An echocardiographic quantitative evaluation of the cardiac and vascular structures is often of critical importance for the diagnosis and management of congenital heart diseases. The authors reviewed the accuracy and limits of published echocardiographic nomograms for cardiac chamber, valve, and main vessel dimensions in children, with special attention to the neonatal age group. A computerized literature search in the National Library of Medicine using the keywords “echocardiographic normal/references values ± children/neonates/newborns” was performed. The research was redefined adding separately the keywords “aortic valve/annulus,” “aortic arch,” “atrio-ventricular valve/annulus,” “left ventricle,” “mitral valve/annulus,” “pulmonary valve/annulus,” “pulmonary artery,” and “tricuspid valve/annulus.” The analysis highlights the accuracy of the latest studies but also underscores that some limitations remain. In many studies, the number of healthy subjects was limited, with poor differentiation among age subgroups, and neonates were fully investigated in a very limited number of studies; moreover, data for many cardiac structures were not numerous, especially for the aortic arch and pulmonary branches. Finally, several methodologic limitations were encountered, including the lack of standardization, the different types of body size measurements used for normalization, and the various ways to express normalized data. As a result, nomograms were heterogeneous and although for some cardiac structures provided comparable confidence intervals, for others showed widely different values. The lack of solid, standardized nomograms, based on a large set of healthy children, may affect accuracy in estimating the severity of defects, especially in neonates, and possibly introduce bias in the clinical decision-making process.
Echocardiography represents the first approach for the diagnosis and management of congenital heart diseases (CHDs). A quantitative assessment of cardiac chamber, valve, and vessel dimensions is often of critical importance to evaluate the severity of any CHD to plan the most appropriate medical, interventional, and/or surgical treatment. A correct judgment of the sizes of cardiac structures in particular may be decisive in the neonatal age group in the case of suspected hypoplasia. Echocardiographic measurements are routinely repeated as part of the follow-up of many CHDs, and normalization of the various parameters according to body size is essential, because cardiac structures, as with the rest of the body, change during growth.
Many echocardiographic nomograms for the pediatric age group have been proposed, from the early 1980s up to the present time. However, the accuracy of some of these previous reports has been limited by the small number of healthy subjects evaluated, the scarce differentiation among groups of age (from birth to 18 years), and the lack of standardization in methods of measurement.
Despite advances in the standardization of echocardiographic measurements and in the number of healthy subjects evaluated, the process of normalization of pediatric echocardiographic values is far from standardized, as recently underscored, thus greatly limiting the diagnostic accuracy of echocardiographic measurements in children.
Available nomograms may produce very different and sometimes confounding results. For instance, for a male child with a body surface area (BSA) of 0.3 m 2 and a mitral valve annulus of 11 mm, significantly different results can be obtained according to the nomogram used, from lower limits of normality (−1.63) up to a clear hypoplasia of −4.84. As a result, grading the severity of mitral hypoplasia is strongly affected by the nomogram used, underscoring that the clinical interpretation of Z scores may be challenging.
Another major limitation to the process of normalization is represented by the extremely small availability or even lack (for some structures) of data in the neonatal age group. For instance, only scattered data are available on the mitral and tricuspid valve annuli, the pulmonary annulus, the main pulmonary artery, the pulmonary branches, and the aortic arch.
Finally, several methodologic points remain unclear, mainly related to the difficulty of normalizing values according to body size. In particular, it is still debated which body size measurement should be used for normalization and how normalized data should be expressed.
The aim of our study was to review the accuracy and limits of published nomograms for echocardiographic measurements of cardiac valve, left ventricular (LV), and main vessel diameters in children, with particular attention to the neonatal age group.
The review was focused on echocardiographic measurements of the following structures: (1) LV end-diastolic and end-systolic diameters evaluated using M-mode imaging; (2) the mitral valve annulus evaluated in the apical four-chamber view and parasternal long-axis view and the tricuspid valve annulus evaluated in the apical four-chamber view; (3) diameters of pulmonary annulus, main pulmonary artery, and pulmonary arteries measured in the parasternal short-axis view; (4) diameters of the aortic annulus, the aortic root at the sinuses of Valsalva, the sinotubular junction, and the ascending aorta evaluated in the parasternal long-axis view; and (5) diameters of the aortic arch performed at different levels: the proximal transverse arch (between the innominate and left common carotid arteries), the distal transverse arch (between the left common carotid and left subclavian arteries), and the aortic isthmus (the narrowest aortic segment distal to the left subclavian artery).
