Published in the December 2011 issue of Today’s Hospitalist
WHEN MIKHAIIL KOSIBOROD, MD, and his cardiologist colleagues at St. Luke’s Mid America Heart and Vascular Institute in Kansas City, Mo., began looking a few years ago at anemia in heart attack patients, they found something startling: Nearly half those patients developed anemia during their hospital stay. What could be causing it?
To find out, Dr. Kosiborod and a team of researchers first focused on bleeding episodes but discovered that only a minority of patients had documented bleeding. They then took a closer look at an everyday event “phlebotomy ” and found a strong link between the volume of blood drawn during a hospitalization and patients’ likelihood of developing hospital-acquired anemia. The results of their study were first published online in August 2011 by the Archives of Internal Medicine.
Dr. Kosiborod and his colleagues culled de-identified data for nearly 17,700 patients in 57 U.S. hospitals who were admitted for acute MI. Twenty percent of those patients were found to develop moderate to severe hospital-acquired anemia.
Those data tracked the date and time of every phlebotomy event and the results of corresponding lab tests. But the data didn’t specify the volume of blood taken at each draw. As a result, Dr. Kosiborod and his team decided to make several conservative assumptions.
For one, they assumed that all tests that could be performed from a single blood tube came from only one tube. They also assumed that clinicians used only the minimum amount of blood needed for each test, and that no blood was wasted during blood draws.
But even with conservative estimates, researchers found dramatic differences between mean blood-draw volumes for patients with hospital-acquired anemia and those without: 139.3 mL for patients who developed anemia in the hospital vs. 52 mL for patients who didn’t develop the condition. Overall, the data showed that for every 50 mL of blood drawn, a patient’s risk of developing moderate to severe hospital-acquired anemia jumped 18%.
Dr. Kosiborod talked to Today’s Hospitalist about those results and about ways to reduce phlebotomy volumes.
What prompted you to look at the association between phlebotomy volume and hospital-acquired anemia?
We were seeing a lot of patients coming in with normal hemoglobin, but then the hemoglobin would drop during hospitalization without a clear explanation, such as bleeding. We wondered if this was merely an anecdotal observation or a common problem. We found it to be common, and we discovered the relationship between developing anemia in the hospital and poor patient outcomes. This made us take a closer look at potential risk factors such as phlebotomy.
Have other studies targeted phlebotomy as a possible risk factor?
There were some small studies in other patient populations, but the issue really has not been highlighted in the literature. We focused on phlebotomy because it made sense clinically. What could be going on in a hospital that may potentially increase patients’ risk for developing anemia, if it’s not bleeding?
Of course, it’s a multifactorial issue. Some risk factors for developing anemia in the hospital are patient-specific. Patients are older and have more comorbidities including inflammatory disorders, renal disease and diabetes.
Also, some hospital-based processes of care may contribute, and phlebotomy is an obvious suspect.
What are some consequences for heart patients who develop anemia?
They feel worse, they have worse health status after they leave the hospital, and they have higher mortality than patients who do not develop anemia. Whether these adverse outcomes are caused by anemia remains to be established, however.
What did you find in terms of the risk of hospital-acquired anemia and length of stay?
The longer you stay in the hospital, the more likely you are to develop anemia. There can be all kinds of reasons for that, including the fact that patients who stay in the hospital longer tend to be sicker “and sicker patients are more vulnerable to anemia.
But the longer the length of stay, the more phlebotomy events patients will have, and overall phlebotomy volume will increase.
The study showed big differences across hospitals in the amount of blood drawn. Why is that?
It’s probably a combination of differences in patients and in processes of care that hospitals use for diagnostic blood work.
Some hospitals may be seeing patients who are older, have more comorbidities and are more acutely ill, and therefore stay in the hospital longer. All those can affect phlebotomy volumes.
On the other hand, there probably are significant variations across hospitals in terms of who draws the blood, how they draw it, and whether there are safeguards in place to see that all blood work is necessary and not duplicative. Without staying by the patient’s bedside, it’s very hard to say what is necessary and what’s not in terms of blood draws. But anecdotally, if you ask most physicians, they will say that probably there is some blood work that is not absolutely necessary. Hospitals also vary on whether they use large or small tubes on most patients.
How can you reduce phlebotomy volumes?
One of the easiest things is simply using smaller tubes. They’re widely available, and they don’t cost more than larger ones. For the majority of the routine tests we do on a daily basis, smaller tubes are just fine.
You can also reduce the number of blood draws. A significant number of blood draws are necessary “but as patients get better, we often don’t need routine daily draws. We can be more selective: If patients’ blood work has been stable for a couple of days and they’re going home in the next day or two, we probably don’t need to continue to draw blood every day.
What’s happening at your hospital?
We already use smaller tubes for many adult patients, but we’re working to make that more systematic. We’re putting together an intervention in cardiology patients to try to reduce phlebotomy volumes by using smaller tubes and implementing a protocol to reduce the number of blood draws.
The next logical step is to determine whether this intervention will minimize drops in hemoglobin and reduce the number of patients developing anemia in the hospital. We also want to see if it will have an impact on the use of blood products.
If these interventions reduce anemia, the next step would be to determine whether preventing anemia actually has an effect on patient outcomes. And we need to do similar investigations in other patient populations. It would be very interesting to try to replicate these findings in patients hospitalized in medical and surgical units for a variety of conditions.
Laura Putre is a freelance health care writer based in Lakewood, Ohio.