Automated prediction of early blood transfusion and mortality in trauma patients

Colin F. Mackenzie, Yulei Wang, Peter F. Hu, Shih Yu Chen, Hegang H. Chen, George Hagegeorge, Lynn G. Stansbury, Stacy Shackelford, Amechi Anazodo, Steven Barker, John Blenko, Chein I. Chang, Theresa Dinardo, Joseph DuBose, Raymond Fang, Yvette Fouche, Linda Goetz, Tom Grissom, Victor Giustina, Anthony Herrera & 13 others John Hess, Cris Imle, Matthew Lissauer, Jay Menaker, Karen Murdock, Mayur Narayan, Tim Oates, Sarah Saccicchio, Thomas Scalea, Robert Sikorski, Lynn Smith, Deborah Stein, Chris Stephens

Research output: Contribution to journalArticle

25 Citations (Scopus)

Abstract

Background: Prediction of blood transfusion needs and mortality for trauma patients in near real time is an unrealized goal. We hypothesized that analysis of pulse oximeter signals could predict blood transfusion and mortality as accurately as conventional vital signs (VSs). Methods: Continuous VS data were recorded for direct admission trauma patients with abnormal prehospital shock index (SI = heart rate [HR] / systolic blood pressure) greater than 0.62. Predictions of transfusion during the first 24 hours and in-hospital mortality using logistical regression models were compared with DeLong's method for areas under receiver operating characteristic curves (AUROCs) to determine the optimal combinations of prehospital SI and HR, continuous photoplethysmographic (PPG), oxygen saturation (SpO2), and HR-related features. Results: We enrolled 556 patients; 37 received blood within 24 hours; 7 received more than 4 U of red blood cells in less than 4 hours or "massive transfusion" (MT); and 9 died. The first 15 minutes of VS signals, including prehospital HR plus continuous PPG, and SpO2 HR signal analysis best predicted transfusion at 1 hour to 3 hours, MT, and mortality (AUROC, 0.83; p < 0.03) and no differently (p = 0.32) from a model including blood pressure. Predictions of transfusion based on the first 15 minutes of data were no different using 30 minutes to 60 minutes of data collection. SI plus PPG and SpO2 signal analysis (AUROC, 0.82) predicted 1-hour to 3-hour transfusion, MT, and mortality no differently from pulse oximeter signals alone. Conclusion: Pulse oximeter features collected in the first 15 minutes of our trauma patient resuscitation cohort, without user input, predicted early MT and mortality in the critical first hours of care better than the currently used VS such as combinations of HR and systolic blood pressure or prehospital SI alone. Level of Evidence: Therapeutic/prognostic study, level II.

Original languageEnglish (US)
Pages (from-to)1379-1385
Number of pages7
JournalJournal of Trauma and Acute Care Surgery
Volume76
Issue number6
DOIs
StatePublished - Jan 1 2014

Fingerprint

Blood Transfusion
Vital Signs
Heart Rate
Mortality
Blood Pressure
Wounds and Injuries
ROC Curve
Patient Admission
Hospital Mortality
Resuscitation
Shock
Erythrocytes
Oxygen

Keywords

  • Automated decision assist
  • blood transfusion
  • mortality
  • photopletysmograph

ASJC Scopus subject areas

  • Surgery
  • Critical Care and Intensive Care Medicine

Cite this

Mackenzie, C. F., Wang, Y., Hu, P. F., Chen, S. Y., Chen, H. H., Hagegeorge, G., ... Stephens, C. (2014). Automated prediction of early blood transfusion and mortality in trauma patients. Journal of Trauma and Acute Care Surgery, 76(6), 1379-1385. https://doi.org/10.1097/TA.0000000000000235

Automated prediction of early blood transfusion and mortality in trauma patients. / Mackenzie, Colin F.; Wang, Yulei; Hu, Peter F.; Chen, Shih Yu; Chen, Hegang H.; Hagegeorge, George; Stansbury, Lynn G.; Shackelford, Stacy; Anazodo, Amechi; Barker, Steven; Blenko, John; Chang, Chein I.; Dinardo, Theresa; DuBose, Joseph; Fang, Raymond; Fouche, Yvette; Goetz, Linda; Grissom, Tom; Giustina, Victor; Herrera, Anthony; Hess, John; Imle, Cris; Lissauer, Matthew; Menaker, Jay; Murdock, Karen; Narayan, Mayur; Oates, Tim; Saccicchio, Sarah; Scalea, Thomas; Sikorski, Robert; Smith, Lynn; Stein, Deborah; Stephens, Chris.

In: Journal of Trauma and Acute Care Surgery, Vol. 76, No. 6, 01.01.2014, p. 1379-1385.

