Development and validation of a dynamic prediction model for unplanned ICU admission and mortality in hospitalized patients

Research output: Working paperPreprintResearch

Standard

Development and validation of a dynamic prediction model for unplanned ICU admission and mortality in hospitalized patients. / Placido, D.; Thorsen-Meyer, H.-C.; Kaas-Hansen, B. S.; Reguant, R.; Brunak, Søren.

2022.

Research output: Working paperPreprintResearch

Harvard

Placido, D, Thorsen-Meyer, H-C, Kaas-Hansen, BS, Reguant, R & Brunak, S 2022 'Development and validation of a dynamic prediction model for unplanned ICU admission and mortality in hospitalized patients'. https://doi.org/10.1101/2022.08.30.22279381

APA

Placido, D., Thorsen-Meyer, H-C., Kaas-Hansen, B. S., Reguant, R., & Brunak, S. (2022). Development and validation of a dynamic prediction model for unplanned ICU admission and mortality in hospitalized patients. https://doi.org/10.1101/2022.08.30.22279381

Vancouver

Placido D, Thorsen-Meyer H-C, Kaas-Hansen BS, Reguant R, Brunak S. Development and validation of a dynamic prediction model for unplanned ICU admission and mortality in hospitalized patients. 2022. https://doi.org/10.1101/2022.08.30.22279381

Author

Placido, D. ; Thorsen-Meyer, H.-C. ; Kaas-Hansen, B. S. ; Reguant, R. ; Brunak, Søren. / Development and validation of a dynamic prediction model for unplanned ICU admission and mortality in hospitalized patients. 2022.

Bibtex

@techreport{b4a9d575502f4c85a1a580d5f5af6f6b,
title = "Development and validation of a dynamic prediction model for unplanned ICU admission and mortality in hospitalized patients",
abstract = "Frequent assessment of the severity of illness for hospitalized patients is essential in clinical settings to prevent outcomes such as in-hospital mortality and unplanned ICU admission. Classical severity scores have been developed typically using relatively few patient features, especially for intensive care. Recently, deep learning-based models demonstrated better individualized risk assessments compared to classic risk scores such as SOFA and NEWS, thanks to the use of aggregated and more heterogeneous data sources for dynamic risk prediction. We investigated to what extent deep learning methods can capture patterns of longitudinal change in health status using time-stamped data from electronic health records. We used medical history data, biochemical measurements, and the clinical notes from all patients admitted to non-intensive care units in 12 hospitals in Denmarks Capital Region and Region Zealand during 2011-2016. Data from a total of 852,620 patients and 2,241,849 admissions were used to predict the composite outcome of unplanned ICU transfer and in-hospital death at different time points after admission to general departments. We subsequently examined feature interpretations of the models. The best model used all data modalities with an assessment rate of 6 hours and a prediction window of 14 days, with an AUPRC of 0.287 and AUROC of 0.898. These performances are comparable to the current state of the art and make the model suitable for further prospective validation as a risk assessment tool in a clinical setting.",
keywords = "health informatics",
author = "D. Placido and H.-C. Thorsen-Meyer and Kaas-Hansen, {B. S.} and R. Reguant and S{\o}ren Brunak",
year = "2022",
doi = "10.1101/2022.08.30.22279381",
language = "English",
type = "WorkingPaper",

}

RIS

TY - UNPB

T1 - Development and validation of a dynamic prediction model for unplanned ICU admission and mortality in hospitalized patients

AU - Placido, D.

AU - Thorsen-Meyer, H.-C.

AU - Kaas-Hansen, B. S.

AU - Reguant, R.

AU - Brunak, Søren

PY - 2022

Y1 - 2022

N2 - Frequent assessment of the severity of illness for hospitalized patients is essential in clinical settings to prevent outcomes such as in-hospital mortality and unplanned ICU admission. Classical severity scores have been developed typically using relatively few patient features, especially for intensive care. Recently, deep learning-based models demonstrated better individualized risk assessments compared to classic risk scores such as SOFA and NEWS, thanks to the use of aggregated and more heterogeneous data sources for dynamic risk prediction. We investigated to what extent deep learning methods can capture patterns of longitudinal change in health status using time-stamped data from electronic health records. We used medical history data, biochemical measurements, and the clinical notes from all patients admitted to non-intensive care units in 12 hospitals in Denmarks Capital Region and Region Zealand during 2011-2016. Data from a total of 852,620 patients and 2,241,849 admissions were used to predict the composite outcome of unplanned ICU transfer and in-hospital death at different time points after admission to general departments. We subsequently examined feature interpretations of the models. The best model used all data modalities with an assessment rate of 6 hours and a prediction window of 14 days, with an AUPRC of 0.287 and AUROC of 0.898. These performances are comparable to the current state of the art and make the model suitable for further prospective validation as a risk assessment tool in a clinical setting.

AB - Frequent assessment of the severity of illness for hospitalized patients is essential in clinical settings to prevent outcomes such as in-hospital mortality and unplanned ICU admission. Classical severity scores have been developed typically using relatively few patient features, especially for intensive care. Recently, deep learning-based models demonstrated better individualized risk assessments compared to classic risk scores such as SOFA and NEWS, thanks to the use of aggregated and more heterogeneous data sources for dynamic risk prediction. We investigated to what extent deep learning methods can capture patterns of longitudinal change in health status using time-stamped data from electronic health records. We used medical history data, biochemical measurements, and the clinical notes from all patients admitted to non-intensive care units in 12 hospitals in Denmarks Capital Region and Region Zealand during 2011-2016. Data from a total of 852,620 patients and 2,241,849 admissions were used to predict the composite outcome of unplanned ICU transfer and in-hospital death at different time points after admission to general departments. We subsequently examined feature interpretations of the models. The best model used all data modalities with an assessment rate of 6 hours and a prediction window of 14 days, with an AUPRC of 0.287 and AUROC of 0.898. These performances are comparable to the current state of the art and make the model suitable for further prospective validation as a risk assessment tool in a clinical setting.

KW - health informatics

U2 - 10.1101/2022.08.30.22279381

DO - 10.1101/2022.08.30.22279381

M3 - Preprint

BT - Development and validation of a dynamic prediction model for unplanned ICU admission and mortality in hospitalized patients

ER -

ID: 318827527