Coumadin Dosing Initiation Explained By an Experienced Hospitalist.

A reader asked me this question:  In a follow-up progress note (soap note) for a hospitalized patient, if things are relatively stable with them but we have more than enough physical exam "bullets," and enough data points / diagnoses that we're watching, does routine monitoring of the patient's PT/INR  or anticoagulation therapy  qualify for the high risk category of medical decision making?

First of all, I'm not a certified coding expert, but I have studied CPT coding for years.  I recommend all medical professionals obtain their own copy of the AMA's CPT 2018 Standard Edition for review as the definitive authority on CPT® coding. The reader is referring to the billing of a high level risk in the E/M coding process.    I consider anticoagulation therapy with warfarin to be a high risk medication monitoring process that requires intensive monitoring.

I would definitely consider ordering a daily protime or INR (International Normalized Ratio) as a form of intensive monitoring for toxicity.   Intensive is a relative word.  I have never seen it defined anywhere.  Does intensive refer to weekly checks?  Does it refer to monthly checks?  Does it refer to hourly checks?   Surely a daily INR would be considered intensive for a drug that can kill you if not prescribed with caution in hospitalized patients who are sick with  multiple comorbidities.

Everyday I take care of anticoagulation therapy in patients who's INR has accelerated out of control due to interactions with other common medications used in the hospital setting.  Some of the most common medications I see that cause wild and rapid rises in a patient's previously stable INR are  Levaquin, Bactrim, Amiodarone and Fluconazole. I would have no problem trying to justify high risk drug management with coumadin in the hospitalized patient.  If they aren't high risk, I don't know who is. 

I find my self adjusting my patients' coumadin dose up or down every day.     I see hospitalists and subspecialists ordering bolus doses of coumadin every day.    Published research suggests a bolus coumadin dosing does not improve outcomes but does increase the risk of bleeding complications by statistically significant amounts.  

How do I initiate coumadin dosing?    After seven years of seeing the horrors of excess anticoagulation in the hospital, I have adopted a highly conservative approach to initiating anticoagulation.  I base my approach on  sound clinical principles  and  clinical experience.   The standard starting dose I use is 6 mg for everyone under the age of 50.  As far as I know, there is no randomized clinical trial that uses an age based scale for initiating coumadin dosing.   All I have is seven years of experience treating thousands of patients.  Here is my age based scale for coumadin dosing and warfarin dosing:

Less than 50 years old:  6 mg
Less than 65 years old 5 mg
Less than 70 years old 4 mg
Less than 75 years old 3.5 mg
Less than 80 years old 3 mg
Less than 85 years old 2.5 mg
Less than 90 years old 2 mg
Over 90 years old, I'm not starting it without a compelling reason.  

I do not know the data or if there even is any data on prescribing short term or long term anticoagulation therapy in the 90+ year old population.  Before I entered medical school, I worked for a year as a lab technician  at a phase three clinical trial company.  The only people being studied were young, healthy nonsmokers.  That's how these drugs clear their clinical trials.  There were no 90 year old participants getting tested on the pharmacokinetics of warfarin dosing in their age group.   All we have as physicians is clinical experience and post marketing data.  That's as good as it's ever going to get.

Took warfarin.  Didn't bleed!
Finding the right dose  of Coumadin can be difficult, especially with genetic factors involved.  Genetic tests are now available, such as the Amplichip CYP450 at a cost $500, to determine whether patients carry genes that affect metabolism of Coumadin.  I don't find them clinically relevant in the hospital setting. Their turn around time is too long.  By the time I get my patient's genetic coumadin profile, they will already be therapeutic from my clinically determined coumadin dose.   

My experience says Coumadin in a patient over 90 years old is a recipe for disaster. I try to avoid  using it if at all possible, even if the guidelines say I should.  My experience says the guidelines were not made for the 90 year old body with twelve medical problems.   I often recommend  prolonged courses of subcutaneous Lovenox or Arixtra for anticoagulation instead of Coumadin in many post fracture elderly patients who need short term venous thromboembolism prophylaxis.  Warfarin management in these folks is simply too erratic.

After age, I next review the patients medication profile.  Many medications will interact with warfarin dosing through the cytochrome P450 enzyme system.  What effect does the cytochrome P450 enzyme system have on Coumadin  metabolism?  Some common drugs  (quinolones such as Levaquin or levafloxacin, Cipro or ciprofloxacin, amiodarone, Bactrim, Diflucan or fluconazole,  Biaxin or clarithromycin, Flagyl or metronidazole) are inhibitors of the  P450 enzyme system. That means they make the patient require less Coumadin.  In other words, if the doctor fails to reduce the Coumadin dose after starting these common medications, the patient's INR can often rise dangerously high.    

Some common drugs (Tegretol or carbamazepine, Dilantin or phenytoin, phenobarbitol, rifampin and even tobacco) are stimulators of the P450 enzyme system.  That means they make the patient require more warfarin dosing).  If doctors fail the increase the warfarin, the patient's INR may fall below the therapeutic range and increase thrombosis risk for the patient.

Once I have reviewed the patients age and medication profile, I make an approximate 30% adjustment in coumadin dosing (my own estimate)  when drug interactions are obvious.  I know of no published data other than clinical experience that can guide the clinical hospitalist towards dosing adjustment based on drug-drug interactions.

Once I have adjusted for age and drug interactions, I want to know what the patient's protein status is.  Coumadin is highly bound to albumin. A low albumin level will mean the patient requires less Coumadin  for long term anticoagulation therapy.  If the patient's albumin level is really low, less than 2.5 for example,   I will cut my expected Coumadin dose in half until I know how the patient responds to their initial dosing regimen. 

For a healthy 40 year old on no medications I will start warfarin at 6 mg a day with no bolus dosing.  For an 85 year old on amiodarone and an albumin level of 2, I may chose to start warfarin at 1 mg a day until I know how the patient responds.  With all that said,  I reiterate that I consider coumadin dosing in the hospital to be considered a high risk drug management process requiring  intensive monitoring for toxicity.  

The same could be said for heparin. I consider heparin drip management to be high risk on the Marshfield Clinic audit tool used by many Medicare carriers for determining the level of risk on E/M visits.   Are you adjusting insulin dosing with close monitoring of glucose levels?  Are you checking Dilantin levels as well?  How about Vancomycin levels?  These are all drugs the require intensive monitoring for toxicity and I would consider management decisions to be high risk. 


EM Pocket Reference Cards Using Marshfield Clinic Point Audit

Click image for high definition view

Print Friendly and PDF