EUSER: Point-of-care ultrasonography in patients admitted with respiratory symptoms
Reference: Laursen CB, Sloth E, Lassen AT, et al. Point-of-care ultrasonography in patients admitted with respiratory symptoms: a single-blind, randomised controlled trial. Lancet Respir Med. 2014;2(8):638-46.
Why did they do this study?
Wanted to see if doing US during initial evaluation decreased
time to the correct diagnosis. Randomized patients in Denmark with
respiratory complaints to either receive US as part of their initial
management or to not receive US. Primary end point was correct diagnosis after 4 hrs of their initial presentation.
When ultrasound is added to standard diagnostics when evaluating a pt with respiratory distress, you get to the correct diagnosis faster.
Inclusion criteria: Adults w/ RR >20, O2 sat <95%, those requiring O2, dyspnea, cough, or chest pain. Randomly assigned to standard diagnostic
strategy (control group) or to standard diagnostic tests w/US of the heart,
lungs, legs (DVT). US group = 158, Control group = 157.
Excluded patients referred to cardiology, trauma patients and those who
could not get ultrasounded within 1 hr of initial presentation.
Ultrasounds were performed by one physician.
Percentage of pts with correct diagnosis at 4 hrs
Length of hospital stay
In-hospital mortality rate
Mortality within 30 days
Time to diagnostic/therapeutic thoracentesis
Percentage of pts who received:
-Correct diagnosis and correct treatment
-Transfer to an ICU
-Readmission within one month
“Masked audit” after patient discharge.
Pts who received ultrasound in their initial assessment had an absolute increase in correct diagnosis within 4 hrs of 24%. This corresponds to a number-needed-to-scan of 1 in 4. Also, there was an absolute increase of 21% of pts who received correct treatment within 4 hrs.
What does it all mean?
This study suggests that if you use ultrasound in the evaluation of patients with respiratory symptoms, you’ll arrive to the diagnosis faster, and therefore initiate the appropriate treatment faster. Intuitively, it would make sense that this would translate to better patient outcomes. Unfortunately, this study does not prove this, and this is likely from the fact that the N is approximately 160 in each group. Larger numbers are needed.