Potential studies were identified from a systematic search in the National Library of Medicine (i.e., PubMed access to MEDLINE citations; ) conducted in November 2010. The search strategy included a mix of Medical Subject Headings and free-text terms for the key concepts, starting with “echocardiographic normal/references values ± children, neonate, newborns.” The research was then refined adding separately the keywords “aorta,” “aortic annulus,” “aortic arch,” “atrio-ventricular annulus,” “left ventricle,” “mitral valve/annulus,” “pulmonary valve/annulus,” “pulmonary artery,” and “tricuspid valve/annulus.” Additionally, we identified potentially relevant studies using a manual search of references from all eligible studies, review articles, and Science Citation Index Expanded on the Web of Science. All identified reports were assessed independently by two reviewers, and a consensus was reached.
Titles and abstracts of all articles identified by the search strategy were evaluated and rejected on initial screening if they (1) did not consider populations widely constituted of normal subjects or mixed adults with children (three works), (2) did not report correlations between echocardiographic measures and BSA expressed using regression equations (eight works; five studies considered relations of echocardiographic measurements with weight instead of BSA, and three did not report regression equations) (we made exceptions for six works reporting values according to weight in the neonatal age group), and (3) were written in languages other than English (two works).
The search identified 44 reports. Thirteen did not meet the aforementioned criteria, leaving 30 reports for analysis.
We first discuss methodologic issues related to measurement adjustment according to body size and how to express normalized values; we then provide a detailed analysis of the available nomograms for the different cardiac structures.
Methodologic Limitations for Adjusting Measurements According to Body Size: BSA, Z Score, and Confounding Factors
For a correct evaluation of cardiac measurements at different ages, adjustment for body size according to a healthy pediatric population is of paramount importance.
In the past, different variables have been used for correction, including age, weight, height, BSA, and body mass index. It is now accepted that BSA is superior to age, height, or weight alone. However, some mathematical limitations in the use of BSA still persist, including a dependence of BSA-adjusted measurements on BSA and high variability of results according to the formula used, particularly for lower height and weight. The DuBois formula is currently the most used, but it has been recently challenged by the Haycock formula.
The DuBois formula in particular tends to increasingly underestimate for BSAs < 0.7 m 2 compared with the Haycock formula. A similar problem of underestimation at low BSA values has also been observed for other formulas.
The use of BSA, furthermore, remains very controversial in neonates, because an increase in body weight leads to a minimal increase in body surface area at this age ; indeed, the majority of the studies on neonates adjusted echocardiographic values according to body weight.
A further issue with the use of BSA is the need for a mathematical descriptor able to maintain a stable and constant variance across the range of body sizes. A feasible solution, used in various works, could be to perform a logarithmic, square root, or exponential transformation of measurements and/or BSA within a linear or nonlinear regression equation. All of these methods, however, do not completely solve the problem of heteroscedasticity. For instance, the use of the square root transformation allows for only a suboptimal linear relationship between the measurements and the transformed BSA, while the logarithmic transformation exhibits limitations for the highest and lowest BSA values.
Normalized data may be expressed in different ways, including the percentage of the mean normal value, percentile charts, and Z scores. The latter approach has become an increasingly popular approach to account for the effects of body size and age. The Z score of a measurement is the number of standard deviations of that value from the mean value. If a measurement is equal to the population mean, the Z score is 0; a Z score of +2 or −2 corresponds to the 95th percentile (i.e., 2 standard deviations above or below the mean). The main improvement of using Z scores has been the absence of any dependence on a predetermined association between the size of a structure and BSA. The mean square error (MSE), however, which is essential for calculating Z scores, is not always indicated, as detailed below.
Other methodologic limitations are represented by the difficulty of defining a “normal” population solely on the basis of “normal” results on echocardiography and by the scarce attention to confounders, such as gender and race, and interobserver and intraobserver variability. With regard to gender, only two studies provided separate nomograms for male and female subjects, while in several other articles, the percentages of male and female subjects in the study populations were not even specified.
The issue of interobserver and intraobserver variability has been scarcely considered, with only a few studies reporting the number of operators, and blinded controls to assess measurement reproducibility have been rarely performed.
