Coronary z scores from four published models side by side, the AHA 2024 classification and risk level with its long-term management, a model-disagreement flag, and the Son 2019 aneurysm risk score
The main level is the classification of the maximum z score ever recorded in any branch; the sub-level comes from the current study (same classification: current or persistent; lower: regressed). Classification on Boston z, as in the AHA statements. Table 2 applies once coronary dimensions are stable or improving; it gives no class of recommendation or level of evidence.
Each cell shows the z score and its AHA classification for that artery. A ≠ in a column header means the models classify that artery differently. The highlighted Boston row drives the classification and risk level above. Rows marked [unverified] use coefficients that could not be checked against the primary paper; they are shown for comparison only.
| Model | LMCA | LAD | RCA | LCx | Most severe | BSA |
|---|---|---|---|---|---|---|
| Enter the patient and at least one diameter. | ||||||
| Classification | z score (normalized to BSA) |
|---|---|
| 1No involvement | Always <2 |
| 2Dilation only | 2 to <2.5 |
| 3Small aneurysm | ≥2.5 to <5 |
| 4Medium aneurysm | ≥5 to <10, and absolute dimension <8 mm |
| 5Large or giant aneurysm | ≥10, or absolute dimension ≥8 mm |
AHA 2024 Table 1 (unchanged from the 2017 statement). Each artery is classified on its unrounded z and its internal diameter, then the most severe artery is reported. The 8 mm rule is applied as written, whatever the z score.
| Model | Coefficients taken from | Status |
|---|---|---|
| PHN (Lopez 2017) | Primary paper, Table 2 coronary rows and the Z equation in Results (PMC author manuscript) | Verified |
| Kobayashi 2016 LMS | The authors' official Z Score Project calculator, version 4 (named in the paper's abstract). The paper's own tables were not retrieved. | Verified against the official calculator |
| Boston (McCrindle 2007) | Primary paper, Methods (p. 175): every coefficient matches, with Haycock BSA. The ZSP calculator v4 (mm form) and the NCT00841789 trial protocol (cm form) give the same digits. | Checked against the primary paper |
| Dallaire & Dahdah 2011 | ZSP calculator v4 only (one secondary implementation). Dallaire 2011 full text could not be retrieved. | [unverified] |
| AHA z-score classification | AHA 2024 Table 1 (corrected full text); the same thresholds as the 2017 statement (its z-score classification list and Table 8, read in full). | Read in the primary documents |
| AHA 2024 risk levels and management | AHA 2024 Table 2 and its footnotes, read from the corrected full text (corrections of February and March 2025). | Read in the primary document |
| Son 2019 score | Primary paper (PMC6585355): Results and Table 4 | Verified |
All 24 LAD z scores in Kim 2022 Table 5 (two boys, four models, three diameters each, from the ZSP calculator) are reproduced exactly to two decimals by this page, and the PHN worked example in Lopez 2017 (mitral diameter 11 mm at BSA 0.3 m² gives Z −1.0) is reproduced by the same allometric function.
BSA [m2] = 0.024265 x weight[kg]^0.5378 x height[cm]^0.3964 (Haycock 1978)
Lopez 2017 states PHN used Haycock. Kim 2022 Table 2 lists Haycock for McCrindle, Dallaire and Kobayashi; the ZSP calculator applies Haycock to every model. A different BSA formula shifts every z score.
Z = (D[mm] / BSA^alpha - mean) / SD
alpha mean SD
LMCA 0.45 2.95 0.57
LAD 0.45 1.90 0.34
RCA 0.45 2.32 0.55 (no LCx row)
Healthy non-obese children ≤18 years from 19 North American centres; coronary arteries were an optional measurement set (available in 78–91% of studies). Robinson 2021 found PHN LAD z scores systematically higher than Boston (median 1.6 vs 0.3), so switching from Boston to PHN reclassified at least one artery in 59.7% of children and changed the recommended antithrombotic strategy in 13.4%.
