GiViTI

(Italian Group for the Evaluation of

Interventions in Intensive Care Medicine)

IRCCS-Institute for Pharmacological

Research ‘Mario Negri’

Villa Camozzi, Via GB Camozzi, 3

24020 Ranica (BG) - Italy

Tel. +39 035.45 35 313

Fax.+ 39 035.45 35 371

E-mail: giviti@marionegri.it

Skype: giviti

The project is supported by the European Commission through the Seventh Framework Programme

under Grant Agreement no. 602714.

CREACTIVE Collaborative REsearch on

ACute Traumatic brain

Injury in intensiVe care medicine in Europe

GiViTI (Italian Group for the

Evaluation of Interventions in

Intensive Care Medicine)

Institute for Pharmacological

Research ‘Mario Negri’

Ranica (BG) - Italy

Semmelweis University

Budapest - Hungary

Medical University of Warsaw

Warsaw - Poland

General Hospital Novo Mesto

Novo Mesto - Slovenia

University Medical Center

Ljubljana - Slovenia

University of Nicosia - Nicosia

General Hospital

Nicosia - Cyprus

Ben-Gurion University of the

Negev

Beer Sheva - Israel

University of Crete Medical

School - University Hospital of

Heraklion

Heraklion - Greece

Orobix Srl

Bergamo - Italy

Partners

Contacts

GiViTI Coordinating Centre

IRCCS - INSTITUTE FOR PHARMACOLOGICAL RESEARCH

‘MARIO NEGRI’

To consolidate the existing PROSAFE

network, based on a prospective data

collection in intensive care units (ICUs).

To describe the epidemiology of

moderate-to-severe TBI in 7 countries.

To evaluate the consequences of TBI in

children, through a multidimensional

study of their outcomes.

To establish a centralized biobank

(blood and derived fluids, CSF) and a

bank of clinical imaging for patients with

TBI.

To build a prognostic model based on

clinical and biological data to predict short

and medium term outcomes of TBI

patients.

To identify the most effective clinical

interventions for optimally treating TBI

patients.

To recognize centres of excellence in

treating TBI patients.

To share data with other international

research groups adhering to the InTBIR

network.

Aims of the study

The project involves 7 partners from

Cyprus, Greece, Hungary, Israel, Italy,

Poland and Slovenia.

CREACTIVE started on 1st October, 2013

and will run until 30th September, 2018.

All ICUs already participating in the PROSAFE project

and hospitalized patients with traumatic brain injury are invited to take part in CREACTIVE. For each

patient with TBI, the centres have to fill out the case report form, which is already part of the PROSAFE database.

Follow-up will be performed six months after the trauma event, and will be two-tiered:

1. by telephone call for all patients;

2. by full patient examination to be performed in a selected subgroup of ICUs only.

In a subset of ICUs, clinical images carried out during routine clinical practice will be collected and centralized for all recruited patients. Images

will undergo computer analysis, with a view to developing automatic or semi-automatic reading systems, thereby facilitating evaluation of the

brain injury.

Biological samples will be collected and centralized in a subset of ICUs for phenotypic biomarker analysis. The choice of markers to be

studied will be made after analysis of the images has suggested the most promising pathophysiological process to investigate.

Analysis

An annual report of the results will be

produced in both aggregate and

customized form for each participating

centre. This will enable the centres to

compare their data with those of the

entire data set . Accordingly, different

quality indicators will be developed (SMR,

Calibration Belt, VLAD) in order to

statistically adjust comparisons for

differences in patient severity. An analysis

plan based on the propensity score will be

used to identify the most effective

therapeutic interventions.

Based on the data collected, it is

expected to enrol 7000 patients with

moderate-to-severe traumatic brain

injury in approximately 125 units, of

which at least 80 Italian, by the end of

the study, scheduled for September 2018.

GiViTI calibration belt to evaluate and develop prognostic models with a dichotomous outcome

GiViTI

(Gruppo italiano per la Valutazione degli

interventi in Terapia Intensiva)

IRCCS-Istituto di Ricerche Farmacologiche

Mario Negri

Villa Camozzi, Via GB Camozzi, 3

24020 Ranica (BG)

Tel. +39 035.45 35 313

Fax.+ 39 035.45 35 371

E-mail: giviti@marionegri.it

Skype: giviti

Il progetto è finanziato dalla Commissione Europea nell’ambito

del VII Programma Quadro Europeo con contratto nr. 602714.

CREACTIVE Collaborative REsearch on

ACute Traumatic brain

Injury in intensiVe care medicine in Europe

GiViTI (Gruppo italiano per la

Valutazione degli interventi in

Terapia Intensiva)

IRCCS - Istituto di Ricerche

Farmacologiche Mario Negri

Ranica (BG) - Italia

Semmelweis University

Budapest - Ungheria

Medical University of Warsaw

Varsavia - Polonia

General Hospital Novo Mesto

Novo Mesto - Slovenia

University Medical Center

Lubliana - Slovenia

University of Nicosia - Nicosia

General Hospital

Nicosia - Cipro

Ben-Gurion University of the

Negev

Beer Sheva - Israele

University of Crete Medical

School - University Hospital of

Heraklion

Heraklion - Grecia

Orobix Srl

Bergamo - Italia

Partners

Contatti

Centro di Coordinamento GiViTI IRCCS - ISTITUTO DI RICERCHE

FARMACOLOGICHE MARIO NEGRI

Consolidare la rete preesistente

PROSAFE, basata su una raccolta dati

prospettica nei reparti di terapia intensiva

(TI).

