Dynamic and explainable machine learning prediction of mortality in patients in the intensive care unit: a retrospective study of high-frequency data in electronic patient records
Many mortality prediction models have been developed for patients in intensive care units (ICUs); most are based on data available at ICU admission. We investigated whether machine learning methods using analyses of time-series data improved mortality prognostication for patients in the ICU by provi...
Saved in:
Published in | The Lancet. Digital health Vol. 2; no. 4; pp. e179 - e191 |
---|---|
Main Authors | , , , , , , , , , , , , , , |
Format | Journal Article |
Language | English |
Published |
England
Elsevier Ltd
01.04.2020
Elsevier |
Subjects | |
Online Access | Get full text |
Cover
Loading…
Abstract | Many mortality prediction models have been developed for patients in intensive care units (ICUs); most are based on data available at ICU admission. We investigated whether machine learning methods using analyses of time-series data improved mortality prognostication for patients in the ICU by providing real-time predictions of 90-day mortality. In addition, we examined to what extent such a dynamic model could be made interpretable by quantifying and visualising the features that drive the predictions at different timepoints.
Based on the Simplified Acute Physiology Score (SAPS) III variables, we trained a machine learning model on longitudinal data from patients admitted to four ICUs in the Capital Region, Denmark, between 2011 and 2016. We included all patients older than 16 years of age, with an ICU stay lasting more than 1 h, and who had a Danish civil registration number to enable 90-day follow-up. We leveraged static data and physiological time-series data from electronic health records and the Danish National Patient Registry. A recurrent neural network was trained with a temporal resolution of 1 h. The model was internally validated using the holdout method with 20% of the training dataset and externally validated using previously unseen data from a fifth hospital in Denmark. Its performance was assessed with the Matthews correlation coefficient (MCC) and area under the receiver operating characteristic curve (AUROC) as metrics, using bootstrapping with 1000 samples with replacement to construct 95% CIs. A Shapley additive explanations algorithm was applied to the prediction model to obtain explanations of the features that drive patient-specific predictions, and the contributions of each of the 44 features in the model were analysed and compared with the variables in the original SAPS III model.
From a dataset containing 15 615 ICU admissions of 12 616 patients, we included 14 190 admissions of 11 492 patients in our analysis. Overall, 90-day mortality was 33·1% (3802 patients). The deep learning model showed a predictive performance on the holdout testing dataset that improved over the timecourse of an ICU stay: MCC 0·29 (95% CI 0·25–0·33) and AUROC 0·73 (0·71–0·74) at admission, 0·43 (0·40–0·47) and 0·82 (0·80–0·84) after 24 h, 0·50 (0·46–0·53) and 0·85 (0·84–0·87) after 72 h, and 0·57 (0·54–0·60) and 0·88 (0·87–0·89) at the time of discharge. The model exhibited good calibration properties. These results were validated in an external validation cohort of 5827 patients with 6748 admissions: MCC 0·29 (95% CI 0·27–0·32) and AUROC 0·75 (0·73–0·76) at admission, 0·41 (0·39–0·44) and 0·80 (0·79–0·81) after 24 h, 0·46 (0·43–0·48) and 0·82 (0·81–0·83) after 72 h, and 0·47 (0·44–0·49) and 0·83 (0·82–0·84) at the time of discharge.
The prediction of 90-day mortality improved with 1-h sampling intervals during the ICU stay. The dynamic risk prediction can also be explained for an individual patient, visualising the features contributing to the prediction at any point in time. This explanation allows the clinician to determine whether there are elements in the current patient state and care that are potentially actionable, thus making the model suitable for further validation as a clinical tool.
Novo Nordisk Foundation and the Innovation Fund Denmark. |
---|---|
AbstractList | Background: Many mortality prediction models have been developed for patients in intensive care units (ICUs); most are based on data available at ICU admission. We investigated whether machine learning methods using analyses of time-series data improved mortality prognostication for patients in the ICU by providing real-time predictions of 90-day mortality. In addition, we examined to what extent such a dynamic model could be made interpretable by quantifying and visualising the features that drive the predictions at different timepoints. Methods: Based on the Simplified Acute Physiology Score (SAPS) III variables, we trained a machine learning model on longitudinal data from patients admitted to four ICUs in the Capital Region, Denmark, between 2011 and 2016. We included all patients older than 16 years of age, with an ICU stay lasting more than 1 h, and who had a Danish civil registration number to enable 90-day follow-up. We leveraged static data and physiological time-series data from electronic health records and the Danish National Patient Registry. A recurrent neural network was trained with a temporal resolution of 1 h. The model was internally validated using the holdout method with 20% of the training dataset and externally validated using previously unseen data from a fifth hospital in Denmark. Its performance was assessed with the Matthews correlation coefficient (MCC) and area under the receiver operating characteristic curve (AUROC) as metrics, using bootstrapping with 1000 samples with replacement to construct 95% CIs. A Shapley additive explanations algorithm was applied to the prediction model to obtain explanations of the features that drive patient-specific predictions, and the contributions of each of the 44 features in the model were analysed and compared with the variables in the original SAPS III model. Findings: From a dataset containing 15 615 ICU admissions of 12 616 patients, we included 14 190 admissions of 11 492 patients in our analysis. Overall, 90-day mortality was 33·1% (3802 patients). The deep learning model showed a predictive performance on the holdout testing dataset that improved over the timecourse of an ICU stay: MCC 0·29 (95% CI 0·25–0·33) and AUROC 0·73 (0·71–0·74) at admission, 0·43 (0·40–0·47) and 0·82 (0·80–0·84) after 24 h, 0·50 (0·46–0·53) and 0·85 (0·84–0·87) after 72 h, and 0·57 (0·54–0·60) and 0·88 (0·87–0·89) at the time of discharge. The model exhibited good calibration properties. These results were validated in an external validation cohort of 