A comparison of random survival forest and Cox regression for prediction of mortality in patients with hemorrhagic stroke
Abstract Objective To evaluate RSF and Cox models for mortality prediction of hemorrhagic stroke (HS) patients in intensive care unit (ICU). Methods In the training set, the optimal models were selected using five-fold cross-validation and grid search method. In the test set, the bootstrap method wa...
Main Authors: | , , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
BMC
2023-10-01
|
Series: | BMC Medical Informatics and Decision Making |
Subjects: | |
Online Access: | https://doi.org/10.1186/s12911-023-02293-2 |
_version_ | 1797559296055050240 |
---|---|
author | Yuxin Wang Yuhan Deng Yinliang Tan Meihong Zhou Yong Jiang Baohua Liu |
author_facet | Yuxin Wang Yuhan Deng Yinliang Tan Meihong Zhou Yong Jiang Baohua Liu |
author_sort | Yuxin Wang |
collection | DOAJ |
description | Abstract Objective To evaluate RSF and Cox models for mortality prediction of hemorrhagic stroke (HS) patients in intensive care unit (ICU). Methods In the training set, the optimal models were selected using five-fold cross-validation and grid search method. In the test set, the bootstrap method was used to validate. The area under the curve(AUC) was used for discrimination, Brier Score (BS) was used for calibration, positive predictive value(PPV), negative predictive value(NPV), and F1 score were combined to compare. Results A total of 2,990 HS patients were included. For predicting the 7-day mortality, the mean AUCs for RSF and Cox regression were 0.875 and 0.761, while the mean BS were 0.083 and 0.108. For predicting the 28-day mortality, the mean AUCs for RSF and Cox regression were 0.794 and 0.649, while the mean BS were 0.129 and 0.174. The mean AUCs of RSF and Cox versus conventional scores for predicting patients’ 7-day mortality were 0.875 (RSF), 0.761 (COX), 0.736 (SAPS II), 0.723 (OASIS), 0.632 (SIRS), and 0.596 (SOFA), respectively. Conclusions RSF provided a better clinical reference than Cox. Creatine, temperature, anion gap and sodium were important variables in both models. |
first_indexed | 2024-03-10T17:43:19Z |
format | Article |
id | doaj.art-f5483cbdb2c743bcaacf17499d0096c1 |
institution | Directory Open Access Journal |
issn | 1472-6947 |
language | English |
last_indexed | 2024-03-10T17:43:19Z |
publishDate | 2023-10-01 |
publisher | BMC |
record_format | Article |
series | BMC Medical Informatics and Decision Making |
spelling | doaj.art-f5483cbdb2c743bcaacf17499d0096c12023-11-20T09:38:12ZengBMCBMC Medical Informatics and Decision Making1472-69472023-10-0123111110.1186/s12911-023-02293-2A comparison of random survival forest and Cox regression for prediction of mortality in patients with hemorrhagic strokeYuxin Wang0Yuhan Deng1Yinliang Tan2Meihong Zhou3Yong Jiang4Baohua Liu5Department of Social Medicine and Health Education, School of Public Health, Peking UniversityDepartment of Social Medicine and Health Education, School of Public Health, Peking UniversityDepartment of Social Medicine and Health Education, School of Public Health, Peking UniversityDepartment of Social Medicine and Health Education, School of Public Health, Peking UniversityDepartment of Neurology, Beijing Tiantan Hospital, Capital Medical UniversityDepartment of Social Medicine and Health Education, School of Public Health, Peking UniversityAbstract Objective To evaluate RSF and Cox models for mortality prediction of hemorrhagic stroke (HS) patients in intensive care unit (ICU). Methods In the training set, the optimal models were selected using five-fold cross-validation and grid search method. In the test set, the bootstrap method was used to validate. The area under the curve(AUC) was used for discrimination, Brier Score (BS) was used for calibration, positive predictive value(PPV), negative predictive value(NPV), and F1 score were combined to compare. Results A total of 2,990 HS patients were included. For predicting the 7-day mortality, the mean AUCs for RSF and Cox regression were 0.875 and 0.761, while the mean BS were 0.083 and 0.108. For predicting the 28-day mortality, the mean AUCs for RSF and Cox regression were 0.794 and 0.649, while the mean BS were 0.129 and 0.174. The mean AUCs of RSF and Cox versus conventional scores for predicting patients’ 7-day mortality were 0.875 (RSF), 0.761 (COX), 0.736 (SAPS II), 0.723 (OASIS), 0.632 (SIRS), and 0.596 (SOFA), respectively. Conclusions RSF provided a better clinical reference than Cox. Creatine, temperature, anion gap and sodium were important variables in both models.https://doi.org/10.1186/s12911-023-02293-2Hemorrhagic strokeRandom survival forestCox regressionIntensive care unit |
spellingShingle | Yuxin Wang Yuhan Deng Yinliang Tan Meihong Zhou Yong Jiang Baohua Liu A comparison of random survival forest and Cox regression for prediction of mortality in patients with hemorrhagic stroke BMC Medical Informatics and Decision Making Hemorrhagic stroke Random survival forest Cox regression Intensive care unit |
title | A comparison of random survival forest and Cox regression for prediction of mortality in patients with hemorrhagic stroke |
title_full | A comparison of random survival forest and Cox regression for prediction of mortality in patients with hemorrhagic stroke |
title_fullStr | A comparison of random survival forest and Cox regression for prediction of mortality in patients with hemorrhagic stroke |
title_full_unstemmed | A comparison of random survival forest and Cox regression for prediction of mortality in patients with hemorrhagic stroke |
title_short | A comparison of random survival forest and Cox regression for prediction of mortality in patients with hemorrhagic stroke |
title_sort | comparison of random survival forest and cox regression for prediction of mortality in patients with hemorrhagic stroke |
topic | Hemorrhagic stroke Random survival forest Cox regression Intensive care unit |
url | https://doi.org/10.1186/s12911-023-02293-2 |
work_keys_str_mv | AT yuxinwang acomparisonofrandomsurvivalforestandcoxregressionforpredictionofmortalityinpatientswithhemorrhagicstroke AT yuhandeng acomparisonofrandomsurvivalforestandcoxregressionforpredictionofmortalityinpatientswithhemorrhagicstroke AT yinliangtan acomparisonofrandomsurvivalforestandcoxregressionforpredictionofmortalityinpatientswithhemorrhagicstroke AT meihongzhou acomparisonofrandomsurvivalforestandcoxregressionforpredictionofmortalityinpatientswithhemorrhagicstroke AT yongjiang acomparisonofrandomsurvivalforestandcoxregressionforpredictionofmortalityinpatientswithhemorrhagicstroke AT baohualiu acomparisonofrandomsurvivalforestandcoxregressionforpredictionofmortalityinpatientswithhemorrhagicstroke AT yuxinwang comparisonofrandomsurvivalforestandcoxregressionforpredictionofmortalityinpatientswithhemorrhagicstroke AT yuhandeng comparisonofrandomsurvivalforestandcoxregressionforpredictionofmortalityinpatientswithhemorrhagicstroke AT yinliangtan comparisonofrandomsurvivalforestandcoxregressionforpredictionofmortalityinpatientswithhemorrhagicstroke AT meihongzhou comparisonofrandomsurvivalforestandcoxregressionforpredictionofmortalityinpatientswithhemorrhagicstroke AT yongjiang comparisonofrandomsurvivalforestandcoxregressionforpredictionofmortalityinpatientswithhemorrhagicstroke AT baohualiu comparisonofrandomsurvivalforestandcoxregressionforpredictionofmortalityinpatientswithhemorrhagicstroke |