Finally, an additional methodologic limitation pertaining to the lack of an international consensus on the standardized way to perform echocardiographic measurements was also considered, as guidelines have only recently become available. Some measurements were therefore performed using different approaches, at various times of the cardiac cycle, and frequently simply transposing adult recommendations to the pediatric age group.
Nomograms for Different Cardiac Structures
We discuss here the limitations of the current nomograms, divided according to the different cardiac structures.
Overall, these limitations are due mainly to a lack of data, to difficulties in translating adult knowledge to the pediatric age group, to the scarce differentiation among age and gender and race subgroups, and to the complexity of adjusting values according to body size and their normalization.
In particular, especially for the neonatal age group, robust nomograms for the atrioventricular valves, pulmonary arteries, and aortic arch are lacking. Hypoplasia of these structures nevertheless is a common cause of many CHDs, and an accurate quantitative assessment is required for planning the correct therapeutic approach, especially in neonates.
To underscore the differences among various nomograms, we computed Z scores for every cardiovascular structure with different nomograms, using an arbitrary value and a given BSA of 0.3 m 2 .
Measurements of LV size and function may play an important role in the evaluation and management of children with congenital and acquired heart diseases.
Adult recommendations for chamber quantification have often been adopted for children, but their accuracy and reproducibility have been barely investigated.
In adults, LV diameters are usually evaluated in the long-axis view, allowing a perpendicular orientation between the measurement and the LV long axis, limiting the risk for overestimating minor-axis diameters (as may happen in the short-axis view).
The use of a short-axis view, on the contrary, used in several studies, may offer some advantages in children, as it accounts for the lateral motion of the left ventricle observed in many children and allows multiple diameter calculations (with the subsequent choice of the one with the best blood-endocardium interface).
The biplane Simpson method to estimate LV volumes, although commonly used in adults, has been scarcely used in children.
Nomograms for LV dimension by M-mode echocardiography were reported by most of the studies evaluated.
Nine works reported the regression equation in relation to BSA and are detailed in Table 1 .
|Gutgesell et al.||145 (80 M, 65 F)||1 d to 19 y (30 neonates)||D||Yes||y = 37.75 + 12.88 × ln(BSA) |
BSA = 0.3: 22.24 (18.91–25.58)
|Henry et al.||105 (52 M, 53 F)||1 d to 23 y (13 neonates)||mD||NR||y = −6.6 + 45.2 × (BSA) 1/3||y = −3.8 + 28.3 × (BSA) 1/3|
|Oberhansli et al.||53 (27 M, 26 F)||1 d to 1 y||NR||Yes||y = 10.0 + 38.6 × BSA |
BSA = 0.3: 21.6 (21.3–21.9)
|y = 7.2 + 23.5 × BSA |
BSA = 0.3: 14.2 (13.9–14.6)
|Akiba et al.||110 (67 M, 43 F)||2–15 y||H||NR||y = 41.4 × BSA 0.49||y = 28.1 × BSA 0.49|
|Ichida et al.||153 (97 M, 76 F)||1 d to 15 y||BT||Yes||y = 3.861 + 1.365 × ln(BSA) |
BSA = 0.3: 22.2 (18.2–26.2)
|Huwez et al.||127 (77 M, 50 F)||7 mo to 19.5 y||D||NR||y = 1.3988 + 1.0295 × BSA||y = 2.2326 + 1.59 × BSA|
|Kampmann et al.||2,036||0–18 y (450 neonates)||D||NR||y = 38.537 × BSA 0.4509 |
BSA = 0.3: 22.9 (18.0–25.8)
|y = 24.231 × BSA 0.4469 |
BSA = 0.3: 14.8 (10.8–18.8)
|Sluysmans et al.||496||0–20 y||H||NR||y = 3.935 × BSA0.444|
|Pettersen et al.||782||1 d to 18 y (82 neonates)||NR||Yes||ln( y ) = 0.105 + 2.859 × BSA − 2.119 × BSA 2 + 0.552 × BSA 3 |
BSA = 0.3: 22.0 (18.0–26.8)
|ln( y ) = −0.371 + 2.833 × BSA − 2.081 × BSA 2 + 0.538 × BSA 3 |
BSA = 0.3: 13.6 (10.5–17.5)
Measurements have been commonly performed with reproducible and standardized protocols; nevertheless, the accuracy of these nomograms could be impaired by different factors, including poor differentiation among age subgroups and limited attention to confounders such as gender, race, and interobserver and intraobserver variability (with only three works reporting the number of operators ).