Research output: Contribution to journalArticle

Mackenzie, CF, Wang, Y, Hu, PF, Chen, SY, Chen, HH, Hagegeorge, G, Stansbury, LG, Shackelford, S, Anazodo, A, Barker, S, Blenko, J, Chang, CI, Dinardo, T, DuBose, J, Fang, R, Fouche, Y, Goetz, L, Grissom, T, Giustina, V, Herrera, A, Hess, J, Imle, C, Lissauer, M, Menaker, J, Murdock, K, Narayan, M, Oates, T, Saccicchio, S, Scalea, T, Sikorski, R, Smith, L, Stein, D & Stephens, C 2014, 'Automated prediction of early blood transfusion and mortality in trauma patients', Journal of Trauma and Acute Care Surgery, vol. 76, no. 6, pp. 1379-1385. https://doi.org/10.1097/TA.0000000000000235
Mackenzie, Colin F. ; Wang, Yulei ; Hu, Peter F. ; Chen, Shih Yu ; Chen, Hegang H. ; Hagegeorge, George ; Stansbury, Lynn G. ; Shackelford, Stacy ; Anazodo, Amechi ; Barker, Steven ; Blenko, John ; Chang, Chein I. ; Dinardo, Theresa ; DuBose, Joseph ; Fang, Raymond ; Fouche, Yvette ; Goetz, Linda ; Grissom, Tom ; Giustina, Victor ; Herrera, Anthony ; Hess, John ; Imle, Cris ; Lissauer, Matthew ; Menaker, Jay ; Murdock, Karen ; Narayan, Mayur ; Oates, Tim ; Saccicchio, Sarah ; Scalea, Thomas ; Sikorski, Robert ; Smith, Lynn ; Stein, Deborah ; Stephens, Chris. / Automated prediction of early blood transfusion and mortality in trauma patients. In: Journal of Trauma and Acute Care Surgery. 2014 ; Vol. 76, No. 6. pp. 1379-1385.
@article{7048176e0cf1487c959f8f5e3cc0de6a,
title = "Automated prediction of early blood transfusion and mortality in trauma patients",
abstract = "Background: Prediction of blood transfusion needs and mortality for trauma patients in near real time is an unrealized goal. We hypothesized that analysis of pulse oximeter signals could predict blood transfusion and mortality as accurately as conventional vital signs (VSs). Methods: Continuous VS data were recorded for direct admission trauma patients with abnormal prehospital shock index (SI = heart rate [HR] / systolic blood pressure) greater than 0.62. Predictions of transfusion during the first 24 hours and in-hospital mortality using logistical regression models were compared with DeLong's method for areas under receiver operating characteristic curves (AUROCs) to determine the optimal combinations of prehospital SI and HR, continuous photoplethysmographic (PPG), oxygen saturation (SpO2), and HR-related features. Results: We enrolled 556 patients; 37 received blood within 24 hours; 7 received more than 4 U of red blood cells in less than 4 hours or {"}massive transfusion{"} (MT); and 9 died. The first 15 minutes of VS signals, including prehospital HR plus continuous PPG, and SpO2 HR signal analysis best predicted transfusion at 1 hour to 3 hours, MT, and mortality (AUROC, 0.83; p < 0.03) and no differently (p = 0.32) from a model including blood pressure. Predictions of transfusion based on the first 15 minutes of data were no different using 30 minutes to 60 minutes of data collection. SI plus PPG and SpO2 signal analysis (AUROC, 0.82) predicted 1-hour to 3-hour transfusion, MT, and mortality no differently from pulse oximeter signals alone. Conclusion: Pulse oximeter features collected in the first 15 minutes of our trauma patient resuscitation cohort, without user input, predicted early MT and mortality in the critical first hours of care better than the currently used VS such as combinations of HR and systolic blood pressure or prehospital SI alone. Level of Evidence: Therapeutic/prognostic study, level II.",
keywords = "Automated decision assist, blood transfusion, mortality, photopletysmograph",
author = "Mackenzie, {Colin F.} and Yulei Wang and Hu, {Peter F.} and Chen, {Shih Yu} and Chen, {Hegang H.} and George Hagegeorge and Stansbury, {Lynn G.} and Stacy Shackelford and Amechi Anazodo and Steven Barker and John Blenko and Chang, {Chein I.} and Theresa Dinardo and Joseph DuBose and Raymond Fang and Yvette Fouche and Linda Goetz and Tom Grissom and Victor Giustina and Anthony Herrera and John Hess and Cris Imle and Matthew Lissauer and Jay Menaker and Karen Murdock and Mayur Narayan and Tim Oates and Sarah Saccicchio and Thomas Scalea and Robert Sikorski and Lynn Smith and Deborah Stein and Chris Stephens",
year = "2014",
month = "1",
day = "1",
doi = "10.1097/TA.0000000000000235",
language = "English (US)",
volume = "76",
pages = "1379--1385",
journal = "Journal of Trauma and Acute Care Surgery",
issn = "2163-0755",
publisher = "Lippincott Williams and Wilkins",
number = "6",

}

TY - JOUR

T1 - Automated prediction of early blood transfusion and mortality in trauma patients

AU - Mackenzie, Colin F.

AU - Wang, Yulei

AU - Hu, Peter F.

AU - Chen, Shih Yu

AU - Chen, Hegang H.