Z = ((D / M)^L - 1) / (L x S) (L -> 0: Z = ln(D / M) / S) M(b) = c6 b^6 + c5 b^5 + c4 b^4 + c3 b^3 + c2 b^2 + c1 b + c0 Male M: c6..c0 L S RCA 0.0183 -0.1867 0.7132 -1.1982 0.6038 1.6558 0.7517 0.292 0.174 LMCA 0.076 -0.6549 2.1792 -3.3446 1.916 1.5869 0.933 0.336 0.151 LAD 0.0755 -0.5537 1.4885 -1.6204 0.155 2.0239 0.633 0.6481 b - 0.3205 S1(b) LCx 0.078 -0.5761 1.5481 -1.6499 0.0767 1.9436 0.5294 0.4536 b - 0.2542 S2(b) Female RCA 0.0363 -0.3503 1.2615 -2.0255 1.1301 1.4256 0.7727 0.607 0.17 LMCA 0.1738 -1.4226 4.5095 -6.7646 4.3972 0.662 1.0021 0.363 S3(b) LAD 0.1611 -1.2491 3.683 -4.9471 2.5568 1.1944 0.7109 0.204 0.168 LCx 0.0495 -0.4345 1.4492 -2.188 1.1905 1.1956 0.6348 0.6683 b - 0.2618 0.2 S1(b) = 0.0298 b^6 - 0.2189 b^5 + 0.5974 b^4 - 0.7135 b^3 + 0.2947 b^2 + 0.0411 b + 0.1459 S2(b) = 0.0011 b^6 - 0.0081 b^5 + 0.0239 b^4 - 0.0351 b^3 + 0.0267 b^2 - 0.0037 b + 0.2009 S3(b) = -0.0002 b^6 + 0.0014 b^5 - 0.0032 b^4 + 0.0032 b^3 + 0.0021 b^2 - 0.0367 b + 0.1819
3,851 healthy Japanese children 0 to 18 years. M is the median diameter (Z = 0). L crosses zero for male LAD (BSA 0.49 m²), male LCx (0.56) and female LCx (0.39); the page uses the logarithmic limit there. Derived in a Japanese population; z scores in other populations may differ.
Z = (D[cm] - mean) / SD LMCA mean = 0.31747 x BSA^0.36008 - 0.02887 SD = 0.03040 + 0.01514 x BSA LAD mean = 0.26108 x BSA^0.37893 - 0.02852 SD = 0.01465 + 0.01996 x BSA RCA mean = 0.26117 x BSA^0.39992 - 0.02756 SD = 0.02407 + 0.01597 x BSA (no LCx)
Normal group of 221 healthy children aged 0 to 18 years (McCrindle 2007, Methods), BSA by Haycock. The ZSP calculator gives the same equations in mm (every coefficient ×10). Both AHA statements base their z-score classification on these equations (AHA 2017, Classification of Coronary Artery Abnormalities; AHA 2024, Cardiac Imaging in Acute KD), so Boston z drives the classification and the risk level on this page.
Z = (D[mm] - (k1 x sqrt(BSA) - k0)) / (s0 + s1 x sqrt(BSA))
k0 k1 s0 s1
LMCA 0.1817 2.9238 0.1801 0.2530
LAD 0.1502 2.2672 0.1709 0.2293
RCA 0.3039 2.7521 0.1626 0.2881 (proximal RCA)
LCx 0.2716 2.3458 0.1142 0.3423
1,033 healthy children; square-root-of-BSA regression with weighted least squares (abstract). Coefficients as implemented in the ZSP calculator only.
Classification (AHA 2024 Table 1): no involvement, always z <2; dilation only, 2 to <2.5; small aneurysm, ≥2.5 to <5; medium aneurysm, ≥5 to <10 and absolute dimension <8 mm; large or giant aneurysm, ≥10 or absolute dimension ≥8 mm. Luminal dimensions, normalized to BSA as z scores; the z-score normal values describe proximal segments.
