Truace
Subscribe

Predictive performance of vital sign–based prediction models for in-hospital cardiac arrest: a systematic review and meta-analysis

Abstract Background In-hospital cardiac arrest (IHCA) is a major patient-safety event often preceded by detectable clinical deterioration. Vital sign–based cardiac arrest prediction models have been developed to identify hospitalized patients at increased risk of IHCA, but their predictive performance has not been quantitatively synthesized across studies. Methods This PRISMA-based systematic review and meta-analysis was registered in PROSPERO. Ovid Embase, Ovid MEDLINE, and the Cochrane Central Register of Cont…

By the Truvace desk1 min read1 sourcePeer-reviewedTRV-2026-1595· cite & verify
Predictive performance of vital sign–based prediction models for in-hospital cardiac arrest: a systematic review and meta-analysis
Construction site tour of the Skilled Nursing Residence, in Edgartown, on the island of Martha's Vineyard, in Dukes County, Maryland on August 5, 2024 - 11 by USDAgov. Public domain

In brief

A systematic review searched Embase, MEDLINE and CENTRAL and screened 10,645 records. Six studies met the criteria for models that use routine bedside vital signs to predict cardiac arrest in general ward patients, including CART, DEWS and DeepCARS.

The finding matters because in-hospital cardiac arrest is often preceded by detectable deterioration, so early warning tools could support patient safety. The pooled result is only exploratory, however, and it is unclear whether the models would work well in practice, with acceptable alert volumes, outside the studies that built them.

Main points

  1. Six studies published between 2012 and 2024 were included from 10,645 records identified.
  2. Three studies reported cardiac arrest-specific AUROC estimates usable in the primary meta-analysis.
  3. The exploratory pooled AUROC was 0.89 with a 95% CI of 0.88–0.91 and I² of 92%.
  4. Risk of bias was low in two studies, high in three and unclear in one.

The problem

The evidence base is weak: half of included studies had high risk of bias, heterogeneity was substantial, and calibration, alert burden and clinical utility remain untested.

The rundown

The review was PRISMA-based and registered in PROSPERO. It included peer-reviewed studies of vital sign–based models for hospitalized general ward patients. Models whose core predictors were laboratory values, imaging or other non-bedside variables were excluded.

AUROC was the primary performance measure, and AUPRC was extracted and summarized where studies reported it. Risk of bias was assessed with PROBAST + AI.

The authors call for standardized outcome definitions, external validation, and prospective evaluation of calibration, alert burden and clinical utility before broad clinical implementation.

Sources

  1. Peer-reviewedBMC Cardiovascular Disorders2026-10-15

The debate