Descrivere l’epidemiologia del trauma

cranico (TBI) moderato-grave in 7 paesi.

Studiare le conseguenze del TBI nei

bambini, attraverso una valutazione

multidimensionale degli esiti.

Realizzare una biobanca (sangue e fluidi

derivati, liquor) e una banca immagini

cliniche per i pazienti con TBI.

Costruire un modello prognostico basato

su dati clinici e biologici per prevedere i

risultati a breve e lungo termine nei

pazienti con TBI.

Identificare gli interventi clinici più

efficaci per il trattamento dei pazienti con

TBI.

Riconoscere i centri di eccellenza nel

trattamento dei pazienti con TBI.

Condividere i dati raccolti con altri

gruppi di ricerca internazionali aderenti

alla rete InTBIR.

Obbiettivi dello studio

Il progetto coinvolge 7 partner da Cipro,

Grecia, Israele, Italia, Polonia, Slovenia e

Ungheria.

CREACTIVE ha avuto inizio il 1° Ottobre

2013 e terminerà il 30 Settembre 2018.

Tutte le TI che partecipano al progetto PROSAFE e

che ricoverano pazienti con trauma cranico sono

invitate a partecipare a CREACTIVE. Per ciascun

paziente con trauma cranico si dovrà compilare la scheda di raccolta dati che è parte del database PROSAFE.

A distanza di sei mesi dal trauma cranico, verrà effettuato un follow-up dei pazienti, per valutare l’esito a medio termine. Sono previsti due livelli:

1.telefonico su tutti i pazienti; 2.ambulatoriale su un sottogruppo di pazienti e di TI.

In un sottogruppo di reparti verranno raccolte e centralizzate le immagini cliniche effettuate durante la normale pratica clinica, per tutti i

pazienti reclutati. L’obiettivo è quello di sviluppare sistemi di lettura automatici o semi-automatici, che facilitino la valutazione della lesione cerebrale.

In un sottogruppo di reparti verranno raccolti e centralizzati campioni biologici per le analisi dei biomarcatori fenotipici. La scelta dei marcatori da

analizzare verrà effettuata dopo che l’analisi delle immagini avrà suggerito quale sia il processo fisiopatologico più promettente da indagare.

Le analisi

Verrà preparato un report annuale sui

risultati conseguiti, che sarà prodotto sia

in forma aggregata sia personalizzata per

ogni centro partecipante. I centri avranno

quindi la possibilità di confrontare i

propri dati con quelli dell’intero

collettivo. Verranno sviluppati diversi

indicatori di qualità dell’assistenza (SMR,

banda di calibrazione, VLAD) al fine di

aggiustare statisticamente i confronti per

le differenze nella gravità dei pazienti. Per

quanto riguarda l’individuazione degli

interventi terapeutici più efficaci, verrà

utilizzato uno schema di analisi basato sul

propensity score.

Sulla base dei dati raccolti, si prevede

l’arruolamento di 7.000 pazienti con

trauma cranico moderato-grave in circa

125 reparti, di cui almeno 80 italiani,

entro la fine dello studio.

Banda di calibrazione GiViTI per la valutazione e lo sviluppo di modelli prognostici con esito dicotomico.

Στόχοι Προγράμματος

Στο πρόγραμμα συμμετέχουν 7 χώρες: Κύπρος, Ελλάδα, Ουγγαρία, Ισραήλ, Ιταλία, Πολωνία και Σλοβενία.

Το πρόγραμμα CREACTIVE ξεκίνησε την 1η Οκτωβρίου 2013 και αναμένεται να ολοκληρωθεί στις 30 Σεπτεμβρίου 2018.

Όλες οι Μονάδες Εντατικής Θεραπείας που ήδη συμμετέχουν στο δίκτυο PROSAFE και νοσηλεύουν ασθενείς με ΚρανιοΕγκεφαλική Κάκωση είναι ευπρόσδεκτες να συμμετάσχουν στο CREACTIVE. Για τον κάθε ασθενή με ΚΕΚ, τα κέντρα πρέπει να συμπληρώνουν το έντυπο αναφοράς περιστατικού που βρίσκεται στην ήδη υπάρχουσα βάση δεδομένων PROSAFE.

Θα γίνεται αξιολόγηση της έκβασης της νόσου 6 και 12 μήνες μετά το συμβάν με δύο τρόπους: 1. Τηλεφωνική επικοινωνία και συμπλήρωση ερωτηματολογίου. 2. Πλήρης κλινική αξιολόγηση που θα περιλαμβάνει κλινική εξέταση και εξειδικευμένες εξετάσεις.

Εδραίωση υφιστάμενου δικτύου PROSAFE βάσει της συνεχούς συλλογής δεδομένων στις Μονάδες Εντατικής Θεραπείας (ΜΕΘ). Περιγραφή επιδημιολογίας μέτριας και βαριάς κρανιοεγκεφαλικής κάκωσης σε 7 χώρες. Αξιολόγηση συνεπειών της κρανιοεγκεφαλικής κάκωσης σε παιδιά μέσω πολυδιάστατης διερεύνησης της έκβασης της νόσου. Εγκαθίδρυση κεντρικής βιοτράπεζας (αίματος και παράγωγων υγρών, CSF) και τράπεζας κλινοεργαστηριακών και απεικονιστικών δεδομένων για ασθενείς με κρανιοεγκεφαλικές κακώσεις. Δημιουργία προγνωστικού μοντέλου βασιζόμενου σε κλινικά και βιολογικά δεδομένα, για την πρόβλεψη των βραχυπρόθεσμων και μακροπρόθεσμων επιπλοκών ασθενών με κρανιοεγκεφαλική κάκωση. Εξεύρεση αποτελεσματικών παρεμβάσεων για τη βέλτιστη θεραπεία ασθενών με κρανιοεγκεφαλική κάκωση. Αναγνώριση κέντρων αριστείας που αφορούν στη θεραπεία ασθενών με κρανιοεγκεφαλική κάκωση. Δημιουργία κοινής βάσης δεδομένων με άλλες διεθνείς ερευνητικές ομάδες που συμμετέχουν στο δίκτυο InTBIR.

