Preoperative risk assessment tools for morbidity after cardiac surgery: A systematic review

Julie Sanders; Nicole Makariou; Adam Tocock; Rosalie Magboo; Ashley Thomas; Leanne M. Aitken

doi:10.1093/eurjcn/zvac003

Preoperative risk assessment tools for morbidity after cardiac surgery: A systematic review

Julie Sanders^*, Nicole Makariou, Adam Tocock, Rosalie Magboo, Ashley Thomas, Leanne M. Aitken

^*Corresponding author for this work

Digital Health & Applied Tech Assessment

Research output: Contribution to journal › Review article › peer-review

5 Citations (Scopus)

Abstract

Background: Postoperative morbidity places considerable burden on health and resources. Thus, strategies to identify, predict, and reduce postoperative morbidity are needed. Aims: To identify and explore existing preoperative risk assessment tools for morbidity after cardiac surgery. Methods: Electronic databases (including MEDLINE, CINAHL, and Embase) were searched to December 2020 for preoperative risk assessment models for morbidity after adult cardiac surgery. Models exploring one isolated postoperative morbidity and those in patients having heart transplantation or congenital surgery were excluded. Data extraction and quality assessments were undertaken by two authors. Results: From 2251 identified papers, 22 models were found. The majority (54.5%) were developed in the USA or Canada, defined morbidity outcome within the in-hospital period (90.9%), and focused on major morbidity. Considerable variation in morbidity definition was identified, with morbidity incidence between 4.3% and 52%. The majority (45.5%) defined morbidity and mortality separately but combined them to develop one model, while seven studies (33.3%) constructed a morbidity-specific model. Models contained between 5 and 50 variables. Commonly included variables were age, emergency surgery, left ventricular dysfunction, and reoperation/previous cardiac surgery, although definition differences across studies were observed. All models demonstrated at least reasonable discriminatory power [area under the receiver operating curve (0.61-0.82)]. Conclusion: Despite the methodological heterogeneity across models, all demonstrated at least reasonable discriminatory power and could be implemented depending on local preferences. Future strategies to identify, predict, and reduce morbidity after cardiac surgery should consider the ageing population and those with minor and/or multiple complex morbidities.

Original language	English
Pages (from-to)	655-664
Number of pages	10
Journal	European Journal of Cardiovascular Nursing
Volume	21
Issue number	7
DOIs	https://doi.org/10.1093/eurjcn/zvac003
Publication status	Published - 1 Sept 2022

Keywords

Cardiac surgery
Morbidity outcome
Postoperative morbidity
Preoperative risk
Risk prediction models

Access to Document

10.1093/eurjcn/zvac003

Cite this

@article{bf01d78f15114714a9f94464a2bd2c5f,

title = "Preoperative risk assessment tools for morbidity after cardiac surgery: A systematic review",

abstract = "Background: Postoperative morbidity places considerable burden on health and resources. Thus, strategies to identify, predict, and reduce postoperative morbidity are needed. Aims: To identify and explore existing preoperative risk assessment tools for morbidity after cardiac surgery. Methods: Electronic databases (including MEDLINE, CINAHL, and Embase) were searched to December 2020 for preoperative risk assessment models for morbidity after adult cardiac surgery. Models exploring one isolated postoperative morbidity and those in patients having heart transplantation or congenital surgery were excluded. Data extraction and quality assessments were undertaken by two authors. Results: From 2251 identified papers, 22 models were found. The majority (54.5%) were developed in the USA or Canada, defined morbidity outcome within the in-hospital period (90.9%), and focused on major morbidity. Considerable variation in morbidity definition was identified, with morbidity incidence between 4.3% and 52%. The majority (45.5%) defined morbidity and mortality separately but combined them to develop one model, while seven studies (33.3%) constructed a morbidity-specific model. Models contained between 5 and 50 variables. Commonly included variables were age, emergency surgery, left ventricular dysfunction, and reoperation/previous cardiac surgery, although definition differences across studies were observed. All models demonstrated at least reasonable discriminatory power [area under the receiver operating curve (0.61-0.82)]. Conclusion: Despite the methodological heterogeneity across models, all demonstrated at least reasonable discriminatory power and could be implemented depending on local preferences. Future strategies to identify, predict, and reduce morbidity after cardiac surgery should consider the ageing population and those with minor and/or multiple complex morbidities.",

keywords = "Cardiac surgery, Morbidity outcome, Postoperative morbidity, Preoperative risk, Risk prediction models",

author = "Julie Sanders and Nicole Makariou and Adam Tocock and Rosalie Magboo and Ashley Thomas and Aitken, {Leanne M.}",

