Thursday, July 11, 2013

Oligopharmacy and Nopharmacy: "He is the best physician who knows the worthlessness of the most medicines."

Refrigerator Magnets of Heavily Promoted Drugs
When I was an intern, I read in a pocketbook (my coat pockets were overflowing with guides and manuals in those halcyon days) called Drug Prescribing in Renal Failure that in elderly patients, 65% of hospital admissions are caused by or contributed to by an adverse drug reaction.  At first blush, this seems like a gross overestimation (like the 100,000 deaths per year from medical errors which probably IS a gross overestimation), but experience bears out this wisdom.  As the number of drugs increases, so does the potential for interactions and side effects - multiplicatively.  There are two related practical implications of this observation.

First, it is my practice when admitting a patient to the ICU or hospital to stop all nonessential medications.  This includes outpatient medications and many medications begun by other physicians during the admission.  It would include medications such as:
  • All psychotropic medications, antidepressants,  narcotics, muscle relaxants, benzodiazepines (I've gotten away from those drugs entirely in the ICU except in alcohol withdrawal).  The caveat here is some medications have associated withdrawal syndromes, such as gabapentin/Neurontin.  Remarkably, I have yet to see opioid withdrawal, in spite of having large numbers of patients on high doses.
  • All medications used for chronic disease management and prevention such as statins (and the loathesome Vytorin), oral hypoglycemics, anti-hypertensives.  I usually continue aspirin and plavix.  Mostly I stop Coumadin but it's a judgment call.
  • All nonsensical PRN "bowel elixirs" such as magnesium, aluminum, Mylanta, this, that and the other thing.  These things just clutter the medication list and make it harder for me to focus on the essentials.  In patients whose disease is likely to include nausea, I will leave Zofran or phenergan or both.
  • Tylenol.  I hate this drug for fever.  If patients are having fever I want to know about it, not have it masked.  But, if you have a headache you're not getting dilaudid, you're getting two extra strength tylenol.  So don't ask.
  • Often I find myself stopping Albuterol and Atrovent.  It's a great irony - as the pulmonologist on duty I have a net negative effect on the use of these drugs in the hospitals I staff.  (Another irony is that I cancel more ABGs and CTPAs and PRBCs than I order.)  Most patients aren't wheezing, just because you're on the vent doesn't mean you need bronchodilators, Albuterol may be harmful in ALI and ARDS, the drugs have no disease modifying ability in COPD and are for symptom relief alone (asthma is different), and Respiratory Therapy loves me for stopping them and/or making them PRN.  I also find myself drastically reducing steroid doses in COPD patients - like from 125 mg solumedrol several times a day to 60 mg once a day, usually as oral prednisone.
The reasons to do this are sound, even if you don't believe the wisdom imparted by my intern pocket manual.  First, in acute illness with dehydration, hypoperfusion, and perturbations in renal function it is reasonable to assume that medications have accumulated to supratherapeutic levels.  (We often see direct evidence of this with Coumadin, if patients are taking it.). For this reason, medications that affect renal function are high on the list of medications to stop, especially since other nephrotoxic things are frequently administered in the hospital, like the pernicious CT scan with IV contrast.  Second, the logic of multiplicative and difficult to discern interactions is inescapable.  Third, I believe there is a serious tendency to miss things when the medication list swells.  It's hard to focus on what the patient is actually being treated for and what they need when you're staring at a list of 20+ meds.  And are you really looking to see what PRN medications are actually being given and at what interval?  My guess is that most physicians are not paying that close of attention.  So PRNs just confuse the entire picture.  Fourth, a swollen medication list creates the impression of a dependent sick person with complex physiology that needs manipulated.  When the patient gets better on few meds there can be no other conclusion but that the primary process has abated - it is not that the patient is being propped up by a complex potpourri of pills and drips.

An interesting twofold observation can be made after this practice is instituted: the first is that patients do remarkably well with a nominal number of inpatient medications.  (They probably do better with fewer medications but I won't stake so audacious a claim - yet.) The second and more astonishing observation is that the first observation is missed by nearly everyone.  (This applies to drawing frequent labs and replacing electrolytes too.)  I think it should be startling to patients, nurses, and other practitioners that I can admit a patient to the ICU, decimate their medication list, and they get better nonetheless.

