What is the PRIMaCY score?

The PRIMaCY score (Precision Medicine for Cardiomyopathy) is a validated risk prediction tool that estimates the 5-year probability of a sudden cardiac death (SCD) event in children and adolescents with hypertrophic cardiomyopathy (HCM). It was developed from an international multicenter observational cohort led by researchers at the Hospital for Sick Children (Toronto) and published by Miron and colleagues in Circulation (2020).

Unlike adult HCM risk calculators (such as the ESC HCM Risk-SCD score for patients older than 16 years), PRIMaCY was built specifically for pediatric phenotype-positive HCM and incorporates risk factor relationships that differ from adult data, including nonlinear effects of wall thickness and an unexpected inverse association between very high left ventricular outflow tract (LVOT) gradients and SCD risk.

The tool outputs an individualized 5-year SCD event risk (expressed as a percentage) and supports shared decision-making about primary prevention implantable cardioverter-defibrillator (ICD) implantation. It does not, by itself, mandate ICD placement or replace specialist judgment.

Clinical context: SCD in pediatric hypertrophic cardiomyopathy

HCM is the most common cardiomyopathy in children and a leading cause of SCD in adolescents and young adults. SCD in HCM is typically due to ventricular fibrillation or sustained ventricular tachycardia (VT). ICDs can terminate lethal arrhythmias and are offered for secondary prevention after cardiac arrest or sustained VT, or for primary prevention when future SCD risk is judged high enough to outweigh device-related harms.

In children, ICD decision-making is especially difficult. Pediatric patients have higher rates of device complications (lead fractures, inappropriate shocks, infection) than adults, and traditional guideline-based risk factors are often applied inconsistently. Many children receive ICDs based on a single risk marker, while others with substantial risk remain unprotected. In the PRIMaCY derivation cohort, the 5-year cumulative proportion of SCD events was 9.1%, yet only a minority of events occurred in patients who had already received a primary prevention ICD at the time of the event.

PRIMaCY was created to provide a quantitative, individualized 5-year risk estimate that integrates multiple age-appropriate predictors rather than relying on binary single-factor thresholds.

Patient population and eligibility

Use PRIMaCY in patients who meet the original study inclusion criteria:

• Age under 18 years at the time of HCM diagnosis.
• Primary (isolated) HCM with echocardiographic evidence of left ventricular hypertrophy (typically maximal wall thickness z-score ≥ 2 using Boston z-scores).
• Phenotype-positive at assessment (genotype-positive patients without hypertrophy were excluded).
• No SCD event before diagnosis.

Do not apply the score in patients with known or suspected secondary causes of hypertrophy (RASopathies such as Noonan syndrome, metabolic or mitochondrial disorders, neuromuscular disease, hypertension, or structural heart defects), or in adults diagnosed after age 18. The score is intended for use at or near diagnosis when baseline clinical, echocardiographic, and (when available) genetic data are collected.

Outcome predicted

The primary outcome is a composite SCD event within 5 years of diagnosis, defined as:

• Sudden cardiac death
• Resuscitated sudden cardiac arrest (including sustained VT or ventricular fibrillation)
• Aborted SCD: appropriate ICD shock in a patient who received a primary prevention ICD

Non-SCD death was treated as a competing risk in model development. Administrative censoring at 5 years was applied so predictions reflect clinically relevant short-term risk for ICD counseling at diagnosis.

Calculator inputs

The PRIMaCY calculator requires the following variables, entered at diagnosis (or the index assessment used for risk stratification):

Anthropometrics

• Age (years)
• Height (cm)
• Weight (kg)

Height and weight are used with echocardiographic dimensions to compute Boston z-scores for septal thickness, posterior wall thickness, and left atrial (LA) size, standardizing measurements for body size in growing children.

Echocardiographic measurements

• Maximum interventricular septal diameter (IVSD) (mm)
• Maximum left ventricular posterior wall diameter (LVPWD) (mm)
• Left atrial diameter at end systole (mm), measured in parasternal long-axis view at end systole (standardized measurement protocol from the study should be followed for consistency)
• Peak resting LVOT gradient (mmHg)

Clinical history (prior 6 months)

• Nonsustained ventricular tachycardia (NSVT): defined as ≥ 3 consecutive ventricular ectopic beats at a rate > 120 bpm and duration < 30 seconds on ambulatory ECG monitoring
• Unexplained syncope: syncope without an identified non-cardiac cause within 6 months before diagnosis

Genetic testing (optional enhanced model)

• Genotype status: whether the patient has undergone genetic testing. When testing has been performed, the clinical/genetic model incorporates presence of a pathogenic or likely pathogenic variant in an HCM-associated gene compared with confirmed absence of a causal variant. Patients not tested are handled per the calculator implementation (clinical-only model or imputed approach as specified in your tool).