5827 patients with 6748 admissions: MCC 0·29 (95% CI 0·27–0·32) and AUROC 0·75 (0·73–0·76) at admission, 0·41 (0·39–0·44) and 0·80 (0·79–0·81) after 24 h, 0·46 (0·43–0·48) and 0·82 (0·81–0·83) after 72 h, and 0·47 (0·44–0·49) and 0·83 (0·82–0·84) at the time of discharge. Interpretation: The prediction of 90-day mortality improved with 1-h sampling intervals during the ICU stay. The dynamic risk prediction can also be explained for an individual patient, visualising the features contributing to the prediction at any point in time. This explanation allows the clinician to determine whether there are elements in the current patient state and care that are potentially actionable, thus making the model suitable for further validation as a clinical tool. Funding: Novo Nordisk Foundation and the Innovation Fund Denmark. Many mortality prediction models have been developed for patients in intensive care units (ICUs); most are based on data available at ICU admission. We investigated whether machine learning methods using analyses of time-series data improved mortality prognostication for patients in the ICU by providing real-time predictions of 90-day mortality. In addition, we examined to what extent such a dynamic model could be made interpretable by quantifying and visualising the features that drive the predictions at different timepoints. Based on the Simplified Acute Physiology Score (SAPS) III variables, we trained a machine learning model on longitudinal data from patients admitted to four ICUs in the Capital Region, Denmark, between 2011 and 2016. We included all patients older than 16 years of age, with an ICU stay lasting more than 1 h, and who had a Danish civil registration number to enable 90-day follow-up. We leveraged static data and physiological time-series data from electronic health records and the Danish National Patient Registry. A recurrent neural network was trained with a temporal resolution of 1 h. The model was internally validated using the holdout method with 20% of the training dataset and externally validated using previously unseen data from a fifth hospital in Denmark. Its performance was assessed with the Matthews correlation coefficient (MCC) and area under the receiver operating characteristic curve (AUROC) as metrics, using bootstrapping with 1000 samples with replacement to construct 95% CIs. A Shapley additive explanations algorithm was applied to the prediction model to obtain explanations of the features that drive patient-specific predictions, and the contributions of each of the 44 features in the model were analysed and compared with the variables in the original SAPS III model. From a dataset containing 15 615 ICU admissions of 12 616 patients, we included 14 190 admissions of 11 492 patients in our analysis. Overall, 90-day mortality was 33·1% (3802 patients). The deep learning model showed a predictive performance on the holdout testing dataset that improved over the timecourse of an ICU stay: MCC 0·29 (95% CI 0·25-0·33) and AUROC 0·73 (0·71-0·74) at admission, 0·43 (0·40-0·47) and 0·82 (0·80-0·84) after 24 h, 0·50 (0·46-0·53) and 0·85 (0·84-0·87) after 72 h, and 0·57 (0·54-0·60) and 0·88 (0·87-0·89) at the time of discharge. The model exhibited good calibration properties. These results were validated in an external validation cohort of 5827 patients with 6748 admissions: MCC 0·29 (95% CI 0·27-0·32) and AUROC 0·75 (0·73-0·76) at admission, 0·41 (0·39-0·44) and 0·80 (0·79-0·81) after 24 h, 0·46 (0·43-0·48) and 0·82 (0·81-0·83) after 72 h, and 0·47 (0·44-0·49) and 0·83 (0·82-0·84) at the time of discharge. The prediction of 90-day mortality improved with 1-h sampling intervals during the ICU stay. The dynamic risk prediction can also be explained for an individual patient, visualising the features contributing to the prediction at any point in time. This explanation allows the clinician to determine whether there are elements in the current patient state and care that are potentially actionable, thus making the model suitable for further validation as a clinical tool. Novo Nordisk Foundation and the Innovation Fund Denmark. |
Author | Nielsen, Anna P Strøm, Thomas Heimann, Marc Toft, Palle Chmura, Piotr J Belling, Kirstine Thorsen-Meyer, Hans-Christian Nielsen, Annelaura B Schierbeck, Jens Perner, Anders Spangsege, Lasse Kaas-Hansen, Benjamin Skov Hulsen, Patrick Brunak, Søren Dybdahl, Lars |
Author_xml | – sequence: 1 givenname: Hans-Christian surname: Thorsen-Meyer fullname: Thorsen-Meyer, Hans-Christian organization: Novo Nordisk Foundation Center for Protein Research, Faculty of Health and Medical Sciences, University of Copenhagen, Copenhagen, Denmark – sequence: 2 givenname: Annelaura B surname: Nielsen fullname: Nielsen, Annelaura B organization: Novo Nordisk Foundation Center for Protein Research, Faculty of Health and Medical Sciences, University of Copenhagen, Copenhagen, Denmark – sequence: 3 givenname: Anna P surname: Nielsen fullname: Nielsen, Anna P organization: Novo Nordisk Foundation Center for Protein Research, Faculty of Health and Medical Sciences, University of Copenhagen, Copenhagen, Denmark – sequence: 4 givenname: Benjamin Skov surname: Kaas-Hansen fullname: Kaas-Hansen, Benjamin Skov organization: Novo Nordisk Foundation Center for Protein Research, Faculty of Health and Medical Sciences, University of Copenhagen, Copenhagen, Denmark – sequence: 5 givenname: Palle surname: Toft fullname: Toft, Palle organization: Department of Anesthesiology and Intensive Care, Odense University Hospital, Odense, Denmark – sequence: 6 givenname: Jens surname: Schierbeck fullname: Schierbeck, Jens organization: Department of Anesthesiology and Intensive Care, Odense University Hospital, Odense, Denmark – sequence: 7 givenname: Thomas surname: Strøm fullname: Strøm, Thomas organization: Department of Anesthesiology and Intensive Care, Odense University Hospital, Odense, Denmark – sequence: 8 givenname: Piotr J surname: Chmura fullname: Chmura, Piotr J organization: Novo Nordisk Foundation Center for Protein Research, Faculty of Health and Medical Sciences, University of Copenhagen, Copenhagen, Denmark – sequence: 9 givenname: Marc surname: Heimann fullname: Heimann, Marc organization: Centre for IT, Medical Technology and Telephony Services, Capital Region of Denmark, Copenhagen, Denmark – sequence: 10 givenname: Lars surname: Dybdahl fullname: Dybdahl, Lars organization: Daintel, Lyngby, Denmark – sequence: 11 givenname: Lasse surname: Spangsege fullname: Spangsege, Lasse organization: Daintel, Lyngby, Denmark – sequence: 12 givenname: Patrick surname: Hulsen fullname: Hulsen, Patrick organization: Daintel, Lyngby, Denmark – sequence: 13 givenname: Kirstine surname: Belling fullname: Belling, Kirstine organization: Novo Nordisk Foundation Center for Protein Research, Faculty of Health and Medical Sciences, University of Copenhagen, Copenhagen, Denmark – sequence: 14 givenname: Søren surname: Brunak fullname: Brunak, Søren email: soren.brunak@cpr.ku.dk organization: Novo Nordisk Foundation Center for Protein Research, Faculty of Health and Medical Sciences, University of Copenhagen, Copenhagen, Denmark – sequence: 15 givenname: Anders surname: Perner fullname: Perner, Anders organization: Department of Intensive Care, Rigshospitalet, Copenhagen University Hospital, Copenhagen, Denmark |
