So I thought I’d get the ball rolling with an article I found from the Critical Care Medicine journal titled ‘Methods of Blood Pressure Measurement in the ICU’. I thought this was a pretty interesting and useful study from a big journal and it seemed one that I could really get my teeth into as a first posting. I’ve included a copy of the abstract and the reference below.
Objective: Minimal clinical research has investigated the significance of different blood pressure monitoring techniques in the ICU and whether systolic vs. mean blood pressures should be targeted in therapeutic protocols and in defining clinical study cohorts. The objectives of this study are to compare real-world invasive arterial blood pressure with noninvasive blood pressure, and to determine if differences between the two techniques have clinical implications.
Design: We conducted a retrospective study comparing invasive arterial blood pressure and noninvasive blood pressure measurements using a large ICU database. We performed pairwise comparison between concurrent measures of invasive arterial blood pressure and noninvasive blood pressure. We studied the association of systolic and mean invasive arterial blood pressure and noninvasive blood pressure with acute kidney injury, and with ICU mortality.
Setting: Adult intensive care units at a tertiary care hospital.
Patients: Adult patients admitted to intensive care units between 2001 and 2007.
Interventions: None.
Measurements and Main Results: Pairwise analysis of 27,022 simultaneously measured invasive arterial blood pressure/noninvasive blood pressure pairs indicated that noninvasive blood pressure overestimated systolic invasive arterial blood pressure during hypotension. Analysis of acute kidney injury and ICU mortality involved 1,633 and 4,957 patients, respectively. Our results indicated that hypotensive systolic noninvasive blood pressure readings were associated with a higher acute kidney injury prevalence (p = 0.008) and ICU mortality (p < 0.001) than systolic invasive arterial blood pressure in the same range (≤70mm Hg). Noninvasive blood pressure and invasive arterial blood pressure mean arterial pressures showed better agreement; acute kidney injury prevalence (p = 0.28) and ICU mortality (p = 0.76) associated with hypotensive mean arterial pressure readings (≤60mm Hg) were independent of measurement technique.
Conclusions: Clinically significant discrepancies exist between invasive and noninvasive systolic blood pressure measurements during hypotension. Mean blood pressure from both techniques may be interpreted in a consistent manner in assessing patients’ prognosis. Our results suggest that mean rather than systolic blood pressure is the preferred metric in the ICU to guide therapy.
(Crit Care Med 2013; 41:34 40) http://journals.lww.com/ccmjournal/Abstract/2013/01000/Methods_of_Blood_Pressure_Measurement_in_the_ICU_.5.aspx
Design: We conducted a retrospective study comparing invasive arterial blood pressure and noninvasive blood pressure measurements using a large ICU database. We performed pairwise comparison between concurrent measures of invasive arterial blood pressure and noninvasive blood pressure. We studied the association of systolic and mean invasive arterial blood pressure and noninvasive blood pressure with acute kidney injury, and with ICU mortality.
Setting: Adult intensive care units at a tertiary care hospital.
Patients: Adult patients admitted to intensive care units between 2001 and 2007.
Interventions: None.
Measurements and Main Results: Pairwise analysis of 27,022 simultaneously measured invasive arterial blood pressure/noninvasive blood pressure pairs indicated that noninvasive blood pressure overestimated systolic invasive arterial blood pressure during hypotension. Analysis of acute kidney injury and ICU mortality involved 1,633 and 4,957 patients, respectively. Our results indicated that hypotensive systolic noninvasive blood pressure readings were associated with a higher acute kidney injury prevalence (p = 0.008) and ICU mortality (p < 0.001) than systolic invasive arterial blood pressure in the same range (≤70mm Hg). Noninvasive blood pressure and invasive arterial blood pressure mean arterial pressures showed better agreement; acute kidney injury prevalence (p = 0.28) and ICU mortality (p = 0.76) associated with hypotensive mean arterial pressure readings (≤60mm Hg) were independent of measurement technique.
