1. Contexts and Rationale
The Global Burden of Disease working group of the WHO has recently shown that major trauma in its various manifestations, from road traffic accidents, interpersonal violence, self-harm to falls, remains a public health challenge and major source of mortality and handicap around the world 1. Major trauma can be defined as any injury that endangers the life or the functional integrity of a person.
Time-critical management of trauma matters most for the two main causes of death in major trauma i.e., hemorrhage and traumatic brain injury. Expedient management of any major trauma improves survival and functional outcome. Time-sensitive management of major trauma based on standardized and protocol-based care improves mortality and morbidity. To be effective, these protocols require adjustments to the individual patient and clinical context on one hand and to the organizational context and trauma system on the other hand 2 3The typical pathway of a trauma patient in a mature trauma system is summarized by figure 1.
Evidence shows that patient management even in mature trauma systems often exceeds acceptable time frames 4 5 and despite existing guidelines 6 deviations from protocol-based care are often observed. These deviations lead to a high variability in care 7 and are associated with bad outcome 2 such as inadequate hemorrhage control or delayed transfusion. Three factors explain these observations.
First, decision-making in trauma care is particularly demanding, because it requires rapid and complex decisions under time pressure in a very dynamic and multi-player environment characterized by high levels of uncertainty 8. Second, a large body of evidence has shown that human behavior and decision-making can be inconsistent because of cognitive bias or behavioral patterns in particular under stress, uncertainty and time constraint 9 10 11 typical for trauma care. Third, being a complex and multiplayer process, trauma care is affected by fragmentation 12. Fragmentation is often the result of loss of information or deformation of information. This disruptive influence prevents providers to engage with each other and commit to the care process 13.
In order to tackle these factors and enable an innovative response to the public health challenge of major trauma, the present proposal offers an integrative solution, the development of an interactive decision-support and information management platform for trauma management during the first 24 hours, the TRAUMA MATRIX.
This platform will empower all providers involved (dispatchers, paramedics, nurses, anesthetists, radiologists, surgeons, blood bank specialists, etc.) to share critical information and content in a timely manner. Results and information will be displayed in intuitive and ergonomic ways through a range of audio-visual signals. The platform will automatically integrate data from monitoring devices and the environment wherever technically feasible to reduce the need for direct provider input. Among other capacities, the platform will provide real-time, advanced probabilistic analysis of complex information to all providers along the care process. For example the platform will alert to trends (physiologic deterioration), quantify risks (shock, hemorrhage, traumatic brain injury,…) and suggest management protocols. This information system will be able to learn and dynamically adapt its predictions over the course of an individual case as well as learn from evolving epidemiology of a large and growing cohort over time to become a learning information system.
As a result, the platform is meant to streamline the care process in the first 24 hours to make it patient-centered, individualized, goal directed and empowering for all involved providers. Such a tool is not intended to become a substitute to human-decision making, but accompany clinicians and professionals to create a synergy in analogy to board instruments that help a pilot fly the plane.
For the development of the platform, the proposal takes strategic advantage of unrestricted access to an existing, prospective trauma observatory, a network of 15 French Trauma centers, the Traumabase® (traumabase.eu). The database collects detailed high quality clinical data from the scene to the discharge from Critical Care. The granularity of the data collected makes the observatory unique in Europe.
Objective
The objective of the proposal is to perform a proof of concept study. Based on clinical data from a large French trauma database, advanced mathematical prediction models will be integrated into an information platform that provides ergonomic and innovative, real-time decision-support and interactive and adaptive information management to a broad range of clinicians in major trauma management.
Hypothesis
The study group claims that the information platform enhances the clinician-driven decision-making and care process. In addition, it posits that the synergy between the clinician and the platform has a substantial potential to streamline and improve the overall trauma care process in the first 24 hours.
2. Originality and Relevance
Firstly, to the best of our knowledge, such a trauma information platform currently does not exist. To develop and design an interactive, real-time, probabilistic decision-support and information management platform constitutes a major conceptual and scientific innovation. No proof of concept study exists that evaluates this approach on a large scale for complex medical decisions such as trauma care.
