Prognostic role of BNP in children undergoing surgery for congenital heart disease

Analysis of prediction models incorporating standard risk factors

Massimiliano Cantinotti, Raffaele Giordano, Marco Scalese, Sabrina Molinaro, Francesca Della Pina, Simona Storti, Luigi Arcieri, Bruno Murzi, Marco Marotta, Vitali Pak, Vincenzo Poli, Giorgio Iervasi, Shelby Kutty, Aldo Clerico

Research output: Contribution to journalArticle

14 Citations (Scopus)

Abstract

Background: The routine use of brain natriuretic peptide (BNP) in pediatric cardiac surgery remains controversial. Our aim was to test whether BNP adds information to predict risk in pediatric cardiac surgery. Methods: In all, 587 children undergoing cardiac surgery (median age 6.3 months; 1.2-35.9 months) were prospectively enrolled at a single institution. BNP was measured pre-operatively, on every post-operative day in the intensive care unit, and before discharge. The primary outcome was major complications and length ventilator stay >15 days. A first risk prediction model was fitted using Cox proportional hazards model with age, body surface area and Aristotle score as continuous predictors. A second model was built adding cardiopulmonary bypass time and arterial lactate at the end of operation to the first model. Then, peak post-operative log-BNP was added to both models. Analysis to test discrimination, calibration, and reclassification were performed. Results: BNP increased after surgery (p<0.001), peaking at a mean of 63.7 h (median 36 h, interquartile range 12-84 h) post-operatively and decreased thereafter. The hazard ratios (HR) for peak-BNP were highly significant (first model HR=1.40, p=0.006, second model HR=1.44, p=0.008), and the log-likelihood improved with the addition of BNP at 12 h (p=0.006; p=0.009). The adjunction of peak-BNP significantly improved the area under the ROC curve (first model p<0.001; second model p<0.001). The adjunction of peak-BNP also resulted in a net gain in reclassification proportion (first model NRI=0.089, p<0.001; second model NRI=0.139, p=0.003). Conclusions: Our data indicates that BNP may improve the risk prediction in pediatric cardiac surgery, supporting its routine use in this setting.

Original languageEnglish (US)
Pages (from-to)1839-1846
Number of pages8
JournalClinical Chemistry and Laboratory Medicine
Volume53
Issue number11
DOIs
StatePublished - Oct 1 2015

Fingerprint

Brain Natriuretic Peptide
Surgery
Heart Diseases
Thoracic Surgery
Pediatrics
Hazards
Proportional Hazards Models
Intensive care units
Body Surface Area
Mechanical Ventilators
Cardiopulmonary Bypass
ROC Curve
Calibration
Area Under Curve
Intensive Care Units
Lactic Acid
Length of Stay

Keywords

  • brain natriuretic peptide (BNP)
  • congenital heart disease
  • pediatric cardiac surgery

ASJC Scopus subject areas

  • Clinical Biochemistry
  • Biochemistry, medical

Cite this

Prognostic role of BNP in children undergoing surgery for congenital heart disease : Analysis of prediction models incorporating standard risk factors. / Cantinotti, Massimiliano; Giordano, Raffaele; Scalese, Marco; Molinaro, Sabrina; Della Pina, Francesca; Storti, Simona; Arcieri, Luigi; Murzi, Bruno; Marotta, Marco; Pak, Vitali; Poli, Vincenzo; Iervasi, Giorgio; Kutty, Shelby; Clerico, Aldo.

In: Clinical Chemistry and Laboratory Medicine, Vol. 53, No. 11, 01.10.2015, p. 1839-1846.

