event rates as the dependent variable and TTR as the independent variable; and each VKA arm in the analysis being weighted according to person-years of follow-up (akin to meta-analysis, studies with a greater number of person-years of follow-up received greater weight in the regression analysis). However, as the relationship between TTR and major adverse outcomes may not be strictly linear, subgroup analysis was performed by stratifying studies into categories based upon whether they reported a mean/median TTR0.05). However, publication bias was found more likely for the time below and above range
RI
analyses (Egger’s p3 months 28 (63.6) 62 (60 to 65) 3.6 (-1.9 to 9.1) 6 (66.7) 62 (56 to 69) 16.5 (-3.3 to 36.3) 3 months 16 (36.4) 56 (54 to 58) Referent 3 (33.3) 54 (50 to 59) Referent Table 3. Results of Traditional Meta-Analysis and Meta-Regression Analyses of Time in the Therapeutic Range and Proportion of International Normalized ratio Values in Range for Venous Thromboembolism Vitamin K Antagonist Studies AC=anticoagulation; CI=confidence interval; NA=not applicable; NC = non-calculable; No.=number of study arms; PINNR=proportion of international normalized ratio values in range; RCT=randomized controlled trial; TTR=time in the therapeutic range;
29
ACCEPTED MANUSCRIPT
TE
D
MA
Records screened (n = 5326)
Eli gi bil ity
AC CE P
Full-text articles assessed for eligibility (n = 1144)
In cl ud ed
PT
RI
Records after duplicates removed (n = 5326)
SC
Sc re en in g
Additional records identified through other sources (n=20)
Records identified through database searching Medline (n = 3260) Embase (n = 1090) CCTR (n=956)
NU
Id en tif ic ati on
Records excluded (n=4182) Not a RCT/OBS study (n=2901) Not in a VTE population (n=861)