I find myself regularly fascinated by the accuracies (and frequent inaccuracies) of our subjective findings regarding Trauma. Consider how often we hear about details, mechanism, intrusion, extrication time to name a few, and how heavily they guide both pre-hospital and ED workups.  Some interesting studies I have come across are regarding blood loss estimation.  Next time you hear report about blood loss at the scene, contemplate (and feel free to cackle and wisely reference) the following studies: 

The first, published in Prehospital Emergency Care, tested a cohort of EMTs and paramedics on estimating blood volumes spilled on to carpet and vinyl surfaces.  Blood product was poured onto these surfaces and participants estimated total volume demonstrated at six different spill sites.  Mean percent errors of 74% and 56% were calculated for initial estimations!   A similar study published in The Journal of Trauma: Injury, Infection, and Critical Care produced similar results, where only 8% of the 99 providers tested were within 20% of the actual volume, and only 24% were within 50% of actual volume.

Perhaps the more interesting feature of the first study is the post-estimation training that occurred afterwards.  One arm of the cohort returned to the initial scene where actual blood loss was revealed followed by some education on making volume estimates.  The second arm had similar education however this was performed in a classroom using slides instead of returning to the original scene.   Both groups were retested with new scenarios and both demonstrated improvement in mean percent errors to 59% (return to scene) and 45% (slides and classroom), suggesting this may be learn-able skill! 

It turns out we are not much better than our pre-hospital colleagues at blood estimation.  A study from the Western Journal of Emergency Medicine examined a cohort of 56 emergency physicians (mixed residents and attendings).  Participants were tested with 4 different scenarios where specific amounts of blood were poured onto a bedsheet, on gauze, a t-shirt and into a commode.   The mean standard error for all estimates was 116% with a range of 0% to 1233%. Only 8% tested were within 20% of the true value (sound familiar?)!

The next question I arrived at, which you may be thinking yourself, was the fact that these studies do not include scenario or vital signs.  One would expect that would result in more accurate estimates.  Unfortunately, the rabbit hole of my literature search revealed that, even equipped with additional information, both pre-hospital providers and emergency room physicians continued to be poor estimators of blood loss.  A study from The Journal of Trauma tested estimations in set amount of blood amounts (300 mL, 800 mL, and 1500 mL) in a “stable” patient and in an “unstable” patient.  Researchers found that in the stable patient (i.e. normal blood pressure and heart rate) blood loss was underestimated in larger amounts and overestimated in the 300 mL patient.  Remarkably, in the unstable patient, blood loss in both 800 and 1500 mL scenarios were underestimated.  Of the 870 estimates made, 51% were underestimated, 39% were overestimated and only 10% were exact. 

As so nicely asked in the discussion with Frank et al “one has to direct the question if visually estimated blood loss is of any pre-clinical value and worth being mentioned during handover in the emergency unit.“   Perhaps not, but there may be hope with additional training!

 References: 

1. Patton K., Funk DL., McErlean M., Bartfield JM. (2001) Accuracy of estimation of external blood loss by EMS personnel. The Journal of Trauma: Injury, Infection, and Critical Care. 50(5):914-916. 
2. Moscati, R., Billittier, AJ, Marshall, B., Fincher, M., Jehle, D., Braen, GR. (1999) Blood loss estimation by out-of-hospital emergency care providers. Prehospital Emergency Care, Jul-Sep;3(3): 239-42. 
3. Ashburn, J. C., Harrison, T., Ham, J. J., & Strote, J. (2012). Emergency physician estimation of blood loss. The western journal of emergency medicine, 13(4), 376-9. 
4. Frank M., Schmucker U, Stengel D, Fischer L, Lange J, Grossjohann R, Ekkernkamp A, Matthes G. (2010) Proper estimation of blood loss on scene of trauma: tool or tale? The Journal of Trauma. Nov;69(5):1191-5.