Σ’ ένα υποσύνολο Μονάδων Εντατικής Θεραπείας θα συλλέγονται οι απεικονιστικές εξετάσεις για κάθε ασθενή που θα συμμετέχει στο πρόγραμμα, θα τυγχάνουν τεχνικής επεξεργασίας και θα αποθηκεύονται ανώνυμα σε ηλεκτρονικό αρχείο, με σκοπό τη δημιουργία ενός αυτόματου ή ημιαυτόματου συστήματος αξιολόγησης.

Θα γίνεται λήψη βιολογικών δειγμάτων για περαιτέρω ανάλυση του φαινότυπου, καθώς και ανάλυση επιλεγμένων βιοδεικτών με σκοπό τη μελέτη τυχόν προδιαθεσικών παραγόντων για την ανταπόκριση στη θεραπεία και την έκβαση της νόσου.

Ανάλυση Αποτελεσμάτων

Ετήσια αναφορά αποτελεσμάτων τόσο συνολικά όσο και για την εκάστοτε Μονάδα. Επισημαίνεται ότι οι Μονάδες θα έχουν τη δυνατότητα σύγκρισης των αποτελεσμάτων τους με το σύνολο των Μονάδων του δικτύου. Θα δημιουργηθεί ένα σύνολο δεικτών το οποίο θα συμβάλει στην πιο αποτελεσματική σύγκριση των Μονάδων σε συνάρτηση με τη σοβαρότητα της κατάστασης του κάθε ασθενούς.

Η ανάλυση της τάσης των αποτελεσμάτων θα καθορίσει τις πιο αποδοτικές παρεμβάσεις όσον αφορά στη θεραπεία των ασθενών της εκάστοτε ομάδας.

Βάσει των στοιχείων που έχουν συγκεντρωθεί μέχρι το Σεπτέμβριο του 2014, αναμένεται να ενταχθούν στο πρόγραμμα 9000 ασθενείς από 125 Μονάδες Εντατικής Θεραπείας μέχρι το Σεπτέμβριο του 2018.

CREACTIVE(Collaborative REsearch on ACute Traumatic brain

Injury in intensiVe care medicine in Europe)

Προοπτική Μελέτη Συλλογής Δεδομένων & Σύγκρισης Αποτελεσματικότητας σε Ασθενείς με KρανιοΕγκεφαλική Kάκωση (ΚΕΚ) στις Μονάδες

Εντατικής Θεραπείας

Το πρόγραμμα συγχρηματοδοτείται από την Ευρωπαϊκή Επιτροπή μέσω του 7ου

Προγράμματος Πλαισίου υπό τη(συμφωνία επιχορήγησης

αρ. 602714)

GIVITI Coordinating CenterIRCCS - INSTITUTE FOR

PHARMACOLOGICAL RESEARCH‘MARIO NEGRI’

GiViTI(Italian Group for the Evaluation of

Interventions in Intensive Care Medicine)

IRCCS - Institute for PharmacologicalResearch ‘Mario Negri’

Villa Camozzi, Via GB Camozzi, 324020 Ranica (BG) - Italy

Τel.: +39 035.45 35 313Fax.: +39 035.45 35 371

Email.: giviti@marionegri.itSkype: giviti

Επικοινωνία

GiViTI (Italian Group for the Evaluation of Interventions in Intensive Care Medicine) Institute for Pharmacological Research ‘Mario Negri’ Ranica (BG) - Italy

Semmelweis University Budapest - Hungary

Medical University of WarsawWarsaw - Poland

University of Nicosia - NicosiaGeneral Hospital Nicosia - Cyprus

University Medical CenterLjubljana - Slovenia

General Hospital Novo Mesto Novo Mesto - Slovenia

Ben-Gurion University of the NegevBeer Sheva - Israel

Orobix SrlBergamo - Italy

University of Crete Medical School - University Hospital of HeraklionHeraclion - Greece

Συνεργάτες

ου

Editorial

www.thelancet.com/neurology Vol 12 December 2013 1127

See Personal View page 1200

For more on InTBIR see Editorial Lancet Neurol 2012, 11: 651

For more on long-term eff ects of mild TBI see Neuron 2012; 76: 886–99

For the study in TBI paediatric centres see Pediatr Crit Care Med 2013; 14: 811–18

For more on TBI common data elements see http://www.commondataelements.ninds.nih.gov/TBI.aspx#tab=Data_Standards

For more on FITBIR see http://fi tbir.nih.gov/

A rally for traumatic brain injury researchWhen members of the International Initiative for Traumatic Brain Injury Research (InTBIR) met in Vancouver (BC, Canada) on Oct 17–18, 2013, some notable stakeholders from the US National Institutes of Health (NIH) could not join them. The temporary shutdown of the US federal government had just been put to an end the night before. But the attendees’ feeling of frustration about the absence of some collaborators did not cloud their enthusiasm, and the InTBIR network reaffi rmed its pledge to improve patients’ outcomes by 2020.