Furthermore, BSA was estimated using four different formulas, different regression equations were used (four nonlinear and five linear), and MSE was not indicated in five cases.
With regard to sample size, moreover, the numbers of subjects were adequate in only a few works that also evaluated the neonatal age group.
As a result, although available nomograms to calculate Z scores may provide similar data for end-diastolic volumes, widely different results may be obtained for end-systolic diameter ( Table 1 ).
Concerning the neonatal age group, only a few data are available for normal LV dimensions in an adequate number of neonates. Difficulties in the estimation of BSA for low-weight neonates led some investigators to express data in relation to body weight. In the widest study conducted so far in the neonatal age group, by Kampmann et al. , body weight was considered in addition to BSA, by dividing a population of 450 European neonates into four weight groups (from 2 to 4 kg). A significant correlation of cardiac dimensions with weight was demonstrated ( Table 2 ).
|Walther et al. (210 neonates)|
|0.750–1.249 kg||12.6 ± 1.5||8.5 ± 1.2||7.2 ± 0.7|
|1.250–1.749 kg||13.3 ± 1.2||9.1 ± 1.0||8.5 ± 1.0|
|1.750–2.249 kg||15.2 ± 1.5||9.8 ± 1.5||9.6 ± 0.8|
|2.250–2.749 kg||17.3 ± 2.2||11.5 ± 1.4||10.3 ± 0.9|
|2.750–3.249 kg||17.9 ± 2.1||11.7 ± 1.5||10.8 ± 1.0|
|3.250–2.749 kg||18.3 ± 2.0||12.1 ± 1.7||11.6 ± 1.0|
|3.750–4.249 kg||19.3 ± 2.4||12.5 ± 1.7||12.0 ± 1.0|
|4.250 kg||21.3 ± 2.3||14.4 ± 1.6||12.5 ± 0.7|
|Kampmann et al. (450 neonates)|
|2 kg||17.1 (−2 Z 15 +2 Z 19.2)||11 (−2 Z 9.7 +2 Z 12.3)||8.2 (−2 Z 6.9 +2 Z 9.5)|
|2.5 kg||18.1 (−2 Z 15 +2 Z 21.1)||11.7 (−2 Z 9.2 +2 Z 14.2)||8.8 (−2 Z 7.4 +2 Z 10.2)|
|3 kg||18.2 (−2 Z 15.1 +2 Z 21.3)||11.7 (−2 Z 9.2 +2 Z 14.2)||9.1 (−2 Z 7.5 +2 Z 10.7)|
|3.5 kg||18.8 (−2 Z 15.4 +2 Z 22.2)||11.9 (−2 Z 9.5 +2 Z 14.3)||9.3 (−2 Z 7.5 +2 Z 11.1)|
|4 kg||19.9 (−2 Z 16.5 +2 Z 23.3)||12.7 (−2 Z 10.2 +2 Z 15.2)||9.6 (−2 Z 7.6 +2 Z 11.6)|
|Tacy et al. (103 neonates)|
|Mitral annulus 4C||y = 7.2 + 1.2 × WT|
|Mitral annulus PLAX||y = 6.8 + 1.3 × WT|
|Tricuspid annulus 4C||y = 7.3 + 1.4 × WT|
|Aortic annulus||y = 4.3 + 0.9 × WT|
|Pulmonary annulus||y = 5.0 + 1.3 × WT|
Valvular diameters and area have been usually evaluated by means of two-dimensional (2D) imaging. More recently, three-dimensional (3D) imaging has been introduced in adults, leading to better correlation with measurements obtained by computed tomography.
A few works have evaluated atrioventricular valve area in children using both 2D planimetry and 3D imaging, and preliminary data suggest that 3D imaging may give a more accurate estimate of valve area than 2D imaging and Doppler equation methods.
Early reports on annular valve diameters were based on M-mode echocardiographic measurements from the leading edge of the near-field reflector to the leading edge of the far-field reflector.
Since these preliminary works, pediatric nomograms for mitral valve diameters have been commonly calculated using apical four-chamber view (lateral diameter) and/or parasternal long-axis view (anterior-posterior diameter) measurements, while tricuspid valve diameters have been evaluated in the four-chamber view.
