Wednesday, January 4, 2012

When the elderly need a protection order against warfarin-wielding physicians

This is a tree.  If you want to see the forest, back up.
(Added 1/11/12 - This post was prescient:  See this article in today's NYT: Interactive Tools to Assess the Likelihood of Death which reviews this JAMA article: Prognostic Indices for Older Adults; here is the Prognosis Website - In my opinion, it's too complicated to be useful, but it's a start.)

Ever since I was in medical school, I have been hearing the tired old refrain about how warfarin for stroke prevention in atrial fibrillation is underutilized in just about everybody, especially the elderly.  Well, if you want to see uber-high rates of warfarin underutilization, you need look no further than trials of stroke prevention such as the original SPAF trial (Ezekowitz, NEJM, 1992), where more than 93% of screened patients were excluded from the trial and not given warfarin!  There's an interesting contradiction.  Moreover, it is likely that warfarin is not as underutilized as has been stated because of inadequate data on appropriate exclusion criteria (see Srivastava, Thrombosis Journal, 2008).

So I was very interested in the Clinical Crossroads published online November 1, 2011 in JAMA (discussed in the last blog post about H. pylori) because of the apparent enthusiasm for restarting both aspirin and warfarin in an elderly woman who had just had a life-threatening hemorrhage associated with these drugs.  (That enthusiasm persisted in the follow-up discussion of the case in the December 7, 2011 issue of JAMA.)  However, I was not surprised.  I constantly see elderly people admitted to the hospital after complications of warfarin therapy or for other conditions which should cause warfarin use to be re-evaluated, such as failure to thrive, inability to walk, new onset gait instability and fall risk, advanced dementia, poor functional status with residence in a care center (I think this is the latest euphemism for nursing home), polypharmacy with drug-drug interactions, supratherapeutic INR, etc.  Many of these poor folks have precious little remaining time to live, and thus have precious little time to "benefit" from warfarin, but plenty of time to get immediate complications from it.  Many of them might be better off, and would certainly prefer, to shake off the throng of doctors stalking them with warfarin prescriptions and appointments to the "coumadin clinic".  (I am reminded of a recent observation that I made about a frail octogenerian (still driving a motor vehicle) with untreated sleep apnea (he refused CPAP) who had been started on warfarin during an admission for CHF and AF:  he would be more likely to die in an MVA driving himself to the coumadin clinic than to have a stroke prevented by warfarin.)

Alas, my reader response to the Clinical Crossroads, which highlights these concerns, was just published on the JAMA website.  There are a couple of general principles that I reason apply not only to warfarin, but also to the use of many other drugs, therapies, and procedures.  In some cases, the medical community has awakened to these realities (such as limiting screening for colon and breast cancer in the elderly) whereas in others, status iatrogenicus is the norm (see: What if they had had to pay? on the Not Running a Hospital Blog).
  1. With advancing age or terminal illness, the opportunity to benefit from preventative therapy declines.  Perhaps simvastatin and aspirin as primary prevention in an 86 year-old nursing home resident should be reconsidered, especially if the patient dislikes taking medicines or has conplications from them.  Likewise, a 72 year-old man with advanced lung cancer hardly needs an intensive and complicated insulin regimen for his diabetes.  And the baseline unresponsive patient with advanced dementia I admitted from a care center last week CERTAINLY does not need Aricept (donepezil).  Really?  What insentient robot doctor wrote that order?
  2. With changing general health or acute medical conditions, the patient's medical regimen should be reviewed and revised in light of the new information.  Here is where we need to recognize that circumstances have changed, and if we fail to do this, we can never get to number 1 above.  I think an obvious opportunity for this is each and every time a patient is admitted.  In some cases, it seems obvious:  the second admission for hyperkalemia should lead to reconsideration of Aldactone (spironolactone) and digoxin in a patient with congestive heart failure and renal insufficiency.  In other cases, the changes may be more subtle - the insidious onset of frailty and unsteady gait that should lead to inquiry about falls and the home environment may only be noted if the patient is observed during physical therapy.
  3. The patient's preferences need to be incorporated into weighty decisions such as warfarin therapy in advanced age, and their preferences can only be known if they are given the data.  A 2% or even a 5% annual reduction in the risk of stroke may be huge for a 55 year-old man with paroxysmal atrial fibrillation, but it may be negligible for an 88 year-old.  Preferences, like expected risks and benefits, can vary with age and change over time.
I wager that, if these three principles alone were followed, 10-20% of the ICU admissions I do could be avoided, medication lists could be reduced and simplified by 50%, and patients would be happier and maybe even healthier.


  1. People and practitioners are catching onto this concept. Here's an atricle in the January 26, 2012 NEJM about deactivating and not replacing AICD defibrillators based on changes in age, values, and preferences:

  2. Here's another relevant article on bone density screening intervals, NEJM Jan 19th 2012:

  3. Those are some well thought ideeas. I'd love to see them applied. What is your opinion on Pradaxa as a substitute for warfarin? From what I heard it can lead to some serious injuries such as internal bleeding

  4. Pradaxa is an evil conspiracy to make money for Big Pharma.