1. Palubiski L & Crizzle AM CARSP Conference Monday June 6, 2016
Predictors of Driving Performance and Resumption of Driving in Persons with Traumatic Brain Injury: An Evidence-Based Review
Palubiski L & Crizzle AM
CARSP Conference
Monday June 6, 2016

2. What is TBI?
Trauma (or external force) causing altered neurological function
Effects: Minor to major symptoms
Symptoms can be detected using standard clinical tests 3 months post-injury (Belanger et al. 2005)
Can impact IADL’s and ADL’s such as driving

3. Impaired Driving in TBI
Compared to controls, those with TBI
Perform more poorly on indicators of driving
hazard perceptions tests (Preece et al. 2010)
information processing speed (Stokx et al. 1986)
reaction time (Stokx et al. 1986)
Inattention (Rike et al. 2011)
being an aggressive driver (Ilie et al. 2015)
Have higher crash rates (Ilie et al. 2015)

4. Resumption of Driving Limited number of studies
On-road driver rehabilitation followed by on-road assessment was associated with a higher probability of return to driving after TBI (Ross et al,. 2015)
Posttraumatic amnesia duration, along with the presence of visual and physical impairment classified 88% of the pass group and 71% of those needing rehabilitation (Ross et al,. 2015)

5. Study Purpose Determine Levels of Evidence
Clinical tests used for driving assessment
Guidelines for return to driving
Provide recommendations for using various clinical tests to assess driving impairment and return to driving in TBI drivers

6. Methods Searched Various Databases (ie., Pudmed, CINAHL)
1652 citations
Reviewed 33 full-texts
Included 20 studies based on following
Published between 2000 and 2015
Primary studies
Quantitative
Driving was the primary outcome measure
Samples only composed of TBI drivers
Used AAN criteria (Edlund et al. 2004)

7. AAN Rating by Class Class I Class II Class III Class IV
Prospective study in a broad spectrum (N≥100)
Criterion standard for
the case definition
Blinded evaluation
Compared to controls
Prospective study of a narrow spectrum of persons (N<100 )
Retrospective study of a broad spectrum
Criterion standard
Retrospective study where either persons with the established condition or controls are of a narrow spectrum (N < 100)
The reference standard, if not objective, is applied by someone other than the person performing the test.
Any design where the test is not applied in an independent evaluation OR evidence provided by the expert opinion alone or in descriptive case series (without controls).

8. AAN Recommendations Level A Level B Level C Level U
Rating by Recommendation
Recommendation: Established as effective/ useful/ or predictive or not. “Should be done, or should not be done”
Recommendation: Probably effective/ useful/ or predictive, or not. “Should be considered, or should not be considered”
Recommendation: Possibly effective/ useful/ or predictive, or not. “May be considered, or may not be considered”
No recommendation
Condition for rating by recommendation
Requires two consistent Class I studies, or one Class I study where the magnitude of the effect is large, and all criteria have been met.
Requires at least one Class I study, or two consistent Class II studies.
Requires at least one Class II study, or two consistent Class III studies.
Data inadequate or conflicting. Given the current knowledge or test, the treatment is unproven.

9. On-Road Studies Results Conclusions. Recommendations.
three class II studies (Ross et al. 2015; Lew et al, 2005; Neyens et al. 2015) and one class III study (Novack et al. 2006)
The sample sizes ranged from 27 (Lew et al. 2005) to 207 (Ross et al. 2015)
mild to severe TBI (Ross et al. 2015; Neyens et al. 2015) or moderate to severe TBI (Lew et al. 2005; Novack et al. 2006) with a time (mean) post-TBI that ranged from 8 months [Lew et al. 2005) to 12.7 years (Neyens et al. 2015).
Conclusions.
Class II:
poorer performance on a driving simulator was predictive of on-road driving in individuals with moderate to severe TBI (Lew et al. 2005)
driver distraction (selecting a CD, radio tuning or coin sorting) did not significantly impair on-road driving (Neyens et al. 2015)
Class III
PTA duration, the presence of physical and/or visual impairment, and slower reaction time predicted failing the on-road test. However, scores on the GCS did not predict on-road driving (Ross et al. 2015)
A class III study found that younger age and slower times on the TMT-B and UFOV-2 test predicted failing a road test; whereas TMT-A did not (Novack et al. 2006)
Recommendations.
Level C: PTA duration, visual/physical impairment, reaction time and performance on a driving simulator
Conversely, GCS scores are possibly not predictive
Level U: age, TMT A & B and UFOV 2

