Statistical machines for trauma hospital outcomes research: Application to the PRospective, Observational, Multi-center Major trauma Transfusion (PROMMTT) study

Sara E. Moore, Anna Decker, Alan Hubbard, Rachael A. Callcut, Erin E. Fox, Deborah J. Del Junco, John B. Holcomb, Mohammad H. Rahbar, Charles E. Wade, Martin Schreiber, Louis H. Alarcon, Karen Brasel, Eileen M. Bulger, Bryan A. Cotton, Peter Muskat, John G. Myers, Herb A. Phelan, Mitchell J. Cohen

    Research output: Contribution to journalArticle

    2 Citations (Scopus)

    Abstract

    Improving the treatment of trauma, a leading cause of death worldwide, is of great clinical and public health interest. This analysis introduces flexible statistical methods for estimating center-level effects on individual outcomes in the context of highly variable patient populations, such as those of the PRospective, Observational, Multi-center Major Trauma Transfusion study. Ten US level I trauma centers enrolled a total of 1,245 trauma patients who survived at least 30 minutes after admission and received at least one unit of red blood cells. Outcomes included death, multiple organ failure, substantial bleeding, and transfusion of blood products. The centers involved were classified as either large or small-volume based on the number of massive transfusion patients enrolled during the study period. We focused on estimation of parameters inspired by causal inference, specifically estimated impacts on patient outcomes related to the volume of the trauma hospital that treated them. We defined this association as the change in mean outcomes of interest that would be observed if, contrary to fact, subjects from large-volume sites were treated at small-volume sites (the effect of treatment among the treated). We estimated this parameter using three different methods, some of which use data-adaptive machine learning tools to derive the outcome models, minimizing residual confounding by reducing model misspecification. Differences between unadjusted and adjusted estimators sometimes differed dramatically, demonstrating the need to account for differences in patient characteristics in clinic comparisons. In addition, the estimators based on robust adjustment methods showed potential impacts of hospital volume. For instance, we estimated a survival benefit for patients who were treated at large-volume sites, which was not apparent in simpler, unadjusted comparisons. By removing arbitrary modeling decisions from the estimation process and concentrating on parameters that have more direct policy implications, these potentially automated approaches allow methodological standardization across similar comparativeness effectiveness studies.

    Original languageEnglish (US)
    Article numbere0136438
    JournalPLoS One
    Volume10
    Issue number8
    DOIs
    StatePublished - Aug 21 2015

    Fingerprint

    Trauma Centers
    Blood
    Outcome Assessment (Health Care)
    Wounds and Injuries
    Public health
    Standardization
    Learning systems
    Statistical methods
    Cells
    death
    blood transfusion
    Multiple Organ Failure
    Decision Support Techniques
    artificial intelligence
    concentrating
    standardization
    Blood Transfusion
    hemorrhage
    Cause of Death
    public health

    ASJC Scopus subject areas

    • Agricultural and Biological Sciences(all)
    • Biochemistry, Genetics and Molecular Biology(all)
    • Medicine(all)

    Cite this

    Statistical machines for trauma hospital outcomes research : Application to the PRospective, Observational, Multi-center Major trauma Transfusion (PROMMTT) study. / Moore, Sara E.; Decker, Anna; Hubbard, Alan; Callcut, Rachael A.; Fox, Erin E.; Del Junco, Deborah J.; Holcomb, John B.; Rahbar, Mohammad H.; Wade, Charles E.; Schreiber, Martin; Alarcon, Louis H.; Brasel, Karen; Bulger, Eileen M.; Cotton, Bryan A.; Muskat, Peter; Myers, John G.; Phelan, Herb A.; Cohen, Mitchell J.

    In: PLoS One, Vol. 10, No. 8, e0136438, 21.08.2015.

