Correspondence

Excessive Blood Drawing for Laboratory Tests

N Engl J Med 1999; 340:1690May 27, 1999

Article

To the Editor:

I was recently hospitalized in a major university hospital for the Guillain–Barré syndrome. While there, I had blood drawn, usually twice a day. During my two weeks in the intensive care unit, my hematocrit dropped from 43 to 31.

As chair of the Department of Laboratory Medicine at a well-known children's teaching hospital, I was shocked at the amount of blood drawn for my tests. I asked the phlebotomist to draw less blood, but she refused, saying that her instructions had to be followed. While I was completely paralyzed, I began to think about why hospitals draw so much blood. I knew that this practice went back 20 years, to when most instruments required large quantities of serum. I decided that when I recovered, I would conduct a survey of blood-drawing practices.

After my recovery, I sent a questionnaire to 24 hospitals in the United States and received responses from 19 — 2 large community hospitals, 10 major university hospitals, and 7 children's hospitals. On the questionnaire, each hospital was asked to list the amount of blood it drew for the following tests: basic metabolic panel (blood urea nitrogen, sodium, potassium, chloride, carbon dioxide, glucose, and creatinine), comprehensive metabolic panel (blood urea nitrogen, sodium, potassium, chloride, carbon dioxide, calcium, glucose, creatinine, bilirubin, albumin, protein, aspartate aminotransferase, and alkaline phosphatase), automated complete blood count, complete blood count with manual differential, and a liver panel (bilirubin, albumin, alanine aminotransferase, aspartate aminotransferase, and alkaline phosphatase). The questionnaire also asked what equipment the hospital used for each of these tests.

The results showed that all the community hospitals and university hospitals drew far more blood for each test than did the children's hospitals. Table 1Table 1Blood-Drawing Practices at 19 Hospitals in the United States. summarizes these practices. For the basic metabolic panel, the amount of blood required by any university or community hospital laboratory was 2.5 to 10 times as much as the maximum required by any children's hospital, even though the tests were the same and the instruments used were the same or similar.

For the comprehensive metabolic panel, the situation was similar. For the automated complete blood count or the complete blood count with manual differential, the community and university hospitals drew 2.5 to 7 times as much blood as did the children's hospitals, again despite the fact that all the hospitals used the same or similar equipment. For the liver panel, the community and university hospitals drew 2.5 to 10 times as much blood as did the children's hospitals, again using the same equipment for the analyses.

I am concerned that in the United States we are drawing far more blood from adults than is necessary. This issue is of particular importance for the increasing number of older persons. It is the responsibility of all physicians and laboratorians to change this practice.

Jocelyn M. Hicks, Ph.D., F.R.C.Path.
Children's National Medical Center, Washington, DC 20010

Citing Articles (5)

Citing Articles

1. 1

   L. Pabla, E. Watkins, H. A. Doughty. (2009) A study of blood loss from phlebotomy in renal medical inpatients. Transfusion Medicine 19:6, 309-314
   CrossRef

2. 2

   J. M. Blondeau. (2008) What have we learned about antimicrobial use and the risks for Clostridium difficile-associated diarrhoea?. Journal of Antimicrobial Chemotherapy 63:2, 238-242
   CrossRef

3. 3

   Luis I. Cortinez, Jacques Somma, Kerri M. Robertson, John C. Keifer, David R. Wright, Yung-Wei Hsu, David B. MacLeod, Eugene W. Moretti. (2005) Changes in hematocrit based on incremental blood sampling: mathematical models perform poorly. Canadian Journal of Anesthesia/Journal canadien d'anesthésie 52:4, 374-378
   CrossRef

4. 4

   L. Kyne, M. B. Hamel, R. Polavaram, C. P. Kelly. (2002) Health Care Costs and Mortality Associated with Nosocomial Diarrhea Due to Clostridium difficile. Clinical Infectious Diseases 34:3, 346-353
   CrossRef

5. 5

   Takeo Kumura, Masayuki Hino, Takahisa Yamane, Noriyuki Tatsumi. (2000) Argatroban as an anticoagulant for both hematologic and chemical tests. Journal of Clinical Laboratory Analysis 14:4, 136-140
   CrossRef