BackLink | https://www.ncbi.nlm.nih.gov/pubmed/33328078$$D View this record in MEDLINE/PubMed |
BookMark | eNqFkU1v1DAQhi1UREvpTwD5CIeAP5J1wgWh8lWpEgfgbI3tya6rxA6OtyK_iz-Is1tW3DjNaPzO49H7PiVnIQYk5DlnrznjmzffRNN2lWoYeynYK8kYbyvxiFycxmf_9Ofkap7vGGNCcKmUekLOpZSiZaq9IL8_LAFGbykER_HXNIAPYAakI9idD0gHhBR82NIpofM2-xho7OkYU4bB54X6QCfIHkOe1z7vsJSMYfb3SC0kpPvg81sKNGFOcZ6wQMrTnPduWVE7v91VfcKfewx2oQ4yrCAcijDFUG574BeAjcnNz8jjHoYZrx7qJfnx6eP36y_V7dfPN9fvbyvbdJtcWSONQ6wNctU4ZsE2UlrsoMYWNwxq2zrFaiVE3Za2EViLppfS8EYxXsSX5ObIdRHu9JT8CGnREbw-DGLaakjZ2wG1NNBJNOAMa2vuZNt3hisnHW6MqDtWWM2RZYsFc8L-xONMr5nqQ6Z6DUwLpg-ZalH2Xhz3pr0Z0Z22_iZYBO-OAixO3HtMerbFLFvCKnblcqr_zxd_AKcdt6Q |
CitedBy_id | crossref_primary_10_1016_j_eclinm_2023_101838 crossref_primary_10_1186_s12879_022_07462_8 crossref_primary_10_1186_s12874_022_01540_w crossref_primary_10_1371_journal_pone_0270973 crossref_primary_10_1109_JSTARS_2021_3103754 crossref_primary_10_3389_fmed_2022_960296 crossref_primary_10_1016_j_mcpsp_2023_100395 crossref_primary_10_1001_jamanetworkopen_2022_37970 crossref_primary_10_1186_s13054_023_04683_4 crossref_primary_10_1155_2022_5682599 crossref_primary_10_1007_s10729_024_09673_8 crossref_primary_10_3389_fmed_2022_837382 crossref_primary_10_1016_j_hlc_2021_09_007 crossref_primary_10_1186_s13098_023_01020_1 crossref_primary_10_1371_journal_pone_0277135 crossref_primary_10_3389_fmed_2021_607952 crossref_primary_10_3390_jpm12091507 crossref_primary_10_7759_cureus_49723 crossref_primary_10_12688_wellcomeopenres_16008_2 crossref_primary_10_1055_a_1470_1377 crossref_primary_10_1186_s12967_024_04915_3 crossref_primary_10_1038_s41598_021_82403_0 crossref_primary_10_1177_09544119221123431 crossref_primary_10_1371_journal_pdig_0000116 crossref_primary_10_2196_43734 crossref_primary_10_3390_jpm12020166 crossref_primary_10_1200_CCI_23_00264 crossref_primary_10_1016_j_patter_2022_100602 crossref_primary_10_1016_j_measen_2024_101053 crossref_primary_10_2196_26426 crossref_primary_10_1016_j_compbiomed_2022_106188 crossref_primary_10_2139_ssrn_3955672 crossref_primary_10_3390_brainsci13040557 crossref_primary_10_1007_s12028_021_01325_x crossref_primary_10_1016_j_compbiomed_2022_106466 crossref_primary_10_1016_j_ajhg_2021_03_010 crossref_primary_10_1002_pds_5756 crossref_primary_10_1016_j_arr_2023_102144 crossref_primary_10_1016_j_chaos_2020_110338 crossref_primary_10_3389_fcvm_2022_977747 crossref_primary_10_31744_einstein_journal_2024AO0328 crossref_primary_10_1038_s41598_020_73558_3 crossref_primary_10_2196_29812 crossref_primary_10_1001_jamanetworkopen_2022_14514 crossref_primary_10_1007_s10877_023_01108_z crossref_primary_10_1007_s11547_024_01785_z crossref_primary_10_1016_j_compbiomed_2021_104954 crossref_primary_10_1016_j_athoracsur_2021_08_040 crossref_primary_10_1016_j_compbiomed_2024_108844 crossref_primary_10_1016_j_ibmed_2022_100071 crossref_primary_10_3389_ijph_2023_1604789 crossref_primary_10_1016_j_cmpbup_2022_100083 crossref_primary_10_1186_s13054_024_04866_7 crossref_primary_10_1016_j_eswa_2023_120966 crossref_primary_10_3390_diagnostics13020264 crossref_primary_10_1109_ACCESS_2021_3110336 crossref_primary_10_3390_a17010006 crossref_primary_10_1109_ACCESS_2020_3043221 crossref_primary_10_7189_jogh_12_04044 crossref_primary_10_1016_j_ibmed_2023_100089 crossref_primary_10_1007_s10462_023_10561_w crossref_primary_10_1186_s13040_022_00309_7 crossref_primary_10_1038_s41598_022_24404_1 crossref_primary_10_1109_JBHI_2021_3089287 crossref_primary_10_1038_s41598_024_61851_4 crossref_primary_10_1016_j_cmpb_2022_107236 crossref_primary_10_1016_j_ijmedinf_2022_104776 crossref_primary_10_3748_wjg_v27_i26_4088 crossref_primary_10_3390_healthcare11060910 crossref_primary_10_1007_s41666_023_00143_4 crossref_primary_10_3389_fcvm_2023_1126888 crossref_primary_10_1097_CCE_0000000000000279 crossref_primary_10_3389_fpsyt_2023_1128862 crossref_primary_10_3390_brainsci12070938 crossref_primary_10_1109_ACCESS_2023_3257406 crossref_primary_10_1016_j_patter_2023_100892 crossref_primary_10_2196_41142 crossref_primary_10_1002_clc_23963 crossref_primary_10_1097_CCM_0000000000006176 crossref_primary_10_3389_fneur_2021_790682 crossref_primary_10_1016_j_icte_2023_07_007 crossref_primary_10_1016_S2589_7500_21_00040_6 crossref_primary_10_1016_j_ijmedinf_2023_105084 crossref_primary_10_2196_21798 crossref_primary_10_1016_j_jbi_2021_103980 crossref_primary_10_1038_s41398_022_01937_7 crossref_primary_10_1007_s10877_021_00703_2 crossref_primary_10_1016_j_ebiom_2023_104937 crossref_primary_10_1111_pan_14736 crossref_primary_10_1016_j_jbi_2023_104310 crossref_primary_10_1038_s41591_023_02332_5 crossref_primary_10_1016_j_compbiomed_2024_107934 crossref_primary_10_1016_j_ccc_2023_03_002 crossref_primary_10_1007_s12553_021_00538_6 crossref_primary_10_1038_s41598_024_63339_7 crossref_primary_10_1097_BRS_0000000000004861 crossref_primary_10_1080_14779072_2023_2223978 crossref_primary_10_3390_app12041926 crossref_primary_10_1145_3531326 crossref_primary_10_1016_j_artmed_2022_102283 crossref_primary_10_1016_j_chest_2022_09_046 crossref_primary_10_1097_MPA_0000000000001762 crossref_primary_10_4258_hir_2023_29_4_301 crossref_primary_10_1186_s13054_022_03915_3 crossref_primary_10_1016_j_suc_2022_11_004 crossref_primary_10_2196_47592 crossref_primary_10_1093_bib_bbac253 crossref_primary_10_2196_49643 crossref_primary_10_1016_S2589_7500_20_30056_X crossref_primary_10_1016_j_ijthermalsci_2023_108499 crossref_primary_10_2196_42452 crossref_primary_10_3390_jcm10102172 crossref_primary_10_1016_S2589_7500_21_00022_4 crossref_primary_10_1038_s41598_020_78184_7 crossref_primary_10_1136_bmjopen_2023_079243 crossref_primary_10_1016_j_jcrc_2022_154106 crossref_primary_10_2147_CLEP_S404580 crossref_primary_10_1055_s_0041_1742218 crossref_primary_10_1097_EA9_0000000000000014 crossref_primary_10_1109_JAS_2023_123123 crossref_primary_10_17816_psaic1547 crossref_primary_10_1007_s10462_022_10304_3 crossref_primary_10_1038_s41591_021_01614_0 crossref_primary_10_1016_j_cmpb_2023_107723 