Intended for healthcare professionals


Australian MPs demand more data on dabigatran for atrial fibrillation

BMJ 2013; 346 doi: (Published 15 January 2013) Cite this as: BMJ 2013;346:f303

Re: Australian MPs demand more data on dabigatran for atrial fibrillation

Recently unsealed court documents [1] have reignited the debate regarding the supposedly improved safety characteristics of new generation blood thinner medications.

Specifically, post-approval adverse events (AE) linked to Pradaxa (dabigatran) are again in the spotlight. It has been alleged that the drug may trigger more fatal bleeding events than was indicated in a clinical study that was pivotal in the Food and Drug Administration’s (FDA) decision to approve the drug [1]. Relatedly, over 2,000 lawsuits have been filed that contend that Boehringer executives knew dabigatran posed a deadly risk to consumers when it won FDA approval.

New orally active anticoagulants (NOACs), like dabigatran, are replacing the use of older drugs like wafarin (Coumadin). Popular opinion would indicate that they work as well as wafarin, have a better therapeutic index, more predictable pharmacokinetics, do not require sometimes complex dietary restrictions, and offer an improved safety profile because of a reduction in major bleeding complications, particularly of the intracranial variety [2].

The other major advantage often cited in the press and in marketing campaigns is that, unlike wafarin, NOACs do not require routine monitoring during their use. This point, however, has been called into question [3].

A major danger for patients taking either NOACs or older drugs like Coumadin is that the medications increase the time for blood to clot. Therefore, normally manageable bleeding from trauma, surgery, or other emergency situations, becomes a very serious concern for patients on these drugs. Coumadin, however, has an “antidote.” The NOACs do not.

Because blood-thinning drugs like dabigatran represent some of the most widely used medicines in the world, we analyzed post-marketing safety signals across this drug class.

Post-marketing surveillance of drug safety necessarily involves the analysis of numerous, and often disparate, AE types. Focusing on only one specific AE type provides good specificity but poor sensitivity. Therefore, signal detection is often enhanced by identifying links between a given drug and a group of AEs associated with either a medical condition or type of outcome.

The Medical Dictionary for Regulatory Activities (MedDRA®*) [4] and the Council for International Organization of Medical Sciences (CIOMS) created and validated categories of related AEs known as “Standardized Queries” (SMQ) [5]. These groupings help improve signal detection from AE case reports contained in the FDA’s Adverse Event Reporting System (FAERS) [6]. SMQs are developed and maintained by clinical experts with the goal of defining clusters of MedDRA terms that are associated with a medical condition.

The Reporting Odds Ratio (ROR) is a disproportionality measure commonly used by drug safety professionals to help identify drug-associated AEs that are reported more frequently than expected. The method compares expected reporting frequencies (based upon all drugs and all AEs in the FAERS database) with the amount of a given AE reported for a specific drug. High RORs suggest disproportional reporting of a given drug-side effect combination.

We examined AE signals associated with main anticoagulant drugs in the FAERS from November 1, 1997 through December 31, 2012 or December 10, 2013 (the most current data we have access to). We used a data mining platform [7] to generate case report counts and reporting ratios for multiple SMQ categories. We derived RORs and Confidence Intervals (CI) by the use of standard formulas [8 9] for “primary suspect” case reports from the drug’s approval date to the most recent FAERS data we have access to.


Table 1 lists the number of “primary suspect” case reports for each drug and how many of those reports fall into corresponding SMQ adverse event categories.

When compared with three other anticoagulants, warfarin had elevated disproportionality results (10.2 to 12.1) for the hemorrhage-related categories, but it showed lower results (0.8 to 1.9) than its’ peers across embolic and thrombotic AE groupings. Both dabigatran and rivaroxaban had hemorrhage-related AEs in the 8.0 range. Apixaban, however, was in the 4.0 range. For embolic and thrombotic AE categories dabigatran and apixaban had similar DA results in the 1.4 – 3.6 range, while rivaroxaban showed higher results of 5.6 and 13.5.


