Blood Transfusion Decisions In The Hospital: The Wild Wild West Of Hospitalist Medicine.

When should I transfuse my patient in the hospital? Is there a magic transfusion hemoglobin or hematocrit number? And if so, does that number have evidence to support patient benefit or does that number just make use feel better as doctors? As a hospitalist, I watch transfusion decisions being made everyday in the hospital and have come to the conclusion that these important clinical decisions are being made based on the random chance of which doctor you get and what kind of mood they are in on any given day.

When I was in training ten years ago, a nice article detailing risks and benefits of restrictive vs liberal transfusion parameters in the ICU had just come out. What they found changed ICU medicine forever, except for the folks who haven't read the article. Lower mortality and fewer complications of transfusion were experienced when euvolemic critical care patients who were allowed to maintain their hemoglobin counts in the 7-9 grams per deciliter instead of 10-12.

Yet, to this day, doctors got worried and pull the trigger day after day, year after year, decade after decade, transfusing blood to patients who aren't acutely bleeding, who aren't complaining of fatigue and who have perfectly good bone marrow that can make their own blood if we would simply stop taking so much blood from them and allow their bodies an opportunity to do what they were designed to do.

Unfortunately, so much is stacked against doctors and patients when it comes to conservative transfusion practice. Our brains, like government, are often wired to do something. Often, the best thing to do is nothing. Unfortunately, that takes practice and confidence.

Every hospital is mandated by accrediting agencies to define critical lab values in their institution. These decisions must ultimately pass through a whole bunch of really smart people in lots of committees before they become a part of the policies and procedures of any hospital. For years, I practiced under a policy of defining a critical hemoglobin as any value 8 grams per deciliter or less. That meant every lab tech was required to call every nurse who was then required to call every doctor with a hemoglobin less than or equal to 8 g/dl.

By default, at many institutions, hospitalists are being called with critical hemoglobins ordered by other physicians because, in the eyes of many nurses, we are doctors of convenience. I will not allow this practice on my service or on any patients I am covering. The ordering physician should always be responsible for managing the lab they have ordered. This is especially true in the days following the post operative period where surgeons have an obligation to manage acute anemia which may be a complication of their surgery or procedure. Hospitalists have an obligation, in the interest of patient safety, to defer clinical decisions in these situations to the operating team.

What do you think is going to happen if a doctor gets called a "critical" lab value of 7.7 g/dl in a completely stable patient without any complaints when a hospital defines their cut off for criticals as 8 g/dl or below? The patient is most likely going to get blood. Why wouldn't they? The hospital has defined that value as critical. Can you imagine defending yourself in a lawsuit if there was a ridiculously small chance that a patient had a bad outcome because their doctor failed to respond with forceful transfusion after a critical hgb was called.

I know, I know, it all sounds so silly. And it is. In the interest of patient safety, all of these lab draws must have defined critical values. That means, if a hospital chooses to use 9 or 8 or 7 as their cut off for critical hemoglobin, the doctors are, I suspect, more likely to use that number as the threshold for transfusion, regardless of what the patient is experiencing clinically. In fact, as an unintended consequences of defining cut offs for critical lab values, I suspect we are killing our patients. But nobody cares, because we are, at least, on paper, in compliance with the rules.

Transfusion needs should always be defined in the patient's clinical situation. Over the years I've admitted patients with hemoglobins in the 2-3 range. They had stable vital signs and only complained of a little "fatigue". I've admitted patients with hemoglobins of 16 in hypovolemic shock because they were losing blood so fast from ruptured varices that we couldn't save them. Unfortunately, it is impossible to develop a hospital policy for defining critical hemoglobins to account for thousands of clinical settings. So, instead, we define a number and over time hospital culture will usually end up treating a number instead of the patient.

I suspect most transfusions will occur for reasons other than the patient will actually benefit from them. In some instances, we have nursing homes who will simply refuse to accept post operative joint replacement or hip fracture patients who's hemoglobin is not above 8. It's sad. I suspect, it's because their doctors have failed to understand the rationale against aggressive blood transfusions. Or perhaps they had one bad outcome years ago for which they lost a lawsuit because someone had a complication that was blamed on anemia. There is no excuse. If I was a patient, I would be upset at the variability in transfusion that occurs at random in hospital settings.

How many patients do we put into heart failure or cause transfusion related lung injury or any number of other of acute transfusion reactions that get forgotten in favor of treating a number? How much do unnecessary transfusions cost The Medicare National Bank every year? WIll insurance stop paying for blood transfusions without preauthorization someday?

Transfusion decisions are the wild wild west of hospital medicine. Ask 100 doctors how they decide to transfuse and you'll get 100 answers. I see differences even within my own hospitalist practice. Add in the noise of hospital compliance with critical lab determinations and we've created a situation where rarely do decisions get made based on what's best for the patient but rather more likely to make us more comfortable.

There has to be a better way.