    Research output: Contribution to journalArticle

    Moore, SE, Decker, A, Hubbard, A, Callcut, RA, Fox, EE, Del Junco, DJ, Holcomb, JB, Rahbar, MH, Wade, CE, Schreiber, M, Alarcon, LH, Brasel, K, Bulger, EM, Cotton, BA, Muskat, P, Myers, JG, Phelan, HA & Cohen, MJ 2015, 'Statistical machines for trauma hospital outcomes research: Application to the PRospective, Observational, Multi-center Major trauma Transfusion (PROMMTT) study', PLoS One, vol. 10, no. 8, e0136438. https://doi.org/10.1371/journal.pone.0136438
    Moore, Sara E. ; Decker, Anna ; Hubbard, Alan ; Callcut, Rachael A. ; Fox, Erin E. ; Del Junco, Deborah J. ; Holcomb, John B. ; Rahbar, Mohammad H. ; Wade, Charles E. ; Schreiber, Martin ; Alarcon, Louis H. ; Brasel, Karen ; Bulger, Eileen M. ; Cotton, Bryan A. ; Muskat, Peter ; Myers, John G. ; Phelan, Herb A. ; Cohen, Mitchell J. / Statistical machines for trauma hospital outcomes research : Application to the PRospective, Observational, Multi-center Major trauma Transfusion (PROMMTT) study. In: PLoS One. 2015 ; Vol. 10, No. 8.
    @article{a611faca24fc45aeb8d1945785234468,
    title = "Statistical machines for trauma hospital outcomes research: Application to the PRospective, Observational, Multi-center Major trauma Transfusion (PROMMTT) study",
    abstract = "Improving the treatment of trauma, a leading cause of death worldwide, is of great clinical and public health interest. This analysis introduces flexible statistical methods for estimating center-level effects on individual outcomes in the context of highly variable patient populations, such as those of the PRospective, Observational, Multi-center Major Trauma Transfusion study. Ten US level I trauma centers enrolled a total of 1,245 trauma patients who survived at least 30 minutes after admission and received at least one unit of red blood cells. Outcomes included death, multiple organ failure, substantial bleeding, and transfusion of blood products. The centers involved were classified as either large or small-volume based on the number of massive transfusion patients enrolled during the study period. We focused on estimation of parameters inspired by causal inference, specifically estimated impacts on patient outcomes related to the volume of the trauma hospital that treated them. We defined this association as the change in mean outcomes of interest that would be observed if, contrary to fact, subjects from large-volume sites were treated at small-volume sites (the effect of treatment among the treated). We estimated this parameter using three different methods, some of which use data-adaptive machine learning tools to derive the outcome models, minimizing residual confounding by reducing model misspecification. Differences between unadjusted and adjusted estimators sometimes differed dramatically, demonstrating the need to account for differences in patient characteristics in clinic comparisons. In addition, the estimators based on robust adjustment methods showed potential impacts of hospital volume. For instance, we estimated a survival benefit for patients who were treated at large-volume sites, which was not apparent in simpler, unadjusted comparisons. By removing arbitrary modeling decisions from the estimation process and concentrating on parameters that have more direct policy implications, these potentially automated approaches allow methodological standardization across similar comparativeness effectiveness studies.",
    author = "Moore, {Sara E.} and Anna Decker and Alan Hubbard and Callcut, {Rachael A.} and Fox, {Erin E.} and {Del Junco}, {Deborah J.} and Holcomb, {John B.} and Rahbar, {Mohammad H.} and Wade, {Charles E.} and Martin Schreiber and Alarcon, {Louis H.} and Karen Brasel and Bulger, {Eileen M.} and Cotton, {Bryan A.} and Peter Muskat and Myers, {John G.} and Phelan, {Herb A.} and Cohen, {Mitchell J.}",
    year = "2015",
    month = "8",
    day = "21",
    doi = "10.1371/journal.pone.0136438",
    language = "English (US)",
    volume = "10",
    journal = "PLoS One",
    issn = "1932-6203",
    publisher = "Public Library of Science",
    number = "8",

    }

    TY - JOUR

    T1 - Statistical machines for trauma hospital outcomes research

    T2 - Application to the PRospective, Observational, Multi-center Major trauma Transfusion (PROMMTT) study

    AU - Moore, Sara E.

    AU - Decker, Anna

    AU - Hubbard, Alan

    AU - Callcut, Rachael A.

    AU - Fox, Erin E.

    AU - Del Junco, Deborah J.

    AU - Holcomb, John B.

    AU - Rahbar, Mohammad H.