crossref_primary_10_1093_jamia_ocad085 crossref_primary_10_3390_jcm12124095 crossref_primary_10_3390_ijms23137132 crossref_primary_10_1109_JBHI_2024_3379998 crossref_primary_10_3390_app13106138 crossref_primary_10_5492_wjccm_v11_i5_317 crossref_primary_10_1016_j_cmpb_2022_107161 crossref_primary_10_1007_s12028_022_01504_4 crossref_primary_10_1016_j_jacadv_2022_100153 crossref_primary_10_1002_osp4_571 crossref_primary_10_1016_j_compbiomed_2021_105041 crossref_primary_10_4258_hir_2021_27_3_241 crossref_primary_10_1161_JAHA_121_021940 crossref_primary_10_37394_23206_2022_21_85 crossref_primary_10_1016_j_radonc_2020_09_054 crossref_primary_10_3389_fmed_2023_1174429 crossref_primary_10_1007_s00520_021_06774_w crossref_primary_10_1016_j_artmed_2024_102802 crossref_primary_10_1002_ijc_34248 crossref_primary_10_1007_s44163_024_00114_7 crossref_primary_10_1186_s12911_021_01566_y crossref_primary_10_1038_s41598_022_09706_8 crossref_primary_10_1016_S2589_7500_21_00084_4 crossref_primary_10_1017_ice_2021_114 crossref_primary_10_1038_s41746_022_00679_6 crossref_primary_10_1093_jalm_jfad094 crossref_primary_10_1111_ijcp_13685 crossref_primary_10_1016_j_accpm_2022_101167 crossref_primary_10_1093_gerona_glac107 crossref_primary_10_3389_fbinf_2023_1112113 crossref_primary_10_29328_journal_acr_1001072 crossref_primary_10_1038_s41598_021_91297_x crossref_primary_10_2196_20268 |
Cites_doi | 10.1109/JPROC.2015.2501978 10.1016/0005-2795(75)90109-9 10.1016/S2213-2600(18)30300-X 10.1001/jama.1993.03510200084037 10.1038/s41551-018-0304-0 10.1177/0272989X06295361 10.1371/journal.pone.0206862 10.1016/S2589-7500(19)30024-X 10.1046/j.1525-1497.2002.10750.x 10.1097/EDE.0b013e3181c30fb2 10.2147/CLEP.S20247 10.1371/journal.pone.0177678 10.7326/M14-0698 10.1093/jamia/ocy032 10.1002/sim.5941 10.1177/1740774515602688 10.1007/s00134-005-2763-5 10.1007/s00134-005-2762-6 10.1378/chest.100.6.1619 10.1109/TKDE.2008.239 10.1097/00003246-200209000-00008 10.1097/MCC.0000000000000135 10.1111/j.1399-6576.2009.01948.x 10.1001/jama.2019.10306 10.1162/neco.1997.9.8.1735 10.1097/00003246-199811000-00016 |
ContentType | Journal Article |
Copyright | 2020 The Author(s). Published by Elsevier Ltd. This is an Open Access article under the CC BY 4.0 license Copyright © 2020 The Author(s). Published by Elsevier Ltd. This is an Open Access article under the CC BY 4.0 license. Published by Elsevier Ltd.. All rights reserved. |
Copyright_xml | – notice: 2020 The Author(s). Published by Elsevier Ltd. This is an Open Access article under the CC BY 4.0 license – notice: Copyright © 2020 The Author(s). Published by Elsevier Ltd. This is an Open Access article under the CC BY 4.0 license. Published by Elsevier Ltd.. All rights reserved. |
DBID | 6I. AAFTH CGR CUY CVF ECM EIF NPM AAYXX CITATION DOA |
DOI | 10.1016/S2589-7500(20)30018-2 |
DatabaseName | ScienceDirect Open Access Titles Elsevier:ScienceDirect:Open Access Medline MEDLINE MEDLINE (Ovid) MEDLINE MEDLINE PubMed CrossRef DOAJ Directory of Open Access Journals |
DatabaseTitle | MEDLINE Medline Complete MEDLINE with Full Text PubMed MEDLINE (Ovid) CrossRef |
DatabaseTitleList | MEDLINE |
Database_xml | – sequence: 1 dbid: DOA name: Open Access: DOAJ - Directory of Open Access Journals url: https://www.doaj.org/ sourceTypes: Open Website – sequence: 2 dbid: NPM name: PubMed url: https://proxy.k.utb.cz/login?url=http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed sourceTypes: Index Database – sequence: 3 dbid: EIF name: MEDLINE url: https://proxy.k.utb.cz/login?url=https://www.webofscience.com/wos/medline/basic-search sourceTypes: Index Database |
DeliveryMethod | fulltext_linktorsrc |
EISSN | 2589-7500 |
EndPage | e191 |
ExternalDocumentID | oai_doaj_org_article_3ba93ebadb0841d38f9b17d3de6b2490 10_1016_S2589_7500_20_30018_2 33328078 S2589750020300182 |
Genre | Research Support, Non-U.S. Gov't Journal Article |
GrantInformation | Novo Nordisk Foundation and the Innovation Fund Denmark. |
GroupedDBID | .1- .FO 0SF 53G 6I. AAEDW AAFTH AALRI AAXUO ACLIJ AEXQZ AFRHN AITUG AJUYK ALMA_UNASSIGNED_HOLDINGS EBS EJD FDB GROUPED_DOAJ M41 M~E NCXOZ OK1 ROL W2D Z5R 0R~ AAMRU CGR CUY CVF ECM EIF NPM AAYXX CITATION |
ID | FETCH-LOGICAL-c596t-cb3bdee4be175d0cac533ce9a4e8e60a4c8d70472248c8d52e425f33b15701c53 |
IEDL.DBID | DOA |
ISSN | 2589-7500 |
IngestDate | Thu Jul 04 21:10:15 EDT 2024 Thu Sep 26 17:28:00 EDT 2024 Sat Sep 28 08:33:15 EDT 2024 Wed May 17 00:26:12 EDT 2023 |
IsDoiOpenAccess | true |
IsOpenAccess | true |
IsPeerReviewed | true |
IsScholarly | true |
Issue | 4 |
Language | English |
License | This is an open access article under the CC BY license. Copyright © 2020 The Author(s). Published by Elsevier Ltd. This is an Open Access article under the CC BY 4.0 license. Published by Elsevier Ltd.. All rights reserved. |
LinkModel | DirectLink |
MergedId | FETCHMERGED-LOGICAL-c596t-cb3bdee4be175d0cac533ce9a4e8e60a4c8d70472248c8d52e425f33b15701c53 |
OpenAccessLink | https://doaj.org/article/3ba93ebadb0841d38f9b17d3de6b2490 |
PMID | 33328078 |
ParticipantIDs | doaj_primary_oai_doaj_org_article_3ba93ebadb0841d38f9b17d3de6b2490 crossref_primary_10_1016_S2589_7500_20_30018_2 pubmed_primary_33328078 elsevier_sciencedirect_doi_10_1016_S2589_7500_20_30018_2 |
PublicationCentury | 2000 |
PublicationDate | April 2020 2020-04-00 2020-04-01 |
PublicationDateYYYYMMDD | 2020-04-01 |
PublicationDate_xml | – month: 04 year: 2020 text: April 2020 |
PublicationDecade | 2020 |
PublicationPlace | England |
PublicationPlace_xml | – name: England |
PublicationTitle | The Lancet. Digital health |
PublicationTitleAlternate | Lancet Digit Health |
PublicationYear | 2020 |
Publisher | Elsevier Ltd Elsevier |
Publisher_xml | – name: Elsevier Ltd – name: Elsevier |