note = "Publisher Copyright: {\textcopyright} 2022 The Author(s).",

year = "2022",

month = sep,

day = "1",

doi = "10.1093/eurjcn/zvac003",

language = "English",

volume = "21",

pages = "655--664",

journal = "European Journal of Cardiovascular Nursing",

issn = "1474-5151",

publisher = "Elsevier",

number = "7",

}

TY - JOUR

T1 - Preoperative risk assessment tools for morbidity after cardiac surgery

T2 - A systematic review

AU - Sanders, Julie

AU - Makariou, Nicole

AU - Tocock, Adam

AU - Magboo, Rosalie

AU - Thomas, Ashley

AU - Aitken, Leanne M.

PY - 2022/9/1

Y1 - 2022/9/1

N2 - Background: Postoperative morbidity places considerable burden on health and resources. Thus, strategies to identify, predict, and reduce postoperative morbidity are needed. Aims: To identify and explore existing preoperative risk assessment tools for morbidity after cardiac surgery. Methods: Electronic databases (including MEDLINE, CINAHL, and Embase) were searched to December 2020 for preoperative risk assessment models for morbidity after adult cardiac surgery. Models exploring one isolated postoperative morbidity and those in patients having heart transplantation or congenital surgery were excluded. Data extraction and quality assessments were undertaken by two authors. Results: From 2251 identified papers, 22 models were found. The majority (54.5%) were developed in the USA or Canada, defined morbidity outcome within the in-hospital period (90.9%), and focused on major morbidity. Considerable variation in morbidity definition was identified, with morbidity incidence between 4.3% and 52%. The majority (45.5%) defined morbidity and mortality separately but combined them to develop one model, while seven studies (33.3%) constructed a morbidity-specific model. Models contained between 5 and 50 variables. Commonly included variables were age, emergency surgery, left ventricular dysfunction, and reoperation/previous cardiac surgery, although definition differences across studies were observed. All models demonstrated at least reasonable discriminatory power [area under the receiver operating curve (0.61-0.82)]. Conclusion: Despite the methodological heterogeneity across models, all demonstrated at least reasonable discriminatory power and could be implemented depending on local preferences. Future strategies to identify, predict, and reduce morbidity after cardiac surgery should consider the ageing population and those with minor and/or multiple complex morbidities.

AB - Background: Postoperative morbidity places considerable burden on health and resources. Thus, strategies to identify, predict, and reduce postoperative morbidity are needed. Aims: To identify and explore existing preoperative risk assessment tools for morbidity after cardiac surgery. Methods: Electronic databases (including MEDLINE, CINAHL, and Embase) were searched to December 2020 for preoperative risk assessment models for morbidity after adult cardiac surgery. Models exploring one isolated postoperative morbidity and those in patients having heart transplantation or congenital surgery were excluded. Data extraction and quality assessments were undertaken by two authors. Results: From 2251 identified papers, 22 models were found. The majority (54.5%) were developed in the USA or Canada, defined morbidity outcome within the in-hospital period (90.9%), and focused on major morbidity. Considerable variation in morbidity definition was identified, with morbidity incidence between 4.3% and 52%. The majority (45.5%) defined morbidity and mortality separately but combined them to develop one model, while seven studies (33.3%) constructed a morbidity-specific model. Models contained between 5 and 50 variables. Commonly included variables were age, emergency surgery, left ventricular dysfunction, and reoperation/previous cardiac surgery, although definition differences across studies were observed. All models demonstrated at least reasonable discriminatory power [area under the receiver operating curve (0.61-0.82)]. Conclusion: Despite the methodological heterogeneity across models, all demonstrated at least reasonable discriminatory power and could be implemented depending on local preferences. Future strategies to identify, predict, and reduce morbidity after cardiac surgery should consider the ageing population and those with minor and/or multiple complex morbidities.

KW - Cardiac surgery

KW - Morbidity outcome

KW - Postoperative morbidity

KW - Preoperative risk

KW - Risk prediction models

UR - http://www.scopus.com/inward/record.url?scp=85140144948&partnerID=8YFLogxK

U2 - 10.1093/eurjcn/zvac003

DO - 10.1093/eurjcn/zvac003

M3 - Review article

C2 - 35171231

AN - SCOPUS:85140144948

SN - 1474-5151

VL - 21

SP - 655

EP - 664

JO - European Journal of Cardiovascular Nursing

JF - European Journal of Cardiovascular Nursing

IS - 7

ER -

Preoperative risk assessment tools for morbidity after cardiac surgery: A systematic review

Abstract

Keywords

Access to Document

Other files and links

Fingerprint

Cite this