Sometimes, we have the dubious fortune of a natural crossover experiment.  A patient is admitted to the hospital and deteriorates.  They come to the ICU on 15-20+ medications.  I stop half or two thirds of the medications, and they improve.  (Or they don't.)  Two possibilities explain the patient's initial deterioration:  1.)  The patient deteriorated in spite of the medications which were causally irrelevant to the course of the illness or inadequate to reverse it; 2.)  The patient deteriorated because of the many medications.  Two similar possibilities explain the patient's subsequent improvement (or death) in the ICU with oligopharmacy:  1a.)  The patient improved (or died)  in spite of the restrictive use of medications which was causally irrelevant to the course of the illness or was not a sine qua non for its reversal;  2a.)  The patient improved (or died) because of restrictive use of medications.  We have a problem of logical inference, and we can make some guesses about these possibilities.  It hinges on whether most patients improve or die after ICU treatment.  If most die, it's a confusing matrix.  But if most live (as is the case in my ICU where mortality is at or below the national average of 15%), it's more straightforward.  In the case of #1, medications are irrelevant, and it is proved when the patient comes to the ICU and improves on oligopharmacy.  In the case of #2, the patient was deteriorating from polypharmacy, and was saved in the ICU when the medications were reduced to oligopharmacy.  So, if the sequence is floor to ICU, polypharmacy to oligopharmacy, and most patients survive, the logic matrix is self-evident:  oligopharmacy is either beneficial, or has no effect.  In the latter case oligopharmacy still wins because of cost, complexity, and complications.  The 3 Cs, we can call them.

This segues to the next practical implication.

If the patient has recovered and is doing quite well off all his outpatient medications, what does that say about the utility of some/most of those medications?  Shouldn't the necessity of all those be reconsidered now?  (The exception being the medications for chronic disease states with good evidence of long term efficacy such as statins, anti-hypertensives, etc.). This is the perfect opportunity to reevaluate if Seroquel is a good idea (outside of schizophrenia which is ironically a minority subset of patients who are taking it).  It's probably not.  Vytorin?  No way.  Avandia or Actos?  Probably not.  Opioids for chronic pain?  Doubt it.  Coumadin?  Not if a functional status assessment has revealed a previously undetected fall risk.  Aliskiren?  Absolutely not, are you kidding me?  Multiple expensive inhalers for COPD?  You be the judge.  Branded Plavix?  Who knows.  Nexium?  No way, Prilosec is generic!

Yet another irony is that we openly express our amazement when we admit a patient in his 80s who's on no meds.  We interpret this as an unusual finding, one that speaks of the good fortune of his good health (assuming no health problems have been neglected.)   (The diametric opposite is when the meds are brought by a family member in filled grocery sacks.   This causes bemusement rather than amusement.). Well, if getting into old age with oligopharmacy or nopharmacy is a good thing,  perhaps we should not be so quick to order and prescribe so many meds.  But there are powerful pressures to prescribe them.

One is patient expectations.  It is not infrequent that I am hounded by a patient or a family member to restart some or all of the chronic meds, especially the psychotropics and pain meds and sleepers.  Another is the perception of providers that we both can and ought to treat every symptom with medications along with the underlying assumption that doing so is beneficial in the net.  (In internal medicine, this flows from the culture of "thoroughness" that is part of our indoctrination and from which we derive smug pride especially when comparing ourselves to other specialties.). Another is powerful Big Pharma influences - drug reps and the like.  Another is nurses' training to actively elicit symptoms and treat all of them as a gesture of kindness, compassion, and patient advocacy.  Another is care fragmentation that results from lack of records in relation to the reasons the medications were started.  Another: reluctance to second guess another professional's judgment.

In spite of these pressures, I think it worthwhile to consider that Ben Franklin may have been right and that the best physician knows the worthlessness of the most medicines.  If he was, we would do well to begin to whittle away at or just plain ax medications of dubious benefit - both acutely and chronically.

1 comment:

  1. Hello,
    Checking searches for oligopharmacy, which I have also written about, I found your blog. I appreciate your post, and will put you on my blogroll.