References | Matthews (bib25) 1975; 405 Reps, Schuemie, Suchard, Ryan, Rijnbeek (bib31) 2018; 25 Boughorbel, Jarray, El-Anbari (bib26) 2017; 12 Austin, Steyerberg (bib28) 2014; 33 Steyerberg (bib30) 2009 Johnson, Ghassemi, Nemati, Niehaus, Clifton, Clifford (bib8) 2016; 104 Steyerberg, Vickers, Cook (bib29) 2009; 21 Meyer, Zverinski, Pfahringer (bib37) 2018; 6 Hochreiter, Schmidhuber (bib17) 1997; 9 Salluh, Soares (bib6) 2014; 20 Lundberg, Nair, Vavilala (bib22) 2018; 2 Lundberg, Lee (bib11) 2017; 1 Kahneman, Lovallo, Sibony (bib7) 2011; 89 Metnitz, Moreno, Almeida (bib24) 2005; 31 O’Neil (bib41) 2016 Cohen, Amarasingham, Shah, Xie, Lo (bib40) 2014; 33 Lemeshow, Teres, Klar, Avrunin, Gehlbach, Rapoport (bib4) 1993; 270 Sandegaard, Schmidt, Sørensen, Pedersen, Ehrenstein, Schmidt (bib14) 2015; 7 Shapley (bib23) 1988 McGee (bib27) 2002; 17 Ribeiro, Singh, Guestrin (bib12) 2016 Platt (bib32) 1999 Meiring, Dixit, Harris (bib38) 2018; 13 Vickers, Elkin (bib34) 2006; 26 Glance, Osler, Dick (bib1) 2002; 30 Aczon, Ledbetter, Ho (bib9) 2017 Moreno, Metnitz, Almeida (bib15) 2005; 31 Christensen, Johansen, Christiansen, Jensen, Lemeshow (bib35) 2011; 3 Nielsen, Thorsen-Meyer, Belling (bib10) 2019; 1 Moons, Altman, Reitsma (bib13) 2015; 162 Haibo, Garcia (bib20) 2009; 21 Baldi, Brunak (bib21) 2001 Shah, Milstein, Bagley (bib44) 2019 Strand, Søreide, Aardal, Flaatten (bib36) 2009; 53 Guyon (bib16) 2003; 3 Cawley, Talbot (bib18) 2010; 11 Guo, Pleiss, Sun, Weinberger (bib33) 2017 Lachin (bib42) 2016; 13 Pearl (bib43) 2009 Vincent, de Mendonça, Cantraine (bib3) 1998; 26 Chollet (bib19) 2017 (bib39) 2016; 59 Knaus, Wagner, Draper (bib2) 1991; 100 Zimmerman, Draper, Wagner (bib5) 2002 33328073 - Lancet Digit Health. 2020 Apr;2(4):e152-e153 Salluh (10.1016/S2589-7500(20)30018-2_bib6) 2014; 20 Hochreiter (10.1016/S2589-7500(20)30018-2_bib17) 1997; 9 Matthews (10.1016/S2589-7500(20)30018-2_bib25) 1975; 405 Steyerberg (10.1016/S2589-7500(20)30018-2_bib29) 2009; 21 Zimmerman (10.1016/S2589-7500(20)30018-2_bib5) 2002 Lundberg (10.1016/S2589-7500(20)30018-2_bib22) 2018; 2 Vincent (10.1016/S2589-7500(20)30018-2_bib3) 1998; 26 Haibo (10.1016/S2589-7500(20)30018-2_bib20) 2009; 21 Vickers (10.1016/S2589-7500(20)30018-2_bib34) 2006; 26 Johnson (10.1016/S2589-7500(20)30018-2_bib8) 2016; 104 Platt (10.1016/S2589-7500(20)30018-2_bib32) 1999 Nielsen (10.1016/S2589-7500(20)30018-2_bib10) 2019; 1 McGee (10.1016/S2589-7500(20)30018-2_bib27) 2002; 17 Lundberg (10.1016/S2589-7500(20)30018-2_bib11) 2017; 1 Boughorbel (10.1016/S2589-7500(20)30018-2_bib26) 2017; 12 Shah (10.1016/S2589-7500(20)30018-2_bib44) 2019 Cohen (10.1016/S2589-7500(20)30018-2_bib40) 2014; 33 Lachin (10.1016/S2589-7500(20)30018-2_bib42) 2016; 13 Ribeiro (10.1016/S2589-7500(20)30018-2_bib12) 2016 Metnitz (10.1016/S2589-7500(20)30018-2_bib24) 2005; 31 Guyon (10.1016/S2589-7500(20)30018-2_bib16) 2003; 3 Lemeshow (10.1016/S2589-7500(20)30018-2_bib4) 1993; 270 Shapley (10.1016/S2589-7500(20)30018-2_bib23) 1988 (10.1016/S2589-7500(20)30018-2_bib39) 2016; 59 Pearl (10.1016/S2589-7500(20)30018-2_bib43) 2009 Steyerberg (10.1016/S2589-7500(20)30018-2_bib30) 2009 Glance (10.1016/S2589-7500(20)30018-2_bib1) 2002; 30 Austin (10.1016/S2589-7500(20)30018-2_bib28) 2014; 33 Guo (10.1016/S2589-7500(20)30018-2_bib33) 2017 Chollet (10.1016/S2589-7500(20)30018-2_bib19) 2017 Moreno (10.1016/S2589-7500(20)30018-2_bib15) 2005; 31 Reps (10.1016/S2589-7500(20)30018-2_bib31) 2018; 25 Knaus (10.1016/S2589-7500(20)30018-2_bib2) 1991; 100 Cawley (10.1016/S2589-7500(20)30018-2_bib18) 2010; 11 Kahneman (10.1016/S2589-7500(20)30018-2_bib7) 2011; 89 Baldi (10.1016/S2589-7500(20)30018-2_bib21) 2001 Strand (10.1016/S2589-7500(20)30018-2_bib36) 2009; 53 Christensen (10.1016/S2589-7500(20)30018-2_bib35) 2011; 3 Meyer (10.1016/S2589-7500(20)30018-2_bib37) 2018; 6 O’Neil (10.1016/S2589-7500(20)30018-2_bib41) 2016 Moons (10.1016/S2589-7500(20)30018-2_bib13) 2015; 162 Meiring (10.1016/S2589-7500(20)30018-2_bib38) 2018; 13 Aczon (10.1016/S2589-7500(20)30018-2_bib9) 2017 Sandegaard (10.1016/S2589-7500(20)30018-2_bib14) 2015; 7 |
References_xml | – volume: 405 start-page: 442 year: 1975 end-page: 451 ident: bib25 article-title: Comparison of the predicted and observed secondary structure of T4 phage lysozyme publication-title: Biochim Biophys Acta contributor: fullname: Matthews – start-page: 121 year: 2002 end-page: 139 ident: bib5 article-title: Comparing ICU populations: background and current methods publication-title: Evaluating critical care contributor: fullname: Wagner – volume: 33 start-page: 517 year: 2014 end-page: 535 ident: bib28 article-title: Graphical assessment of internal and external calibration of logistic regression models by using loess smoothers publication-title: Stat Med contributor: fullname: Steyerberg – year: 2017 ident: bib33 article-title: On calibration of modern neural networks publication-title: arXiv contributor: fullname: Weinberger – volume: 3 start-page: 1157 year: 2003 end-page: 1182 ident: bib16 article-title: An introduction to variable and feature selection publication-title: J Mach Learn Res contributor: fullname: Guyon – volume: 12 year: 2017 ident: bib26 article-title: Optimal classifier for imbalanced data using Matthews Correlation Coefficient metric publication-title: PLoS One contributor: fullname: El-Anbari – year: 2016 ident: bib41 publication-title: Weapons of math destruction: how big data increases inequality and threatens democracy contributor: fullname: O’Neil – volume: 17 start-page: 647 year: 2002 end-page: 650 ident: bib27 article-title: Simplifying likelihood ratios publication-title: J Gen Intern Med contributor: fullname: McGee – volume: 2 start-page: 749 year: 2018 end-page: 760 ident: bib22 article-title: Explainable machine-learning predictions for the prevention of hypoxaemia during surgery publication-title: Nat Biomed Eng contributor: fullname: Vavilala – volume: 26 start-page: 565 year: 2006 end-page: 574 ident: bib34 article-title: Decision curve analysis: a novel method for evaluating prediction models publication-title: Med Decis Making contributor: fullname: Elkin – volume: 53 start-page: 595 year: 2009 end-page: 600 ident: bib36 article-title: A comparison of SAPS II and SAPS 3 in a Norwegian intensive care unit population publication-title: Acta Anaesthesiol Scand contributor: fullname: Flaatten – volume: 104 start-page: 444 year: 2016 end-page: 466 ident: bib8 article-title: Machine learning and decision support in critical care publication-title: Proc IEEE contributor: fullname: Clifford – volume: 162 start-page: W1 year: 2015 end-page: W73 ident: bib13 article-title: Transparent Reporting of a multivariable prediction model for Individual Prognosis Or Diagnosis (TRIPOD): explanation and elaboration publication-title: Ann Intern Med contributor: fullname: Reitsma – volume: 21 start-page: 128 year: 2009 end-page: 138 ident: bib29 article-title: Assessing the performance of prediction models publication-title: Epidemiology contributor: fullname: Cook – volume: 7 start-page: 449 year: 2015 ident: bib14 article-title: The Danish National Patient Registry: a review of content, data quality, and research potential publication-title: Clin Epidemiol contributor: fullname: Schmidt – year: 2001 ident: bib21 publication-title: Bioinformatics—the machine learning approach contributor: fullname: Brunak – volume: 30 start-page: 1995 year: 2002 end-page: 2002 ident: bib1 article-title: Identifying quality outliers in a large, multiple-institution database by using customized versions of the Simplified Acute Physiology