Research output: Contribution to journalArticle

Cantinotti, M, Giordano, R, Scalese, M, Molinaro, S, Della Pina, F, Storti, S, Arcieri, L, Murzi, B, Marotta, M, Pak, V, Poli, V, Iervasi, G, Kutty, S & Clerico, A 2015, 'Prognostic role of BNP in children undergoing surgery for congenital heart disease: Analysis of prediction models incorporating standard risk factors', Clinical Chemistry and Laboratory Medicine, vol. 53, no. 11, pp. 1839-1846. https://doi.org/10.1515/cclm-2014-1084
Cantinotti, Massimiliano ; Giordano, Raffaele ; Scalese, Marco ; Molinaro, Sabrina ; Della Pina, Francesca ; Storti, Simona ; Arcieri, Luigi ; Murzi, Bruno ; Marotta, Marco ; Pak, Vitali ; Poli, Vincenzo ; Iervasi, Giorgio ; Kutty, Shelby ; Clerico, Aldo. / Prognostic role of BNP in children undergoing surgery for congenital heart disease : Analysis of prediction models incorporating standard risk factors. In: Clinical Chemistry and Laboratory Medicine. 2015 ; Vol. 53, No. 11. pp. 1839-1846.
@article{cf0c008052234ea8aeb1ac251d6e1c65,
title = "Prognostic role of BNP in children undergoing surgery for congenital heart disease: Analysis of prediction models incorporating standard risk factors",
abstract = "Background: The routine use of brain natriuretic peptide (BNP) in pediatric cardiac surgery remains controversial. Our aim was to test whether BNP adds information to predict risk in pediatric cardiac surgery. Methods: In all, 587 children undergoing cardiac surgery (median age 6.3 months; 1.2-35.9 months) were prospectively enrolled at a single institution. BNP was measured pre-operatively, on every post-operative day in the intensive care unit, and before discharge. The primary outcome was major complications and length ventilator stay >15 days. A first risk prediction model was fitted using Cox proportional hazards model with age, body surface area and Aristotle score as continuous predictors. A second model was built adding cardiopulmonary bypass time and arterial lactate at the end of operation to the first model. Then, peak post-operative log-BNP was added to both models. Analysis to test discrimination, calibration, and reclassification were performed. Results: BNP increased after surgery (p<0.001), peaking at a mean of 63.7 h (median 36 h, interquartile range 12-84 h) post-operatively and decreased thereafter. The hazard ratios (HR) for peak-BNP were highly significant (first model HR=1.40, p=0.006, second model HR=1.44, p=0.008), and the log-likelihood improved with the addition of BNP at 12 h (p=0.006; p=0.009). The adjunction of peak-BNP significantly improved the area under the ROC curve (first model p<0.001; second model p<0.001). The adjunction of peak-BNP also resulted in a net gain in reclassification proportion (first model NRI=0.089, p<0.001; second model NRI=0.139, p=0.003). Conclusions: Our data indicates that BNP may improve the risk prediction in pediatric cardiac surgery, supporting its routine use in this setting.",
keywords = "brain natriuretic peptide (BNP), congenital heart disease, pediatric cardiac surgery",
author = "Massimiliano Cantinotti and Raffaele Giordano and Marco Scalese and Sabrina Molinaro and {Della Pina}, Francesca and Simona Storti and Luigi Arcieri and Bruno Murzi and Marco Marotta and Vitali Pak and Vincenzo Poli and Giorgio Iervasi and Shelby Kutty and Aldo Clerico",
year = "2015",
month = "10",
day = "1",
doi = "10.1515/cclm-2014-1084",
language = "English (US)",
volume = "53",
pages = "1839--1846",
journal = "Clinical Chemistry and Laboratory Medicine",
issn = "1434-6621",
publisher = "Walter de Gruyter GmbH & Co. KG",
number = "11",

}

TY - JOUR

T1 - Prognostic role of BNP in children undergoing surgery for congenital heart disease