How the score is calculated

PRIMaCY is not a simple integer point score. It is derived from competing-risk cause-specific hazard regression implemented with a gradient boosting algorithm that models nonlinear relationships between predictors and 5-year SCD risk. Continuous variables (age, IVSD z-score, LVPWD z-score, LA diameter z-score, peak LVOT gradient) are not assumed to have linear effects across their full range.

Two related models are reported in the publication:

• Clinical model (c-statistic 0.75 in derivation): age, IVSD z-score, LVPWD z-score, LA diameter z-score, peak LVOT gradient, NSVT, unexplained syncope.
• Clinical/genetic model (c-statistic 0.76 in derivation): all clinical predictors plus pathogenic/likely pathogenic genotype status when testing has been completed.

The calculator converts raw echocardiographic millimeter measurements to Boston z-scores using patient height and weight, then applies the published algorithm to estimate each patient’s 5-year cumulative probability of an SCD event.

Key predictors and pediatric-specific findings

Categorical predictors

Variable	Association with 5-year SCD events
NSVT (prior 6 months)	Approximately 2.6 to 2.9-fold higher hazard in derivation models
Unexplained syncope (prior 6 months)	Approximately 7.2 to 7.4-fold higher hazard; among the strongest categorical predictors
Pathogenic/likely pathogenic variant (clinical/genetic model)	Approximately 1.3-fold higher hazard versus confirmed genotype-negative on testing
Family history of SCD	Not an independent predictor in the pediatric cohort (unlike many adult frameworks)

Continuous predictors (nonlinear relationships)

• Age at diagnosis: SCD risk increases with older age at diagnosis; adolescent and teenage years remain the highest-risk period for events regardless of whether HCM was diagnosed in early childhood.
• IVSD and LVPWD z-scores: Risk rises with greater wall thickness but plateaus at very high z-scores (approximately > 20), suggesting massive hypertrophy alone does not confer unbounded incremental SCD risk in children the way it may in adults.
• LA diameter z-score: Larger LA size associates with higher SCD risk.
• Peak resting LVOT gradient: Risk is relatively flat for gradients ≤ 100 mmHg. Unlike adult HCM models, higher gradients above 100 mmHg show an inverse association with SCD risk in this pediatric cohort (for example, an increase from 100 to 120 mmHg was associated with roughly 10% lower predicted risk in the published analysis). This finding may reflect confounding by surgical myectomy, spectrum bias, or pediatric-specific physiology and should not be interpreted as implying that severe obstruction is protective.

Risk stratification and interpretation

PRIMaCY outputs a continuous 5-year SCD event probability. In the derivation study, patients were stratified into three risk groups by tertiles of predicted risk for calibration assessment. These tertile cutoffs are useful for contextualizing results:

Predicted 5-year SCD event risk	Risk group (derivation tertiles)	Clinical framing
< 4.7%	Lower tertile	Lowest observed event rates in derivation and validation cohorts; primary prevention ICD generally less favored if ischemic or arrhythmic markers are absent, though follow-up remains essential.
4.7% to 8.3%	Middle tertile	Intermediate risk; individualized discussion weighing ICD benefits against pediatric device complications.
> 8.3%	Upper tertile	Highest observed event rates; stronger consideration for primary prevention ICD when aligned with guideline frameworks and family values, especially if additional concerning features are present.

The graded risk output should be reviewed alongside the absolute percentage rather than tertile label alone. A patient at 7% predicted 5-year risk and a patient at 15% predicted risk both warrant careful counseling, but the magnitude of expected benefit from ICD may differ substantially.

Model performance and validation

In the derivation cohort of 572 patients (2,855 patient-years of follow-up):

• Clinical model c-statistic 0.75
• Clinical/genetic model c-statistic 0.76
• Good calibration between predicted and observed 5-year event rates across risk tertiles

External validation in the independent SHaRe cohort (285 patients) demonstrated:

• Clinical model c-statistic 0.71
• Clinical/genetic model c-statistic 0.72
• Acceptable agreement between predicted and observed events

PRIMaCY is the first pediatric HCM SCD prediction model with published external validation in an independent cohort and discriminatory performance comparable to the adult ESC HCM Risk-SCD calculator used in clinical practice.

