Emergency Medicine · Cardiology

Steinhart Model for Acute Heart Failure in Undifferentiated Dyspnea calculator

Enter the published formula inputs below to calculate and interpret this score.

Clinical inputs

Age

years

Allowed range: 18–110 years

Pretest probability

Allowed range: 1–100 %

The clinician’s estimate, from 1% to 100%, recorded before the NT-proBNP result.

NT-proBNP concentration

pg/mL

Allowed range: 1–100000 pg/mL

The common logarithm of this value, in pg/mL, enters the appendix equation.

Result

3 fields remaining

Anyone can run this calculator. Sign in to save, favourite, or send the result.

Primary source

Steinhart B et al. Improving the diagnosis of acute heart failure using a validated prediction model. J Am Coll Cardiol. DOI: 10.1016/j.jacc.2009.05.065. PMID: 19815122. The online appendix prints Pr(AHF) = 1 / (1 + exp(8 + 0.011×age − 5.9×pretest − 2.3×log10(NT-proBNP) + 0.82×pretest×log10(NT-proBNP))). Pretest is the clinician’s probability as a fraction. Log10 is the common logarithm of NT-proBNP in pg/mL. The printed example is age 65 years, pretest 0.5, and NT-proBNP 10,000 pg/mL: the exponent is −1.795 and the appendix states the probability is 86%. The same coefficients give 85.7539%, which displays as 85.8 and remains at or above 80%. Methods classify probability as low at or below 20%, intermediate from above 20% through below 80%, and high at or above 80%. The group follows that unrounded probability. (2009)

Scope: Adults in the emergency department with undifferentiated dyspnea. The clinician records the pretest probability before the NT-proBNP result. Creatinine above 250 µmol/L was excluded from the derivation cohort.

Last reviewed 2026-10-17

View source

Professional use only

For qualified clinicians. Independently verify inputs, calculations, source guidance, and applicability before any clinical action. This calculator does not replace clinical judgment.

Source: Steinhart B et al. Improving the diagnosis of acute heart failure using a validated prediction model. J Am Coll Cardiol. DOI: 10.1016/j.jacc.2009.05.065. PMID: 19815122. The online appendix prints Pr(AHF) = 1 / (1 + exp(8 + 0.011×age − 5.9×pretest − 2.3×log10(NT-proBNP) + 0.82×pretest×log10(NT-proBNP))). Pretest is the clinician’s probability as a fraction. Log10 is the common logarithm of NT-proBNP in pg/mL. The printed example is age 65 years, pretest 0.5, and NT-proBNP 10,000 pg/mL: the exponent is −1.795 and the appendix states the probability is 86%. The same coefficients give 85.7539%, which displays as 85.8 and remains at or above 80%. Methods classify probability as low at or below 20%, intermediate from above 20% through below 80%, and high at or above 80%. The group follows that unrounded probability. (2009)

Ruqelo is AI and can make mistakes. Check important information.

Clinical inputs

Age (years)Accepted range: 18 to 110
Pretest probability (%)Accepted range: 1 to 100
NT-proBNP concentration (pg/mL)Accepted range: 1 to 100000

Primary source

View source