SLIDE 1 Recovery is in the eye
How visual fixation can predict TBI patients’ outcome
Caroline Arbour, RN, PhD 12th annual conference, Brain Injury Canada September 30th, 2015
SLIDE 2 Annual incidence of TBI in Quebec
Abitibi- Témiscamingue Montérégie
About 400 moderate/severe TBI are admitted in ICUs Of them, 65% will survive their injuries and return home
Outaouais Nord du Québec Baie James Saguenay Lac-St-Jean Côte-Nord Québec Centre Estrie
SLIDE 3 The initial Glasgow score
(Teasdale & Jennett, 1974)
Lowest score in the first 24h post-TBI and prior to intubation
Used to index severity of TBI
Severe - score 3-8 Moderate - score 9-12
More or less performant for the prediction of TBI prognosis
But remain widely used as it is the
- nly marker available within 24h
Eye opening
- Spontaneous
- To speech
- To pain
- None
4 3 2 1 Verbal response
- Oriented
- Confused
- Words (inappropriate)
- Sounds
- None
5 4 3 2 1 Motor response
- Obey commands
- Localise pain
- Normal flexion
- Decortication
- Decerebration
- None
6 5 4 3 2 1
SLIDE 4 Post-traumatic amnesia - PTA
(GOAT: Levin, O’Donnell & Grossman, 1979)
Retrograde amnesia
TBI
APT assessment Recovery of ability to create new memory Coma Glasgow score of ≥9
Known as the most robust predictor of TBI prognosis However, takes several days/weeks to assess
SLIDE 5 Recovery of visual fixation post-TBI
Already assessed regularly by nurses in the study setting
Fixation was considered present…
- If sustained 2sec
- On 2 consecutive times
Otherwise considered absent
Published with patient’s consent
SLIDE 6
Global recovery in the months following TBI? PTA duration?
Two-phase cohort study
Does recovery of visual fixation at 24h post-ICU admission can predict better than the initial Glasgow score…
SLIDE 7
Markers of recovery
Phase 1 – During hospitalisation
PTA duration
≤ 14 days indicated good recovery > 14 days indicated poorer recovery
Phase 2 – 1 to 3 years post-TBI
Glasgow Outcome Scale – Extended (GOSE score) Visual attention task (in sec) Structural brain deficits (cerebellum, thalamus, PFC)
SLIDE 8 181 m/s TBI discharged alive between 2010-2013
Not eligible for further testing n=45
Study sample
Phase 1 Follow up Phase 2
GOSE score n=92 Attention task n=32 MRI n=20 Unlocatable n=89 Refused n=15
SLIDE 9 Comparing the sub-samples considered for GOSE testing
Variables n (%) or mean (SD) Included n=92 Unlocatable n=89 Group differences Gender (Male) Age ICU stay (days) Initial Glasgow score Hrs without fixation in the ICU
- Cum. dose of Fentanyl adm. first 24h
- Cum. dose of Propofol adm. first 24h
CT scan findings Midline shift Compressed cisterns Subdural hemorrhage Subarachnoid hemorrhage Swelling 71 (77%) 34 14 14 13 8 ± 3 175 162 1318 951 1406 1271 10 (11%) 21 (23%) 47 (51%) 40 (43%) 40 (43%) 74 (83%) 35 15 14 13 7 ± 3 143 127 1071 1019 2056 1731 9 (10%) 17 (19%) 47 (53%) 38 (43%) 34 (38%) ns ns ns ns ns ns ns ns ns ns ns ns
SLIDE 10 Comparing the sub-samples considered for attention testing
Variables n (%) or mean (SD) Included n=32 Not eligible n=45 Eligible but refused n=15 Group differences
Gender (Male) Age ICU stay (days) Initial Glasgow score Hrs without fixation in the ICU
- Cum. dose of Fentanyl adm. first 24h
- Cum. dose of Propofol adm. first 24h
CT scan findings Midline shift Compressed cisterns Subdural hemorrhage Subarachnoid hemorrhage Swelling 21 (66%) 34 15 16 15 9 ± 4 184 171 1326 699 1583 1137 3 (9%) 9 (29%) 17 (54%) 11 (33%) 13 (41%) 33 (73%) 36 14 14 13 8 ± 3 182 162 1200 979 1400 910 4 (9%) 11 (24%) 24 (53%) 19 (42%) 17 (38%) 11 (73%) 37 ± 14 8 ± 4 7 ± 2 118 ± 111 1825 ± 530 900 ± 636 1 (7%) 4 (28%) 8 (53%) 6 (40%) 8 (53%) ns ns ns ns ns ns ns ns ns ns ns ns
SLIDE 11 Phase I (N=181)
Variables n (%) or mean (SD) Fixation at 24h PRESENT n=103 Fixation at 24h ABSENT n=78 Group differences Gender (Male) Age ICU stay (days) Initial Glasgow score Hrs without fixation in the ICU
- Cum. dose of Fentanyl adm. first 24h
- Cum. dose of Propofol adm. first 24h
PTA duration (days) 76 (74%) 34 16 7 6 8 ± 3 7 ± 6 1187 1029 1759 1443 6 ± 5 62 (79%) 35 15 20 13 7 ± 3 352 ± 170 1392 1323 1793 1533 23 ± 17 ns ns 0.000 ns 0.000 ns ns 0.000
SLIDE 12 20 40 60 80 100 Courbe ROC pour prédiction APT < 14 jours 20 40 60 80 100 100-Spécificité Sensibilité Hr sans fixation à l'USI - AUC 89% Score de Glasgow initial - AUC 67% 12h 36h 12
At 24h, recovery of visual fixation showed: Sensitivity of 84% Specificity of 82%
sensitivity
ROC curve for the prediction of PTA 14days
100-specificity
- ---- Hrs without fixation in the ICU – AUC 89%
- ---- Initial Glasgow score – AUC 67%
SLIDE 13 Phase II – GOSE score (n=92)
Variables B SE (B) t Sig PTA duration
0.014
0.015 Initial Glasgow score 0.038 0.067 0.100 0.560 ns Hrs without fixation in the ICU
0.001
0.003
SLIDE 14
Phase II – Attention task (n=32)
Variables B SE (B) t Sig PTA duration 1.072 0.428 0.496 2.508 0.021 Initial Glasgow score 0.865 1.832 0.086 0.472 ns Hrs without fixation in the ICU 3.623 1.490 0.441 2.431 0.023
SLIDE 15
SLIDE 16
Post-hoc analysis
What about the prediction of intra-hospital mortality?
We reviewed the medical charts of all TBI patients deceased in the ICU (n=52) over a period of 12 months (Jan-Dec 2013)… Of all patients who died during hospitalization, only one was able to fixate at 24h post-ICU admission
SLIDE 17 Conclusions
This study is the first to show that:
- Visual fixation recovery is a strong predictor of PTA duration
- It was also found to be as performant as PTA for predicting TBI
long term outcomes Visual fixation could be considered by clinicians for the estimation
- f TBI recovery, especially when PTA is under investigation
SLIDE 18
Acknowledgments
Co-authors
Andrée-Ann Baril, PhD (c)
Francis Bernard, MD, FRCSC Marie-Aurélie Bruno, PhD Jean-François Giguère, MD, PhD, FRCSC Danielle Gilbert, MD Nadia Gosselin, PhD Steven Laureys, MD, PhD Gilles Lavigne, DMD, PhD
SLIDE 19
Acknowledgments
We would like to thank the ICU nursing team of Sacred-Heart Hospital in Montreal
Funding agencies
SLIDE 20
Questions