Traumatic brain injury (TBI) research has historically been neglected and underfunded. The major funding agencies behind InTBIR (The European Commission’s Health Directorate, NIH, and the Canadian Institutes of Health Research, and their military partners) have now aligned their national programmes and dedicated more than US$80 million to the initiative. Such an unprecedented coordinated strategy refl ects the urgent need to tackle a global public-health crisis. In Europe alone, TBI causes an estimated 75 000 deaths per year and more than 1 million hospital admissions; but these estimates are thought to be even higher in other continents. Warfare, violence and terrorism, and road traffi c accidents contribute to the huge global burden of TBI; long-term sequelae are both psychological and physical. Furthermore, recurring mild TBI can lead to chronic traumatic encephalopathy—evidence that is starting to aff ect policies of sports organisations.

As documented in a Personal View, adding to the complexity of a disease for which heterogeneity is an intrinsic feature, substantial diff erences in patient management exist even among specialist centres in the setting of clinical trials. Not surprisingly then, given the lack of guidelines based on high-quality evidence, inconsistencies in management of both paediatric and adult patients are pervasive. For instance, a recent study of paediatric TBI centres in France, Spain, the UK, and the USA found a great deal of variability in key therapy goals, such as those set up at individual centres on intracranial hypertension and metabolic therapies, or brain monitoring of partial pressure of oxygen.

In Vancouver, Graham Teasdale (University of Glasgow, UK) reminded attendees of the need to strike the right balance between aiming to improve standardised versus personalised clinical care. Without denying that challenge,

InTBIR members emphasised the opportunities to exploit heterogeneity in presentation, care, and outcomes, by means of rigorous observational studies and comparative eff ectiveness analyses to tackle both standardised and personalised care. Four major InTBIR studies in the civilian population have either started recruitment or will do so soon: the Approaches and Decisions for Acute Pediatric TBI (ADAPT) study, an observational comparative eff ectiveness study of 1000 children with severe TBI; the Collaborative European NeuroTrama Eff ectiveness Research in TBI (CENTER-TBI), a longitudinal study to characterise disease phenotypes and compare clinical care in more than 5000 patients, with a concurrent registry to collect data from up to 40 000 patients; the Collaborative Research on Acute Traumatic Brain Injury in Intensive Medicine in Europe (CREACTIVE) study, an epidemiological and comparative eff ectiveness study of patients with moderate-to-severe TBI in more than 100 intensive care units; and the Transforming Research and Clinical Knowledge in Traumatic Brain Injury (TRACK-TBI) study, which will enrol 3000 patients for further analyses of comparative eff ectiveness and diagnostic and prognostic markers. Although each of these studies (plus a plethora of smaller projects also part of InTBIR) should provide a wealth of data, unique potential lies in their synergies. The integrated analysis of fi ndings will be feasible because data collection is standardised according to NIH common data elements (CDEs). And the newly implemented Federal Interagency TBI Research (FITBIR) Informatics System will equip the initiative with a data sharing platform. However, the diffi culties of establishing a transnational, open-access research culture are not to be underestimated, and the most heated discussions in Vancouver covered the intricacies of the use of CDEs, legal and ethical barriers to data sharing, and the eff ects of data sharing in the current academic system of rewards.

The InTBIR community is nevertheless committed to address these obstacles over the next few months. In doing so, they will not only accelerate progress in TBI, but also facilitate the integration of InTBIR with other brain research networks for swift public health improvements. The next meeting of the consortium to review progress will take place in 2 years—hopefully political disruption in Washington, or elsewhere, won’t hold any investigator back on that occasion. ■ The Lancet Neurology

BRAIN INJURY PROFESSIONAL24

CREACTIVEA EuRopEAn EndEAVoR To ImpRoVE

ouTComE of pATIEnTs wITh TRAumATIC BRAIn InjuRy

BackgroundTraumatic Brain Injury (TBI) is an alteration in brain function, or other evidence of brain pathology, caused by an external force1. For many years, a myriad of epidemiological and clinical studies have documented the huge burden of TBI for patients, their relatives, and the whole of society, in terms of unexpected death, long-lasting or permanent disability in previously healthy persons, di-minished or even a sharp drop in quality of life, massive direct and enormous indirect costs2, 3. Nevertheless, TBI research has historically been neglected and underfunded4, thus it is not sur-prising that the therapeutic armamentarium to improve outcome of TBI patients is far from being satisfactory5.

Recently, however, the International Initiative for Traumatic Brain Injury (InTBIR) has been set up with the long term goal of improving the clinical outcomes of these patients and lessening the global burden of the disease by 2020. InTBIR is an unprece-dented consortium of three major funding agencies (the European Commission’s Health Directorate, the NIH’s National Institute of Neurological Disorders and Stroke, and the Canadian Institutes of Health Research Institute of Neurosciences, Mental Health and Addiction), which opted to align their national programs to accel-erate progress in TBI research. Within this common endeavor, the

three agencies have independently launched a number of funding calls focused on basic and clinical TBI research, for a total invest-ment of about US$90 million4.