Score II and the Mortality Probability Model II0 publication-title: Crit Care Med contributor: fullname: Dick – start-page: 1135 year: 2016 end-page: 1144 ident: bib12 article-title: ‘Why should I trust you?’ publication-title: Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining contributor: fullname: Guestrin – start-page: 31 year: 1988 end-page: 41 ident: bib23 article-title: A value for n-person games publication-title: The Shapley value contributor: fullname: Shapley – volume: 20 start-page: 557 year: 2014 end-page: 565 ident: bib6 article-title: ICU severity of illness scores publication-title: Curr Opin Crit Care contributor: fullname: Soares – volume: 26 start-page: 1793 year: 1998 end-page: 1800 ident: bib3 article-title: Use of the SOFA score to assess the incidence of organ dysfunction/failure in intensive care units: results of a multicenter, prospective study. Working group on ‘sepsis-related problems’ of the European Society of Intensive Care Medicine publication-title: Crit Care Med contributor: fullname: Cantraine – volume: 13 year: 2018 ident: bib38 article-title: Optimal intensive care outcome prediction over time using machine learning publication-title: PLoS One contributor: fullname: Harris – year: 2017 ident: bib9 article-title: Dynamic mortality risk predictions in pediatric critical care using recurrent neural networks publication-title: arXive contributor: fullname: Ho – volume: 33 start-page: 1139 year: 2014 end-page: 1147 ident: bib40 article-title: The legal and ethical concerns that arise from using complex predictive analytics in health care publication-title: Heal Aff Anal Heal Care contributor: fullname: Lo – volume: 1 start-page: 4765 year: 2017 end-page: 4774 ident: bib11 article-title: A unified approach to interpreting model predictions publication-title: Adv Neur In contributor: fullname: Lee – volume: 21 start-page: 1263 year: 2009 end-page: 1284 ident: bib20 article-title: Learning from imbalanced data publication-title: IEEE Trans Knowl Data Eng contributor: fullname: Garcia – volume: 31 start-page: 1336 year: 2005 end-page: 1344 ident: bib24 article-title: SAPS 3—from evaluation of the patient to evaluation of the intensive care unit. Part 1: objectives, methods and cohort description publication-title: Intensive Care Med contributor: fullname: Almeida – start-page: 61 year: 1999 end-page: 74 ident: bib32 article-title: Probabilistic outputs for support vector machines and comparisons to regularized likelihood methods publication-title: Advances in large margin classifiers contributor: fullname: Platt – volume: 100 start-page: 1619 year: 1991 end-page: 1636 ident: bib2 article-title: The APACHE III prognostic system. Risk prediction of hospital mortality for critically ill hospitalized adults publication-title: Chest contributor: fullname: Draper – volume: 89 start-page: 50 year: 2011 end-page: 60 ident: bib7 article-title: Before you make that big decision publication-title: Harv Bus Rev contributor: fullname: Sibony – volume: 25 start-page: 969 year: 2018 end-page: 975 ident: bib31 article-title: Design and implementation of a standardized framework to generate and evaluate patient-level prediction models using observational healthcare data publication-title: J Am Med Informatics Assoc contributor: fullname: Rijnbeek – volume: 270 start-page: 2478 year: 1993 end-page: 2486 ident: bib4 article-title: Mortality probability models (MPM II) based on an international cohort of intensive care unit patients publication-title: JAMA contributor: fullname: Rapoport – volume: 9 start-page: 1735 year: 1997 end-page: 1780 ident: bib17 article-title: Long short-term memory publication-title: Neural Comput contributor: fullname: Schmidhuber – volume: 6 start-page: 905 year: 2018 end-page: 914 ident: bib37 article-title: Machine learning for real-time prediction of complications in critical care: a retrospective study publication-title: Lancet Respir Med contributor: fullname: Pfahringer – year: 2009 ident: bib30 publication-title: Clinical prediction models contributor: fullname: Steyerberg – volume: 3 start-page: 203 year: 2011 end-page: 211 ident: bib35 article-title: Comparison of Charlson comorbidity index with SAPS and APACHE scores for prediction of mortality following intensive care publication-title: Clin Epidemiol contributor: fullname: Lemeshow – year: 2019 ident: bib44 article-title: Making machine learning models clinically useful publication-title: JAMA contributor: fullname: Bagley – volume: 11 start-page: 2079 year: 2010 end-page: 2107 ident: bib18 article-title: On over-fitting in model selection and subsequent selection bias in performance evaluation publication-title: J Mach Learn Res contributor: fullname: Talbot – volume: 1 start-page: e78 year: 2019 end-page: e89 ident: bib10 article-title: Survival prediction in intensive-care units based on aggregation of long-term disease history and acute physiology: a retrospective study of the Danish National Patient Registry and electronic patient records publication-title: Lancet Digital Health contributor: fullname: Belling – volume: 59 start-page: 1 year: 2016 end-page: 88 ident: bib39 article-title: Regulation (EU) 2016/679 of the European Parliament and of the Council of 27 April 2016 on the protection of natural persons with regard to the processing of personal data and on the free movement of such data, and repealing Directive 95/46/EC (General Data Protection Regulation) publication-title: Off J Eur Communities – volume: 13 start-page: 161 year: 2016 end-page: 168 ident: bib42 article-title: Fallacies of last observation carried forward analyses publication-title: Clin Trials contributor: fullname: Lachin – volume: 31 start-page: 1345 year: 2005 end-page: 1355 ident: bib15 article-title: SAPS 3—from evaluation of the patient to evaluation of the intensive care unit. Part 2: development of a prognostic model for hospital mortality at ICU admission publication-title: Intensive Care Med contributor: fullname: Almeida – year: 2009 ident: bib43 publication-title: Causality contributor: fullname: Pearl – year: 2017 ident: bib19 publication-title: Deep learning with Python contributor: fullname: Chollet – volume: 104 start-page: 444 year: 2016 ident: 10.1016/S2589-7500(20)30018-2_bib8 article-title: Machine learning and decision support in critical care publication-title: Proc IEEE doi: 10.1109/JPROC.2015.2501978 contributor: fullname: Johnson – volume: 3 start-page: 1157 year: 2003 ident: 10.1016/S2589-7500(20)30018-2_bib16 article-title: An introduction to variable and feature selection publication-title: J Mach Learn Res contributor: fullname: Guyon – volume: 405 start-page: 442 year: 1975 ident: 10.1016/S2589-7500(20)30018-2_bib25 article-title: Comparison of the predicted and observed secondary structure of T4 phage lysozyme publication-title: Biochim Biophys Acta doi: 10.1016/0005-2795(75)90109-9 contributor: fullname: Matthews – volume: 6 start-page: 905 year: 2018 ident: 10.1016/S2589-7500(20)30018-2_bib37 article-title: Machine learning for real-time prediction of complications in critical care: a retrospective study publication-title: Lancet Respir Med doi: 10.1016/S2213-2600(18)30300-X contributor: fullname: Meyer – volume: 11 start-page: 2079 year: 2010 ident: 10.1016/S2589-7500(20)30018-2_bib18 article-title: On over-fitting in model selection and subsequent selection bias in performance evaluation publication-title: J Mach Learn Res contributor: fullname: Cawley – year: 2017 ident: 10.1016/S2589-7500(20)30018-2_bib33 article-title: On calibration of modern neural networks publication-title: arXiv contributor: fullname: Guo – year: 2016 ident: 10.1016/S2589-7500(20)30018-2_bib41 contributor: fullname: O’Neil – year: 2017 ident: 10.1016/S2589-7500(20)30018-2_bib9 article-title: Dynamic mortality risk predictions in pediatric critical care using recurrent neural networks publication-title: arXive contributor: fullname: Aczon – start-page: 31 year: 1988 ident: 10.1016/S2589-7500(20)30018-2_bib23 article-title: A value for n-person games contributor: fullname: Shapley – volume: 59 start-page: 1 year: 2016 ident: 10.1016/S2589-7500(20)30018-2_bib39 publication-title: Off J Eur Communities – volume: 270 start-page: 2478 year: 1993 ident: 10.1016/S2589-7500(20)30018-2_bib4 article-title: Mortality probability models (MPM II) based on an international cohort of intensive care unit patients publication-title: JAMA doi: 10.1001/jama.1993.03510200084037 contributor: fullname: Lemeshow – volume: 2 start-page: 749 year: 2018 ident: 10.1016/S2589-7500(20)30018-2_bib22 article-title: Explainable machine-learning predictions for the prevention of hypoxaemia during surgery publication-title: Nat Biomed Eng doi: 10.1038/s41551-018-0304-0 contributor: fullname: Lundberg – start-page: 61 year: 1999 ident: 10.1016/S2589-7500(20)30018-2_bib32 article-title: Probabilistic outputs for support vector machines and comparisons to regularized likelihood methods contributor: fullname: Platt – volume: 26 start-page: 565 year: 2006 ident: 10.1016/S2589-7500(20)30018-2_bib34 article-title: Decision curve analysis: a novel method for evaluating prediction models publication-title: Med Decis Making doi: 10.1177/0272989X06295361 contributor: fullname: Vickers – volume: 13 year: 2018 ident: 10.1016/S2589-7500(20)30018-2_bib38 article-title: Optimal intensive care outcome prediction over time using machine learning publication-title: PLoS One doi: 10.1371/journal.pone.0206862 contributor: fullname: Meiring – volume: 1 start-page: e78 year: 2019 ident: 10.1016/S2589-7500(20)30018-2_bib10 article-title: Survival prediction in intensive-care units based on aggregation of long-term disease history and acute physiology: a retrospective study of the Danish National Patient Registry and electronic patient records publication-title: Lancet Digital Health doi: 10.1016/S2589-7500(19)30024-X contributor: fullname: Nielsen – volume: 17 start-page: 647 year: 2002 ident: 10.1016/S2589-7500(20)30018-2_bib27 article-title: Simplifying likelihood ratios publication-title: J Gen Intern Med doi: 10.1046/j.1525-1497.2002.10750.x contributor: fullname: McGee – year: 2017 ident: 10.1016/S2589-7500(20)30018-2_bib19 contributor: fullname: Chollet – volume: 21 start-page: 128 year: 2009 ident: 10.1016/S2589-7500(20)30018-2_bib29 article-title: Assessing the performance of prediction models publication-title: Epidemiology doi: 10.1097/EDE.0b013e3181c30fb2 contributor: fullname: Steyerberg – year: 2009 ident: 10.1016/S2589-7500(20)30018-2_bib30 contributor: fullname: Steyerberg – volume: 3 start-page: 203 year: 2011 ident: 10.1016/S2589-7500(20)30018-2_bib35 article-title: Comparison of Charlson comorbidity index with SAPS and APACHE scores for prediction of mortality following intensive care publication-title: Clin Epidemiol doi: 10.2147/CLEP.S20247 contributor: fullname: Christensen – year: 2001 ident: 10.1016/S2589-7500(20)30018-2_bib21 contributor: fullname: Baldi – volume: 12 year: 2017 ident: 10.1016/S2589-7500(20)30018-2_bib26 article-title: Optimal classifier for imbalanced data using Matthews Correlation Coefficient metric publication-title: PLoS One doi: 10.1371/journal.pone.0177678 contributor: fullname: Boughorbel – volume: 162 start-page: W1 year: 2015 ident: 10.1016/S2589-7500(20)30018-2_bib13 article-title: Transparent Reporting of a multivariable prediction model for Individual Prognosis Or Diagnosis (TRIPOD): explanation and elaboration publication-title: Ann Intern Med doi: 10.7326/M14-0698 contributor: fullname: Moons – start-page: 121 year: 2002 ident: 10.1016/S2589-7500(20)30018-2_bib5 article-title: Comparing ICU populations: background and current methods contributor: fullname: Zimmerman – volume: 25 start-page: 969 year: 2018 ident: 10.1016/S2589-7500(20)30018-2_bib31 article-title: Design and implementation of a standardized framework to generate and evaluate patient-level prediction models using observational healthcare data publication-title: J Am Med Informatics Assoc doi: 10.1093/jamia/ocy032 contributor: fullname: Reps – volume: 33 start-page: 517 year: 2014 ident: 10.1016/S2589-7500(20)30018-2_bib28 article-title: Graphical assessment of internal and external calibration of logistic regression models by using loess smoothers publication-title: Stat Med doi: 10.1002/sim.5941 contributor: fullname: Austin – year: 2009 ident: 10.1016/S2589-7500(20)30018-2_bib43 contributor: fullname: Pearl – start-page: 1135 year: 2016 ident: 10.1016/S2589-7500(20)30018-2_bib12 article-title: ‘Why should I trust you?’ contributor: fullname: Ribeiro – volume: 13 start-page: 161 year: 2016 ident: 10.1016/S2589-7500(20)30018-2_bib42 article-title: Fallacies of last observation carried forward analyses publication-title: Clin Trials doi: 10.1177/1740774515602688 contributor: fullname: Lachin – volume: 31 start-page: 1345 year: 2005 ident: 10.1016/S2589-7500(20)30018-2_bib15 article-title: SAPS 3—from evaluation of the patient to evaluation of the intensive care unit. Part 2: development of a prognostic model for hospital mortality at ICU admission publication-title: Intensive Care Med doi: 10.1007/s00134-005-2763-5 contributor: fullname: Moreno – volume: 7 start-page: 449 year: 2015 ident: 10.1016/S2589-7500(20)30018-2_bib14 article-title: The