Literature DB >> 30059457

Machine learning without borders? An adaptable tool to optimize mortality prediction in diverse clinical settings.

S Ariane Christie1, Alan E Hubbard, Rachael A Callcut, Morad Hameed, Fanny Nadia Dissak-Delon, David Mekolo, Arabo Saidou, Alain Chichom Mefire, Pierre Nsongoo, Rochelle A Dicker, Mitchell Jay Cohen, Catherine Juillard.   

Abstract

BACKGROUND: Mortality prediction aids clinical decision making and is necessary for quality improvement initiatives. Validated metrics rely on prespecified variables and often require advanced diagnostics, which are unfeasible in resource-constrained contexts. We hypothesize that machine learning will generate superior mortality prediction in both high-income and low- and middle-income country cohorts.
METHODS: SuperLearner, an ensemble machine-learning algorithm, was applied to data from three prospective trauma cohorts: a highest-activation cohort in the United States, a high-volume center cohort in South Africa (SA), and a multicenter registry in Cameroon. Cross-validation was used to assess model discrimination of discharge mortality by site using receiver operating characteristic curves. SuperLearner discrimination was compared with standard scoring methods. Clinical variables driving SuperLearner prediction at each site were evaluated.
RESULTS: Data from 28,212 injured patients were used to generate prediction. Discharge mortality was 17%, 1.3%, and 1.7% among US, SA, and Cameroonian cohorts. SuperLearner delivered superior prediction of discharge mortality in the United States (area under the curve [AUC], 94-97%) and vastly superior prediction in Cameroon (AUC, 90-94%) compared with conventional scoring algorithms. It provided similar prediction to standard scores in the SA cohort (AUC, 90-95%). Context-specific variables (partial thromboplastin time in the United States and hospital distance in Cameroon) were prime drivers of predicted mortality in their respective cohorts, whereas severe brain injury predicted mortality across sites.
CONCLUSIONS: Machine learning provides excellent discrimination of injury mortality in diverse settings. Unlike traditional scores, data-adaptive methods are well suited to optimizing precise site-specific prediction regardless of diagnostic capabilities or data set inclusion allowing for individualized decision making and expanded access to quality improvement programming. LEVEL OF EVIDENCE: Prognostic and therapeutic, level II and III.

Entities:  

Mesh:

Year:  2018        PMID: 30059457      PMCID: PMC6225991          DOI: 10.1097/TA.0000000000002044

Source DB:  PubMed          Journal:  J Trauma Acute Care Surg        ISSN: 2163-0755            Impact factor:   3.313


  28 in total

1.  The Major Trauma Outcome Study: establishing national norms for trauma care.

Authors:  H R Champion; W S Copes; W J Sacco; M M Lawnick; S L Keast; L W Bain; M E Flanagan; C F Frey
Journal:  J Trauma       Date:  1990-11

2.  Revised trauma score: a triage tool in the accident and emergency department.

Authors:  D A Gilpin; P G Nelson
Journal:  Injury       Date:  1991-01       Impact factor: 2.586

3.  The utility of the Kampala trauma score as a triage tool in a sub-Saharan African trauma cohort.

Authors:  Bryce Haac; Carlos Varela; Andrew Geyer; Bruce Cairns; Anthony Charles
Journal:  World J Surg       Date:  2015-02       Impact factor: 3.352

4.  Comparison of modified Kampala trauma score with trauma mortality prediction model and trauma-injury severity score: A National Trauma Data Bank Study.

Authors:  Serhat Akay; Ahmet Mucteba Ozturk; Huriye Akay
Journal:  Am J Emerg Med       Date:  2017-02-16       Impact factor: 2.469

5.  Regularization Paths for Generalized Linear Models via Coordinate Descent.

Authors:  Jerome Friedman; Trevor Hastie; Rob Tibshirani
Journal:  J Stat Softw       Date:  2010       Impact factor: 6.440

6.  Time-dependent prediction and evaluation of variable importance using superlearning in high-dimensional clinical data.

Authors:  Alan Hubbard; Ivan Diaz Munoz; Anna Decker; John B Holcomb; Martin A Schreiber; Eileen M Bulger; Karen J Brasel; Erin E Fox; Deborah J del Junco; Charles E Wade; Mohammad H Rahbar; Bryan A Cotton; Herb A Phelan; John G Myers; Louis H Alarcon; Peter Muskat; Mitchell J Cohen
Journal:  J Trauma Acute Care Surg       Date:  2013-07       Impact factor: 3.313

Review 7.  Trauma scoring systems.

Authors:  Rolf Lefering
Journal:  Curr Opin Crit Care       Date:  2012-12       Impact factor: 3.687

8.  Improved predictions from a severity characterization of trauma (ASCOT) over Trauma and Injury Severity Score (TRISS): results of an independent evaluation.

Authors:  H R Champion; W S Copes; W J Sacco; C F Frey; J W Holcroft; D B Hoyt; J A Weigelt
Journal:  J Trauma       Date:  1996-01

Review 9.  Predicting outcome after multiple trauma: which scoring system?

Authors:  M N Chawda; F Hildebrand; H C Pape; P V Giannoudis
Journal:  Injury       Date:  2004-04       Impact factor: 2.586

10.  Variable importance and prediction methods for longitudinal problems with missing variables.

Authors:  Iván Díaz; Alan Hubbard; Anna Decker; Mitchell Cohen
Journal:  PLoS One       Date:  2015-03-27       Impact factor: 3.240

View more
  2 in total

Review 1.  Artificial Intelligence and Surgical Decision-making.

Authors:  Tyler J Loftus; Patrick J Tighe; Amanda C Filiberto; Philip A Efron; Scott C Brakenridge; Alicia M Mohr; Parisa Rashidi; Gilbert R Upchurch; Azra Bihorac
Journal:  JAMA Surg       Date:  2020-02-01       Impact factor: 14.766

2.  Conventional regression analysis and machine learning in prediction of anastomotic leakage and pulmonary complications after esophagogastric cancer surgery.

Authors:  Robert T van Kooten; Renu R Bahadoer; Bouwdewijn Ter Buurkes de Vries; Michel W J M Wouters; Rob A E M Tollenaar; Henk H Hartgrink; Hein Putter; Johan L Dikken
Journal:  J Surg Oncol       Date:  2022-05-03       Impact factor: 2.885

  2 in total

北京卡尤迪生物科技股份有限公司 © 2022-2023.