Machine learning-enabled maternal risk assessment for women with pre-eclampsia (the PIERS-ML model) : a modelling study
Copyright © 2024 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..
BACKGROUND: Affecting 2-4% of pregnancies, pre-eclampsia is a leading cause of maternal death and morbidity worldwide. Using routinely available data, we aimed to develop and validate a novel machine learning-based and clinical setting-responsive time-of-disease model to rule out and rule in adverse maternal outcomes in women presenting with pre-eclampsia.
METHODS: We used health system, demographic, and clinical data from the day of first assessment with pre-eclampsia to predict a Delphi-derived composite outcome of maternal mortality or severe morbidity within 2 days. Machine learning methods, multiple imputation, and ten-fold cross-validation were used to fit models on a development dataset (75% of combined published data of 8843 patients from 11 low-income, middle-income, and high-income countries). Validation was undertaken on the unseen 25%, and an additional external validation was performed in 2901 inpatient women admitted with pre-eclampsia to two hospitals in south-east England. Predictive risk accuracy was determined by area-under-the-receiver-operator characteristic (AUROC), and risk categories were data-driven and defined by negative (-LR) and positive (+LR) likelihood ratios.
FINDINGS: Of 8843 participants, 590 (6·7%) developed the composite adverse maternal outcome within 2 days, 813 (9·2%) within 7 days, and 1083 (12·2%) at any time. An 18-variable random forest-based prediction model, PIERS-ML, was accurate (AUROC 0·80 [95% CI 0·76-0·84] vs the currently used logistic regression model, fullPIERS: AUROC 0·68 [0·63-0·74]) and categorised women into very low risk (-LR <0·1; eight [0·7%] of 1103 women), low risk (-LR 0·1 to 0·2; 321 [29·1%] women), moderate risk (-LR >0·2 and +LR <5·0; 676 [61·3%] women), high risk (+LR 5·0 to 10·0, 87 [7·9%] women), and very high risk (+LR >10·0; 11 [1·0%] women). Adverse maternal event rates were 0% for very low risk, 2% for low risk, 5% for moderate risk, 26% for high risk, and 91% for very high risk within 48 h. The 2901 women in the external validation dataset were accurately classified as being at very low risk (0% with outcomes), low risk (1%), moderate risk (4%), high risk (33%), or very high risk (67%).
INTERPRETATION: The PIERS-ML model improves identification of women with pre-eclampsia who are at lowest and greatest risk of severe adverse maternal outcomes within 2 days of assessment, and can support provision of accurate guidance to women, their families, and their maternity care providers.
FUNDING: University of Strathclyde Diversity in Data Linkage Centre for Doctoral Training, the Fetal Medicine Foundation, The Canadian Institutes of Health Research, and the Bill & Melinda Gates Foundation.
Medienart: |
E-Artikel |
---|
Erscheinungsjahr: |
2024 |
---|---|
Erschienen: |
2024 |
Enthalten in: |
Zur Gesamtaufnahme - volume:6 |
---|---|
Enthalten in: |
The Lancet. Digital health - 6(2024), 4 vom: 20. Apr., Seite e238-e250 |
Sprache: |
Englisch |
---|
Beteiligte Personen: |
Montgomery-Csobán, Tünde [VerfasserIn] |
---|
Links: |
---|
Themen: |
---|
Anmerkungen: |
Date Completed 25.03.2024 Date Revised 03.04.2024 published: Print Citation Status MEDLINE |
---|
doi: |
10.1016/S2589-7500(23)00267-4 |
---|
funding: |
|
---|---|