Secondly, prospective and interventional clinical research becomes ever more difficult to perform, because of logistical and financial constraints and the complexity of the studied interventions. This explains why results are ever so often difficult to translate into clinical practice. Trauma is a perfect example for this dilemma, since it requires complex and multiplayer strategies that do not depend on a single intervention.
The medical community urgently needs to adopt and develop new methods to evaluate therapeutic strategies. For this reason, the medical community relies on the analysis of large amount of data for diagnosis, decision-support and treatment. The project will bring together mathematical, methodological, technological, cognitive and medical expertise and make existing advanced methods available to the medical community. In particular, the project will develop innovative methods to face existing challenges. One of these challenges is the handling of missing and heterogeneous data. Management of missing data with different coding, coupled with a complex data structure is an important research topic. The results will have an impact beyond the scope of the present proposal. In this respect, the project provides an excellent opportunity for trans-disciplinary research and collaboration.
3. Methods and structure of the project
The overall projects extends over several years and is structured into three steps:
I) Innovative and advanced statistical methods, so called machine learning, will be deployed to analyze data from a large clinical observatory of major trauma patients in France, the Traumabase®. This analysis will generate models to predict the probability of pertinent events and needs (shock, brain injury, transfusion, urgent interventions, airway management,….) that impact patient outcome in the management of major trauma. A committee of French Trauma experts will define the list of events and needs to be predicted by the tool. The algorithms will be designed to make predictions not only at a single point in time, but incorporate the trends of clinical evolution in an individual patient, as well as larger trends within the cohort over time to create a so called learning information system (duration 2 years).
II) The data-based models and algorithms developed in part I will be integrated into a user friendly, real-time, adaptive information platform. The probability of clinical events and needs will be displayed to the clinician in an ergonomic way and integrate event- and outcome-specific management recommendations based on official guidelines (2 years).
III) The impact of the deployment of the information platform on the performance of professionals, on patient-centered outcomes (mortality, morbidity, and functional outcomes) and on process variables (under- and over-triage, time on scene, in resuscitation, to theater) needs to be assessed in simulated and real cases (2 years from the end of step II).
5. Impact and expected spin offs
The development of an interactive, real-time decision-support and information management platform would constitute a major conceptual and scientific innovation. The advanced prediction methods deployed in the proposal are usually referred to as Machine Learning, which are expected to operate a paradigm shift in medicine challenging the way it is currently practiced. The disruptive consequences on the medical profession and society of such innovations could be substantial, in a positive and negative sense 15. A proof of concept study is therefore urgently needed to assess such an approach in real-life situations. It is important for the society to help the French scientific and medical community to take leadership in this exploratory process to anticipate and prevent disruptive influences. The project responds to a question with substantial societal impact and provides an excellent opportunity for trans disciplinary research. The approach raises important ethical, regulatory and legal aspects, which require a separate exploration in an ancillary project.
From a public health and clinical point of view, this program is relevant, because the combined effects of the platform could have a substantial impact on the quality of care for major trauma patients. The platform streamlines the integral clinical pathway of major trauma patients from the scene to the ICU in the first 24 hours, makes it patient-centered and goal-directed and facilitates interaction and information exchange between health professionals empowering them to have timely access to critical information. Compliance with management guidelines will increase and lead to a higher standardisation of care. Standardisation facilitates appropriate use of resources and reduces costs. With a successful proof of concept in trauma, use of the platform could be extended to other complex critical care pathologies such as sepsis. The platform could also be deployed in a simplified version to support disaster response and management. As such, the project responds to the societal and public health challenge of major trauma.
