Prognostic models for outcome prediction following in-hospital cardiac arrest using pre-arrest factors: a systematic review, meta-analysis and critical appraisal

Critical Care

Research

Marc Schluep¹  Casey Grandbois van Ravenhorst¹  Robert-Jan Stolker¹  Sanne Elisabeth Hoeks¹  Henrik Endeman² 

[1] Department of Anaesthesia, Erasmus University Medical Centre, P.O. Box 2040, Room Na-1718, 3000 CA, Rotterdam, The Netherlands;Department of Intensive Care Medicine, Erasmus University Medical Centre, Rotterdam, The Netherlands;

关键词: In-hospital cardiac arrest;    Cardiopulmonary resuscitation;    Clinical outcome;    Prognostic model;   

DOI  :  10.1186/s13054-023-04306-y

 received in 2022-11-15, accepted in 2023-01-06,  发布年份 2023

来源: Springer

【 摘 要 】

BackgroundSeveral prediction models of survival after in-hospital cardiac arrest (IHCA) have been published, but no overview of model performance and external validation exists. We performed a systematic review of the available prognostic models for outcome prediction of attempted resuscitation for IHCA using pre-arrest factors to enhance clinical decision-making through improved outcome prediction.MethodsThis systematic review followed the CHARMS and PRISMA guidelines. Medline, Embase, Web of Science were searched up to October 2021. Studies developing, updating or validating a prediction model with pre-arrest factors for any potential clinical outcome of attempted resuscitation for IHCA were included. Studies were appraised critically according to the PROBAST checklist. A random-effects meta-analysis was performed to pool AUROC values of externally validated models.ResultsOut of 2678 initial articles screened, 33 studies were included in this systematic review: 16 model development studies, 5 model updating studies and 12 model validation studies. The most frequently included pre-arrest factors included age, functional status, (metastatic) malignancy, heart disease, cerebrovascular events, respiratory, renal or hepatic insufficiency, hypotension and sepsis. Only six of the developed models have been independently validated in external populations. The GO-FAR score showed the best performance with a pooled AUROC of 0.78 (95% CI 0.69–0.85), versus 0.59 (95%CI 0.50–0.68) for the PAM and 0.62 (95% CI 0.49–0.74) for the PAR.ConclusionsSeveral prognostic models for clinical outcome after attempted resuscitation for IHCA have been published. Most have a moderate risk of bias and have not been validated externally. The GO-FAR score showed the most acceptable performance. Future research should focus on updating existing models for use in clinical settings, specifically pre-arrest counselling.Systematic review registration PROSPERO CRD42021269235. Registered 21 July 2021.

【 授权许可】

CC BY   
© The Author(s) 2023

【 图 表 】

41116_2022_35_Article_IEq230.gif

41116_2022_35_Article_IEq228.gif

41116_2022_35_Article_IEq208.gif