Förderinstitution / Projekttitel: |
|
PPN (Katalog-ID): |
NLM370087208 |
---|
LEADER | 01000caa a22002652 4500 | ||
---|---|---|---|
001 | NLM370087208 | ||
003 | DE-627 | ||
005 | 20240403235755.0 | ||
007 | cr uuu---uuuuu | ||
008 | 240324s2024 xx |||||o 00| ||eng c | ||
024 | 7 | |a 10.1016/S2589-7500(23)00267-4 |2 doi | |
028 | 5 | 2 | |a pubmed24n1363.xml |
035 | |a (DE-627)NLM370087208 | ||
035 | |a (NLM)38519152 | ||
035 | |a (PII)S2589-7500(23)00267-4 | ||
040 | |a DE-627 |b ger |c DE-627 |e rakwb | ||
041 | |a eng | ||
100 | 1 | |a Montgomery-Csobán, Tünde |e verfasserin |4 aut | |
245 | 1 | 0 | |a Machine learning-enabled maternal risk assessment for women with pre-eclampsia (the PIERS-ML model) |b a modelling study |
264 | 1 | |c 2024 | |
336 | |a Text |b txt |2 rdacontent | ||
337 | |a ƒaComputermedien |b c |2 rdamedia | ||
338 | |a ƒa Online-Ressource |b cr |2 rdacarrier | ||
500 | |a Date Completed 25.03.2024 | ||
500 | |a Date Revised 03.04.2024 | ||
500 | |a published: Print | ||
500 | |a Citation Status MEDLINE | ||
520 | |a Copyright © 2024 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. | ||
520 | |a BACKGROUND: Affecting 2-4% of pregnancies, pre-eclampsia is a leading cause of maternal death and morbidity worldwide. Using routinely available data, we aimed to develop and validate a novel machine learning-based and clinical setting-responsive time-of-disease model to rule out and rule in adverse maternal outcomes in women presenting with pre-eclampsia | ||
520 | |a METHODS: We used health system, demographic, and clinical data from the day of first assessment with pre-eclampsia to predict a Delphi-derived composite outcome of maternal mortality or severe morbidity within 2 days. Machine learning methods, multiple imputation, and ten-fold cross-validation were used to fit models on a development dataset (75% of combined published data of 8843 patients from 11 low-income, middle-income, and high-income countries). Validation was undertaken on the unseen 25%, and an additional external validation was performed in 2901 inpatient women admitted with pre-eclampsia to two hospitals in south-east England. Predictive risk accuracy was determined by area-under-the-receiver-operator characteristic (AUROC), and risk categories were data-driven and defined by negative (-LR) and positive (+LR) likelihood ratios | ||
520 | |a FINDINGS: Of 8843 participants, 590 (6·7%) developed the composite adverse maternal outcome within 2 days, 813 (9·2%) within 7 days, and 1083 (12·2%) at any time. An 18-variable random forest-based prediction model, PIERS-ML, was accurate (AUROC 0·80 [95% CI 0·76-0·84] vs the currently used logistic regression model, fullPIERS: AUROC 0·68 [0·63-0·74]) and categorised women into very low risk (-LR <0·1; eight [0·7%] of 1103 women), low risk (-LR 0·1 to 0·2; 321 [29·1%] women), moderate risk (-LR >0·2 and +LR <5·0; 676 [61·3%] women), high risk (+LR 5·0 to 10·0, 87 [7·9%] women), and very high risk (+LR >10·0; 11 [1·0%] women). Adverse maternal event rates were 0% for very low risk, 2% for low risk, 5% for moderate risk, 26% for high risk, and 91% for very high risk within 48 h. The 2901 women in the external validation dataset were accurately classified as being at very low risk (0% with outcomes), low risk (1%), moderate risk (4%), high risk (33%), or very high risk (67%) | ||
