The use of arterial catheters (A-lines) to monitor blood pressure in septic shock isn’t well supported by evidence. There’s never been an RCT showing that A-lines improve outcomes. It remains debated whether invasive blood pressure monitoring is even needed.1
Nonetheless, most clinicians believe that some patients benefit from continuous, invasive arterial BP monitoring (especially those on higher doses of vasopressors). The issue then becomes the best site to insert the A-line. Traditionally, the first-line choice has been the radial artery.
Implicit assumptions versus new concepts
Historically, emphasis has rested on the distinction between noninvasive versus invasive BP (e.g., cuff pressure vs. radial arterial pressure). Attention focused on whether noninvasive oscillometric BP monitoring is adequate. Meanwhile, it has been assumed that all invasive BP measurement sites are created equal:2
However, the anesthesiology literature suggests that this model is wrong. Specifically, the radial and femoral BP diverge among the sickest patients (e.g. patients undergoing cardiopulmonary bypass or liver transplantation).3 In such patients, the radial arterial pressure may be misleadingly low, with a potential to instigate excessive vasopressor and fluid administration.
We routinely do femoral A lines plus radial A-lines during liver transplants. Interesting to observe the discrepancies between the two measurements as vasopressor doses increase.
— Vivek Moitra, MD (@vmoitra) August 2, 2018
How reliable is the radial artery pressure in septic shock? This post will attempt to evaluate the evidence.
Relative accuracy of BP measurement at different sites: Primary literature
Noninvasive BP cuff vs. radial arterial catheter
Riley 2017 may be the best study to directly investigate this question.4 These authors prospectively compared noninvasive versus radial arterial pressure monitoring among 31 ICU patients with septic shock. All patients were on vasopressors, with about a quarter of the patients on multiple vasopressors. Using the radial arterial pressure as the gold standard, noninvasive BP had a bias of -2.5 mm and a 95% confidence interval ranging from -9 mm to +14 mm.
Ribezzo 2013 performed a similar prospective study comparing noninvasive versus invasive radial BP.5 This study utilized a mixed population of 50 ICU patients, of whom only 26% had septic shock. Using radial arterial pressure as the gold standard, noninvasive BP had a bias of -5mm with a 95% confidence interval between -19mm and +9mm:
This data may seem to be incompatible with Riley’s results above. However, most of the discordance occurs among patients with normal or elevated blood pressure. If we ignore patients with a MAP > 85mm and align this data with Riley’s data, the results are consistent:
Dorman 1998 prospectively compared invasive radial vs. femoral artery pressures in 14 postoperative septic patients receiving at least 5 mcg/min norepinephrine (although most received much higher doses, with a mean rate of 86 mcg/min).6 On average, femoral artery MAP was 15 mm higher than that obtained from the radial artery! Femoral A-line insertion often prompted substantial reductions in vasopressor dose after concluding that the radial A-line pressure was underestimating central arterial pressures.
Mignini 2006 prospectively compared radial and femoral A-line pressures among 55 patients in a medical-surgical ICU.7 Most patients (40/55) were receiving significant vasopressor infusions (defined as >10 mcg/kg/min of dopamine or >0.1 ug/kg/min of norepinephrine/epinephrine). Using femoral MAP as the gold standard, the radial MAP was on average 3 mm lower, with a 95% confidence interval ranging between -11 mm to +5 mm. This study may not be perfectly generalizable to septic shock since these patients had a variety of different diagnoses (with only 17/55 patients receiving norepinephrine).
Compton 2008 prospectively compared radial and femoral A-line pressures among 25 medical ICU patients (19/25 of whom were admitted due to respiratory failure).8 Nearly all patients were on vasopressors, most often norepinephrine with an average dose of 0.3 mcg/kg/min. Compared to the femoral MAP, the radial MAP was on average 3 mm lower, with a 95% confidence interval ranging between -17 mm to +11 mm:
Galluccio 2009 prospectively compared radial and femoral A-line pressures among 24 patients in a mixed medical-surgical ICU.9 Most patients (62%) had septic shock. All patients were on vasopressors, most commonly norepinephrine (79%) with a mean dose of 0.26 ug/kg/min. Compared to the femoral MAP, the radial MAP was on average 4 mm lower, with a 95% confidence interval ranging between -12 mm to +4 mm. However, this confidence interval isn’t entirely accurate since the distribution of blood pressure deviations appears skewed (the radial BP is never higher than the femoral BP):
Kim 2013 prospectively compared radial and femoral A-line pressures among 37 patients with septic shock who were initially receiving >0.1 ug/kg/min norepinephrine.10 Compared to the femoral MAP, the radial MAP was on average 5 mm lower, with a 95% confidence interval ranging between -17 mm to +7 mm. Differences tended to decrease as patients were weaned off norepinephrine.