Until a few years ago, there was no agreement on the optimal methodology for valve dimension measurement. Recently, recommendations have been published suggesting that 2D valvular diameters be measured at the moment of maximum expansion (in diastole for the atrioventricular valve and in systole for semilunar valves) and from inner edge to inner edge at the hinge points of the leaflet attachments.
We evaluated six different nomograms of valvular diameters expressing data according to BSA. A linear correlation of valve diameters and area with logarithmic or exponential transformation of BSA, stronger than that with weight or height, has emerged. Gender was indicated in only three works, although sex-related differences have been reported, with boys showing larger annular diameters at any age, even after adjustment for differences in body sizes.
Differences in regression equations (nonlinear vs linear), together with the scarce attention to confounders such as sex and race and poor differentiation among groups of age (from birth to adulthood), limit the comparability of different nomograms and result in significant discrepancies in reference values. Data on normal values for the mitral and tricuspid annulus in neonatal age are extremely limited.
Only one study reported data in relation to BSA in a sufficient size (82 subjects with BSAs < 0.25 m 2 ). The formula used to calculate BSA, however, is not specified, and gender and race were not taken into account ( Table 3 ).
|Study||n||Age||BSA||MSE||Mitral Annulus 4C||Mitral Annulus PLAX||Tricuspid Annulus 4C|
|King et al.||103||1 d to 15 y||NR||Yes||y = 32.3 + 12.47 × ln(BSA) |
BSA = 0.3: 17.3 (13.7–20.9)
|y = 23.9 + 8.56 × ln(BSA) |
BSA = 0.3: 13.6 (10.9–16.3)
|y = 32.4 + 12.29 × ln(BSA) |
BSA = 0.3: 17.6 (13.9–21.3)
|Hanséus et al.||120 (58 M, 62 F)||3 d to 15.5 y||NR||Yes||ln( y ) = −0.209 + 0.302 × ln(BSA) |
BSA = 0.3: 5.6 (3.2–9.9)
|ln( y ) = −0.209 + 0.302 × ln(BSA) |
BSA = 0.3: 5.6 (3.2–9.9)
|Daubeney et al.||125 (69 M, 56 F)||0–17 y||B||Yes||ln( y ) = 0.9651 + 0.4658 × ln(BSA) |
BSA = 0.3: 15.0 (12.5–18.0)
|ln( y ) = 0.9445 + 0.5022 × ln(BSA) |
BSA = 0.3: 14.0 (11.6–17.0)
|ln( y ) = 1.084 + 0.4945 × ln(BSA) |
BSA = 0.3: 16.3 (13.9–19.2)
|Sluysmans et al.||496||0–20 y||H||NR||y = 0.125 + 2.002 × (BSA) 1/2||y = 0.117 + 1.844 × (BSA) 1/2||y = 0.155 + 2.075 × (BSA) 1/2|
|Zilberman et al.||434 M||0–18 y||D||Yes||ln( y ) = 0.765 + 0.425 × ln(BSA) |
BSA = 0.3: 12.9 (9.2–18.1)
|ln( y ) = 0.817 + 0.391 × ln(BSA) |
BSA = 0.3: 14.1 (10.0–19.9)
|314 F||0–18 y||ln( y ) = 0.733 + 0.408 × ln(BSA) |
BSA = 0.3: 12.7 (8.9–18.3)
|ln( y ) = 0.755 + 0.364 × ln(BSA) |
BSA = 0.3: 13.7 (9.5–19.9)
|Pettersen et al.||782||1 d to 18 y (82 neonates)||NR||Yes||ln(y) = −0.271 + 2.446 × BSA − 1.70 × BSA 2 + 0.425 × BSA 3 |
BSA = 0.3: 13.8 (10.2–18.6)
|ln(y) = −0.164 + 2.341 × BSA − 1.596 × BSA 2 + 0.387 × BSA 3 |
BSA = 0.3: 15.0 (10.3–21.9)
Tacy et al. , in 103 preterm and term neonates aged 0 to 10 days, reported mitral and tricuspid annular dimensions in relation to body weight, dividing patients into seven groups according to weight, including very low weight (0.5–4 kg). This work, which was methodologically correct, demonstrated that weight was a good predictor of valve annular dimension, and no significant gender-related differences were noted. Intraobserver variability was also estimated, being on average 13.5% and ranging from a mean of 11.3% for pulmonary valve measurements to 15.8% for the tricuspid valve.