Four major studies and a number of smaller projects have been funded and are now in the starting blocks: the Approaches and Decisions for Acute Pediatric TBI (ADAPT) study; the Collab-orative European NeuroTrauma Effectiveness Research in TBI (CENTER-TBI); the Collaborative REsearch on ACute Traumat-ic brain Injury in intensiVe care medicine in Europe (CREAC-TIVE) study; the Transforming Research and Clinical Knowledge in Traumatic Brain Injury (TRACK-TBI) study. Despite differing in many ways, but being committed to render the resulting da-tabases accessible to the research community, all these studies will collect at least the same core data, the so-called Common Data Elements (CDEs) [6-8], thus laying the foundations for a unique international research collaboration.

Here we present the protocol of CREACTIVE, an observa-tional study that will involve about 100 intensive care units (ICUs) in seven countries: Cyprus, Greece, Hungary, Israel, Italy, Poland, and Slovenia. While mild TBI patients are variably managed in the different health services, most moderate and almost all severe TBI patients who manage to reach a hospital are admitted to an ICU. Although these patients represent only 20% of the total, they carry the main burden of the disease. Some permanent disability is estimated to occur in 10% of mild, 66% of moderate, and 100% of severe TBIs9-11. Estimated in-hospital mortality is <5% in mild TBI, while it increases to 21% in moderate, and 46% in severe

Authors: Guido Bertolini, Giulia Paci, luca Antiga, Akos Csomos, rafael kaps, isaac lazar, malgorzata mikaszewska, matteo mondini, nektaria Xirouchaki, roberto latini, Primoz Gradisek, Joanne fleming, theodoros kyprianou

BRAIN INJURY PROFESSIONAL 25

cases at six months12. Hence, the ICU is in an ideal position to adequately evaluate and monitor the bulk of the burden of the disease, identify and assess the most effective clinical interventions, and recognize excellence in TBI management.

Aims of the CREACTIVE studyThe CREACTIVE project has several aims: 1) to better describe the epidemiology of moderate-to-severe TBI in the participating countries; 2) to establish centralized repositories of clinical data (CDEs+), biological samples (blood and derived fluids, cerebro-spinal fluid) and clinical imaging data, to be exploited for prog-nostic purposes; 3) to build a prognostic model based on clinical and biological data to predict short- and long-term outcome; 4) to identify the most effective clinical interventions for optimally treating TBI patients; 5) to recognize the determinants of optimal vs. suboptimal performance.

Progress beyond the state-of-the-art Epidemiology of TBI The existing PROSAFE network, recently established through EU funding (PHEA 2007331), will be consolidated through the CREACTIVE project. The PROSAFE network was created to export the experience gained with the Italian Margherita Project in establishing clinical audit to improve the quality of care in the ICU setting. In 2012, 242 ICUs, mainly from Italy, had joined PROSAFE, recruiting 85,965 patients in total13. Of these ICUs, 199 admitted at least one TBI patient, totaling 3,344 cases. On the basis of these statistics, the network expects to recruit about 7,000 moderate to severe TBI patients over a period of 4 years.

Two features of the PROSAFE project are particularly valu-able for CREACTIVE from an epidemiological perspective. First, all kinds of healthcare facilities will be involved, from secondary to tertiary, to quaternary care level centers. This is extremely important, considering that the representativeness of the sample matters even more than its size. At the present, a large fraction of moderate to severe TBI patients continue, perhaps inevitably, to be admitted to peripheral, secondary care centers, where it is often hard to decide whether to centralize a patient to a tertiary or even a quaternary care facility. On this premise, if InTBIR is actually to be able to improve outcome of TBI patients, it is mandatory to know what happens in sec-ondary centers. Second, all admitted patients will be registered in the PROSAFE Case Report Form, while those with TBI will be registered in both the CREACTIVE and PROSAFE forms. This is of paramount importance for many reasons: it will provide a general picture of the case mix of the ICUs and their general performance, which can be taken into account in explaining variability in TBI patient management; it will be possible to monitor the participation of each ICU in the study; the software will alert the presence of eligible patients, thereby avoiding any selection bias; and, ultimately, this will foster fa-miliarity with data collection and promote the overall quality of the information gathered.

Finally, from the patient perspective, given the importance of TBI-related disabilities, mortality cannot be considered the only outcome to assess the impact of the condition. Accordingly, a follow-up will be performed six months after the trauma event, and will be two-tiered. The first level will be administered over the phone and consist of the extended version of the Glasgow

Outcome Scale (GOSe) to assess disability and the QOLIBRI-OS to evaluate the health-related quality of life. The second level will encompass a full patient examination. In the case of chil-dren, this will include a dedicated sleep disturbances study. The second level follow-up will be performed in a selected subgroup of ICUs only.

Bio-bank and bank of clinical imagingThe damage occurring after trauma is the consequence of two distinct, partially independent mechanisms, namely primary and secondary injury14. Whilst the first is directly caused by the im-pact, the second is governed by a complex set of cellular processes and biochemical cascades that occur in the minutes to days fol-lowing the trauma. Through different pathways, the most critical results of secondary injury are the progression of the hemorrhag-ic lesion and/or the cytotoxic and vasogenic edema (swelling of the brain), which ultimately cause a rise in intracranial pressure. Since this accounts for the greatest number of TBI deaths occur-ring in hospital14, the main aim of treatment in the acute stage of TBI is to control and lower intracranial hypertension15.

In this framework, it is crucial to better characterize the pro-gression of hemorrhagic lesions and cerebral edema and to in-crease our knowledge on the factors that can influence them. The substantial variability of these two phenomena among subjects not fully accounted for by the initial lesion severity suggests that an individual genetic predisposition can play a role in their mag-nitude and, consequently, in the final patient outcome.