Danish National Patient Registry: a review of content, data quality, and research potential publication-title: Clin Epidemiol contributor: fullname: Sandegaard – volume: 31 start-page: 1336 year: 2005 ident: 10.1016/S2589-7500(20)30018-2_bib24 article-title: SAPS 3—from evaluation of the patient to evaluation of the intensive care unit. Part 1: objectives, methods and cohort description publication-title: Intensive Care Med doi: 10.1007/s00134-005-2762-6 contributor: fullname: Metnitz – volume: 100 start-page: 1619 year: 1991 ident: 10.1016/S2589-7500(20)30018-2_bib2 article-title: The APACHE III prognostic system. Risk prediction of hospital mortality for critically ill hospitalized adults publication-title: Chest doi: 10.1378/chest.100.6.1619 contributor: fullname: Knaus – volume: 21 start-page: 1263 year: 2009 ident: 10.1016/S2589-7500(20)30018-2_bib20 article-title: Learning from imbalanced data publication-title: IEEE Trans Knowl Data Eng doi: 10.1109/TKDE.2008.239 contributor: fullname: Haibo – volume: 89 start-page: 50 year: 2011 ident: 10.1016/S2589-7500(20)30018-2_bib7 article-title: Before you make that big decision publication-title: Harv Bus Rev contributor: fullname: Kahneman – volume: 30 start-page: 1995 year: 2002 ident: 10.1016/S2589-7500(20)30018-2_bib1 article-title: Identifying quality outliers in a large, multiple-institution database by using customized versions of the Simplified Acute Physiology Score II and the Mortality Probability Model II0 publication-title: Crit Care Med doi: 10.1097/00003246-200209000-00008 contributor: fullname: Glance – volume: 20 start-page: 557 year: 2014 ident: 10.1016/S2589-7500(20)30018-2_bib6 article-title: ICU severity of illness scores publication-title: Curr Opin Crit Care doi: 10.1097/MCC.0000000000000135 contributor: fullname: Salluh – volume: 33 start-page: 1139 year: 2014 ident: 10.1016/S2589-7500(20)30018-2_bib40 article-title: The legal and ethical concerns that arise from using complex predictive analytics in health care publication-title: Heal Aff Anal Heal Care contributor: fullname: Cohen – volume: 53 start-page: 595 year: 2009 ident: 10.1016/S2589-7500(20)30018-2_bib36 article-title: A comparison of SAPS II and SAPS 3 in a Norwegian intensive care unit population publication-title: Acta Anaesthesiol Scand doi: 10.1111/j.1399-6576.2009.01948.x contributor: fullname: Strand – volume: 1 start-page: 4765 year: 2017 ident: 10.1016/S2589-7500(20)30018-2_bib11 article-title: A unified approach to interpreting model predictions publication-title: Adv Neur In contributor: fullname: Lundberg – year: 2019 ident: 10.1016/S2589-7500(20)30018-2_bib44 article-title: Making machine learning models clinically useful publication-title: JAMA doi: 10.1001/jama.2019.10306 contributor: fullname: Shah – volume: 9 start-page: 1735 year: 1997 ident: 10.1016/S2589-7500(20)30018-2_bib17 article-title: Long short-term memory publication-title: Neural Comput doi: 10.1162/neco.1997.9.8.1735 contributor: fullname: Hochreiter – volume: 26 start-page: 1793 year: 1998 ident: 10.1016/S2589-7500(20)30018-2_bib3 article-title: Use of the SOFA score to assess the incidence of organ dysfunction/failure in intensive care units: results of a multicenter, prospective study. Working group on ‘sepsis-related problems’ of the European Society of Intensive Care Medicine publication-title: Crit Care Med doi: 10.1097/00003246-199811000-00016 contributor: fullname: Vincent |
SSID | ssj0002213777 |
Score | 2.5514464 |
Snippet | Many mortality prediction models have been developed for patients in intensive care units (ICUs); most are based on data available at ICU admission. We... Background: Many mortality prediction models have been developed for patients in intensive care units (ICUs); most are based on data available at ICU... |
SourceID | doaj crossref pubmed elsevier |
SourceType | Open Website Aggregation Database Index Database Publisher |
StartPage | e179 |
SubjectTerms | Aged Algorithms Area Under Curve Cohort Studies Critical Illness Data Analysis Electronic Health Records Female Hospital Mortality Hospitalization Humans Intensive Care Units Machine Learning Male Middle Aged Models, Biological Prognosis Retrospective Studies Risk Assessment ROC Curve Simplified Acute Physiology Score |
Title | Dynamic and explainable machine learning prediction of mortality in patients in the intensive care unit: a retrospective study of high-frequency data in electronic patient records |
URI | https://dx.doi.org/10.1016/S2589-7500(20)30018-2 https://www.ncbi.nlm.nih.gov/pubmed/33328078 https://doaj.org/article/3ba93ebadb0841d38f9b17d3de6b2490 |
Volume | 2 |
hasFullText | 1 |
inHoldings | 1 |
isFullTextHit | |
isPrint | |
link | http://utb.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwrV3BTtwwELUqTlxQq1K6lFZz6KEcUpzYyTq9FQpCSPTSInGz7HiCtirZVVgkuPBT_GBn7GTZSpW49BJZkeVYmUnmTfLmjRAfa6NRlUpnU4IjmWYZQscsCw5NwbUBZexacv69Or3QZ5fl5VqrL-aEJXngdOMOlHe1Qu-Cl0bnQZm29vk0qICVp9QhZet5uZZM_YqiLqykx7XSRWnqjMKifCrfOfixOvmpkPuKG9NlxV-BKer3_ys-rQWfk5dia0CN8DXt9pV4gd1r8fgtdZMH1wXAu8XvoRAKriNBEmHoCHEFi55_x7AJYN7CdUTchL5h1sGgq3rDY8KCMBsp7cCcMLilJ_4LOOhx2c_HqkyImrS8FIsdZ22f6Nj3wHRTXuipt864PqSPQTfb4uLk-OfRaTb0YMiasq6WWeOVD4jaI-GMIBvXED5ssHYaDVbS6caEaVSc1IaGZYH0EmiV8nk5lTlNfiM2unmHbwVojY32UvtWtzRJujq4HNsKq4rSqKAn4vNoALtIUht2xUFji1m2mC2kjRazxUQcsplWk1kpO54g_7GD_9jn_GcizGhkO4COBCZoqdlz199JTrHagVKK1YXM7v_Y2TuxWXByH2lCe2Jj2d_ie0JAS_8hOjsdzx-O_wBQ7wNg |
link.rule.ids | 315,786,790,870,2115,27957,27958 |
linkProvider | Directory of Open Access Journals |
openUrl | ctx_ver=Z39.88-2004&ctx_enc=info%3Aofi%2Fenc%3AUTF-8&rfr_id=info%3Asid%2Fsummon.serialssolutions.com&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Ajournal&rft.genre=article&rft.atitle=Dynamic+and+explainable+machine+learning+prediction+of+mortality+in+patients+in+the+intensive+care+unit%3A+a+retrospective+study+of+high-frequency+data+in+electronic+patient+records&rft.jtitle=The+Lancet.+Digital+health&rft.au=Hans-Christian+Thorsen-Meyer%2C+MD&rft.au=Annelaura+B+Nielsen%2C+PhD&rft.au=Anna+P+Nielsen%2C+MD&rft.au=Benjamin+Skov+Kaas-Hansen%2C+MD&rft.date=2020-04-01&rft.pub=Elsevier&rft.issn=2589-7500&rft.eissn=2589-7500&rft.volume=2&rft.issue=4&rft.spage=e179&rft.epage=e191&rft_id=info:doi/10.1016%2FS2589-7500%2820%2930018-2&rft.externalDBID=DOA&rft.externalDocID=oai_doaj_org_article_3ba93ebadb0841d38f9b17d3de6b2490 |
thumbnail_l | http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=2589-7500&client=summon |
thumbnail_m | http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=2589-7500&client=summon |
thumbnail_s | http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=2589-7500&client=summon |