Академический Документы
Профессиональный Документы
Культура Документы
Analysis of the evidence for the lower limit of systolic and mean
arterial pressure in children
Ikram U. Haque, MD, FAAP; Arno L. Zaritsky, MD, FAAP, FCCM
LEARNING OBJECTIVES
On completion of this article, the reader should be able to:
1. Identify key factors that contribute to determining the optimal blood pressure goals for resuscitation of critically ill children.
2. Select appropriate blood pressure targets in critically ill children.
3. Recall the relationship between systolic, diastolic, and mean arterial blood pressure.
Both authors have disclosed that they have no financial relationships with or interests in any commercial companies
pertaining to this educational activity.
Lippincott CME Institute, Inc., has identified and resolved all faculty conflicts of interest regarding this educational activity.
Visit the Pediatric Critical Care Medicine Web site (www.pccmjournal.org) for information on obtaining continuing medical
education credit.
Objective: Systolic blood pressure (SBP) and mean arterial pres- lower than our calculated values in adolescents. Clinical formulas
sure (MAP) are essential evaluation elements in ill children, but there for calculation of SBP and MAP (mm Hg) in normal children are as
is wide variation among different sources defining systolic hypoten- follows: SBP (5th percentile at 50th height percentile) ⴝ 2 ⴛ age
sion in children, and there are no normal reference values for MAP. in years ⴙ 65, MAP (5th percentile at 50th height percentile) ⴝ
Our goal was to calculate the 5th percentile SBP and MAP values in 1.5 ⴛ age in years ⴙ 40, and MAP (50th percentile at 50th height
children from recently updated data published by the task force percentile) ⴝ 1.5 ⴛ age in years ⴙ 55.
working group of the National High Blood Pressure Education Pro- Conclusion: We developed new estimates for values of 5th
gram and compare these values with the lowest limit of acceptable percentile SBP and created a table of normal MAP values for
SBP and MAP defined by different sources. reference. SBP is significantly affected by height, which has not
Design: Mathematical analysis of clinical database. been considered previously. Although the estimated lower limits
Methods: The 50th and 95th percentile SBP values from task of SBP are lower than currently used to define hypotension, these
force data were used to derive the 5th percentile value for children values are derived from normal healthy children and are likely not
from 1 to 17 yrs of age stratified by height percentiles. MAP values appropriate for critically ill children. Our data suggest that the
were calculated using a standard mathematical formula. Calculated current values for hypotension are not evidence-based and may
SBP values were compared with systolic hypotension definitions need to be adjusted for patient height and, most important, for
from other sources. Linear regression analysis was applied to create clinical condition. Specifically, we suggest that the definition of
simple formulas to estimate 5th percentile SBP and 5th and 50th hypotension derived from normal children should not be used to
percentile MAP for different age groups at the 50th height percentile. define the SBP goal; a higher target SBP is likely appropriate in
Results: A 9 –21% range in both SBP and MAP values was many critically ill and injured children. Further studies are needed
noted for different height percentiles in the same age groups. The to evaluate the appropriate threshold values of SBP for determin-
5th percentile SBP values used to define hypotension by different ing hypotension. (Pediatr Crit Care Med 2007; 8:138 –144)
sources are higher than our calculated values in children but are KEY WORDS: blood pressure; hypotension; infant; child
M onitoring blood pressure in diatric emergency department, and in- hypotension is associated with an in-
critically ill or injured chil- tensive care unit (1–5). Studies in adults creased risk of mortality (7–10). The
dren is considered one of and children with trauma showed that depth and duration of hypotension appear
the mainstays for patient systolic blood pressure (SBP) is a predic- to have a direct relationship with adverse
evaluation in the prehospital setting, pe- tor of mortality (6) and that prehospital hospital outcome in emergency depart-
ment patients with nontraumatic shock
(11). Moreover, hypotension is used as a
Assistant Professor (IUH), Professor and Chief Copyright © 2007 by the Society of Critical Care criterion for diagnosis of “decompen-
(ALZ), Division of Pediatric Critical Care Medicine, De- Medicine and the World Federation of Pediatric Inten- sated” shock along with other features of
partment of Pediatrics, University of Florida College of sive and Critical Care Societies poor perfusion in pediatric patients (1, 3).
