Dataset Information

Development of early prediction model of in-hospital cardiac arrest based on laboratory parameters.

ABSTRACT:

Background

In-hospital cardiac arrest (IHCA) is an acute disease with a high fatality rate that burdens individuals, society, and the economy. This study aimed to develop a machine learning (ML) model using routine laboratory parameters to predict the risk of IHCA in rescue-treated patients.

Methods

This retrospective cohort study examined all rescue-treated patients hospitalized at the First Medical Center of the PLA General Hospital in Beijing, China, from January 2016 to December 2020. Five machine learning algorithms, including support vector machine, random forest, extra trees classifier (ETC), decision tree, and logistic regression algorithms, were trained to develop models for predicting IHCA. We included blood counts, biochemical markers, and coagulation markers in the model development. We validated model performance using fivefold cross-validation and used the SHapley Additive exPlanation (SHAP) for model interpretation.

Results

A total of 11,308 participants were included in the study, of which 7779 patients remained. Among these patients, 1796 (23.09%) cases of IHCA occurred. Among five machine learning models for predicting IHCA, the ETC algorithm exhibited better performance, with an AUC of 0.920, compared with the other four machine learning models in the fivefold cross-validation. The SHAP showed that the top ten factors accounting for cardiac arrest in rescue-treated patients are prothrombin activity, platelets, hemoglobin, N-terminal pro-brain natriuretic peptide, neutrophils, prothrombin time, serum albumin, sodium, activated partial thromboplastin time, and potassium.

Conclusions

We developed a reliable machine learning-derived model that integrates readily available laboratory parameters to predict IHCA in patients treated with rescue therapy.

SUBMITTER: Ding X

PROVIDER: S-EPMC10698946 | biostudies-literature | 2023 Dec

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Development of early prediction model of in-hospital cardiac arrest based on laboratory parameters.

Ding Xinhuan X Wang Yingchan Y Ma Weiyi W Peng Yaojun Y Huang Jingjing J Wang Meng M Zhu Haiyan H

Biomedical engineering online 20231206 1

<h4>Background</h4>In-hospital cardiac arrest (IHCA) is an acute disease with a high fatality rate that burdens individuals, society, and the economy. This study aimed to develop a machine learning (ML) model using routine laboratory parameters to predict the risk of IHCA in rescue-treated patients.<h4>Methods</h4>This retrospective cohort study examined all rescue-treated patients hospitalized at the First Medical Center of the PLA General Hospital in Beijing, China, from January 2016 to Decemb ...[more]

PMID: 38057823

Similar Datasets

Project description:We aimed to determine the characteristics of quantitative pupillary response parameters other than amplitude of pupillary light reflex (PLR) early after return of spontaneous circulation (ROSC) and their implications for predicting neurological outcomes early after cardiac arrest (CA). Fifty adults resuscitated after non-traumatic out-of-hospital CA from four emergency hospitals were enrolled. Pupil diameters, PLR, constriction velocity (CV), maximum CV (MCV), dilation velocity (DV), latency of constriction, and Neurological Pupil index (NPi) were quantitatively measured at 0, 6, 12, 24, 48, and 72 h post-ROSC using an automated pupillometer. Change over time of each parameter was compared between favorable (Cerebral Performance Category [CPC] 1 or 2) and unfavorable neurological outcome (CPC 3-5) groups. Prognostic values of 90-day favorable outcome by these parameters and when combined with clinical predictors (witness status, bystander cardiopulmonary resuscitation, initial shockable rhythm, implementation of target temperature management) were tested. Thirteen patients achieved favorable outcome. CV, MCV, DV (P < 0.001), and NPi (P = 0.005) were consistently greater in the favorable group than in the unfavorable outcome group. Change over time was not statistically different between the groups in all parameters. CV, MCV, DV (ρ = 0.96 to 0.97, P < 0.001, respectively), and NPi (ρ = 0.65, P < 0.001) positively correlated with PLR. The prognostic value of 0-hour CV (area under the curve, AUC [95% confidence interval]: 0.92 [0.80-1.00]), DV (0.84 [0.68-0.99]), and NPi (0.88 [0.74-1.00]) was equivalent to that of PLR (0.84 [0.69-0.98]). Prognostic values improved to AUC of 0.95-0.96 when 0-hour PLR, CV, DV, or NPi was combined with clinical predictors. The 0-hour CV, MCV, and NPi showed equivalent prognostic values to PLR alone/in combination with clinical predictors. Using PLR among several quantitative pupillary response parameters for early neurological prognostication of post-CA patients is a simple and effective strategy.