520 | |a INTERPRETATION: The PIERS-ML model improves identification of women with pre-eclampsia who are at lowest and greatest risk of severe adverse maternal outcomes within 2 days of assessment, and can support provision of accurate guidance to women, their families, and their maternity care providers | ||
520 | |a FUNDING: University of Strathclyde Diversity in Data Linkage Centre for Doctoral Training, the Fetal Medicine Foundation, The Canadian Institutes of Health Research, and the Bill & Melinda Gates Foundation | ||
650 | 4 | |a Journal Article | |
700 | 1 | |a Kavanagh, Kimberley |e verfasserin |4 aut | |
700 | 1 | |a Murray, Paul |e verfasserin |4 aut | |
700 | 1 | |a Robertson, Chris |e verfasserin |4 aut | |
700 | 1 | |a Barry, Sarah J E |e verfasserin |4 aut | |
700 | 1 | |a Vivian Ukah, U |e verfasserin |4 aut | |
700 | 1 | |a Payne, Beth A |e verfasserin |4 aut | |
700 | 1 | |a Nicolaides, Kypros H |e verfasserin |4 aut | |
700 | 1 | |a Syngelaki, Argyro |e verfasserin |4 aut | |
700 | 1 | |a Ionescu, Olivia |e verfasserin |4 aut | |
700 | 1 | |a Akolekar, Ranjit |e verfasserin |4 aut | |
700 | 1 | |a Hutcheon, Jennifer A |e verfasserin |4 aut | |
700 | 1 | |a Magee, Laura A |e verfasserin |4 aut | |
700 | 1 | |a von Dadelszen, Peter |e verfasserin |4 aut | |
700 | 0 | |a PIERS Consortium |e verfasserin |4 aut | |
700 | 1 | |a Brown, Mark A |e investigator |4 oth | |
700 | 1 | |a Davis, Gregory K |e investigator |4 oth | |
700 | 1 | |a Parker, Claire |e investigator |4 oth | |
700 | 1 | |a Walters, Barry N |e investigator |4 oth | |
700 | 1 | |a Sass, Nelson |e investigator |4 oth | |
700 | 1 | |a Ansermino, J Mark |e investigator |4 oth | |
700 | 1 | |a Cao, Vivien |e investigator |4 oth | |
700 | 1 | |a Cundiff, Geoffrey W |e investigator |4 oth | |
700 | 1 | |a von Dadelszen, Emma C M |e investigator |4 oth | |
700 | 1 | |a Douglas, M Joanne |e investigator |4 oth | |
700 | 1 | |a Dumont, Guy A |e investigator |4 oth | |
700 | 1 | |a Dunsmuir, Dustin T |e investigator |4 oth | |
700 | 1 | |a Hutcheon, Jennifer A |e investigator |4 oth | |
700 | 1 | |a Joseph, K S |e investigator |4 oth | |
700 | 1 | |a Lalji, Sayrin |e investigator |4 oth | |
700 | 1 | |a Lee, Tang |e investigator |4 oth | |
700 | 1 | |a Li, Jing |e investigator |4 oth | |
700 | 1 | |a Lim, Kenneth I |e investigator |4 oth | |
700 | 1 | |a Lisonkova, Sarka |e investigator |4 oth | |
700 | 1 | |a Lott, Paula |e investigator |4 oth | |
700 | 1 | |a Menzies, Jennifer M |e investigator |4 oth | |
700 | 1 | |a Millman, Alexandra L |e investigator |4 oth | |
700 | 1 | |a Palmer, Lynne |e investigator |4 oth | |
700 | 1 | |a Payne, Beth A |e investigator |4 oth | |
700 | 1 | |a Qu, Ziguang |e investigator |4 oth | |
700 | 1 | |a Russell, James A |e investigator |4 oth | |
700 | 1 | |a Sawchuck, Diane |e investigator |4 oth | |
700 | 1 | |a Shaw, Dorothy |e investigator |4 oth | |
700 | 1 | |a Still, D Keith |e investigator |4 oth | |
700 | 1 | |a Ukah, U Vivian |e investigator |4 oth | |
700 | 1 | |a Wagner, Brenda |e investigator |4 oth | |
700 | 1 | |a Walley, Keith R |e investigator |4 oth | |
700 | 1 | |a Hugo, Dany |e investigator |4 oth | |
700 | 1 | |a Gruslin, The Late Andrée |e investigator |4 oth | |
700 | 1 | |a Tawagi, George |e investigator |4 oth | |
700 | 1 | |a Smith, Graeme N |e investigator |4 oth | |
700 | 1 | |a Côté, Anne-Marie |e investigator |4 oth | |
700 | 1 | |a Moutquin, Jean-Marie |e investigator |4 oth | |
700 | 1 | |a Ouellet, Annie B |e investigator |4 oth | |
