I Know it When I See it: Antibiotic Resistance, the Chronically Dying, and Futile Medical Care

Today's issue of the New York Times features (yet another) article sounding alarms about antibiotic resistance and the supposed need for new antimicrobial agents.  There are several common hackneyed reasons why we have the current predicament of drug resistant organisms such as:

• Use of antibiotics in livestock, which accounts for 80% of use in the United States.  This should be the prime target for efforts to reduce resistance, not use in humans.  
• Inappropriate use of antibiotics in outpatients who have mild/minor illness and who have viral illness (not responsive to antibiotics) 75% of the time.
• Inappropriate drug, dose, duration of therapy, etc., in both outpatients and inpatients.

The article calls for reducing the hurdles needed to develop and test new agents.  Well and good.  But I have another idea:  fund testing of antibiotic dosing duration for several common infections such as pneumonia, cellulitis, urinary tract infection, etc.  Right now, it's all over the board, with some patients getting two weeks for pneumonia, some getting five days (my patients).  Moreover it is my suggestion that such trials stratify patients on the basis of the prior probability of the infection being treated.  I would posit that the lower the probability of infection, the more likely it is that a shorter course will suffice.  Clinical improvement criteria could also be built into these studies so that courses could be truncated when patients improve.  If we could cut the duration of exposure from 10 days to 5 days, we would remove selective pressure for development of resistant organisms and opportunistic infections (fungemia; C. diff; ESBL).  I'm not sure we need new toys so much as we need to play properly with the toys we already have.  (And no, I'm not a fan of biomarkers like procalcitonin to guide duration of administration.)

Another part of the problem which has received inadequate (that is, not any) attention is a third rail, an elephant in the room:  the use of antibiotics in low value or futile care, in the chronically dying.  Nobody wants to talk about this, but we need to open the discussion.  When a patient with brain injury and a tracheostomy or advanced age and a non-healing wound and malnutrition......I could go on for pages.......when such a patient is institutionalized and supported by artificial ventilation and nutrition, they are chronically dying.  (Death used to be a mostly acute event, for almost all of human history.  Only recently have we developed the power to turn death into a chronic event.)  And in this state, with the body's normal defenses against infection that are usually bolstered by mobility, walking, coughing, spitting, swallowing, urinating, defecating, maintenance of hygiene, with these defenses impaired by the underlying disease and it's crippling of normal function, the chronically dying patient becomes a breeding ground for resistant organisms.  The NYT article noted that most of these infections develop in the hospital (although recent community outbreaks of PVL Staph aureus and Clostridium difficile are noted).  It is no wonder.  Antibiotic use is intense there, but the hospital is also a frequent place of visit for the chronically dying.

This problem of the chronically ill (where is the line between chronically ill and chronically dying?  That's for you to decide.) receiving repeated courses of antibiotics because of impaired host defenses (mostly mucosal and reflex and barrier defenses, not immunological ones) is just part of the problem that we have created in critical care and indeed all of modern medical care by not being realistic about prognosis.  As I have stated elsewhere, I am convinced that much of the futile care we provide, exceptional cases notwithstanding, is the result of a failure of physicians to provide realistic expectations for patients and their families about what can be achieved with intensive care (especially that which becomes chronic, requiring an artificial feeding tube and a tracheostomy).

A case in point:  I was recently called by the ER about a patient with a cardiac arrest.  The patient had recently diagnosed Stage IIIa non-small cell lung cancer.  I need to hear no more to know what the prognosis is going to be:  very poor.  The median survival of patients with a new diagnosis of stage IIIa lung cancer is just over a year.  One with an out of hospital cardiac arrest who has had 40+ minutes of CPR is probably not going to survive, and even if s/he did survive, they will likely succumb to the underlying disease within a year.  The medical community has struggled to come up with a "definition" of futile care, much as the Supreme Court has struggled with coming up with a definition of pornography.  So I will quote Supreme Court Justice Potter Stewart:  "I know it when I see it."  In any case, the patient died before I arrived at the hospital.  I  nonetheless went to speak with the family to see if they had any questions.  I asked if the oncologist had discussed the median survival of stage IIIa lung cancer, and, it was no surprise to me, s/he had not.  This patient had initiated chemotherapy and radiation therapy for a disease that was likely to kill her within one year, and had not been informed of the prognosis.  I can think of few acts more cruel than to withhold this kind of information from a patient.  This person may have wanted to travel the world or achieve any number of other personal goals before death, and was not afforded the opportunity to make plans to that end, but rather was signed up for chemo and radiation.  Which is not to say that chemo and radiation were not indicated, warranted, or wanted.  But it is clear that an informed choice could not have been made without knowing the prognosis.  (And I don't want to go on another tangent, but that's why this whole thing about autonomy is a charade, the patients don't usually have the information that would allow them to act as autonomous decision makers.)

Sometimes it is not the entire patient that is chronically dying, but just a limb.  How many courses of antibiotics have been given to diabetic patients to "treat" osteomyelitis in a diabetic foot that they can't walk on nor ever will walk on because of other debilities and co-morbidities such as obesity, weakness, poor functional status, deconditioning, heart failure, arthritis?  Why would we admit the patient 5 times over the course of 2 years only to eventually amputate anyway?  To "save the foot"?  Save it for what, the next infection so we can save it again?  If the foot cannot be saved, maybe it would be better to amputate and install the prosthesis and get on with rehabilitation.  If rehabilitation is not possible, at least the patient is not having repeated admissions for antibiotics and complications and the development of ever more resistant organisms.  In many of these cases, it's not whether amputation will take place, but when.

My point here is that, while more antibiotic choices might be nice, we are bewitched by a problem of our own making, and the problem won't go away with more antibiotics.  In addition to political failure to reign in antibiotics for livestock, moral failure of primary care physicians to just to say no to people demanding antibiotics, and lack of courage on the part of inpatient physicians to limit the duration of antibiotic courses,  we have the looming and pervasive failure to discuss prognosis which leads to a large population of chronically dying patients who inevitably develop resistant organisms (which can then spread to other patients) because of repeated exposure to broad spectrum antibiotics.  And we need to start discussing the moral implications of creating an ecological hazard for all of us in our treatment of the chronically dying.