Somebody has done it. The sometimes contentious negotiations that happen at the end of every operation between the surgeon and the anesthesiologist over the case's estimated blood loss has finally been studied scientifically.
John Stover, MSN, a nurse practitioner out of Duke University, conducted a study on the EBL of 60 multilevel spine surgery patients. The methods they used to recollect the blood after being lost from the body are too lengthy and tedious for me to repeat here. You can read it from the linked article. Briefly, I'll just say that it involves adding heparin to the collection cannister and soaking the surgical sponges with saline.
By the end of the study, they found that anesthesiologists overestimated surgical blood loss by 40%. The average estimate by anesthesiologists was 860 mL while the measured loss was 611 mL. The reason this is important is because if anesthesiologists overestimate the EBL, we also run the risk of transfusing patients when it's not necessary. This could potentially lead to disastrous complications like viral infections, transfusion reactions, transfusion pneumonitis, and poor wound healing.
However, the study's methodology would be impractical to replicate in a real world operating room. As Martin London, MD, professor of clinical anesthesia at the University of California, San Francisco, dryly remarked, the processes involved are rather "labor intensive." Even Mr. Stover agreed. Said the author, "I think the problem is that we still use visual estimation of hemoglobin loss, and it’s very difficult to determine the actual hemoglobin concentration. So at this point there’s not a silver bullet that will give us an exact number."
Personally, I don't transfuse a patient based solely on the amount of blood loss I'm seeing. Yes, while 40% overestimation may seem a lot, I would not transfuse a patient just because I think the EBL is 800 mL when in fact she only lost 600 mL. I also have to take into consideration a multitude of other factors including the patient's physiologic stamina, the stability of the vital signs, and the likelihood of large ongoing exsanguination. All these are important when an anesthesiologist decides to give a blood transfusion.
We do not take this task lightly. We are well aware of our responsibilities to our patients when we decide to give blood. The approach to making that decision is extensively discussed during residency training. We don't give blood just based on a cutoff number. So while this study is interesting, its impractical application is of little help to anesthesiologists. We'll continue to transfuse patients based on what decades of transfusion studies have taught us: first do no harm to the patient.