    AU - Wade, Charles E.

    AU - Schreiber, Martin

    AU - Alarcon, Louis H.

    AU - Brasel, Karen

    AU - Bulger, Eileen M.

    AU - Cotton, Bryan A.

    AU - Muskat, Peter

    AU - Myers, John G.

    AU - Phelan, Herb A.

    AU - Cohen, Mitchell J.

    PY - 2015/8/21

    Y1 - 2015/8/21

    N2 - Improving the treatment of trauma, a leading cause of death worldwide, is of great clinical and public health interest. This analysis introduces flexible statistical methods for estimating center-level effects on individual outcomes in the context of highly variable patient populations, such as those of the PRospective, Observational, Multi-center Major Trauma Transfusion study. Ten US level I trauma centers enrolled a total of 1,245 trauma patients who survived at least 30 minutes after admission and received at least one unit of red blood cells. Outcomes included death, multiple organ failure, substantial bleeding, and transfusion of blood products. The centers involved were classified as either large or small-volume based on the number of massive transfusion patients enrolled during the study period. We focused on estimation of parameters inspired by causal inference, specifically estimated impacts on patient outcomes related to the volume of the trauma hospital that treated them. We defined this association as the change in mean outcomes of interest that would be observed if, contrary to fact, subjects from large-volume sites were treated at small-volume sites (the effect of treatment among the treated). We estimated this parameter using three different methods, some of which use data-adaptive machine learning tools to derive the outcome models, minimizing residual confounding by reducing model misspecification. Differences between unadjusted and adjusted estimators sometimes differed dramatically, demonstrating the need to account for differences in patient characteristics in clinic comparisons. In addition, the estimators based on robust adjustment methods showed potential impacts of hospital volume. For instance, we estimated a survival benefit for patients who were treated at large-volume sites, which was not apparent in simpler, unadjusted comparisons. By removing arbitrary modeling decisions from the estimation process and concentrating on parameters that have more direct policy implications, these potentially automated approaches allow methodological standardization across similar comparativeness effectiveness studies.

    AB - Improving the treatment of trauma, a leading cause of death worldwide, is of great clinical and public health interest. This analysis introduces flexible statistical methods for estimating center-level effects on individual outcomes in the context of highly variable patient populations, such as those of the PRospective, Observational, Multi-center Major Trauma Transfusion study. Ten US level I trauma centers enrolled a total of 1,245 trauma patients who survived at least 30 minutes after admission and received at least one unit of red blood cells. Outcomes included death, multiple organ failure, substantial bleeding, and transfusion of blood products. The centers involved were classified as either large or small-volume based on the number of massive transfusion patients enrolled during the study period. We focused on estimation of parameters inspired by causal inference, specifically estimated impacts on patient outcomes related to the volume of the trauma hospital that treated them. We defined this association as the change in mean outcomes of interest that would be observed if, contrary to fact, subjects from large-volume sites were treated at small-volume sites (the effect of treatment among the treated). We estimated this parameter using three different methods, some of which use data-adaptive machine learning tools to derive the outcome models, minimizing residual confounding by reducing model misspecification. Differences between unadjusted and adjusted estimators sometimes differed dramatically, demonstrating the need to account for differences in patient characteristics in clinic comparisons. In addition, the estimators based on robust adjustment methods showed potential impacts of hospital volume. For instance, we estimated a survival benefit for patients who were treated at large-volume sites, which was not apparent in simpler, unadjusted comparisons. By removing arbitrary modeling decisions from the estimation process and concentrating on parameters that have more direct policy implications, these potentially automated approaches allow methodological standardization across similar comparativeness effectiveness studies.

    UR - http://www.scopus.com/inward/record.url?scp=84942844049&partnerID=8YFLogxK

    UR - http://www.scopus.com/inward/citedby.url?scp=84942844049&partnerID=8YFLogxK

    U2 - 10.1371/journal.pone.0136438

    DO - 10.1371/journal.pone.0136438

    M3 - Article

    C2 - 26296088

    AN - SCOPUS:84942844049

    VL - 10

    JO - PLoS One

    JF - PLoS One

    SN - 1932-6203

    IS - 8

    M1 - e0136438

    ER -