Conclusions: Clinically significant discrepancies exist between invasive and noninvasive systolic blood pressure measurements during hypotension. Mean blood pressure from both techniques may be interpreted in a consistent manner in assessing patients’ prognosis. Our results suggest that mean rather than systolic blood pressure is the preferred metric in the ICU to guide therapy.
(Crit Care Med 2013; 41:34 40) http://journals.lww.com/ccmjournal/Abstract/2013/01000/Methods_of_Blood_Pressure_Measurement_in_the_ICU_.5.aspx
So what’s the paper about?
Well the authors have conducted this study to help provide more information on the use and drawbacks of non-invasive blood pressure (NIBP) monitoring of ICU patients. Whilst invasive blood pressure (IBP) monitoring through arterial lines can be clearly seen as having greater accuracy for measuring BP (an thus act as the gold standard), it is obviously not without its own set of limitations and indeed risks. In this article the authors have tried to shed light on the limitations of NIBP in comparison to IBP, but particularly focusing on the use of mean arterial pressure (MAP) compared with systolic blood pressure (SBP) as marker for haemodynamic monitoring. They’ve done this by comparing paired values of the two different methods (i.e. results that were recorded simultaneously by both methods in the same patient) and seeing how close the SBP and MAP values corresponded with each other in these two groups. But they’ve not stopped there. They’ve also gone on to have a look at how the difference in readings might affect different clinical outcomes, namely mortality and risk of acute kidney injury (AKI). In effect they have tried to support 3 hypotheses:
Well the authors have conducted this study to help provide more information on the use and drawbacks of non-invasive blood pressure (NIBP) monitoring of ICU patients. Whilst invasive blood pressure (IBP) monitoring through arterial lines can be clearly seen as having greater accuracy for measuring BP (an thus act as the gold standard), it is obviously not without its own set of limitations and indeed risks. In this article the authors have tried to shed light on the limitations of NIBP in comparison to IBP, but particularly focusing on the use of mean arterial pressure (MAP) compared with systolic blood pressure (SBP) as marker for haemodynamic monitoring. They’ve done this by comparing paired values of the two different methods (i.e. results that were recorded simultaneously by both methods in the same patient) and seeing how close the SBP and MAP values corresponded with each other in these two groups. But they’ve not stopped there. They’ve also gone on to have a look at how the difference in readings might affect different clinical outcomes, namely mortality and risk of acute kidney injury (AKI). In effect they have tried to support 3 hypotheses:
- There is a significant difference between recordings of BP by non invasive methods compared with invasive ones.
- This difference is not as great if MAP values are used instead of SBP values.
- This difference is clinically significant and has implications for patient mortality and development of AKI.
What’ve they found?
Well unsurprisingly they’ve been able to show how NIBP has the potential to measure BP (for which I suppose IBP monitoring can be considered to be the ‘true’ value) inaccurately. What’s most interesting is the clear trend that they have been able to demonstrate. Basically, when using the SBP value, NIBP recording will tend to underestimate the actual BP at higher values and overestimate the actual BP at lower blood pressures ( a systolic of <90mmHg). In contrast, this difference is not as great when the MAP value is used instead and whilst the potential to underestimate the actual BP at higher BP values remains, there is not the same overestimation at lower values.
The clinical significance of this is clear when you consider the group of patients where we want to be able to closely monitor their BP; those who are already hypotensive. The authors have been able to go on and demonstrate this clinical significance with some strong evidence in the second half of the paper. Here they have shown that hypotension (based on both SBP and MAP) has a relationship with increased incidence of ICU mortality and AKI. With hypotension being a risk for developing these clinical outcomes, it follows that use of SBP as a method for assessing this risk has the potential to overestimate the true BP and therefore underestimate the risk. This is again shown by the increased mortality and AKI in patients with the same systolic BP as measured on NIBP compared with IBP.
Well unsurprisingly they’ve been able to show how NIBP has the potential to measure BP (for which I suppose IBP monitoring can be considered to be the ‘true’ value) inaccurately. What’s most interesting is the clear trend that they have been able to demonstrate. Basically, when using the SBP value, NIBP recording will tend to underestimate the actual BP at higher values and overestimate the actual BP at lower blood pressures ( a systolic of <90mmHg). In contrast, this difference is not as great when the MAP value is used instead and whilst the potential to underestimate the actual BP at higher BP values remains, there is not the same overestimation at lower values.