Risk level: AHA 2024 bases long-term management on the maximum coronary z score reached during the illness and keeps the current-or-regressed sub-levels of the 2017 statement (Table 2). As the 2017 statement defines them (Table 8), the main level rests on the maximal z score at any time point and in any branch, and the sub-level on the maximal z score in any branch at the current assessment: the same classification gives x.1, a lower one gives 3.2, 4.2, 4.3, 5.2, 5.3 or 5.4 as described in Table 2. Levels 1 and 2 have no sub-levels. A current classification above the maximum is refused (the maximum includes the current study), as are contradictory answers. From one echo the page asks whether this echo is the highest z so far (or for the maximum), and gives no level until that is answered. The current classification uses Boston z plus the 8 mm rule on every entered artery, LCx included (an absolute diameter needs no z equation). The page asks whether any study showed an internal diameter ≥8 mm, because such a study is large or giant whatever its z score and so sets the maximal classification.
max \ current 1 2 3 4 5 1 1 2 2 2 3 3.2 3.2 3.1 4 4.3 4.3 4.2 4.1 5 5.4 5.4 5.3 5.2 5.1
Table 2 cells condensed from the corrected full text (intervals, drugs and strength wording unchanged), with its footnotes: stress testing earlier if symptoms (*); β-blockers and statins may be considered for levels 4 and 5 (†); anticoagulation in level 5.2 (‡). Follow-up intervals apply after the patient has stable coronaries or improved dimensions at discharge. As printed, the Table 2 row headers read “Z score 2–2.5” for level 2 and “absolute dimension >8 mm” for level 5; the page classifies with Table 1 (2 to <2.5; ≥8 mm), as the 2017 statement does.
The 2025 corrections changed Table 2 at levels 5.2 (anticoagulation “may be considered”, with footnote ‡) and 5.3 (anticoagulation “not indicated”). The correction notices themselves were not read; the corrected full text was.
Superseded: AHA 2017 Tables 10 and 11. The 2017 statement graded each cell with a class of recommendation. The 2024 update changes, among others: stress-test intervals (for example 3.1 every 2–3 y in 2017, every 3–5 y in 2024; 5.2 yearly, now every 2–5 y); 4.1 antiplatelet (2017: aspirin, Class I, with dual antiplatelet therapy “may be considered”, Class IIb; 2024: aspirin plus clopidogrel); 5.1 anticoagulation adds DOACs; 5.2 anticoagulation (2017: not indicated, Class III; 2024: may be considered); 5.3 dual antiplatelet therapy (2017: not indicated, Class III; 2024: may be considered); 3.2 discharge at 5 y if stress test and coronary CTA are normal (new); β-blockers and statins may be considered across levels 4 and 5 (2017: β-blocker not indicated, Class III, at 4.1–4.3 and 5.4). The 2017 tables also covered cardiovascular risk-factor assessment and reproductive counseling, which Table 2 does not. The 2017 dilation criterion “or, if initially <2, a decrease in z during follow-up ≥1” is not in AHA 2024 Table 1 and is not assessed here.
2 points baseline max z of proximal LAD or RCA ≥2.00 (Boston z, McCrindle 2007) 1 point age at fever onset <6 months 1 point Asian race reported (sole or mixed) 1 point baseline CRP ≥13 mg/dL Low 0-1 · Moderate 2 · High 3-5
Checked in October 2026 against the official PHN Z-score calculator and the Kobayashi Z Score Project calculator (version 4, recalculated unmodified). PHN: 17 of 18 at two decimals; the one difference comes from the official page rounding BSA. Kobayashi: 21 of 24 at two decimals, largest difference 7×10−5, because the workbook uses the BSA constant 0.0242654 instead of Haycock’s 0.024265 (with that constant the two agree exactly). AHA classifications 42 of 42; Son 2019 score 6 of 6. Boston z agrees with the ZSP implementation in every case, and its coefficients were checked against the McCrindle 2007 Methods; Dallaire was checked only against that secondary implementation and stays labelled [unverified]. The AHA 2024 risk-level mapping was checked against the corrected Table 2 with simulated cases (every maximum-by-current combination and each z and 8 mm boundary), and every condensed Table 2 cell against the source for its intervals, drugs and strength wording.
z scores are computed in this browser from the published equations; nothing you enter in the calculator is stored or sent. The headline classification and the risk level use Boston z (McCrindle 2007), on which the AHA classification is based; the other models are shown for comparison. Always state the model next to a z score, and use the same z equations over time.
Decision aid only. Anticoagulation, follow-up and IVIG intensification decisions need the full clinical picture and serial imaging.