Medicine, Gainesville, FL. DOI: 10.1097/01.PCC.0000257039.32593.DC Most published guidelines define hypo-
Table 1. Definition of hypotension in children by different sources The demographic information on the
source of blood pressure data of the pop-
Hypotension Guidelines, mm Hg ulation was described in detail in the
original report (30). Table 2 shows the 5th
Pediatric Advanced Life Brain Trauma International Pediatric Sepsis percentile values of SBP from age 1–18
Age Group Support Foundationa Consensus Conferenceb yrs for males and females; the table also
details the range of SBP for each age
0 days–1 wk ⬍60 ⬍65 ⬍59
1 week–1 mo ⬍60 ⬍65 ⬍79 group ranging from the 5th percentile to
1 mo–1 yr ⬍70 ⬍65 ⬍75 95th percentile for height. As seen in this
⬎1–5 yrs 70 ⫹ 2 ⫻ yrs ⬍70–75 ⬍74 table, the SBP in the same age group
6–12 yrs 70 ⫹ 2 ⫻ yrs (up to 10 yrs) ⬍80–90 ⬍83 varies by approximately 8 –9 mm Hg be-
⬍90 (⬎10 yrs)
tween the 5th and 95th height percentile
13–18 yrs ⬍90 ⬍90 ⬍90
in males and by approximately 6 –7 mm
a
Brain Trauma Foundation last accessed October 14, 2005 (http://www2.braintrauma.org/ Hg in females. Relative to the mean SBP,
guidelines/downloads/btf_guidelines_prehospital.pdf); b corrected values, letter to the editor, Gold- the magnitude of the difference between
stein et al. Pediatr Crit Care Med 2005; 6:500 –501. the 5th and 95th height percentile com-
Age, Yrs M F M F M F M F M F
DISCUSSION
1 62 66 65 68 67 68 70 71 72 73
2 67 68 70 70 70 71 72 71 74 73 The early detection and rapid treat-
3 68 68 71 71 73 71 76 74 77 76 ment of hypotension are important in
4 70 71 73 73 75 74 78 74 79 76
the management of critically ill and in-
5 72 71 76 74 78 76 78 77 80 79
6 73 74 76 76 78 77 81 79 83 81 jured children. Moreover, MAP is often
7 74 76 77 78 79 79 81 79 83 82 used as a therapeutic end point in shock
8 77 78 80 78 82 81 82 82 84 84 management. Thus, it is important to
9 77 78 80 81 82 83 85 84 87 86 define the normal ranges for both vari-
10 79 80 83 83 85 85 85 86 89 88
11 81 82 85 85 87 85 87 88 89 90 ables in children. These normal values
12 83 85 86 87 89 87 91 90 93 92 represent a starting point for thinking
13 87 87 88 89 90 90 92 92 94 92 about the desired blood pressure target,
14 88 89 91 89 94 92 96 93 98 95 but the desired target blood pressure in
15 92 90 95 92 95 93 97 93 99 95
16 93 91 96 93 98 93 101 96 103 98 a critically ill or injured child is likely
17 97 91 98 93 100 93 102 96 104 98 higher than these lower limits of nor-
mal blood pressure obtained in healthy
children.
pared with the 50th height percentile and were chosen because they are easier Although not evidence based, hypo-
SBP by age was 7–14.9% in boys and to recall. tension is typically defined as an SBP
5.4 –11.1% in females. This variation is less than the 5th percentile for age.
seen graphically in Figure 1. SBP (5th percentile at 50th height Since it is often difficult to recall blood
The calculated values for MAP at the pressure data listed in tables by age,
percentile) ⫽ 2 ⫻ age in years ⫹ 65 various formulas have been used to es-
5th, 50th, and 95th percentiles stratified
by different height percentiles in the [2] timate the lower limit of acceptable
same age group are shown in Table 3. A blood pressure. For example, the for-
difference of approximately 5– 6 mm Hg SBP (50th percentile at 50th height mula used in PALS training materials
is noted between 5th and 95th height (1) states that the lower limit of accept-
percentile) ⫽ 2 ⫻ age in years ⫹ 85 able SBP is 70 ⫹ 2 ⫻ (age in years) for
percentiles for the same age group in
both males and females. This results in a [3] children from 1 to 10 yrs of age. For
range of MAP across different heights of children ⬍1 yr, the lowest acceptable
8 –13% in girls and 9 –21% in boys. There MAP (5th percentile at 50th height SBP is 60 mm Hg up to 1 month and 70
is wide variation, however, in the normal mm Hg from 1 month to 1 yr. For
percentile) ⫽ 1.5 ⫻ age in years children ⬎10 yrs of age, the lowest ac-
MAP values across the 5th and 95th per-
centiles for a given height. ⫹ 40 [4] ceptable SBP is 90 mm Hg; the latter is
Linear regression was applied to the consistent with Advanced Cardiac Life
50th height percentile group for the 5th MAP (50th percentile at 50th height Support guidelines for defining hypo-
and 50th percentile values of SBP and the tension in adults (31). A review of dif-
percentile) ⫽ 1.5 ⫻ age in years ferent sources revealed substantial vari-
5th and 50th percentile MAP (data not
shown). Simplified formulas for estimat- ⫹ 55 [5] ation in their definition of systolic
ing the 5th percentile SBP and the 5th hypotension, and current sources do