We will study the evolution of the focal lesion volume (hemorrhagic and/or perilesional edema) through the analysis of serial computed tomography (CT) images, circulating and cerebrospinal fluid (CSF) biomarkers, and possible underly-ing genetic predisposition. There is an increasing volume of evidence that highlights the association between outcome and CT imaging16, biomarkers17, and genetic polymorphisms18. But these three important aspects have always been analyzed inde-pendently from each other, and never been collected together in the same population.

We will implement centralized repositories of detailed clini-cal data, biological samples (Bio-bank), and repeated clinical im-aging, on the same large sample of adult patients (up to 2,000). This will provide a unique opportunity for the integrated analy-sis of secondary injury following TBI.

Starting from the automated analysis of imaging data, we will evaluate the variability in the progression of both the hemor-rhagic lesion and the perilesional edema, also assessing their asso-ciation with the outcome. We will then concentrate the analysis on circulating and CSF biomarkers and genetic predisposition to the most promising mechanism: either trauma-induced co-agulopathy underlying hemorrhagic progression, or regulation of both blood-brain barrier permeability behind vasogenic edema and ischemic events underlying cytotoxic edema19.

Given the complexity of the entire picture and the expected new knowledge that may be available when the samples are ready for analysis, an ad hoc, independent, scientific advisory board (SAB) will finalize the protocol of the various determinations. Depending on the results of the imaging analysis and the new available evidence, the SAB will jointly identify which already recognized circulating and CSF biomarkers, and which other in-novative ones will be tested, in addition to the most promising genetic analysis to be performed.

BRAIN INJURY PROFESSIONAL26

Building a prognostic modelOf all the CREACTIVE project aims, the most important is defi-nitely to build an accurate prognostic model for TBI patients. In-deed, the other aims can only be achieved if such model becomes available.

In the critical care medicine field, several “severity of illness” scoring systems have been proposed, but they have inevitably proved to lack generalizability. This is particularly true for TBI, where in 2006 prognostic models were still inadequate, as reported by a sys-tematic review20, and a more recent model still proved to calibrate poorly when applied to an external cohort21. At present, a number of problems hamper the development of a reliable global prognos-tic model. First, a scoring system developed in a specific geograph-ic, economic and social context, produces biased estimates when applied to other areas. This is because unmeasured context-specific variables influence the weight of the prognostic factors included in the model22. Second, the development of prognostic models is so complicated that once built, they are unlikely to be updated for many years. At present, severity scores constructed many years ago, such as SAPS II, APACHE II, PIM2 and PRISM2, to cite just the general ones, are still commonly used. This generates a temporal bias related to the improvement of health care quality and changes in case mix over time that the model cannot account for23. Finally, the internal validity of currently used severity scores has never been adequately tested in important subgroups and so miscalibration of the model cannot be ruled out23. Consequently, when severity scores are applied to populations from the developmental sample with different case mixes, their predictive power deteriorates.

The aim of the CREACTIVE consortium is to overcome these problems. In essence, a large number of ICUs will be collecting data on a regular basis, providing enough statistical power for prognostic modeling; a prognostic model will be developed every year to avoid the temporal bias; a second yearly model will take

into account the context-specific variables, together with the pa-tients’ characteristics, to increase generalizability; the uniformity of fit of the models will be assured through the GiViTI Calibra-tion Belt24, 25 in subsets identified by a high number of prognostic variables, to assure internal validity.

Comparative effectiveness of clinical interventionsThe main goals in managing patients with moderate to severe TBI are (a) to maintain life in spite of the primary tissue injury, (b) to protect the brain from secondary neuronal injury, and (c) to pre-vent secondary injury in other organs resulting in multi-organ dys-function syndrome1. Some of the pathophysiological mechanisms that affect the outcome of these patients include brain edema, increased intracranial pressure, hyper- and hypotension, hypoxia, brain ischemia, hypercapnia, temperature regulation, hyper and hypoglycemia (peripheral or regional), hyper- and hyponatremia, seizures, sepsis, etc. A number of interventions designed to manage these derangements have been a topic of controversy, mainly due to the lack of high quality randomized-controlled trials.

Firstly, ad hoc analyses (e.g., cluster analysis) will be performed to identify the most frequently used bundles of interventions (in-cluding individual treatments). For each bundle, a propensity score (i.e., the probability of bundle assignment according to observed baseline characteristics) will be developed through a logistic regres-sion model. Comparative effectiveness analysis will be performed through both covariate adjustment and stratification, using the propensity score.

Identification of excellence in treating TBI patientsIn the literature on quality of care assessment, one of the most widely used indicators is the standardized mortality ratio (SMR), which is the ratio between observed and expected mortality in a subgroup, according to the benchmark26. When a prognostic

model is used as benchmark for the centers forming its de-velopment sample, the SMRs of the individual centers will be distributed around one. However, some centers will exhibit a statistically signifi-cant deviation from one. This means that their better- or worse-than-benchmark out-come is not due to chance variation but to potentially different performance. More-over, SMRs that do not devi-ate significantly from 1 may hide statistically significant differences from benchmark in specific classes of risk (say, patients at low risk of death), or in specific subgroups (say, patients with severe infec-tion).