700 | 1 | |a Lee, Shoo K |e investigator |4 oth | |
700 | 1 | |a Duan, Tao |e investigator |4 oth | |
700 | 1 | |a Zhou, Jian |e investigator |4 oth | |
700 | 1 | |a Haniff, The Late Farizah |e investigator |4 oth | |
700 | 1 | |a Mahajan, Swati |e investigator |4 oth | |
700 | 1 | |a Noovao, Amanda |e investigator |4 oth | |
700 | 1 | |a Karjalainend, Hanna |e investigator |4 oth | |
700 | 1 | |a Kortelainen, Alja |e investigator |4 oth | |
700 | 1 | |a Laivuori, Hannele |e investigator |4 oth | |
700 | 1 | |a Ganzevoort, J Wessel |e investigator |4 oth | |
700 | 1 | |a Groen, Henk |e investigator |4 oth | |
700 | 1 | |a Kyle, Phillipa M |e investigator |4 oth | |
700 | 1 | |a Moore, M Peter |e investigator |4 oth | |
700 | 1 | |a Pullar, Barbra |e investigator |4 oth | |
700 | 1 | |a Bhutta, Zulfiqar A |e investigator |4 oth | |
700 | 1 | |a Qureshi, Rahat N |e investigator |4 oth | |
700 | 1 | |a Sikandar, Rozina |e investigator |4 oth | |
700 | 1 | |a Bhutta, The Late Shereen Z |e investigator |4 oth | |
700 | 1 | |a Cloete, Garth |e investigator |4 oth | |
700 | 1 | |a Hall, David R |e investigator |4 oth | |
700 | 1 | |a van Papendorp, The Late Erika |e investigator |4 oth | |
700 | 1 | |a Steyn, D Wilhelm |e investigator |4 oth | |
700 | 1 | |a Biryabarema, Christine |e investigator |4 oth | |
700 | 1 | |a Mirembe, Florence |e investigator |4 oth | |
700 | 1 | |a Nakimuli, Annettee |e investigator |4 oth | |
700 | 1 | |a Allotey, John |e investigator |4 oth | |
700 | 1 | |a Thangaratinam, Shakila |e investigator |4 oth | |
700 | 1 | |a Nicolaides, Kypros H |e investigator |4 oth | |
700 | 1 | |a Ionescu, Olivia |e investigator |4 oth | |
700 | 1 | |a Syngelaki, Argyro |e investigator |4 oth | |
700 | 1 | |a de Swiet, Michael |e investigator |4 oth | |
700 | 1 | |a Magee, Laura A |e investigator |4 oth | |
700 | 1 | |a von Dadelszen, Peter |e investigator |4 oth | |
700 | 1 | |a Akolekar, Ranjit |e investigator |4 oth | |
700 | 1 | |a Walker, James J |e investigator |4 oth | |
700 | 1 | |a Robson, Stephen C |e investigator |4 oth | |
700 | 1 | |a Broughton-Pipkin, Fiona |e investigator |4 oth | |
700 | 1 | |a Loughna, Pamela |e investigator |4 oth | |
700 | 1 | |a Vatish, Manu |e investigator |4 oth | |
700 | 1 | |a Redman, Christopher W G |e investigator |4 oth | |
700 | 1 | |a Barry, Sarah J E |e investigator |4 oth | |
700 | 1 | |a Kavanagh, Kimberley |e investigator |4 oth | |
700 | 1 | |a Montgomery-Csobán, Tunde |e investigator |4 oth | |
700 | 1 | |a Murray, Paul |e investigator |4 oth | |
700 | 1 | |a Robertson, Chris |e investigator |4 oth | |
700 | 1 | |a Tsigas, Eleni Z |e investigator |4 oth | |
700 | 1 | |a Woelkers, Douglas A |e investigator |4 oth | |
700 | 1 | |a Lindheimer, Marshall D |e investigator |4 oth | |
700 | 1 | |a Grobman, William A |e investigator |4 oth | |
700 | 1 | |a Sibai, Baha M |e investigator |4 oth | |
700 | 1 | |a Merialdi, Mario |e investigator |4 oth | |
700 | 1 | |a Widmer, Mariana |e investigator |4 oth | |
773 | 0 | 8 | |i Enthalten in |t The Lancet. Digital health |d 2019 |g 6(2024), 4 vom: 20. Apr., Seite e238-e250 |w (DE-627)NLM300580312 |x 2589-7500 |7 nnns |
773 | 1 | 8 | |g volume:6 |g year:2024 |g number:4 |g day:20 |g month:04 |g pages:e238-e250 |
856 | 4 | 0 | |u http://dx.doi.org/10.1016/S2589-7500(23)00267-4 |3 Volltext |
912 | |a GBV_USEFLAG_A | ||
912 | |a GBV_NLM | ||
951 | |a AR | ||
952 | |d 6 |j 2024 |e 4 |b 20 |c 04 |h e238-e250 |