Tell me if this sounds like your department – some folks, every time you receive sign-out, the radiologist has billed for a new BMW. Other times, there’s an uncomfortable disuse of CT, even in the ticking time bombs of the elderly. We all have that anecdotal feel for the variation in practice, and these authors have gone ahead and quantified it.
This is a descriptive, observational study of a single-center in Virginia evaluating individual physician CT scan use. Overall, the department performed CT scans on 23.8% of the nearly 200,000 patient visits during the study period. Across the 49 Emergency Physicians tracked during the study period – yes, the variation is exactly as one would expect. The most frequent utilizers ordered CTs on nearly one third of patients, while the least frequent users only 1 in 8.
There are a couple individually coded information graphics embedded in the paper that demonstrate some extremely striking variation. For example, as coded by chief complaint – a few physicians ordered CT scans on >60% of headache patients, while an even greater number ordered zero. 60% vs. zero! While some practice variation is obviously acceptable in evolving practice and heterogenous patient substrate, this clearly reflects some element of underlying low-quality care.
As poorly implemented as they may be, the conception behind prior quality measures for CT use certainly has merit. Quality measures aimed at increasing testing yield – ideally coupled with liability protections for Emergency Physicians – are likely to have increasing roles moving forward.
“Variation in use of all types of computed tomography by emergency physicians”
http://www.ncbi.nlm.nih.gov/pubmed/23998807