Project description:Background: Early prognostication is a major challenge after out-of-hospital cardiac arrest (OHCA). Aims: We hypothesized that a genome-wide analysis of blood gene expression could offer new prognostic tools and lines of research. Methods: Sixty-nine patients were enrolled from an ancillary study of the clinical trial NCT00999583 that tested the effect of erythropoietin (EPO) after OHCA. Blood samples were collected in comatose survivors of OHCA at hospital admission and 1 and 3 days after resuscitation. Gene expression profiles were analyzed (Illumina HumanHT-12 V4 BeadChip; >34,000 genes). Patients were classified into two categories representing neurologically favorable outcome (Cerebral Performance Category [CPC] = 1-2) vs unfavorable outcome (CPC>2) at Day 60 after OHCA. Differential and functional enrichment analyses were performed to compare transcriptomic profiles between these two categories. Results: Among the 69 enrolled patients, 33 and 36 patients were treated or not by EPO, respectively. Among them, 42% had a favorable neurological outcome in both groups. EPO did not affect the transcriptomic response at Day-0 and 1 after OHCA. In contrast, 76 transcripts differed at Day-0 between patients with unfavorable vs favorable neurological outcome. This signature persisted at Day-1 after OHCA. Functional enrichment analysis revealed a down-regulation of adaptive immunity with concomitant up-regulation of innate immunity and inflammation in patients with unfavorable vs favorable neurological outcome. The transcription of many genes of the HLA family was decreased in patients with unfavorable vs favorable neurological outcome. Concomitantly, neutrophil activation and inflammation were observed. Up-stream regulators analysis showed the implication of numerous factors involved in cell cycle and damages. A logistic regression including a set of genes allowed a reliable prediction of the clinical outcomes (Specificity=88%; Hit Rate=83%). Conclusions: A transcriptomic signature involving a counterbalance between adaptive and innate immune responses is able to predict neurological outcome very early after hospital admission after OHCA. This deserves confirmation in a larger population.

Project description:BackgroundOut-of-hospital cardiac arrest (OHCA) is a serious public health issue, and predicting the prognosis of OHCA patients can assist clinicians in making decisions about the treatment of patients, use of hospital resources, or termination of resuscitation.ObjectiveThis study aimed to develop a time-adaptive conditional prediction model (TACOM) to predict clinical outcomes every minute.MethodsWe performed a retrospective observational study using data from the Korea OHCA Registry in South Korea. In this study, we excluded patients with trauma, those who experienced return of spontaneous circulation before arriving in the emergency department (ED), and those who did not receive cardiopulmonary resuscitation (CPR) in the ED. We selected patients who received CPR in the ED. To develop the time-adaptive prediction model, we organized the training data set as ongoing CPR patients by the minute. A total of 49,669 patients were divided into 39,602 subjects for training and 10,067 subjects for validation. We compared random forest, LightGBM, and artificial neural networks as the prediction model methods. Model performance was quantified using the prediction probability of the model, area under the receiver operating characteristic curve (AUROC), and area under the precision recall curve.ResultsAmong the three algorithms, LightGBM showed the best performance. From 0 to 30 min, the AUROC of the TACOM for predicting good neurological outcomes ranged from 0.910 (95% CI 0.910-0.911) to 0.869 (95% CI 0.865-0.871), whereas that for survival to hospital discharge ranged from 0.800 (95% CI 0.797-0.800) to 0.734 (95% CI 0.736-0.740). The prediction probability of the TACOM showed similar flow with cohort data based on a comparison with the conventional model's prediction probability.ConclusionsThe TACOM predicted the clinical outcome of OHCA patients per minute. This model for predicting patient outcomes by the minute can assist clinicians in making rational decisions for OHCA patients.

Dataset Information

Development of early prediction model of in-hospital cardiac arrest based on laboratory parameters.

Background

Methods

Results

Conclusions

Publications

Development of early prediction model of in-hospital cardiac arrest based on laboratory parameters.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets