The Anatomy of Dying

(Some details changed to protect identity.)

At age 68, Mr. A walked into the emergency room. He was a little short of breath but 2L of oxygen by nasal cannula relieved his difficulty breathing. In childhood, he had rheumatic heart disease, and as a result, his aortic and mitral valves had failed. About twenty years ago, he had the valves replaced with mechanical valves, and those had served him for far longer than the artificial valves were expected to last. He was admitted to the hospital for presumed heart failure; perhaps the valves were finally giving out and his heart was having difficulty pumping blood. His other medical conditions included systemic lupus erythematosus (SLE).

The assumption was that as the heart failed to pump blood forward, it was backing up into his lungs, causing pulmonary edema. Our treatment was aggressive diuresis, to cause him to urinate out all the extra fluid. However, day after day, he was not improving; his oxygen requirement slowly went up and he was unable to sleep lying flat, a symptom called orthopnea which often characterizes congestive heart failure. One day, his symptoms and oxygen requirement drastically shot up to 10L by nasal cannula.

Chest X-ray and a CT scan didn't show any good causes for the increasing dyspnea (shortness of breath). On a ventilation-perfusion scan, a defect was identified that showed "high probability of pulmonary embolism." A pulmonary embolism or PE is a clot thrown to the lungs that obstructs blood flow and can cause low oxygen levels. Interestingly though, Mr. A was on an anticoagulant at supratherapeutic levels (INR 3.5-4.0) for his mechanical heart valves. Someone at that level of anticoagulation should not be clotting; in fact, he would be at high risk for bleeds. However, Mr. A had one condition that did not make this impossible; lupus (SLE) is sometimes associated with a prothrombotic state ("lupus anticoagulant," a misnomer) that can break through therapeutic anticoagulation (the other possibility was an occult cancer).

This conundrum confounded us for days. How could he clot on anticoagulation? And given a probable PE despite anticoagulation, how should we manage him? We consulted numerous services: hematology, pulmonary, rheumatology, intensive care, anesthesiology, and the congestive heart failure service.

Meanwhile, he was getting worse and worse. He required high-flow oxygen (15L) or a non-rebreather mask. He could only tolerate sitting straight up at 90 degrees. He spoke in two to three word sentences. He was getting no sleep. He was stressed, tired, and despairing. Originally, he and his wife thought they could be out of the hospital by Thanksgiving. But things were getting worse and worse.

We began heparin in conjunction with warfarin, two powerful anticoagulants with a considerable bleeding risk. But weighing the risks and benefits, we knew that another PE would kill this patient. We sent off fancy lab tests: chromogenic factor X and markers of prothrombotic states in an attempt to fully characterize what was happening.

We began talking about "goals of care," perhaps a euphemism for admitting that medical care was not making this patient better. This was a tough discussion, perhaps the hardest among those I've experienced. Between the lines, we were saying that he was dying and we wanted to know how far to go. He didn't want prolonged intubation (being on a breathing machine) or heroic measures of resuscitation. But he wanted to get better; his son from Texas was visiting for Thanksgiving, and his wife was by his bedside every day, unfailingly.

Given the situation, we decided to act. A trans-thoracic echocardiogram was not giving us valuable information. The patient could not tolerate going to a scanner or a trans-esophageal echocardiogram (a better test at looking at the heart). We kept diuresing and diuresing but we could not wean his oxygen. We needed to look at the lungs. We intubated Mr. A and brought him into the intensive care unit.

A bronchoscopy showed his lungs were full of blood - not water. Instead of severe pulmonary edema from congestive heart failure, he was bleeding profusely into his airways. Our aggressive anticoagulation probably made things worse. But, we said, he had a PE. How was he clotting and bleeding at the same time? We were also losing vascular access. Nurses couldn't get an IV started because he bled with every stick.

We stopped the anticoagulation and began aggressive ICU support. Sedation kept Mr. A comfortable, but we could not wean him off of the ventilator. His renal function began declining (Cr baseline 0.7, now up to 1.6). His blood pressure began to drop, and one day, his liver function enzymes skyrocketed (AST/ALT in the thousands). This was probably shock liver, maybe from ischemia, maybe from a clot. And now several organ systems were shot - cardiovascular, pulmonary, renal, hepatic.

Multi-system organ dysfunction has a very poor prognosis. After numerous discussions with the wife, we decided to withdraw care. After stopping the pressors (cardiac medications to maintain blood pressure), Mr. A passed away in less than half an hour. I learned a whole lot from this case. From a medical standpoint, I realized how difficult it is to operate beyond the realm of data; much of what we were doing had never been studied. I learned how important it is to re-evaluate constantly; we treated a long time for pulmonary edema, and perhaps there was some component of it, but it did not explain the entire picture. From a patient care standpoint, I realized the import of the end of life, how to approach it, how to accept it.