Aorta and Aortic Arch
Quantitative evaluations of aortic root and ascending aorta diameters are important both in suspected hypoplasia and in the case of dilatation, such as in follow-up of patients with connective tissue disease (Marfan syndrome) or bicuspid aortic valves.
We selected 15 works providing nomograms for the aortic annulus, aortic root, and sinotubular junction according to BSA; only five also provided normal values for the aortic arch.
As with atrioventricular valves, gender-related differences have been reported for the aortic annulus, the aortic root, and sinotubular junction, with male subjects presenting higher values than female subjects ( Table 4 ).
|Study||n||Age||BSA||MSE||Aortic Annulus||Aortic Root||Sinotubular Junction|
|Henry et al.||105 (52 M, 53 F)||1 d to 23 y (13 neonates)||mD||NR||y = −6.1 + 26.8 × (BSA) 1/3|
|Oberhansli et al.||53 (27 M, 26 F)||1 d to 1 y||NR||Yes||y = 6.4 + 18.1 × BSA |
BSA = 0.3: 11.8 (11.5–12.1)
|Ichida et al.||153 (97 M, 76 F)||1 d to 15 y||BT||Yes||y = 0.084 + 1.944 × (BSA) 1/2 |
BSA = 0.3: 11.5 (7.5–15.5)
|Hanséus et al.||120 (58 M, 62 F)||3 d to 15.5 y||NR||Yes||ln( y ) = 0.232 + 0.455 × ln(BSA) |
BSA = 0.3: 7.3 (4.9–10.9)
|Roman et al.||52 (28 M, 24 F)||1 mo to 15 y||NR||Yes||y = 1.02 + 0.98 × BSA |
BSA = 0.3: 13.1 (4.7–21.6)
|y = 0.87 + 0.80 × BSA |
BSA = 0.3: 11.1 (2.9–19.3)
|Huwez et al.||127 (77 M, 50 F)||7 mo to 19.5 y||D||NR||y = 0.62119 + 0.8248 × BSA||y = 1.05 + 0.8522 × BSA|
|Daubeney et al.||125 (69 M, 56 F)||0 d to 17 y||B||Yes||ln( y ) = 0.5183 + 0.5347 × ln(BSA) |
BSA = 0.3: 8.8 (7.7–10.1)
|ln( y ) = 0.7224 + 0.5082 × ln(BSA) |
BSA = 0.3: 11.2 (9.7–12.9)
|ln( y ) = 0.5417 + 0.5490 × ln(BSA) |
BSA = 0.3: 8.9 (7.5–10.6)
|Kampmann et al.||2,036||0–18 y (450 neonates)||D||NR||y = 19.443 × BSA 0.4578 |
BSA = 0.3: 11.3 (9.0–13.6)
|Poutanen et al.||168 (85 M, 83 F)||2–27 y||D||NR||y = 0.896 + 0.716 × BSA||y = 1.207 + 0.902 × BSA||y = 0.949 + 0.757 × BSA|
|Sluysmans et al.||496||0 d to 20 y||H||NR||y = 0.015 + 1.51 × (BSA) 1/2||y = −0.002 + 2.003 × (BSA) 1/2||y = −0.029 + 1.723 × (BSA) 1/2|
|Kaldararova et al.||702 (365 M, 337 F)||0-20 y (60 neonates)||Mo||NR||y = 6.46 + 8.76 × BSA + 2.23 × (BSA − 0.89) 2 |
BSA = 0.3: 8.8 (6.8–10.6)
|y = 8.28 + 11.49 × BSA-2.92 × (BSA − 0.89) 2 |
BSA = 0.3: 10.8 (7.9–13.7)
|y = 7.02 + 9.87 × BSA + 3.12 × (BSA − 0.89) 2 |
BSA = 0.3: 9.0 (6.4–11.5)
|Pettersen et al.||782||1 d to 18 y (82 neonates)||NR||Yes||ln( y ) = −0.874 + 2.708 × BSA − 1.841 × BSA 2 + 0.452 × BSA 3 |
BSA = 0.3: 8.1 (6.6–9.9)
|ln( y ) = −0.5 + 2.537 × BSA − 1.707 × BSA 2 + 0.42 × BSA 3 |
BSA = 0.3: 11.3 (9.0–14.0)
|ln( y ) = −0.759 + 2.643 × BSA − 1.797 × BSA 2 + 0.442 × BSA 3 |
BSA = 0.3: 8.9 (6.8–11.6)
|Gautier et al.||194 M||2–18 y||D||Yes||ln( y ) = 2.78 + 0.47 × ln(BSA) |
BSA = 0.3: 9.2 (7.5–11.2)
|ln( y ) = 3.1 + 0.49 × ln(BSA) |
BSA = 0.3: 12.3 (10.1–15.0)
|ln( y ) = 2.9 + 0.47 × ln(BSA) |
BSA = 0.3: 10.3 (8.4–12.6)
