Long Tail Diagnosis: Everyone Has It, You Just Have To Look For It.

There are times in medicine that can make you feel good. That makes you appreciate why you do what you do. Like any other field, when you become really good at what you do, it can become monotonous. Every day can run together. The people are all different. Personalities run the gamut from pleasantly demented to antisocial drug addicts. But their diseases are often treated in rigourous guideline like fashion. The people are different. Diseases are diseases.

The guidelines for treating pyelonephritis are the same in the vast majority of patients. The guidelines to treat community acquired pneumonia are the same in the vast majority of patients. The guidelines for stroke treatment are the same in the vast majority of patients through fairly rigid protocol driven guidelines. However, it is the nuances of treatment that make doctors doctors. The deviation from protocols that aren't appreciated by all of the rigid government performance standards being presented to the public as quality care. That is the value in physician level training.

I found the discussion on the long tail of medicine (here and here) at DB's Medical Rants fascinating. Essentially, he is saying the 15% of long tail medicine is what makes doctors doctors. He says that 85% of what we do can be robotic. I would argue, to some degree, that is true. BUT, I also believe that every patient is a long tail. That every patient is different, even though their diseases are the same, even the common ones. And that's why protocol driven medicine is so difficult across most of the medical treatment and evaluation sphere.

It's called medical judgement. An art that we learn to practice through intense medical school and residency trained education. It is an art that is nearly impossible to be emulated by extenders, nurses and other health care personal. That's not meant to be insulting, demeaning condescending or arrogant. It's just a fact. That's what medical school and residency does to you. It is very difficult to explain unless you have experienced it.

Tonight, I had a long tail presentation of a common medical condition: deep venous thrombosis (DVT).

The standard of care for DVT is initiation of warfarin (vitamin K antagonist) plus initiation with a few days (about 4-5 days) of low molecular weight heparin (lovenox) subcutaneous 1 mg/kg twice daily or 1.5 mg/kg daily or until the INR is greater than 2. Now this would be the robotic response to treatment. That 85% that robots follow. But that's not what happened tonight. Let me give you this scenario.

35 year old Mr. Peel presented to the emergency room 12 hours ago with a few day history of a mildly swollen right leg. Otherwise asymptomatic. His mother has a history of a blood clot and she was worried. Mr. Peel has no physician and no insurance. And no money. He's dirt poor. In the ED a venous doppler confirmed the presence of a right lower extremity distal DVT (dorsalis pedis posterior tibial vein) with sub acute features. In other words days to weeks old and sitting right at the ankle bone. The patient's d-dimer (a marker of active clot turn over) was barely above normal range (indicating the clot is not very active and therefor less likely to extend).

The ER gave Mr Peel a dose of lovenox and recommended hospital admission to the hospital. In my experience, distal DVT's can be managed as an outpatient. The patient refused admission, was given a script for lovenox and coumadin dosing and set up with a primary care physician at our federally subsidized clinic (many many days from now).

When the patient went to fill his lovenox, he was shocked to learn it would cost $700. Of course, several hours later, he made the decision to return to the ER to be admitted. He tells me he was scared out of his wits for the blood clot that was presented to him as a possible death calling. At this point, I was called to admit the patient. This was my first contact with the patient.

In my residency years, a time of incredible knowledge base solidification, I remember reading an article about the risks of pulmonary embolism from DVT. In essence, the farther up the leg the clot is, the higher the risk of pulmonary embolism. I seem to remember iliac clots carried a 1 in 4 or higher chance of clinically significant pulmonary embolism. The risk of a clot below the knee was on the order of 1% or less. The exact numbers are not important, only the general risk involved. That's why I have a much lower threshold to admit a proximal DVT, at least for 24 hours of observation until anticoagulants can be administered to stabilize the clot. But to admit a distal (below the knee) DVT is almost always unnecessary.

Where does the long tail come in to play in my patient with a common presentation? It's not just clot: present or absent. The site of the clot, proximal or distal, has a lot to do with the risk assessment of my patient. How much of a risk do they represent for pulmonary embolism? And does that risk out weight the risk of bleeding from anticoagulation? We also have the social situation. No money. No lovenox.

When I walked into the room I explained in my first sentence that I wanted to find a way to avoid hospitalization. I explained his $700 prescription for lovenox would pale in comparison to a 5 day hospital stay for about $8,000-$10,000. He categorically agreed with me. I talked with our social worker. I explained my need for setting up Mr Peel for drug assistance outpatient. But this would not happen at 9pm at night. That he would have to call tomorrow (personal responsibility) to get it worked out.