ICD decision-making and shared decision-making

American College of Cardiology / American Heart Association guidelines emphasize that ICD decisions in HCM should incorporate individual clinical judgment and informed shared decision-making regarding the strength of evidence, benefits, and risks. PRIMaCY supports that process by translating multiple variables into a single estimated 5-year risk rather than treating each risk factor as an automatic ICD indication.

Important principles when using PRIMaCY at the bedside:

• Do not base ICD decisions on a single risk factor. The model was designed because binary single-marker approaches perform poorly in children.
• The score does not provide treatment recommendations by itself. Combine PRIMaCY output with electrophysiology consultation, holter findings, exercise testing where indicated, family preferences, and institutional experience.
• Secondary prevention ICD remains indicated after survived cardiac arrest or hemodynamically significant sustained VT, independent of calculated primary prevention risk.
• Reassess over time. HCM phenotype evolves during growth; a low risk at young diagnosis may change with new syncope, NSVT, wall thickening, or LA enlargement on follow-up echocardiography.
• Pediatric ICD complications (inappropriate shocks, lead failure, psychosocial impact) must be weighed explicitly in every conversation, especially in small children and pre-teen patients.

Comparison with adult risk tools

The ESC HCM Risk-SCD calculator applies to patients over 16 years of age and uses a different variable set (including family history of SCD, which was not predictive in the pediatric PRIMaCY cohort). PRIMaCY should be used for childhood-onset HCM risk assessment at diagnosis in eligible pediatric patients; adult tools should not be extrapolated to young children without understanding these differences.

Practical measurement and data quality tips

• Use the Boston z-score methodology consistently for wall thickness and LA size; PRIMaCY was trained on z-scores, not raw millimeter values alone.
• Measure LA diameter at end systole in parasternal long-axis view per the study protocol.
• Document whether NSVT and syncope occurred within the defined 6-month window before diagnosis.
• Record peak resting LVOT gradient (not provoked gradient alone) unless your institutional protocol specifies otherwise.
• When genetic testing is available, distinguish pathogenic/likely pathogenic variants from variants of uncertain significance, which are not treated as genotype-positive in the clinical/genetic model.
• Missing echocardiographic data were imputed in the original study (LA diameter in 21%, LVOT gradient in 29%); prefer complete real-world measurements when possible rather than relying on imputation.

Important limitations

• Retrospective design: Subject to missing data, survivor bias, and incomplete family SCD history.
• Genetic testing uptake: Only about half the derivation cohort had genetic testing; the incremental value of the genetic model may increase as testing becomes universal.
• Does not include cardiac MRI late gadolinium enhancement or other emerging markers, which may add prognostic information in selected patients.
• Excludes syndromic and metabolic HCM; results do not apply to those populations.
• LVOT gradient paradox: The inverse association at very high gradients may not be causal; do not withhold appropriate obstruction management based on gradient alone.
• Dynamic childhood phenotype: A single time-point prediction at diagnosis may underestimate future risk if the phenotype progresses; serial reassessment is important.
• Appropriate ICD shocks as events: Including aborted SCD in the outcome composite may slightly overestimate “fatal” risk in patients already implanted, though this reflects the study endpoint definition.

Documentation and workflow tips

When using PRIMaCY in practice, document all input variables (age, height, weight, echocardiographic measurements with z-scores if calculated, NSVT, syncope, genetic result), the predicted 5-year SCD event percentage, which model was used (clinical vs clinical/genetic), and the content of the shared decision-making discussion regarding ICD. Note competing considerations (exercise restrictions, medical therapy, family screening) that extend beyond SCD risk alone.

Using this CalcMD calculator

Enter patient age, height, weight, echocardiographic IVSD, LVPWD, LA diameter at end systole, and peak LVOT gradient, plus histories of NSVT and unexplained syncope in the prior 6 months, and genotype status when known. The tool computes Boston z-scores, applies the validated PRIMaCY algorithm, and displays the estimated 5-year SCD event risk with risk-group context to support education and shared decision-making regarding primary prevention ICD.

Use the output alongside pediatric cardiology and electrophysiology expertise. It does not replace clinical judgment, guideline recommendations, or emergency management of hemodynamically unstable arrhythmias.