The clinical significance of this is clear when you consider the group of patients where we want to be able to closely monitor their BP; those who are already hypotensive. The authors have been able to go on and demonstrate this clinical significance with some strong evidence in the second half of the paper. Here they have shown that hypotension (based on both SBP and MAP) has a relationship with increased incidence of ICU mortality and AKI. With hypotension being a risk for developing these clinical outcomes, it follows that use of SBP as a method for assessing this risk has the potential to overestimate the true BP and therefore underestimate the risk. This is again shown by the increased mortality and AKI in patients with the same systolic BP as measured on NIBP compared with IBP.
So what does this mean for us?
Well it would be great if we could use this evidence to adjust the SBP figures that we get from NIBP, e.g. knock 10mmHg off the value recorded. However, as they’ve pointed out in the article, the discrepancy between the values is highly variable across the whole range of blood pressure values. For example, the 95% intervals of discrepancy at a SBP of 70mmHg were +33.46mmHg and -19.37mmHg. But don’t worry, there is good news and it’s in the form of the authors’ main conclusion. That is the evidence from this study that if you use the MAP as a measure of cardiovascular status instead of SBP, then the discrepancy, particularly at low BP values, is much less. This is mirrored by the hard clinical outcome data which shows that the MAP values obtained by NIBP and IBP monitoring have pretty much the same ability to predict risk of death and AKI. And as we’d expect, the underestimation discrepancy that occurs at higher BP values doesn’t have any detectable clinical significance (I’m not going to be too worried by a discrepancy of 20 or even 30mmHg if their MAP is 100mmHg).
Well it would be great if we could use this evidence to adjust the SBP figures that we get from NIBP, e.g. knock 10mmHg off the value recorded. However, as they’ve pointed out in the article, the discrepancy between the values is highly variable across the whole range of blood pressure values. For example, the 95% intervals of discrepancy at a SBP of 70mmHg were +33.46mmHg and -19.37mmHg. But don’t worry, there is good news and it’s in the form of the authors’ main conclusion. That is the evidence from this study that if you use the MAP as a measure of cardiovascular status instead of SBP, then the discrepancy, particularly at low BP values, is much less. This is mirrored by the hard clinical outcome data which shows that the MAP values obtained by NIBP and IBP monitoring have pretty much the same ability to predict risk of death and AKI. And as we’d expect, the underestimation discrepancy that occurs at higher BP values doesn’t have any detectable clinical significance (I’m not going to be too worried by a discrepancy of 20 or even 30mmHg if their MAP is 100mmHg).
Is the study any good?
Time to do a bit of critical appraisal then. Well what’s impressed me most about the study is their use of hard clinical endpoints to reinforce the value discrepancy that they have measured. I think this really helps highlight the importance of using the MAP rather than SBP for these sorts of patients, as it is actually going to have a palpable difference to patient outcomes. There were many other good points about the study that I think are worth mentioning. First is the sample size that they’ve use; a whopping 27,022 recorded values from 852 patients. Not bad. Another plus point is that the population group that the used is probably a pretty representative one in terms of how well we would be able to apply the results to our own patients. Admittedly, as they have pointed out, it is only looking at the patients who are sick enough to require an arterial line. However, I doubt this would impact on the pattern of any discrepancy in the study, maybe just the total numbers who are hypotensive. In addition, with the one exception that I’ve mentioned below, they have been pretty thorough in their description of their selection process, as well as in their testing of possible confounding factors (use of vasopressors). I think this lends an increased level of weight to the conclusions that they are able to draw at the end.
The weaknesses of the study have mostly been covered in the authors’ own discussion section at the end. One of them which has unclear significance to me is the use of a limited variety of NIBP monitoring equipment (just the 2 types and both from Philips). Now I’m not sure what the differences are like for the different algorithms etc that these oscillometric monitors use, and I imagine it’s not great, but I don’t know enough about this sort of equipment to comment fully on it. Maybe those monitors from a different company would be much more/less accurate, but using just the data from this study we can only speculate. I think this has the potential to limit the applicability of the results, but again I’m not knowledgeable enough about oscillometric technique algorithms/equipment to comment.