I had a very long talk about risk benefit. I researched guidelines for treatment of distal DVT. There really are none. There are no randomized trials that state one way or another whether anticoagulation is even necessary. Uptodate recommended 6-12 weeks of anticoagulation in the absence of hard data. And if anticoagulation was not possible for any reason, they recommended a follow up Doppler ultrasound in about 2 weeks to verify lack of proximal extension of the clot.

I discussed this data with the patient. I stated, while I felt the risk of proximal extension and symptomatic pulmonary embolism was very low, the data can not support no anticoagulation. I gave him the option of coumadin alone with out a lovenox bridge. I explained the risk of protein C and S deficiency as about 1 in 10,000 folks in which coumadin could actually increase the risk of clot in the initiation phase without using the lovenox bridging method for 5 days. I explained that I was willing to treat him with coumadin alone due to his finances, if he was willing to accept the small risk of hypercoagulable state without lovenox. He gladly accepted that risk.

So now I have walked the patient through my thoughts on risk benefit.

1. The distal nature of the clot represents a very low risk of proximal extension to symptomatic DVT or pulmonary embolism
2. The 1/10,000 risk of hypercoagulable state by starting coumadin without lovenox.
3. The d-dimer near normal range and subacute nature of the clot means it is not likely unstable

At this point, the plan seemed reasonable. Risk benefit analysis in the setting of his lack of money and no insurance determined that we would use coumadin with not lovenox. He was willing to accept the risks, as was I. This was his longtail presentation of a commonly diagnosed condition. But, hold on. It gets better.

He drops a bombshell. Mr Peel has been treated for "leukemia" for 10 years. I did notice a white count of about 15. I ask him if it's CLL. He says yes. (I'm no oncologist, but this whole story sounds strange). He tells me he has rectal bleeding for 3 days, every month, as if he's having a period. I ask him about colonoscopies and hemorrhoids. The answer was yes to the first, no to the second. My hunch is AV malformations, but those words didn't ring a bell with him. Perhaps male menstrual AVMs. Reportable perhaps?

At this point, all bets are off. I explained I would never anticoagulate him in the setting of recurrent GI bleeds. I do not see a strong correlation between his "leukemia" and his new onset DVT. If in fact he has had leukemia, and 10 years later develops a dorsalis pedis local clot, that, in my medical judgment, does not represent a strong clinical correlation. 10 years folks. If his "leukemia" was hypercoagulable he would have presented years ago with a DVT.

Now, if his DVT was proximal, he would be admitted for an inferior vena cava filter to prevent proximal extension and pulmonary embolism. I would still not anticoagulate him. In my opinion, his recurrent GI bleeds represent a far greater risk than a DVT with an IVC filter to offer some protection against pulmonary embolism.

At this point, the patient, in my history taking made the decision for me. I would not offer him lovenox. I would not offer him coumadin. We would go with straight aspirin, which as far as I can tell, has no data either. But at least I feel better about it. The risk of proximal extension is very low. There are no clinical guidelines for distal DVT treatment. The risk of bleeding far outweighs the benefit of a low complication DVT.

I did not recommend follow up ultrasound. I recommended clinical change. I told him if his leg gets more swollen; if he got pain or redness to call his comprehensive care doc for a possible repeat ultrasound. But simply ordering it for 2 weeks in a guy with no money seems silly, in a low risk clot.

We went through my thoughts, the risks and the benefits. And we came to an agreement of acceptable risk. It turns out to be the most financially palatable for the patient. And he was one happy camper.

1. Baby aspirin indefinitely
2. Call comprehensive care doc if worsening symptoms

Instead of admitting him for 5 days of lovenox because he couldn't afford it, I played doctor tonight and used my clinical expertise and knowledge base to move outside that robotic box. Every patient is a long tail. You just have to look for it.

This took a very comprehensive discussion. And time. Lot's of it. Something the ER docs don't have. Nor do I expect them to. These are the types of talks necessary to foster a decrease in resource utilization in every aspect of care. In the current payment system doctors operate in, volume will trump this necessary process every single time. You can't financially survive spending one hour (as I did) in these talks and expect to make a living. Our government has sabotaged the cost effective practice of comprehensive care. Paying a doctor to answer more phone calls or fill out more FMLA papers won't save money. Medical Homes won't pay you to see fewer patients. They pay you to do more paper work. Time with patients will save money. And until we incentivize payment for spending more time with patients all the quality programs in the world will be gamed and manipulated to the benefit of the players. I saved over $10,000 for Mr Peel today. I saved him months of coumadin, PT INR testing. I saved our hospital a $10,000 no pay subsidy. Every one wins tonight. Now. Multiply this time and thought process by hundreds of times a year for just me. One hospitalist.

And people question the value of hospitalists. Huh. I can't wait for the gainsharing to begin. I'm going to be a gazillionare.