Trauma Matrix is a project supported by http://www.cvt-athena.fr/actualites/evenements-a-venir/163-traumabase-peps-maths-shs.html
https://www.traumatrix.fr/
Bibliography
1. GBD 2016 DALYs and HALE Collaborators. Global, regional, and national disability-adjusted life-years (DALYs) for 333 diseases and injuries and healthy life expectancy (HALE) for 195 countries and territories, 1990-2016: a systematic analysis for the Global Burden of Disease Study 2016. Lancet Lond Engl 2017; 390: 1260–344
2. Rice TW, Morris S, Tortella BJ, Wheeler AP, Christensen MC. Deviations from evidence-based clinical management guidelines increase mortality in critically injured trauma patients*. Crit Care Med 2012; 40: 778–86 T
3. Khan S, Allard S, Weaver A, Barber C, Davenport R, Brohi K. A major haemorrhage protocol improves the delivery of blood component therapy and reduces waste in trauma massive transfusion. Injury [Internet] 2012 [cited 2013 Jan 26];
4. Yeguiayan J-M, Garrigue D, Binquet C, et al. Medical pre-hospital management reduces mortality in severe blunt trauma: a prospective epidemiological study. Crit Care 2011; 15: R34 a
5. Hamada SR, Gauss T, Duchateau F-X, et al. Evaluation of the performance of French physician-staffed emergency medical service in the triage of major trauma patients. J Trauma Acute Care Surg 2014; 76: 1476–83
6. Rossaint R, Bouillon B, Cerny V, et al. The European guideline on management of major bleeding and coagulopathy following trauma: fourth edition. Crit Care Lond Engl 2016; 20: 100
7. Hamada SR, Gauss T, Pann J, Dünser M, Leone M, Duranteau J. European trauma guideline compliance assessment: the ETRAUSS study. Crit Care Lond Engl 2015; 19: 423
8. deMattos PC, Miller DM, Park EH. Decision making in trauma centers from the standpoint of complex adaptive systems. Ribeiro Soriano D, editor. Manag Decis 2012; 50: 1549–69
9. Tversky A, Kahneman D. Judgment under Uncertainty: Heuristics and Biases. In: Wendt D, Vlek C, editors. Util Probab Hum Decis Mak [Internet] Dordrecht: Springer Netherlands; 1975 [cited 2017 Oct 20]. p. 141–62 M
10. Gigerenzer G, Brighton H. Homo heuristicus: why biased minds make better inferences. Top Cogn Sci 2009; 1: 107–43
11. Catchpole K, Ley E, Wiegmann D, et al. A human factors subsystems approach to trauma care. JAMA Surg 2014; 149: 962–8
12. Elhauge E. The fragmentation of U.S. health care: causes and solutions. New York: Oxford University Press; 2010.
13. Stange KC. The Problem of Fragmentation and the Need for Integrative Solutions. Ann Fam Med 2009; 7: 100–3
14. Lelaidier R, Balança B, Boet S, et al. Use of a hand-held digital cognitive aid in simulated crises: the MAX randomized controlled trial. Br J Anaesth 2017
15. Cabitza F, Rasoini R, Gensini GF. Unintended Consequences of Machine Learning in Medicine. JAMA 2017; 318: 517
16. Perez P, Harrois A, Raux M, Hamada S, Nadal JP, Paugam-Burtz, Gauss T. Prediction of Haemorrhagic shock by a Bayes Network in comparison to experienced trauma experts., Group Traumabase; presented conference SFAR 2015
Study Group
Pr Jean-Pierre Nadal (JPN)
Directeur de recherche au CNRS & Directeur d’études à l’EHESS
Laboratoire de Physique Statistique (LPS, UMR CNRS-ENS-UPMC-Univ. Paris Diderot)
Centre d’Analyse et de Mathématique Sociales (CAMS, UMR CNRS-EHESS)
75006 Paris
Pr Julie Josse (JJ)
Centre des Mathématiques Appliqués
Ecole Polytechnique
Route de Saclay,
91128 Palaiseau
Pr Catherine Paugam-Burtz (CPB)
Professeur des Universités-Praticien Hospitalier
Anesthésie et Réanimation Chirurgicale Polyvalente - Hôpital Beaujon, APHP
Hôpitaux Universitaires Paris Nord Val de Seine, 92110 Clichy
Dr Tobias Gauss (TG)
Praticien Hospitalier
Anesthésie et Réanimation Chirurgicale Polyvalente - Hôpital Beaujon, APHP
Hôpitaux Universitaires Paris Nord Val de Seine, 92110 Clichy
Pr Romain Pirracchio
Professeur des Universités - Praticien Hospitalier
Anesthésie-Réanimation – Hôpital Européen Georges Pompidou – APHP
CRESS U1153, équipe ECSTRA, Hôpital Saint Louis – 75011 Paris