and 50th percentile MAP for 50th percen- The 5th percentile SBPs are plotted as not consider the variation in blood
tile of height were derived from these shown in Figure 1 for both males and pressure due to height differences
values, as shown next. A simple calcula- females along with the PALS, Brain within an age group. Analysis of our
tion can be used to estimate 5th percen- Trauma Foundation, and International derived data from the updated blood
tile SBP adjusted for height percentile for Pediatric Sepsis Consensus Conference pressure data from the task force report
each quartile above or below 50th height definitions and our estimated formulas does not agree with any of the recom-
percentile. Our calculations suggest that for SBP determined from regression anal- mended threshold blood pressures to de-
for males ⫾2 mm Hg for each height ysis. The graph shows that systolic blood fine hypotension, making it difficult for the
quartile and for females ⫾1.5 mm Hg for pressure limits defined by both PALS and clinician to know which source to use and
each quartile of height can approximate Brain Trauma Foundation are high com- remember. Furthermore, it is interesting
the 25th to 75th percentile SBP values. pared with our calculated values, and the to note that the threshold blood pressure
The line of best fit for MAP was exponen- PALS limit often exceeds the 5th percen- values used in the Paediatric Logistic Or-
tial, but the linear equations produce es- tile SBP even in children at the 95th gan Dysfunction score are much higher
timates that are within the range of val- percentile of height. Moreover, the graph than any of the current threshold blood
ues across the range of height percentiles illustrates that most normal healthy ad- pressure values that define hypotension
Mean Arterial Blood Pressure for Boys and Girls, Percentile for Height
1 5 30 35 33 37 34 37 36 39 37 40
50 49 53 52 54 53 55 54 57 56 58
95 69 71 70 72 72 73 73 74 74 76
2 5 35 39 38 41 39 42 40 42 41 44
50 54 57 56 58 57 59 59 60 60 62
95 73 75 75 76 76 77 77 78 79 80
3 5 39 42 41 44 42 44 44 46 45 47
50 58 60 60 61 61 62 62 64 64 65
95 77 78 78 79 80 80 81 81 82 83
4 5 42 45 43 46 46 47 47 47 48 49
50 61 63 63 64 64 65 66 65 67 67
95 79 80 82 82 83 83 84 84 86 85
5 5 45 46 47 48 49 49 49 50 51 52
50 63 64 66 66 67 67 68 68 69 69
95 82 82 84 83 85 85 87 86 88 87
6 5 47 49 49 50 50 51 52 52 53 54
50 66 66 67 68 69 69 70 69 71 71
95 84 84 86 85 87 86 88 87 90 89
7 5 51 50 50 51 52 52 53 53 54 55
50 67 68 69 69 70 70 72 71 73 72
95 83 85 88 87 89 88 90 89 92 90
8 5 50 52 53 52 54 54 55 55 56 56
50 69 70 71 70 72 71 73 72 75 74
95 87 87 89 88 91 89 92 90 93 91
9 5 51 53 53 54 55 55 56 56 58 57
50 70 71 72 71 73 73 75 74 76 75
95 88 89 91 89 92 90 93 91 94 93
10 5 52 54 55 55 56 56 56 57 59 59
50 71 72 73 73 75 74 75 75 77 76
95 90 90 92 90 93 92 94 93 96 94
11 5 54 55 56 56 57 57 58 59 59 60
50 72 73 74 74 75 75 76 76 78 78
95 91 91 92 92 94 93 95 94 96 95
12 5 54 57 57 58 58 58 60 60 61 61
50 73 75 75 75 77 76 78 78 79 79
95 92 92 94 93 95 94 96 95 98 97
13 5 56 58 57 59 59 60 60 61 61 62
50 75 76 76 77 77 78 79 79 80 80
95 93 94 95 94 96 95 97 97 99 98
14 5 59 60 59 60 61 61 62 62 63 64
50 75 77 78 78 79 79 80 80 82 81
95 91 95 96 96 97 97 99 98 100 99
15 5 58 61 61 61 62 62 63 63 64 64
50 77 78 79 79 80 80 82 81 83 82
95 96 96 98 97 99 98 100 99 102 100
16 5 60 61 62 62 63 63 65 63 66 66
50 79 79 81 80 82 81 83 82 85 84
95 98 96 99 98 101 99 102 100 104 101
17 5 63 61 63 62 65 63 67 65 69 66
50 81 79 83 80 84 81 85 82 87 84
95 100 96 102 98 103 99 104 100 106 101
(32). These empirically derived blood pres- blood pressure in children. Using a re- It is important for clinicians to know
sure values suggest that our current gression model, we developed new simple that there is a fairly large difference on a
thresholds are too low. formulas that give a more accurate ap- percentile basis between the lower limit
Our study analyzed a large database of proximation of 5th percentile SBP for dif- of SBP for a short vs. a tall child. This
blood pressures in children to provide ferent age groups in normal children variation has not previously been reported
objective evidence for the lowest accept- from 1 to 17 yrs of age: or used to adjust the goal for assessment or
able SBP and MAP. Our data show that blood pressure target. We do not know the
the definition of the 5th percentile SBP SBP (5th percentile at 50th height clinical or pathophysiologic significance of
from all the currently used sources is this observation. The estimated SBP for
often higher than the values calculated percentile) ⫽ 2 ⫻ age in years ⫹ 65 25th and 75th percentile height children at
based on the largest available database of [6] a given age can be estimated by adding or