— Ognjen Gajic (@ogi_gajic) August 28, 2017
Traditional assumptions vs. evidence
To summarize the above studies:
- Cuff vs. radial A-line: The 95% confidence interval of the MAP measured using an oscollometric noninvasive BP cuff is roughly +/- 12 mm when compared to a radial A-line.
- Radial A-line vs. femoral A-line: The 95% confidence interval of the MAP measured with a radial A-line appears to range between ~5 mm higher and ~15 mm lower that measured with a femoral artery A-line. Among sicker patients on higher vasopressor doses, correlation may be worse.
Overall, the correlation between cuff and radial artery MAP isn’t hugely different from that observed between radial and femoral artery A-line pressures (figure above; in both cases, the 95% confidence interval spans ~20 mmHg). This seems to dispel the traditional belief that all invasive BP sites are created equal:
This uncovers a problem in how we’ve been measuring blood pressure for decades:
- Most evidence and experience about blood pressure involves radial pressures. Therefore, one could argue that to be consistent with this historical evidence basis, we should continue using radial pressures.
- From a physiologic standpoint, central pressures ought to be better. Central pressures best describe perfusion of critical visceral organs (heart, kidneys, etc.). Alternatively, the blood pressure in the hand is unimportant.
Ideally, RCTs would be performed to determine the impact of using femoral vs. radial pressures. Ideally, landmark studies on BP targets in septic shock should be replicated using central BP targets. Lacking such evidence, most writers assume that central arterial pressures are preferred. I agree with this assumption, but it’s worth noting that it is an assumption.
Evidence-based approach to invasive Bp monitoring?
The first question is who needs invasive BP monitoring. There’s little evidence about this. Many patients don’t seem to require invasive BP monitoring (patients who are clinically improving and perfusing well on low-dose vasopressor support). Such patients certainly require meticulous ICU-level monitoring, but this can be achieved noninvasively. If the patient is responding to therapy and improving clinically, then there is little rationale to add invasive BP monitoring.
Invasive BP monitoring is reasonable for patients who are responding poorly to therapy or requiring higher-dose vasopressors. In this situation, a femoral arterial catheter seems more logical for several reasons:
- If a cuff pressure is deemed insufficiently accurate, then a radial A-line is also insufficiently accurate. A radial A-line might not be dramatically superior to the cuff pressure in this situation. If definitive BP monitoring is needed for a septic patient on high-dose vasopressors, central arterial pressure is ideal.
- Femoral lines can generally be inserted faster among critically ill patients with hypotension and vasoconstriction. Alternatively, it’s possible to waste an enormous amount of time fiddling around with radial A-lines, delaying other critical interventions.
- Femoral arterial lines are more reliable over time and less likely to suddenly quit working.11
- With good technique and ultrasonography, femoral A-lines have a similar safety profile compared with a radial A-lines (with one possible exception being coagulopathy: femoral A-lines do carry a risk of retroperitoneal hemorrhage, whereas the radial site is easily compressible).
Love it. My 2 cents for fem art lines over radial
1. Central BP more accurate in really sick
2. Bigger target for Ultrasoud
3. Less painful than radial
4. Safer during cardiac arrest
5. Sure for VA ECMO (if you do that type of thing)
— Haney Mallemat (@CriticalCareNow) August 2, 2018
In short, if you’re going to perform a procedure to measure the arterial pressure, you might as well perform the definitive procedure. The main drawback of a femoral arterial catheter may be infection, but septic patients will often improve within a few days allowing arterial line removal.
Fem A lines are the only good option in shock. Radial line BP unreliable, often underestimate central bp (up to 30% – will dig up reference). Dangerous. Seen a few pts almost killed by over-vasoconstriction. Levo dropped by >70% when fem line placed. Radials only ok for gases.
— Philippe Rola (@ThinkingCC) August 3, 2018
A few tips on safe placement of a femoral A-line
Intuitively it might seem that the femoral arterial line is a more invasive procedure that should be associated with greater complications. However, available evidence suggests that the femoral A-line is a safe procedure, with a similar complication rate compared to radial A-lines.12 13 14 Furthermore, meticulous placement with full sterility and ultrasound guidance makes the femoral A-line procedure safer today than it has been historically.15 Instead of running away from this procedure, we need to learn how to perform it safely.