AU - Hagegeorge, George

AU - Stansbury, Lynn G.

AU - Shackelford, Stacy

AU - Anazodo, Amechi

AU - Barker, Steven

AU - Blenko, John

AU - Chang, Chein I.

AU - Dinardo, Theresa

AU - DuBose, Joseph

AU - Fang, Raymond

AU - Fouche, Yvette

AU - Goetz, Linda

AU - Grissom, Tom

AU - Giustina, Victor

AU - Herrera, Anthony

AU - Hess, John

AU - Imle, Cris

AU - Lissauer, Matthew

AU - Menaker, Jay

AU - Murdock, Karen

AU - Narayan, Mayur

AU - Oates, Tim

AU - Saccicchio, Sarah

AU - Scalea, Thomas

AU - Sikorski, Robert

AU - Smith, Lynn

AU - Stein, Deborah

AU - Stephens, Chris

PY - 2014/1/1

Y1 - 2014/1/1

N2 - Background: Prediction of blood transfusion needs and mortality for trauma patients in near real time is an unrealized goal. We hypothesized that analysis of pulse oximeter signals could predict blood transfusion and mortality as accurately as conventional vital signs (VSs). Methods: Continuous VS data were recorded for direct admission trauma patients with abnormal prehospital shock index (SI = heart rate [HR] / systolic blood pressure) greater than 0.62. Predictions of transfusion during the first 24 hours and in-hospital mortality using logistical regression models were compared with DeLong's method for areas under receiver operating characteristic curves (AUROCs) to determine the optimal combinations of prehospital SI and HR, continuous photoplethysmographic (PPG), oxygen saturation (SpO2), and HR-related features. Results: We enrolled 556 patients; 37 received blood within 24 hours; 7 received more than 4 U of red blood cells in less than 4 hours or "massive transfusion" (MT); and 9 died. The first 15 minutes of VS signals, including prehospital HR plus continuous PPG, and SpO2 HR signal analysis best predicted transfusion at 1 hour to 3 hours, MT, and mortality (AUROC, 0.83; p < 0.03) and no differently (p = 0.32) from a model including blood pressure. Predictions of transfusion based on the first 15 minutes of data were no different using 30 minutes to 60 minutes of data collection. SI plus PPG and SpO2 signal analysis (AUROC, 0.82) predicted 1-hour to 3-hour transfusion, MT, and mortality no differently from pulse oximeter signals alone. Conclusion: Pulse oximeter features collected in the first 15 minutes of our trauma patient resuscitation cohort, without user input, predicted early MT and mortality in the critical first hours of care better than the currently used VS such as combinations of HR and systolic blood pressure or prehospital SI alone. Level of Evidence: Therapeutic/prognostic study, level II.

AB - Background: Prediction of blood transfusion needs and mortality for trauma patients in near real time is an unrealized goal. We hypothesized that analysis of pulse oximeter signals could predict blood transfusion and mortality as accurately as conventional vital signs (VSs). Methods: Continuous VS data were recorded for direct admission trauma patients with abnormal prehospital shock index (SI = heart rate [HR] / systolic blood pressure) greater than 0.62. Predictions of transfusion during the first 24 hours and in-hospital mortality using logistical regression models were compared with DeLong's method for areas under receiver operating characteristic curves (AUROCs) to determine the optimal combinations of prehospital SI and HR, continuous photoplethysmographic (PPG), oxygen saturation (SpO2), and HR-related features. Results: We enrolled 556 patients; 37 received blood within 24 hours; 7 received more than 4 U of red blood cells in less than 4 hours or "massive transfusion" (MT); and 9 died. The first 15 minutes of VS signals, including prehospital HR plus continuous PPG, and SpO2 HR signal analysis best predicted transfusion at 1 hour to 3 hours, MT, and mortality (AUROC, 0.83; p < 0.03) and no differently (p = 0.32) from a model including blood pressure. Predictions of transfusion based on the first 15 minutes of data were no different using 30 minutes to 60 minutes of data collection. SI plus PPG and SpO2 signal analysis (AUROC, 0.82) predicted 1-hour to 3-hour transfusion, MT, and mortality no differently from pulse oximeter signals alone. Conclusion: Pulse oximeter features collected in the first 15 minutes of our trauma patient resuscitation cohort, without user input, predicted early MT and mortality in the critical first hours of care better than the currently used VS such as combinations of HR and systolic blood pressure or prehospital SI alone. Level of Evidence: Therapeutic/prognostic study, level II.

KW - Automated decision assist

KW - blood transfusion

KW - mortality

KW - photopletysmograph

UR - http://www.scopus.com/inward/record.url?scp=84901975343&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84901975343&partnerID=8YFLogxK

U2 - 10.1097/TA.0000000000000235

DO - 10.1097/TA.0000000000000235

M3 - Article

VL - 76

SP - 1379

EP - 1385

JO - Journal of Trauma and Acute Care Surgery

JF - Journal of Trauma and Acute Care Surgery

SN - 2163-0755

IS - 6

ER -