In general, post-marketing data may be subject to biases such as underreporting, stimulated reporting, and confounding by comorbidities [6 10]. A specific potential limitation to this analysis relates to the fact that warfarin has been on the market for so long that it is possible that physicians may particularly under-report AEs for the drug because they assume all of its side effects are well-known and properly documented. The disproportionality measure used herein, however, compares specific AEs for a drug with its overall AE reports, and therefore may help minimize such reporting influences.


Our review of FAERS post-marketing data suggest disproportionally elevated reporting of bleeding-related AEs linked to both newer anticoagulants (dabigatran and rivaroxaban) as well as older drugs in the class (warfarin). Our analysis also suggests that the commonly held assumption that dabigatran is generally safer than warfarin may not be entirely accurate.


*MedDRA®, the Medical Dictionary for Regulatory Activities terminology is the international medical terminology developed under the auspices of the International Conference on Harmonization of Technical Requirements for Registration of Pharmaceuticals for Human Use (ICH). The MedDRA® trademark is owned by IFPMA on behalf of ICH.


1. Feeley J, Cortez MF. Boehringer Kept Pradaxa Analysis From FDA, Records Show. 2014. Accessed February 2014.
2. Gomez-Outes A, Terleira-Fernandez AI, Calvo-Rojas G, et al. Dabigatran, Rivaroxaban, or Apixaban versus Warfarin in Patients with Nonvalvular Atrial Fibrillation: A Systematic Review and Meta-Analysis of Subgroups. Thrombosis 2013;2013:640723 doi: 10.1155/2013/640723.
3. McCarthy M. Drug company staff fretted when in-house paper's conclusion clashed with marketing claims. BMJ (Clinical research ed) 2014;348:g1505 doi: 10.1136/bmj.g1505.
4. MedDRA. Medical Dictionary for Regulatory Activities and the Maintenance and Support Services. 2013. Accessed February 2014.
5. MedDRA. Standardised MedDRA Queries. 2014. Accessed February 2014.
6. FDA. FDA Adverse Event Reporting System (FAERS) (formerly AERS). 2012. Accessed February 2014.
7. Hoffman KB, Overstreet, B.M., Doraiswamy, P.M. A Drug Safety ePlatform for Physicians, Pharmacists and Consumers based on Post-Marketing Adverse Events. Drugs and Therapy Studies 2013;3(e4)
8. Bate A, Evans SJ. Quantitative signal detection using spontaneous ADR reporting. Pharmacoepidemiology and drug safety 2009;18(6):427-36 doi: 10.1002/pds.1742.
9. Liu M, McPeek Hinz ER, Matheny ME, et al. Comparative analysis of pharmacovigilance methods in the detection of adverse drug reactions using electronic medical records. Journal of the American Medical Informatics Association : JAMIA 2013;20(3):420-6 doi: 10.1136/amiajnl-2012-001119.
10. Szarfman A, Tonning JM, Doraiswamy PM. Pharmacovigilance in the 21st century: new systematic tools for an old problem. Pharmacotherapy 2004;24(9):1099-104

Competing interests: Hoffman, Demakas, Erdman, and Dimbil are employees of AdverseEvents, Inc. (AEI). AEI, however, nor its officers or employees currently hold, and will not hold, a direct or indirect investment position (long or short) in any of the companies mentioned in this report. Neither AdverseEvents, nor its officers or employees have been directly compensated by any party for the preparation of this report. The inclusion of a particular company, drug, class or indication in this report is determined wholly by our quantitative signaling systems along with our qualitative analysis work. The inclusion or exclusion of any drug, company, or indication has not, and will not, be influenced by any third party, including any clients of AdverseEvents.

27 February 2014
Keith B. Hoffman
Scientific Affairs
Andrea Demakas, Colin B. Erdman, Mo Dimbil
Adverse Events, Inc.
3663 N. Laughlin Road