The GIVITI calibration belt does instead provide this detailed information24, 25, 27. The use of such a tool will thus allow identification, on

figure 1

In this example, the GiViTI Calibration Belt shows that unit A has major problems in less severe patients (those with expected mortality lower than 0.45), where the observed mortality is higher than expected (the belt is above the bisector). Conversely, ICU B has weaknesses in more severe patients (the belt is above the bisector) and strengths in less severe ones (where the observed mortality is lower than expected, and the belt is below the bisector).

ICU A ICU B

Ratio O/E = 1.25Con�dence interval O/E (95%) = (1.1,1.41)

Total patients = 210 Total expected deaths (E) = 65.4

Total observed deaths (O) = 82

Calibration belt

Polynomial degree:1GiViTI Cal. Test, p 0.003

Obs

erve

d m

orta

lity

Expected mortality

0.0 0.2 0.4 0.6 0.8 1.0

Con�dence Over Under Level the bisector the bisector

80% 0.00 - 0.56 Never95% 0.03 - 0.45 Never

1.0

0.8

0.6

0.4

0.2

0.0

Ratio O/E = 0.99Con�dence interval O/E (95%) = (0.9,1.09)

Total patients = 1108 Total expected deaths (E) = 210.7

Total observed deaths (O) = 209

Calibration belt

Polynomial degree:1GiViTI Cal. Test, p <0.001

Obs

erve

d m

orta

lity

1.0

0.8

0.6

0.4

0.2

0.0

Expected mortality

0.0 0.2 0.4 0.6 0.8 1.0

Con�dence Over Under Level the bisector the bisector

80% 0.40 - 1.00 0.00 - 0.1695% 0.48 - 1.00 0.00 - 0.12

BRAIN INJURY PROFESSIONAL 27

the one hand, of centers with the highest performance in TBI treatment and, on the other, of those with poor results. In short, the GiViTI Calibration Belt is the confidence band that compares the observed with the expected probability of a dichotomous out-come (according to a given model). A statistically significant devia-tion from the null hypothesis of perfect calibration, i.e. when the model predicts what is actually observed, occurs when the 95% confidence boundary of the GiViTI Calibration Belt does not encompass the bisector (i.e. the ideal line of perfect calibration). Once a model has been demonstrated to calibrate well in all sub-sets of patients defined by the covariates considered, any deviation from the bisector by the Calibration Belt of an individual ICU means that the outcome observed in that ICU is not in line with the average performance (see Figure 1).

This approach will furnish ICUs with appropriate tools to self-evaluate weaknesses and strengths in their care performance and to enable identification of ICUs of excellence, permitting good exchange of practices and ultimately quality improvement.

DiscussionThe CREACTIVE project’s partners, fully embracing the InTBIR Initiative objectives, envisage the formation of an ever growing, open access, true international database that will include thou-sands of patients from the whole spectrum of Traumatic Brain Injury and hundreds of clinical, radiological, biochemical, hor-monal, genetic, cognitive parameters/derived parameters, provid-ing the research community with a valuable tool for TBI research. Limitations of the project’s epidemiological approach would pos-sibly include: a) failure to recruit a valid and representative num-ber of patients with moderate to severe TBI from each country, thus limiting their interpretation, b) failure to achieve high-level of compliance in follow up studies, rendering research based on outcome indices and construction of a reliable predictive model, difficult to pursue. Regarding collection of clinical, physiological, biological and radiological data, this attempt constitutes a serious world-leading step towards the establishment of a BIG DATA fa-cility, dedicated to Traumatic Brain Injury. Entering the daunting era of BIG data, offers unique opportunities but also creates an unprecedented bulk of null hypotheses! Running diverse experi-ments on this huge testing bed and checking for pathophysiologi-cal interrelations, many research lines in basic, translational as well as industrial research shall be abandoned and new will emerge.

InTBIR will soon face the ultimate challenge to deploy the most appropriate IT infrastructure and service architecture(s) re-quired to link, share or combine disparate datasets, as well as the correct data structure(s) to enable natural language processing, trend analysis and outcome prediction. It will also be called to face the diverse ethical andlegal frameworks for the international/trans-atlantic transfer of medical-clinical data. At the same time, the research community across the scientific disciplines dealing with Traumatic Brain Injury, should identify areas of greatest need that would benefit most from or are most amenable to big data cogni-tive analytical approaches and specific projects should be designed.

RefeRences1. Maas, A.I., N. Stocchetti, and R. Bullock, Moderate and severe traumatic brain injury in

adults. Lancet Neurol, 2008. 7(8): p. 728-41.

2. Corso, P., et al., Incidence and lifetime costs of injuries in the United States. Inj Prev, 2006. 12(4): p. 212-8.

3. Corrigan, J.D., A.W. Selassie, and J.A. Orman, The epidemiology of traumatic brain in-jury. J Head Trauma Rehabil, 2010. 25(2): p. 72-80.

4. The Lancet, N., A rally for traumatic brain injury research. Lancet Neurol, 2013. 12(12): p. 1127.

5. Maas, A.I., et al., Re-orientation of clinical research in traumatic brain injury: report of an international workshop on comparative effectiveness research. J Neurotrauma, 2012. 29(1): p. 32-46.

6. Adelson, P.D., et al., Common data elements for pediatric traumatic brain injury: recom-mendations from the working group on demographics and clinical assessment. J Neu-rotrauma, 2012. 29(4): p. 639-53.

7. Kaloupek, D.G., et al., Common data elements for posttraumatic stress disorder re-search. Arch Phys Med Rehabil, 2010. 91(11): p. 1684-91.

