5 Part Series: Improving the Quality of Chronic Oral Anticoagulation with Evidence-Based Guidelines

Part 1: Adverse Drug Events Related to Anticoagulation/VKA

Coumadin (warfarin) is a tough sell. As someone who has directed some of the largest Coumadin clinics in the country over the last 20 years, it is true that Coumadin is a drug that is impossible to dose, tough to monitor, causes death when used properly, and then lifesaving, when it’s not trying to kill you.

It shouldn’t be surprising that warfarin is the number one cause of adverse drug-related events in older Americans requiring hospitalization. In a fairly recent study (New England Journal of Medicine, 2011; 365:2002-12), adverse drug events (ADE’s) associated with warfarin occurred more often than oral antiplatelet agents, oral hypoglycemic agents, opioid analgesics, Digoxin as well as other high-risk drugs.

Warfarin is associated with a third of all ADE’s in older Americans. It’s a widespread and systemic problem. Because of this, there are multiple anticoagulant related ADE metrics as part of the Federal Surveillance System. In both in-patient and outpatient settings, ADE metrics for anticoagulants includes:

  • outpatient clinic visits
  • emergency department visits
  • hospitalizations using relevant ICD9 criteria which accounts for bleeding and other adverse events

In my view, some of the most important federal surveillance measures for anticoagulation includes process measures such as:

  • time spent in therapeutic range (TTR) for a patient on warfarin
  • patient on Coumadin with out-of-range international normalized ration (INR) values
  • patients with the need for reversal agents such as vitamin K orders
  • INR monitoring frequency

These are important concepts, because, we should keep in mind that warfarin lies within a very narrow therapeutic range that is measured by the INR. Once the INR dips below 2, there is a sharp rise in thrombotic events, and once the INR goes above 3, specifically above 5, we start seeing a sharp rise in bleeding events.


In order to maximize the beneficial effects of warfarin, one has to keep this drug within a very narrow therapeutic INR range, usually, by convention, between 2 and 3. Systematic review of 67 studies of anticoagulation management showed the average TTR of usual medical care was 56.7%.

Looking at the chart below, warfarin treatment in patients with atrial-fibrillation, observing outcomes associated with varying levels of INR controls, show the expected survival rate stroke for patients with moderate to high level risks.


The turquois blue line shows patients not on warfarin. The yellow and red line show the patients that have less than optimal TTR, i.e. a TTR less than 60%. We’re not doing much better than nothing with respect to improving survival with patients with AF, likely by subjecting patients to all of the potential harms of warfarin therapy without the benefits of stroke risk reduction. We see that ONLY when TTR’s approach 60% as we see in the green line, and specifically over 70% as we see in the blue line, do we maximize the beneficial effects of warfarin.

This shows the concept of TTR is quite important from the population survival point of view. I would also like you to know that if patients have poor warfarin control as depicted by TTR’s under 40% (as seen in the yellow or red lines), then actually, we are killing more patients by warfarin-associated adverse events than helping them by preventing strokes associated with atrial-fibrillation. This has major quality implications from a public health perspective.

IMAGE3 Warfarin Adverse Events

Indeed, if we take this concept of TTR one step further, and go to the individual patient level, what we see are patients that have very poor TTR metrics (TTR <60%) have an over two-fold increase of thromboembolic events and major bleed events compared to patients with good control (TTR >65%). This has implications not just from a population point of view, but also at the individual patient level.

We see on average, that 25-33% of patients have extremely poor TTR metrics. From a public health perspective, what we see is a 10% improvement in TTR has been associated with a 10% reduction in adverse event rates. Optimal INR management over usual medical care should prevent 7 strokes, MI (myocardial infarctions), major bleeds, or death per 100/patients per year.

It is for this reason, that the Center for Medicare and Medicaid Services (CMS) has garnered TTR as a Meaningful Use 2 goal, specifically a TTR < 65% as needing improvement, a TTR >= 65 – 70% as acceptable, and a TTR > 70% as high quality.


What are you and your clinic doing to reduce adverse events and preparing to meet MU-2 requirements for TTR? I would love to hear your comments.

Also, please follow this blog for part 2 of my 5 part series on Improving the Quality of Chronic Oral Anticoagulation with Evidence-Based Guidelines where I will discuss clinical decision support tools in the dosing of both warfarin and non-vitamin K antagonist oral anticoagulants (NOACs).

Topics: Anticoagulation, Hamilton Nomogram, Warfarin, DOAC, Adverse Drug Events