Finally, one area that I’ve not entirely sussed is how the sample group for describing the difference between the types of measurement was obtained. It may just be me but the appendix has a great explanation of how the groups for analysing AKI and mortality were selected from the large original database, and yet I’m still not convinced I’ve grasped how they sampled for the first part of the study. The sampling/exclusion criteria that they have described is less restrictive than those they used for the AKI and mortality cohorts, and yet the patient numbers are smaller.
Time to do a bit of critical appraisal then. Well what’s impressed me most about the study is their use of hard clinical endpoints to reinforce the value discrepancy that they have measured. I think this really helps highlight the importance of using the MAP rather than SBP for these sorts of patients, as it is actually going to have a palpable difference to patient outcomes. There were many other good points about the study that I think are worth mentioning. First is the sample size that they’ve use; a whopping 27,022 recorded values from 852 patients. Not bad. Another plus point is that the population group that the used is probably a pretty representative one in terms of how well we would be able to apply the results to our own patients. Admittedly, as they have pointed out, it is only looking at the patients who are sick enough to require an arterial line. However, I doubt this would impact on the pattern of any discrepancy in the study, maybe just the total numbers who are hypotensive. In addition, with the one exception that I’ve mentioned below, they have been pretty thorough in their description of their selection process, as well as in their testing of possible confounding factors (use of vasopressors). I think this lends an increased level of weight to the conclusions that they are able to draw at the end.
The weaknesses of the study have mostly been covered in the authors’ own discussion section at the end. One of them which has unclear significance to me is the use of a limited variety of NIBP monitoring equipment (just the 2 types and both from Philips). Now I’m not sure what the differences are like for the different algorithms etc that these oscillometric monitors use, and I imagine it’s not great, but I don’t know enough about this sort of equipment to comment fully on it. Maybe those monitors from a different company would be much more/less accurate, but using just the data from this study we can only speculate. I think this has the potential to limit the applicability of the results, but again I’m not knowledgeable enough about oscillometric technique algorithms/equipment to comment.
Finally, one area that I’ve not entirely sussed is how the sample group for describing the difference between the types of measurement was obtained. It may just be me but the appendix has a great explanation of how the groups for analysing AKI and mortality were selected from the large original database, and yet I’m still not convinced I’ve grasped how they sampled for the first part of the study. The sampling/exclusion criteria that they have described is less restrictive than those they used for the AKI and mortality cohorts, and yet the patient numbers are smaller.
That’s a lot of talking, summarise it for me.
Okay then:
Okay then:
- Non invasive measurement of systolic blood pressure is, on average, likely to overestimate the actual systolic value in hypotensive patients.
- This overestimation is clinically significant, reducing the sensitivity of the value as a marker to detect those at risk of death and AKI.
- Use of the non invasive MAP value does not demonstrate a significant overestimation at low blood pressures.
- The MAP can therefore be used as a suitable measure for assessing cardiovascular status in patients, especially those who are hypotensive.
So an overall learning point?
Organ perfusion (or lack of it) is the reason we get so excited about blood pressure values in critical care. MAP is a better measure of this than SBP, so we should be using it.
Well I hope you found that relatively interesting and informative. I thought that overall it was a fairly straightforward study that had some useful learning points extractable from it. It’s certainly updated my own clinical practice and I’ve enjoyed having some further practice at critically appraising a paper. Let me know your thoughts.
Tom Heaton
Organ perfusion (or lack of it) is the reason we get so excited about blood pressure values in critical care. MAP is a better measure of this than SBP, so we should be using it.
Well I hope you found that relatively interesting and informative. I thought that overall it was a fairly straightforward study that had some useful learning points extractable from it. It’s certainly updated my own clinical practice and I’ve enjoyed having some further practice at critically appraising a paper. Let me know your thoughts.
Tom Heaton
RSS Feed