T2 - Analysis of prediction models incorporating standard risk factors

AU - Cantinotti, Massimiliano

AU - Giordano, Raffaele

AU - Scalese, Marco

AU - Molinaro, Sabrina

AU - Della Pina, Francesca

AU - Storti, Simona

AU - Arcieri, Luigi

AU - Murzi, Bruno

AU - Marotta, Marco

AU - Pak, Vitali

AU - Poli, Vincenzo

AU - Iervasi, Giorgio

AU - Kutty, Shelby

AU - Clerico, Aldo

PY - 2015/10/1

Y1 - 2015/10/1

N2 - Background: The routine use of brain natriuretic peptide (BNP) in pediatric cardiac surgery remains controversial. Our aim was to test whether BNP adds information to predict risk in pediatric cardiac surgery. Methods: In all, 587 children undergoing cardiac surgery (median age 6.3 months; 1.2-35.9 months) were prospectively enrolled at a single institution. BNP was measured pre-operatively, on every post-operative day in the intensive care unit, and before discharge. The primary outcome was major complications and length ventilator stay >15 days. A first risk prediction model was fitted using Cox proportional hazards model with age, body surface area and Aristotle score as continuous predictors. A second model was built adding cardiopulmonary bypass time and arterial lactate at the end of operation to the first model. Then, peak post-operative log-BNP was added to both models. Analysis to test discrimination, calibration, and reclassification were performed. Results: BNP increased after surgery (p<0.001), peaking at a mean of 63.7 h (median 36 h, interquartile range 12-84 h) post-operatively and decreased thereafter. The hazard ratios (HR) for peak-BNP were highly significant (first model HR=1.40, p=0.006, second model HR=1.44, p=0.008), and the log-likelihood improved with the addition of BNP at 12 h (p=0.006; p=0.009). The adjunction of peak-BNP significantly improved the area under the ROC curve (first model p<0.001; second model p<0.001). The adjunction of peak-BNP also resulted in a net gain in reclassification proportion (first model NRI=0.089, p<0.001; second model NRI=0.139, p=0.003). Conclusions: Our data indicates that BNP may improve the risk prediction in pediatric cardiac surgery, supporting its routine use in this setting.

AB - Background: The routine use of brain natriuretic peptide (BNP) in pediatric cardiac surgery remains controversial. Our aim was to test whether BNP adds information to predict risk in pediatric cardiac surgery. Methods: In all, 587 children undergoing cardiac surgery (median age 6.3 months; 1.2-35.9 months) were prospectively enrolled at a single institution. BNP was measured pre-operatively, on every post-operative day in the intensive care unit, and before discharge. The primary outcome was major complications and length ventilator stay >15 days. A first risk prediction model was fitted using Cox proportional hazards model with age, body surface area and Aristotle score as continuous predictors. A second model was built adding cardiopulmonary bypass time and arterial lactate at the end of operation to the first model. Then, peak post-operative log-BNP was added to both models. Analysis to test discrimination, calibration, and reclassification were performed. Results: BNP increased after surgery (p<0.001), peaking at a mean of 63.7 h (median 36 h, interquartile range 12-84 h) post-operatively and decreased thereafter. The hazard ratios (HR) for peak-BNP were highly significant (first model HR=1.40, p=0.006, second model HR=1.44, p=0.008), and the log-likelihood improved with the addition of BNP at 12 h (p=0.006; p=0.009). The adjunction of peak-BNP significantly improved the area under the ROC curve (first model p<0.001; second model p<0.001). The adjunction of peak-BNP also resulted in a net gain in reclassification proportion (first model NRI=0.089, p<0.001; second model NRI=0.139, p=0.003). Conclusions: Our data indicates that BNP may improve the risk prediction in pediatric cardiac surgery, supporting its routine use in this setting.

KW - brain natriuretic peptide (BNP)

KW - congenital heart disease

KW - pediatric cardiac surgery

UR - http://www.scopus.com/inward/record.url?scp=84944055654&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84944055654&partnerID=8YFLogxK

U2 - 10.1515/cclm-2014-1084

DO - 10.1515/cclm-2014-1084

M3 - Article

VL - 53

SP - 1839

EP - 1846

JO - Clinical Chemistry and Laboratory Medicine

JF - Clinical Chemistry and Laboratory Medicine

SN - 1434-6621

IS - 11

ER -