Based on its ease of cannulation and low record of complications, the femoral artery has been called the vessel of choice for arterial access… some hospitals use femoral arterial lines almost exclusively -Roberts & Hedges’ 6th edition.
Negative example: How to damage the femoral artery
- The needle is inserted at a sharp angle, causing it to back-end through the vessel.
- The wire is forcibly advanced, causing it to kink several times within the soft tissue behind the vessel.
- Eventually, the wire must be removed. However, the kinked end of the wire prevents it from being smoothly withdrawn through the artery. Instead, the kinked wire lacerates both walls of the artery as it is dragged through them.
Complications may be minimized with some basic safeguards:
- External rotation of the legs (“frog-leg positioning”) is often helpful.
- If there is a pannus it must be retracted (typically with tape) until the skin is pulled taut. If pannus retraction cannot be achieved, consider placing axillary arterial line (more on this below).
The catheter should be placed within the common femoral artery for several reasons:
- It’s a larger target, so it’s easier to cannulate and wire successfully.
- As a larger artery, it’s less likely for it to become occluded if there is some minor trauma.
- A common femoral puncture allows the A-line to be converted into a site for intra-aortic balloon pump, ECMO catheterization, or interventional radiology procedures.16
Ultrasound should be used to identify the location where the common femoral artery splits into the superficial and deep femoral arteries. The femoral arterial catheter should be placed a few centimeters proximal to this split.17
3) Try to center the needle on the vessel
Using ultrasound guidance, insert the needle so that it punctures the center of the artery. If the needle punctures the edge of the artery this will produce blood return, but it may be difficult to advance the wire (the wire may impinge on the arterial side-wall). Your goal with cannulation isn’t merely to get the needle somewhere inside the artery – the goal is to get the needle into the center of the artery so that the wire threads easily.
4) Don’t insert the wire unless blood is spurting out of the needle
If the needle is positioned within the artery lumen, blood should be spurting out. If blood stops spurting, that means the needle has slipped into a suboptimal position (either outside vessel or hugging the vessel wall). Don’t insert the wire if blood isn’t gushing out. Stop, take a deep breath, evaluate needle position with ultrasound, and make small adjustments to the needle until blood starts spurting out again – then insert the wire.
5) Don’t force the wire
If the wire doesn’t seem to be advancing smoothly, don’t force it. This usually means that the wire is suboptimally placed. Forcing the wire will often cause the wire to kink – which risks vascular damage and also requires opening a new line kit. It’s safer and faster to remove the wire and try again.
Parting shot: The ultrasound-guided axillary arterial line?
Another nice option here is the ultrasound-guided axillary arterial catheter. This basically involves ultrasound-guided placement of a long catheter (use a femoral A-line kit) into the axillary artery close to the armpit.18 19 This is essentially “the femoral A-line of the arm,” with many of the same advantages as a femoral A-line:
- Placement of an axillary A-line is generally faster and easier than a radial arterial line.
- A long catheter placed via the axillary artery will lie in the subclavian artery and thus transduce central arterial pressure.20
The axillary A-line has some advantages compared to the femoral A-line. The greatest advantage is probably that an axillary A-line is easier to insert in patients with supermorbid obesity – if the patient’s hand is positioned behind their head, this will generally open up an ultrasonographic window allowing axillary A-line placement. One potential concern is that air bubbles could theoretically cause cerebral embolism, but this may be avoided by preferential use of the left axillary artery and flushing the A-line slowly.
Where should the axillary A-line fit within this rubric? It’s hard to say. There isn’t an enormous volume of evidence with axillary A-lines, which makes it difficult to clarify their precise role. Overall the axillary site seems like an excellent option that might become more widely utilized in the future.
- Historically BP monitoring has focused on the difference between noninvasive cuff monitoring vs. invasive BP monitoring, with the assumption that any site of invasive monitoring would be equivalent.
- Accumulating evidence shows that the radial A-line may inaccurately estimate central MAP in extremely sick patients. As such, a radial A-line may not represent a huge benefit over a noninvasive brachial BP cuff.
- Femoral A-lines offer several advantages in critical illness: accurate measurement of central BP, greater ease of insertion, and more reliable functioning. The femoral site may be preferable in most patients.
- Ultrasound-guided axillary A-lines are an emerging option which may allow safe insertion and accurate measurement of central blood pressure.
- EMCrit Arterial Lines: Part 1 & Part 2
- The Dirty Double (PulmCrit)
- The anatomy of femoral vascular access (Taming the SRU)
Acknowledgement: Thanks to Dr. Gilman Allen for thoughtful comments on this post.