10. Simulator Studies Results Conclusions: Recommendations:
One class III study (Beaulieu-Bonneau et al ) and one class IV study (Gamache et al. 2011)
The sample sizes ranged from one (Gamache et al. 2011) to 44 (Beaulieu-Bonneau et al ). The class III study included individuals with moderate to severe TBI with a time (mean) post-TBI of 53 months (Beaulieu-Bonneau et al )
The class IV study was a case study with one individual with severe TBI, followed over four months of training (Gamache et al. 2011)
Conclusions:
Simulated driving performance did not differ significantly between controls and TBI groups in one class III study (Beaulieu-Bonneau et al )
Recommendations: Level U: There is no data (or evidence) due to the limited number of studies examining simulated driving performance
Recommendations:
Level U: There is no data (or evidence) due to the limited number of studies examining simulated driving performance

11. Off-Road Screening Tests
Results:
Two class III studies (Preece et al. 2010; 2011)
The sample sizes ranged from 55 (Preece et al 2011) to 85 (Preece et al. 2010)
mild (20) and mild to severe (Preece et al. 2010)
Time post-TBI of 10.2 hours (Preece et al. 2010) and days (Preece et al. 2011)
Conclusions:
Both studies found that HPT scores were significantly worse in the TBI group than controls
GCS scores were not related with performance on the HPT (response time)
Recommendations:
Level C: GCS scores are possibly not predictive

12. Return to Driving Results Conclusions: Recommendations:
Eight primary studies related to driving status (Baker et al. 2015; Leon-Carrion et al. 2006; Novack et al. 2010; Rapport et al. 2006; Coleman et al. 2002; Labbe et al. 2014; Hawley, 2001; Pietrapiana et al. 2005)
Conclusions:
One Class III study found that lower GCS scores were related to return to driving (driving status) at one, two, and five years post-TBI (Novack et al. 2010) but not in another study (Coleman et al. 2002)
TMT A & B related to return to driving in one Class III study (Labbe et al. 2014)
Higher scores (indicating better performance) on the Matrix Reasoning Test on the Wecshler Adult Intelligence Scale (WAIS) was related to driving status in a Class III study (Coleman et al. 2002)
Two Class III studies found that higher scores on the Functional Independence Measure (FIM) - Functional Assessment Measure (FAM; indicating less severity) were related to an increased likelihood of driving (Novack et al. 2010; Hawley, 2001)
One class III study found that time to complete home maze test was predictive of return-to-drive 2 weeks after sustaining a TBI (Baker et al. 2015)
In a class III study, caregiver perceptions were related to non-driving (Rapport et al. 2006)
Recommendations:
Level C: FIM scores are possibly predictive of driving status based on two Class III studies; Level U: Evidence concerning GCS score, Trails A & B scores, WAIS, Drive Home Maze test and caregiver perceptions

13. Discussion
Limited information on how to assess for fitness to drive among TBI patients
sample characteristics (e.g. TBI severity)
sample size
clinical tests performed
outcome measures
length of follow-up periods
Evidence based assessment tools
Based on research
Time efficient
Cost effective

14. Best Advice
Refer for a Comprehensive Driving Evaluation to a DRS or CDRS
Further studies are needed to address impact of TBI on driving performance
Class I studies needed
Level A recommendations needed

15. Questions or Comments?

16. Contact Information
Dr. Alexander M. Crizzle, PhD, MPH
Assistant Professor
School of Public Health and Health Systems
University of Waterloo