|159 F||2–18 y||ln( y ) = 2.75 + 0.44 × ln(BSA) |
BSA = 0.3: 9.2 (7.5–11.3)
|ln( y ) = 3.1 + 0.44 × ln(BSA) |
BSA = 0.3: 13.1 (10.9–15.7)
|ln( y ) = 2.9 + 0.42 × ln(BSA) |
BSA = 0.3: 10.9 (9.2–13.1)
Most of the limitations previously discussed were encountered also for these nomograms. These included the use of different regression equations (11 linear, three nonlinear), the lack of MSE in six works, the use of six different formulas to calculate BSA (when given), the lack of differentiation among sex and race subgroups, limited numbers of patients, and methodologic differences in data acquisition.
According to the latest American Society of Echocardiography recommendations, aortic diameters should be measured from one intraluminar inner edge to the opposite inner intraluminar edge perpendicular to the long axis of the vessel at the moment of maximum expansion (i.e., in midsystole), because vessel dimensions in systole are larger than in diastole.
Thus, the choice of cardiac cycle may represent a limitation if the cardiac cycle is not indicated or if different cycles are used for aortic root and aortic annular measurements.
Scattered data have been collected for the ascending aorta and the sinotubular junction.
Limited information is available on the aortic arch, accounting for 836 healthy children on the whole, from neonatal age to adulthood (with no specification of the number of neonates evaluated in every single study). According to current guidelines, multiple measurements of the aortic arch should be performed at different levels: the proximal transverse arch (between the innominate and left common carotid arteries), the distal transverse arch (between the left common carotid and left subclavian arteries), and the aortic isthmus (narrowest aortic segment distal to the left subclavian artery). Complete evaluations of the aortic arch, however, were obtained in only two studies, while others performed single or two-site measurements. As a result, the number of healthy subjects for every diameter of the aortic arch is extremely limited ( Table 5 ).
|Study||n||Age||BSA||MSE||Transverse Arch Between IA and LCA||Transverse Arch Between LCA and LSA||Arch After LSA|
|Snider et al.||110||1–18 y (10 neonates)||B||Yes||y = −0.0356 + 1.343 × (BSA) 1/2|
|Hanséus et al.||120 (58 M, 62 F)||3 d to 15.5 y||NR||Yes||ln( y ) = 0.205 + 0.349 × ln(BSA) |
BSA = 0.3: 8.1 (5.2–12.6)
|Poutanen et al.||168 (85 M, 83 F)||2–27 y||D||NR||y = 0.854 + 0.681 × BSA||y = 0.751 + 0.613 × BSA||y = 0.646 + 0.550 × BSA|
|Sluysmans et al.||496||0 d to 20 y||H||NR||y = −0.032 + 1.533 × (BSA) 1/2||y = 0.089 + 1.134 × (BSA) 1/2|
|Pettersen et al.||782||1 day to 18 y (82 neonates)||NR||Yes||ln( y ) = −0.790 + 3.020 × BSA − 2.484 × BSA 2 + 0.712 × BSA 3 |
BSA = 0.3: 9.2 (6.8–12.4)
|ln( y ) = −1.072 + 2.539 × BSA − 1.627 × BSA 2 + 0.368 × BSA 3 |
BSA = 0.3: 6.4 (4.6–8.9)
|ln( y ) = −0.976 × 2.469 × BSA − 1.746 × BSA 2 + 0.445 × BSA 3 |
BSA = 0.3: 6.8 (5.0–9.4)