As many Emergency Physicians can probably attest, one of the curious practices of critical care is to catheterize every potential organ system - as though the presence of these catheters in some way improves outcomes. And, the theory is - the non-invasive numbers are not accurate enough upon which to base treatment options.
So, this is a simple study performed in an intensive care unit in which patients with arterial blood pressure monitoring receive non-invasive measurements at the arm, ankle, and thigh (not everyone in the ICU will have an accessible arm). And, essentially, the results show - even in the critically ill, even on vasopressors - that the mean arterial pressure in the arm is probably a accurate measurement, with a mean bias of 3.4 mmHg. The systolic and diastolic numbers, as well as the ankle and thigh values, were not quite as precise or accurate.
For the Emergency Department, it probably tells you it's OK to do what you probably already do - critically ill patients get arterial lines only if there is a luxury of time available. Someone else with half an hour to spare can poke around fruitlessly in the radial wrist before surrendering to the femoral....
"Noninvasive monitoring of blood pressure in the critically ill: Reliability according to the cuff site (arm, thigh, or ankle)"