8. Maas, A.I., et al., Common data elements for traumatic brain injury: recommendations from the interagency working group on demographics and clinical assessment. Arch Phys Med Rehabil, 2010. 91(11): p. 1641-9.

9. Frey, L.C., Epidemiology of posttraumatic epilepsy: a critical review. Epilepsia, 2003. 44 Suppl 10: p. 11-7.

10. Styrke, J., et al., Traumatic brain injuries in a well-defined population: epidemiological aspects and severity. J Neurotrauma, 2007. 24(9): p. 1425-36.

11. Tagliaferri, F., et al., A systematic review of brain injury epidemiology in Europe. Acta Neurochir (Wien), 2006. 148(3): p. 255-68; discussion 268.

12. Andriessen, T.M., et al., Epidemiology, severity classification, and outcome of moder-ate and severe traumatic brain injury: a prospective multicenter study. J Neurotrauma, 2011. 28(10): p. 2019-31.

13. Rossi, C., et al., MargheritaPROSAFE project. Promoting patient safety research and quality improvement in critical care medicine. Report 2012. 2013, Bergamo (IT): Sestante Edizioni.

14. Ghajar, J., Traumatic brain injury. Lancet, 2000. 356(9233): p. 923-9.

15. Farahvar, A., et al., Response to intracranial hypertension treatment as a predictor of death in patients with severe traumatic brain injury. J Neurosurg, 2011. 114(5): p. 1471-8.

16. Hunter, J.V., et al., Emerging imaging tools for use with traumatic brain injury research. J Neurotrauma, 2012. 29(4): p. 654-71.

17. Hergenroeder, G.W., et al., Biomarkers in the clinical diagnosis and management of trau-matic brain injury. Mol Diagn Ther, 2008. 12(6): p. 345-58.

18. Weaver, S.M., et al., Genetic polymorphisms and traumatic brain injury: the contribu-tion of individual differences to recovery. Brain Imaging Behav, 2012.

19. Donkin, J.J. and R. Vink, Mechanisms of cerebral edema in traumatic brain injury: therapeutic developments. Curr Opin Neurol, 2010. 23(3): p. 293-9.

20. Perel, P., et al., Systematic review of prognostic models in traumatic brain injury. BMC Medical Informatics and Decision Making, 2006. 6(1): p. 38.

21. Steyerberg, E.W., et al., Predicting outcome after traumatic brain injury: development and international validation of prognostic scores based on admission characteristics. PLoS Med, 2008. 5(8): p. e165; discussion e165.

22. Poole, D. and G. Bertolini, Outcome-based benchmarking in the ICU Part I: Statistical tools for the creation and validation of severity scores, in Patient Safety and Quality of Care in Intensive Care Medicine, J.-D. Chiche, et al., Editors. 2009, MWV: Berlin. p. 141-149.

23. Poole, D. and G. Bertolini, Outcome-based benchmarking in the ICU Part II: Use and limitation of severity scores in critical care, in Patient Safety and Quality of Care in Intensive Care Medicine, J.-D. Chiche, et al., Editors. 2009, MWV: Berlin. p. 151-159.

24. Finazzi, S., et al., Calibration belt for quality-of-care assessment based on dichotomous outcomes. PLoS One, 2011. 6(2): p. e16110.

25. Nattino, G., S. Finazzi, and G. Bertolini, A new calibration test and a reappraisal of the calibration belt for the assessment of prediction models based on dichotomous out-comes. Stat Med, 2014.

26. Rothman, K.J., S. Greenland, and T.L. Lash, Modern Epidemiology. Second edition ed. 2008, Philadelphia, PA: Lippincott Williams & Wilkins.

27. Poole, D., et al., Comparison between SAPS II and SAPS 3 in predicting hospital mortal-ity in a cohort of 103 Italian ICUs. Is new always better? Intensive Care Med, 2012. 38(8): p. 1280-8.

ABOUT THe AUTHORsGuido Bertolini, Giulia Paci, Joanne Fleming - GiViTI Coordinating Cen-ter, IRCCS-Istituto di Ricerche Farmacologiche “Mario Negri”: Centro di Ricerche Cliniche per le Malattie Rare Aldo e Cele Daccò, Ranica, Ber-gamo – Italy; Luca Antiga - Orobix, Bergamo – Italy; Akos Csomos - Sem-melweis University, Budapest – Hungary; Rafael Kaps - General Hospital Novo Mesto – Slovenia; Isaac Lazar - Soroka University Medical Center and the Faculty of Health Sciences, Ben Gurion University – Israel; Mal-gorzata Mikaszewska - Medical University of Warsaw – Poland; Matteo Mondini1, Nektaria Xirouchaki - University of Crete Medical School, University Hospital of Heraklion – Greece; Roberto Latini - IRCCS – Is-tituto di Ricerche Farmacologiche “Mario Negri”: Dept of Cardiovascular Diseases, Milano – Italy; Primoz Gradisek - University Medical Center Ljubljana, Ljubljana - Slovenia; Theodoros Kyprianou - Medical School, University of Nicosia & Nicosia General Hospital – CyprusThe study was funded by the European Commission in the context of the 7° Framework Program (contract n. 602714)To contact the corresponding author: Guido Bertolini, GiViTI Coordinat-ing Center, Istituto di Ricerche Farmacologiche Mario Negri, Centro di Ricerche Cliniche per le Malattie Rare Aldo, e Cele Daccò, 24020 Ranica (Bergamo), Italy. E-mail: guido.bertolini@marionegri.it.